SITE MANAGEMENT SIG Message # 288 0 replies Reply to # 279 Left by: PHILLIP CREWS On 26-May-89 At : AMEX HSG 704 333 1276 Subject: SHUTDOWN FROM BATCH You can always shutdown a system (from Batch, a network task, or an interactive process) with: $ MCR OPCCRASH If you want to dismount disks first (to avoid quota and extent rebuilds on boot), use $ DISMOUNT/ABORT/NOUNLOAD for each one before the above command. If you want to cause an automatic reboot, add the following command: $ DEFINE OPC$REBOOT Y Beware: processes are not stopped normally, queues are not terminated with grace, and file modification dates may not be updated for open files if ACP_WRITEBACK is set to 1. Using OPCCRASH also avoids messy warning messages letting those irritating users know what is going on. ;-) SITE MANAGEMENT SIG Message # 289 0 replies Reply to # 285 Left by: PHILLIP CREWS On 26-May-89 At : AMEX HSG 704 333 1276 Subject: PAGE READ ERROR Page read errors are due to disk errors. These may be in either the pagefile or any image or section file. Sounds like some bad blocks are appearing on your disk. Check error log for bad block replacements] and disk errors and call field service. CLUSTERING/NETWORKING SIG Message # 393 0 replies Reply to # 390 Left by: PHILLIP CREWS On 26-May-89 At : AMEX HSG 704 333 1276 Subject: BACKUP VIA DECNET Re: BACKUP handling tape errors. Almost all bad spots on tapes will be passed over by the magtape driver. Backup may be somewhat more forgiving, but if tapes are that bad, you really don't want to use them anyway. BACKUP creates XOR redundancy records within a saveset regardless of its destination (unless you specify /GROUP=0). These redundancy records will be written to the tape along with the rest of the data, and will be used in a restore if any single record is unreadable. VMS version 5.2 will include improvements to the BACKUP utility so that it will avoid the CRC instruction, which is emulated on all MicroVAX and 6xxx processors. This should speed up backups that use /CRC considerably. VMS SIG Message # 957 0 replies Reply to # 951 Left by: PHILLIP CREWS On 26-May-89 At : AMEX HSG 704 333 1276 Subject: DETACHED PROCESS PARAMETERS Several possibilities: 1. Create a file with the parameters in it, and specify the file as input to the detached image: $ CREATE TEMP.TXT parameter 1 parameter 2 $ RUN/DETACHED/INPUT=TEMP.TXT image-name The program can read the records from SYS$INPUT as its parameters. 2. Use a mailbox. It must be created by a program (no DCL command to do it), but then DCL can write the parameters to the mailbox, and the detached image can read them. 3. Define /SYSTEM or /GROUP logical names with parameter values. The program can translate the logicals and use the results as the parameters. I hope this helps. MASS STORAGE SIG Message # 258 0 replies Reply to # 256 Left by: BERN FOX On 26-May-89 At : (313) 994-1200 Subject: SPURIOUS MOUNT VERIFIES... ...sounds like Kitty has pointed you in the right direction on this problem Gabriel. Is DUB0: the only drive on this QD21? Has the error logger noted any problems, other than the usual init sequence failures? Seems to me that the Fugi might have some sort of a status/ fault LEDs on the logic board, try to monitor these for voltage fault and/or seek error condidtions, either of which might render the drive unavailable to the controller. Try: ANA/DISK/READ DUB0:, if any glitches, repeat ANA/DISK/READ DUB0:/REPAIR. If problems persist, a waltz through the Emulex format & verify routine may in order. Good luck! - bern - VMS SIG Message # 958 0 replies Left by: DAVID MCCALLIE On 26-May-89 At : (617) 735-8380 Subject: PSEUDO-TERMINAL-DRVR Hello. I need some advice on how I can have one process pass information to another process where the second process is expecting input from a terminal type device (and is in fact in 'passthru' mode.) I tried mailboxes but the system complains that the MBAn: is not a terminal device. I need something like the 'pty' pseudoterminals that Unix offers. I suspect that I can take advantage of the 'virtual terminal' (VTA:) device, but I don't know how to create a process that is expecting input from a VTA device (other than to login), nor how to read/write to that processs if I got it running with a VTA (I assume I can just QIO to it, but I haven't been able to try yet. How does DCL implement the 'disconnect' and 'connect' commands ? Is there a system service that I can use from within a program to do similar tricks? Thanks in advance to anyone who can help ! VMS SIG Message # 959 0 replies Left by: DAVID MCCALLIE On 26-May-89 At : (617) 735-8380 Subject: LATS APP-TO-APP ? I know that with the LATS protocol one can communicate from terminal server to a dedicated application, and conversly from an application to a dedicated port on a terminal server - but - can LATS communicate from one application to another _application_ ? In other words, can LATS be used as a substitute for DECNET task-to-task communication ? Thanks. RSTS SIG Message # 382 0 replies Reply to # 369 Left by: KELVIN SMITH On 27-May-89 At : (203) 357-0504 Subject: COMING LIVE TO YOU FROM... Dave, I live and work in Stamford, that silly city that can't decide whether it wants to be a suburb of New York or a corporate center in its own right. As one might expect, we've got the worst of both worlds: The streets are the right size for a small town, while the traffic is in amounts for a large city, etc. For future reference, my first name is Kelvin, not Kevin. Don't feel bad about misreading it; I think half the people I come in contact with think the name is either Kevin or Calvin, depending on whether they see it or hear it first. Just think of Kelvinator refrigerators and the Kelvin temperature scale (0 degrees Kelvin is absolute zero). I can't claim credit for either use of the name. RSX SIG Message # 275 1 replies Reply to # 273 Left by: KELVIN SMITH On 27-May-89 At : (203) 357-0504 Subject: INFO FROM ATLANTA Your best option to find out about a seminar at the Atlanta DECUS symposium may be to get an audio tape. There's a firm in Maryland somewhere that handles it all for DECUS, I believe by the name of Chesapeake Communications. I don't remember their address, and I'm not at my office, but I'm sure if you contacted DECUS headquarters they could tell you (address and phone number in the back of some operating systems manuals and the Software Dispatch). The tapes are $8, a cheap price if it's an important topic. The problem is that sometimes the speakers make a lot of references to handouts or session notes, which may be harder to find post-Symposium. However, session notes are combined for an entire SIG, so you may be able to find someone with session notes (if any) for your seminar who doesn't even want them. RSX SIG Message # 276 0 replies Reply to # 275 Left by: KELVIN SMITH On 27-May-89 At : (203) 357-0504 Subject: DECUS SPEAKER ACCESS Just re-read the note I was responding to--you wanted to get in touch with the actual speaker. The best way would be to contact the honchos of the sponsoring SIG. If you get the DECUS SIGs Newsletter, their names, addresses, and phone numbers are in the back of the Newsletter each month. If not, I'd suggest you call DECUS headquarters to get the name of the SIG chair or someone else like that. Since the SIGs actually schedule the seminars, they'd be most likely to have all the details on the speakers, although DECUS HQ may have it as well (if they're allowed to give it out). RSTS SIG Message # 383 0 replies Reply to # 378 Left by: KELVIN SMITH On 27-May-89 At : (203) 357-0504 Subject: CLE IS SAFE (I THINK) We've been using CLE for close to a year, and I've never noticed any problems, except for building up a slew of 1K user libraries. The only memory problems that I remember that caused system crashes were associated with making certain changes to the size (or existence, perhaps) of a virtual disk; when you brought up the system and initialized DV0:, the system crashed. However, that's been quite a while; I don't remember the precise circumstances, nor which version of RSTS was affected, nor if it has since been fixed. If it's bugging you, call me during the day and I'll go through my SPR files. I'm very impressed with RSTS's lack of crashing. I don't think I've had a system crash on me (aside from power failure) in the past 2 years. Maybe this will get some juices flowing, but I haven't heard of a lot of VAX shops with such uninterrupted service. I tend to think that no faults is better than "fault tolerant," although I recognize that for some situations you have to assume the worst. --- It's coming to me now: If you move a RSTS SIL from one machine to a second with a larger amount of memory, and INIT automatically resizes the memory, then if your virtual disk is defined in the section of memory that didn't exist on the original machine, the system crashes when you initialize DV0:. The workaround is to go through the memory allocation subsection of DEFAULT in INIT. I don't remember if you actually had to make changes, or if just going through DEFAULT was sufficient. In any case, that's the only memory problem I remember that crashed my system. Whether others with different configurations (e.g., on a DECnet) would have other problems, I don't know. RSTS SIG Message # 384 0 replies Reply to # 383 Left by: Phil Anthony On 30-May-89 At : Here Subject: VIRTUAL DISKS AND CRASHES Dave, check out Messages 350, 352, and 359 in the RSTS SIG for Brett and Kelvin's war stories - information on system crashes after messing around with the size of a virtual disk. Kelvin, I recall I owe you some information, but I can't remember off the top of my head what it was. I went searching for your letter, and I'm sure it's on the stack of papers that reliable sources tell me has a desk under- neath it, but I can't put my hands on it right now. Tickle my memory - was it information on the Epsilon editor, which is available from (among others) the Programmer's Shop, 5-D Pond Park Road, Hingham, Mass. 02043 - (800) 421-8006? (That's where I got my copy.) A fairly recent issue of Dr. Dobb's Journal has their price at $149, but check it out before ordering. VMS SIG Message # 960 1 replies Left by: BARRY YOUNG On 30-May-89 At : JEDDAH (02)-647-0773 Subject: DECSERVERS/AUTOLOGIN HAVING JUST RECEIVED 11 DECSERVER 200'S I AM VAINLY ATTEMPTING TO CONFIGURE THEM INTO OUR TERMINAL NETWORK, BUT THERE SEEMS TO BE AN ENORMOUS GAP BETWEEN THE VAX DOCUMENTATION SET AND THE DECSERVER MANUALS. DOES ANYONE KNOW OF A METHOD OF MAINTAINING AUTOMATIC LOGIN FACILITIES VIA DECSERVER PORTS. I'VE TRIED USING ALFMAINT, BUT THAT EXPECTS A FIXED TERMINAL ID WHICH DOES NOT EXIST WHEN USING DECSERVERS. IS IT POSSIBLE TO CREATE PERMANENT TERMINAL ID'S ON A DECSERVER. IS THERE A WAY OF CONNECTING THE SERVICE NAMES OFFERED BY THE VAX HOST SYSTEM TO A LOGIN USERNAME ACCOUNT, THEREFORE WHEN CONNECTING TO A SERVICE THE USER IS AUTOMATICALLY LOGGED IN UNDER A PARTICULAR USERNAME. ANY HELP ON THIS SUBJECT WOULD BE MUCH APPRECIATED AS THE DEC SOFTWARE SUPPORT IN THE MIDDLE EAST STINKS !! REGARDS VMS SIG Message # 961 0 replies Left by: BOB ZISEK On 30-May-89 At : (203)798-5240 Subject: CHARACTER ATTRIBUTES Because of the lack of time to fool around with this, I am inquiring about the way to include the enhanced character attributes (bold,reverse video,underline, etc.) in a Mail message. I know all about the different escape code definitions, however, I'm not sure how to get them into a mail message. This comes in handy when making an important point in mail. Also, is there a way to incorporate these attributes into help files? Thanks in advance. Bob Zisek Boehringer Ingelheim Pharmaceuticals Inc. 90 East Ridge Ridgefield, Ct. 06877 VMS SIG Message # 962 0 replies Reply to # 960 Left by: ROBERT G. SCHAFFRATH On 30-May-89 At : 914-335-9792 Subject: Autologins from a DECserver (Second Printing) Sometime ago I had reported the method of doing autologins on a DECserver. It is not documented anywhere and not currently supported due to the fact that ALFMAINT.COM has not been modified to handle the naming format for DECserver ports. First, you MUST be on a V5.0 or above system. Secondly, determine which DECservers and ports you are going to have 'autologin'. For example, I have a DECserver called DS200R and I want ports 5 and 6 to have autologin capability. The VMS specification for these ports would be 'DS200R/PORT_5' and 'DS200R/PORT_6' respectively. This information can be obtained by logging in on the port you want to have autologin and typing, in DCL, 'WRITE SYS$OUTPUT F$GETDVI("TT","TT_ACCPORNAM")'. This will return the information that VMS will se when the port connects. Next, you will have to make a modified copy of ALFMAINT.COM. In this modified .COM file (call it PORT_ALFMAINT), remove all references to F$GETDVI and make sure that ALL symbols which use information returned by F$GETDVI (other than EXISTS) are equated properly. One location for example, says something like 'FDEVNAM = F$GETDVI(DEVNAM,"FULLDEVNAM")'. You must replace this with 'FDEVNAM = DEVNAM'. Once this is done, type @PORT_ALFMAINT and start answering the questions. When you are asked for a terminal, respond with 'DS200R/PORT_5'. As you can see, this is not a device name and for that reason all F$GETDVI checks must be removed or you will get errors. Finally, on your DECservers, don't forget to DEDICATE the ports to the appropriate service. Otherwise your users will get a 'Local>' prompt. Any questions or if you want a pre-modified ALFMAINT.COM, feel free to call me. Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 P.S. Sorry folks for the long 'letter'! SITE MANAGEMENT SIG Message # 290 0 replies Reply to # 284 Left by: MARTY JOHNSON On 30-May-89 At : (203)666-1737 Subject: IMAGE ACCOUNTING Why don't you just turn on image accounting? (SET ACCOUNT/ENABLE=IMAGE) This will log who, when and what image in the accounting log file. Take care as this can chew up over 4000 blocks a day on a busy system! Use the accounting utility, especially the /SUMMARY and /REPORT qualifiers, to tailor a report to fit your needs. Hope this helps, Marty MISCELLANEOUS SIG Message # 510 0 replies Left by: DAVID SOMMERS On 30-May-89 At : (206) 525-3130 Subject: 9-TRACK FOR VAX-3100 DOES ANYONE KNOW IF THERE IS A 9-TRACK TAPE DRIVE FOR THE VAXSTATION 3100?? I UNDERSTAND THAT THERE IS A DRIVE MADE BY KENNEDY WHICH HAS A SCSSI INTERFACE BUT I CAN'T FIND ANYONE WITH DEVICE DRIVERS FOR IT. ANYONE?? CLUSTERING/NETWORKING SIG Message # 394 0 replies Left by: BRIAN HULSE On 31-May-89 At : 313 468-3252 Subject: SI'S CLUSTOR I saw the note from Jared Hecker for SI's CLUSTOR product and found it very interesting. I've spoken to SI, but was hoping to find more comments from current users. If there are any others out there using this product, what are your experiences? (Jared, are you still as enthusiastic as you were in Oct?) Thanks. VMS SIG Message # 963 0 replies Reply to # 961 Left by: MARTY JOHNSON On 31-May-89 At : (203)666-1737 Subject: MAIL As far as I know, you can't display escape sequences in mail. If you embed them in your message (using EDT or whatever), MAIL displays the escape sequence, displaying a "$" for the escape character. For example, the escape sequence for bolding, "[1m", is displayed as "$[1m". However, if you EXTRACT the document, even to the terminal (EXTRACT TT:), the escape sequences are preserved. Which means, as long as you are reading mail messages within the mail utility, you only have UPPER CASE, and the **** imaginative use **** of your keyboard, for emphasis. :-) Marty CLUSTERING/NETWORKING SIG Message # 395 0 replies Left by: DUANE SADOWSKI On 31-May-89 At : (212)577-0543 Subject: PC LAN - VAX LAN I'm exploring ways of allowing users of a Novell PC LAN at one site to use VAXs on an Ethernet DECnet at a different site. One suggestion I received was to put an X.25 switch on the PC LAN and to use TCP/IP and X.25 protocols across a dedicated phone line. My question is what sort of hardware and software would be necessary at the VAX site under such a scheme. Would the VAXs themselves have to perform the conversion between TCP/IP - X.25 and DECnet protocols, or could a personal computer attached to the DECnet perform such conversion? If a PC could do the job, would it need to run LAT or DECnet? Also, what sorts of management problems, if any, would the X.25 scheme introduce that do not exist with asynchronous terminals? FORUM - 3RD PARTY SUPPLIERS SIG Message # 102 0 replies Reply to # 94 Left by: RICH NARDIELLO On 01-Jun-89 At : 609-764-0100 Subject: LANPAR TUBES Dave, Thanks for the reply. After begging and pleading with my superiors regarding our poor performance with these tubes, they have done the illogical: Ordered 24 more! I didn't think the 3210's were available, and for that matter have heard that Lanpar Technologies went belly-up. If you hear anything further, I'd appreciate the info. Thanks, Rich Nardiello APPLICATIONS SIG Message # 333 1 replies Left by: RICH NARDIELLO On 01-Jun-89 At : 609-764-0100 Subject: USERBASE 3.3 Looking for some information regarding the usage of the database package Userbase from Userware Int'l based in Escondido, Cal. Specifically, I'd like to create a discussion of Userbase "users" on the East Coast. Most applications appear to be installed west of the Miss. . We are running this beauty on an 8250, 4.7 VMS. No Vendor replies, please. VMS SIG Message # 964 1 replies Left by: NICHOLAS ABBOT On 01-Jun-89 At : (914) 347 2700 Subject: MODEMS, AND DTR DROPPING VAX 8530, DECserver 200/MC, modem control is enabled on a port, Telebit T1000 modem on port. DTR drops every 30 seconds if CD is low or 2 minutes if CD is high. Now if an incoming call happens to connect at the same time that VMS polls and drops DTR your disconnected and have to redial and hope that you connect and initiate a login before VMS polls again. So, depending on the interval there is a chance that a user will 'connect' and be dropped, and have to redial and that's a pain in the baud. Now that I've set the stage let's talk about my problem. There is some question as to whether this is a 'bug' or a 'feature', I say it's a feature. So let's put it to a vote. All of you who think it's a BUG call the number above ask for Hilmar and say "BUG", and if you think it's a feature, say "FEATURE", However, if you believe it is a feature and it's suppose to be that way, you must supply an explaination, if you happen to be at a loss for words a simple "because" will do. All of you who are undecided can call me and I will sway you to the "feature" side, and any of you who are on the "bug" side can call me and try to sway me and/or tell me a way to make DTR aways high and still have the modem the ability to hangup. Incidently is there a way to regulate the timing, and does TTY_SCANDELTA control that? From the front line, - NIK VMS SIG Message # 965 1 replies Reply to # 964 Left by: NICHOLAS ABBOT On 01-Jun-89 At : (914) 347 2700 Subject: ...MORE 1) How do I get out of ARIS 2) I can see that this port does not hangup during a logout... WHY? VMS SIG Message # 966 0 replies Reply to # 965 Left by: On 02-Jun-89 At : Subject: DATA/FILE TRANSFER SIG Message # 550 1 replies Left by: JOHN S. MIELE On 02-Jun-89 At : (203) 656-5149 Subject: KERMIT - VT2XX EMULATION I have installed MS/KERMIT ON MY IBM/AT COMPATIBLE. I HAVE INSTALLED THE PARADISE/VGA CARD ON THIS P.C. I AM USING KERMIT TO DIAL UP OUR VAX 8200 AT WORK USING A DIGITAL SCHOLAR MODEM. WHEN RUNNING IN TERMINAL EMULATION MODE, I AM UNABLE TO GET 132 COLUMN SCREEN WIDTH. I AM NOT SURE, BUT I RECALL HEARING OR READING THAT KERMIT TERMINAL EMULATION CAN HANDLE 132 COLUMN SCREEN WIDTHS. I CAN'T SEEM TO FIND THE DOCUMENTATION. I'D APPRECIATE ANY INSIGHT AS TO KERMIT CAPABILITIES IN TERMINAL EMULATION. ALSO, ANYONE USING ANY TYPE OF EMULATION, IF THEY COULD LET ME KNOW OF PLUSSES AND MINUSES. SINCERELY, JOHN S. MIELE COMMUNICATIONS SIG Message # 283 0 replies Reply to # 282 Left by: RICHARD B. GILBERT On 02-Jun-89 At : (609) 452-5130 Subject: HP-1000 / HP TERMINALS Stephen, You neglected to say how you would like to connect your HP1000 and your VAX. You also neglected to say what you wanted to use the connection for. We might be able to help you a little bit more if we had a little more information to work with. There are a number of possibilities, among them: "sneaker net" (carry physical media (tape or disk) between the two machines), a serial line and Kermit, and Ethernet with a TCP/IP package on each end. Unless your HP150 terminal does a DEC VT100 emulation, you will be much better off buying DEC or DEC compatible terminals to use with your VAX. HP terminals use a protocol of their own which is not the same as anyone elses and most emphaticly not the ANSI standard. DEC terminals use a superset of the ANSI standard. If you want anything more than "glass teletype" functionality from your terminal go with the DEC terminals. They are required to take full advantage of DEC's screen editors, the Phone utility, the mail utility and probably a few other things as well. DEC makes a token effort to support foreign terminals, but that is all it is. VMS SIG Message # 967 0 replies Reply to # 958 Left by: RICHARD B. GILBERT On 02-Jun-89 At : (609) 452-5130 Subject: PTY DRIVERS Kevin Carroso has written a pseudo terminal driver known as PTYDRIVER. The latest version can be built to run under either VMS V4 or VMS V5. I believe that you can get it off one of the recent DECUS VAXSIG Symposium tapes. I you can't find it, I will send you a copy if you will send me a tape and return postage. DATA/FILE TRANSFER SIG Message # 551 0 replies Reply to # 550 Left by: Phil Anthony On 02-Jun-89 At : Here Subject: VT??? EMULATION ON PCS John, I've never heard of a version of MS-Kermit that would handle 132-column VTxxx emulation. In the past, at least, the authors of MS-Kermit have tried to write as generically, with as few hardware dependencies as possible. In fact, there are darned few communications packages that will. It's a very hardware-dependent function (only EGA, VGA, and Herc boards can handle 132-column screens), and each of those options does it rather differently. Implementing 132-column mode would add quite a bit of code for a feature that most people don't think is very important. Now, there may be a new version of MS-Kermit that I haven't seen. But so far as I know, people who need 132-column VTxxx emulation on PCs have had to turn to specialized software (PolyCom comes to mind) designed specifically for DEC users. I know for a fact that the latest versions of CrossTalk, Mirror, ProComm Plus, Telix, and SmartCom don't provide it; nor is it in- cluded in the Greenleaf Communications Library for people who want to write their own comm software on PCs. What else would you like to know about emulations? MASS STORAGE SIG Message # 259 0 replies Reply to # 256 Left by: PHILLIP A CREWS On 02-Jun-89 At : AMEX HSG 704 333 1276 Subject: DISK DROPPING READY ___ It appears that your ESDI disk is dropping the RDY line to the controller periodically. Ive seen this happen on ESDI if the entire disk is not formatted or if there is a seeking problem when accessing close to the spindle (on the smallest tracks). Or it could be the controller; I've seen QD33s and QD34s do similar things when getting ready to "really" fail. VMS SIG Message # 968 0 replies Reply to # 980 Left by: PHILLIP A CREWS On 02-Jun-89 At : AMEX HSG 704 333 1276 Subject: AUTO-LOGIN DS PORTS I think DEC finally did document the procedure for changing ALFMAINT to handle DECserver ports. I'm pretty sure I saw it on DSIN (the Digital Software Information Network), which you get access to if you purchase Basic service (BSS) or higher for VMS. VMS SIG Message # 969 0 replies Reply to # 962 Left by: BARRY YOUNG On 04-Jun-89 At : JEDDAH (02)-647-0773 Subject: DECSERVERS/AUTOLOGIN Many thanks for this suggestion and a very prompt reply. Maybe I should cancel our DEC software support contract & use ARIS instead. I'll try this modified ALFMAINT and see how we get on. Best regards, Barry Young, Systems Manager, Marine Transport International, Jeddah, Saudi Arabia. CLUSTERING/NETWORKING SIG Message # 396 0 replies Reply to # 395 Left by: DUANE SADOWSKI On 04-Jun-89 At : (212)577-0543 Subject: After I sent the original message, the user who suggested using TCP/IP suggested the possibility of using a PC terminal emulator package instead of TCP/IP. I'm curious to know whether this alternative would reduce the amount of equipment needed at the VAX site or change the feasibility of the link between the two LANs. PC'S SIG Message # 215 2 replies Left by: LARRY GHERARDI On 05-Jun-89 At : (516)957-2300 Subject: VT220 EMULATOR FOR RAINBOW Does anyone know of a VT220 Emulator for the rainbow. The built in emulator on the rainbow emulates a vt102 and we require vt220 to use with Oracle's SQL FORMS Package on the VAX. Larry FORUM - DEC SIG Message # 65 0 replies Left by: PATRICK MAHAN On 05-Jun-89 At : (714) 720-7868 Subject: CDA/DDIF APPLICATIONS/EXAMPLES Does anyone know of who or where I can turn to for CDA/DDIF help. I'm on a quest for support and/or examples of how to utilize the CDA and the DDIF formats provided by Digital. I'm the lead programmer on a project that requires the ability to display graphic images brought in by scanner and from our existing CAD databases. We are using DECwindows to build the user's interface and needed a "generic" file format to store our graphics images. From looking at the CDA manual it seems as if this will do the trick. However, the documentation is hard to understand the examples are poor (plus no example of a complete front & back end). Plus I sent one of my programmers to the DECwindows class which states in the description of the class that CDA will be covered, when the programmer returned he stated that the instructor did not have the faintes (sic) idea of what the CDA was much less how to utilize it. I need help pretty fast as our prototype deadline is by summers end. PC'S SIG Message # 216 1 replies Reply to # 215 Left by: MARTY JOHNSON On 05-Jun-89 At : (203) 666-1737 Subject: VT220 EMULATION Poly-COM (the TRM program) provides VT220 emulation (from Polygon, Inc). However, I don't know if they are still marketing the product. It was supposed to be replaced by Poly-STAR, but I haven't heard of Poly-STAR for the Rainbow. The emulation is good. We have a customer who has been using Poly-COM for the last three years to do LEX-11 word processing on a VAX 8200 using a Rainbow as a terminal. Hope this helps. RSTS SIG Message # 385 0 replies Left by: ARTHUR WALSH On 05-Jun-89 At : 408 371-0575 Subject: REMOTE DISCONNECT ABORTS JOB We have a PDP-11/73 with Hayes 2400 and 1200 modems running RSTS/E v9.5. Our application programs are written in DIBOL. Many of the programs open the printer directly for reporting purposes. When these type of jobs are started remotely and we disconnect with "+ + + A T H " the process is interupted and returned to DCL ^C state. Comparable jobs that write the report file to disk and then spool are not interupted by the same procedure. Aside from rewriting the direct printing programs, are there any other solutions??? PC'S SIG Message # 217 1 replies Reply to # 215 Left by: ANDREW DUGGAN On 05-Jun-89 At : Subject: VT220 EMULATOR FOR RAINBOW Larry, I use a package from DECUS called Turbo-Comm and it works very well. It does not do UDKs (Shift Fnn), or selective erase. It is written in Turbo Pascal from Borland which has/had, at least at V3.0, a Rainbow version. It comes with the source code and is easily mod- ified. PC'S SIG Message # 218 0 replies Reply to # 216 Left by: LARRY GHERARDI On 06-Jun-89 At : (516)957-2300 Subject: VT220 EMULATION FOR RAINBOW We have POLY-TRM here but when I looked through the manual it said nothing about vt220. Maybe we have an earlier version? Thanks. Larry PC'S SIG Message # 219 0 replies Reply to # 217 Left by: LARRY GHERARDI On 06-Jun-89 At : (516)957-2300 Subject: VT220 EMULATION FOR RAINBOW That sounds good. I'll ask our vax manager to check it out. Thanks for your help. Larry MISCELLANEOUS SIG Message # 511 0 replies Reply to # 510 Left by: Phil Anthony On 06-Jun-89 At : Here Subject: 9-TRACK FOR VAX-3100 From the lack of responses over the past week, it's likely that nobody here knows of such a driver off the top of their head. Two possible sources come to mind, though. First, the DECUS tape library has an amazing amount of stuff in it. Check in with your local DECUS chapter, if you haven't al- ready, and ask around. Second, have you talked to Kennedy? They may not supply such a driver, but their sales reps may be able to track down a site where somebody is doing exactly what you want and is willing to share it. Sales reps will go to amazing lengths sometimes to close a sale! PC'S SIG Message # 220 0 replies Reply to # 218 Left by: MARTY JOHNSON On 06-Jun-89 At : (203)666-1737 Subject: Polygon never claims full VT220 emulation for Poly-TRM. When we tested several products, Poly-TRM performed everything we wanted without any problems. Poly-XFR and Poly-HST could transfer LEX documents - the only product we tested to do so, which is why we chose Poly-COM (which consists of Poly-TRM, Poly-XFR and Poly-HST). The major problem we found was the lack of products for the Rainbow. There may be more products available now (our tests were run 3 years ago), but we've been satisfied with Poly-COM and have no need to look at other packages. CLUSTERING/NETWORKING SIG Message # 397 0 replies Reply to # 81 Left by: BRAD MILLER On 06-Jun-89 At : 703-824-0100 Subject: LAT OVERHEAD I have found that all 8 ports on a decserver 200 will slow down to 4800 baud after 4 of the ports have connected to a host. Emulex makes a LAT compatible terminal server that does not step down to 4800 no matter how many connections are made to a host. This causes your session to seem incredibally slow and may give the impression that there is a tremendous amount of overhead on the net when there really isn't. UNIX/ULTRIX SIG Message # 39 0 replies Reply to # 7 Left by: BRAD MILLER On 06-Jun-89 At : 703-824-0100 Subject: ULTRIX MONITORING I have used the accton and last commands along with du to track system usage and disk storage. There are some software packages available that track system usage fairly well on UNIX based systems. Call if you want phone numbers or rep names. LASER PRINTERS SIG Message # 70 0 replies Reply to # 59 Left by: ROLAND BURGESS On 06-Jun-89 At : 302 886-4083 Subject: MASS-11 SUPPORTS POSTSCRIPT MASS-11 SUPPORTS POSTSCRIPT BEAUTIFULLY. IT IS A GREAT WORD PROCESSOR THAT RUNS ON VAX AND PC'S. VERY GOOD SUPPORT FOR USING ALL POSTSCRIPT FONTS AND POINT SIZES, AS WELL AS GRAPHICS IN POSTSCRIPT. FOLLOWS COMPOUND DOC. ARCHITECTURE STANDARDS TOO. CAN'T BE BEAT. BY THE WAY, WE ALSO HAVE A LN03 POSTSCRIPT PRINTER CONNECTED TO OUR VAXSATION RUNNING INTERLEAF TPS. VMS SIG Message # 970 0 replies Left by: HARVEY RARBACK On 06-Jun-89 At : (516)282-3758 Subject: LOGICAL NAME SYSGEN PARAMETER I remember reading about improving the performance of logical name searches in a recent DEC PRO or VAX PRO, but I can't seem to locate the issue. When I installed VWS on a MicroVAX II, the pagefile quotas had to be signficantly increased to get processes (not using VWS) to run. I believe this is because of the large increase in the number of system logical names and the attendant virtual memory required to resolve the names. If I remember right, there is a SYSGEN parameter which improves the efficiency of the hash table searches. Can anyone help? BTW, DEC (Atlanta) was no help. They said the virtual memory required was "normal", but I am suspicious... Internet: rarback@bnl.gov BITNET:RARBACK@BNL HEPnet:44166::RARBACK MISCELLANEOUS SIG Message # 512 0 replies Left by: JERRY KILGORE On 07-Jun-89 At : 512-331-0508 Subject: DECUS SESSION NOTES NEEDED I urgently need to get a copy of the session notes for the RSX session RX-036, "LOADABLE DIRECTIVES", given by Brian McCarthy (DEC RSX Engg) at the recent DECUS meeting in Atlanta. I have purchased the audio recording of the session but the session notes are necessary to make sense of the tape. If you have a copy (or know of someone who has) please call me collect at 512-331-0508. I will be happy to defray any costs of duplicating and shipping, etc. Thanks, Jerry Kilgore VMS SIG Message # 971 2 replies Left by: RUTH MERMELSTEIN On 07-Jun-89 At : HERE Subject: DCL .COM PROCEDURES I wrote a really simple captive menu-type login procedure for some data entry users who do not need access to a DCL prompt. Among other things, it allows them to send mail and reset their password. The relevent code looks like this: $LOOP_: $ WRITE SYS$OUTPUT " 1 Mail" $ WRITE SYS$OUTPUT " 2 Set password" . . . $ INQUIRE CHOICE "Choice ?" $ IF CHOICE .EQ. 1 THEN MAIL $ IF CHOICE .EQ. 2 THEN SET PASSWORD . . . $ GOTO LOOP_ The problem is that they can only choose MAIL or SET PASSWORD once per login. I know I could get around this by generating subprocesses to do the job, but we're short on process slots these days, so that's not an option. Any suggestions? CLUSTERING/NETWORKING SIG Message # 398 1 replies Left by: RICK CURRY On 07-Jun-89 At : (714) 921-7683 Subject: DYNAMIC ASYNC DECNET HAS ANYONE EVER TRIED TO BRING UP A DYNAMIC ASYNC DECNET CONNECTION USING MICROCOM AX/9624c MODEMS???? THEY ARE ERROR-CORRECTING MODEMS WITH SPEEDS UP TO 9600 BPS. ANY SUGGESTIONS WOULD BE GREATLY APPRECIATED. PROGRAMMING LANGUAGES/AI SIG Message # 397 0 replies Left by: DAVID ROBISON On 07-Jun-89 At : 702-364-8633 Subject: VAXC ATEXIT() I am having problems using the VAX C RTL routine atexit(). This routine should set up an exit handler, but when the program exit I get an access violation. Has any one used this function? Are there any particular tricks involved? Thanks... Dave PROGRAMMING LANGUAGES/AI SIG Message # 398 0 replies Left by: STEPHEN TATRAI On 07-Jun-89 At : 201 524-1325 Subject: DATATRIEVE DOMAIN PROTECTION I need to provide access to a subset of a larger data file. I know I can do a view domain but I still have to grant access to the main data file also i have to give VMS read privileges to the main data set. How can I only allow acces to the subset? Thank you. S. Tatrai RSTS SIG Message # 386 0 replies Reply to # 385 Left by: BRUCE A. GREMBOWSKI On 07-Jun-89 At : 408/291-3364 Subject: INTERRUPTED PRINTING Arthur, It sounds like the job is seeing its terminal log off, so it just stops. The only thing I can think of is setting the modem port to /NODIALUP. A call would then have to be manually terminated, and RSTS may at that time still cause your job to stop. If not, then the next person calling on that line may attach to the job that's printing. Other than this, maybe starting the program that prints directly from a batch job would work? In any case, good luck! Regards, Bruce RSTS SIG Message # 387 1 replies Left by: BRUCE A. GREMBOWSKI On 07-Jun-89 At : 408/291-3364 Subject: JOB NO. FROM KB NO. I'm trying to find out what job, if any, a keyboard is running on in a BASIC-Plus program using a SYS() call. Following the RSTS/E Programming Manual V9.1 (we're running V9.2), I make a SYS call with 6+16+0+kb#+ 16 0's (bytes 5-20) + 128 (byte 21-return perm chars) + 7 0's. According to the manual, the job number can be determined from byte 3 by ANDing it with 126% and dividing by two. I do this and I get the KB number divided by two! Does anyone know where I can get the job number? If not from this SYS() call, maybe from a PEEK() combination? Thanks in advance, Bruce SITE MANAGEMENT SIG Message # 291 0 replies Left by: STEVE J. DYER On 07-Jun-89 At : 812-853-1333 Subject: ANYONE USING RSM? I'd like to hear general comments from those using RSM ( Remote System Manager ) on VMS systems. I'm particularly intersted in its backup features and capability to do an inital system load. Can it do image backups and restores ? Thanks for any help. I'm looking at using it on 13 microVAXes as clients and a 3500 or 3800 as the server. SITE MANAGEMENT SIG Message # 292 0 replies Left by: FODULYAN On 07-Jun-89 At : 818-843-120 SITE MANAGEMENT SIG Message # 293 0 replies Left by: FODULYAN On 07-Jun-89 At : 818-843-1200 Subject: IDLE PROCESS KILLER Has anybody out there used a program like WATCHDOG to kill idle processes under VMS 4.5: any prblems? What about VMS v4.6 and 4.7. And what about vms v5.0? Any help is appreciated. Thank you. VMS SIG Message # 972 0 replies Reply to # 971( Left by: PATRICK MAHAN On 07-Jun-89 At : (714) 720-7868 Subject: DCL .COM PROCEDURES I don't seem to understand your problem. Your examples code looks good. We are running a number of captured accounts with dcl menus that allow the user access to a variety of programs and dcl products (mail and set password). RSTS SIG Message # 388 0 replies Reply to # 387 Left by: Phil Anthony On 07-Jun-89 At : Here Subject: JOB NO. FROM KB NO. Bruce, the following - from Mike Maysfield's "RSTS/E Monitor Internals" for V9.6 - ought to get you the information you want. Needless to say, you can always combine the separate PEEKs into a single statement: DIM MTAB1% (30%) KB% = CHANGE SYS (CHR$(6%) + CHR$ (-3%) + STRING$ (28%, 0%)) TO MTAB1% . ! Get monitor tables, Part I DEVPTR% = MTAB1% (7%) + SWAP% (MTAB1% (8%)) . ! Address of device table for all . ! legal devices DEVTBL% = PEEK (DEVPTR% + 32%) ! Address of table listing DDB . ! addresses for all configured KBs DDB% = PEEK (DEVTBL% + (KB% * 2%)) . ! Address of DDB for keyboard you're . ! interested in DDJBNO% = PEEK (DDB% + 2%) AND 255% . ! Job number times 2 of keyboard you . ! want - maybe IF (DDJBNO% AND 1%) THEN . ELSE . JOBNO% = DDJBNO% / 2% ENDIF One warning on this. If the keyboard is the console keyboard for a detached job that has relinquished control, DDJBNO% will not reflect this fact. Any attempt to reattach by the detached job will run into trouble if you use the keyboard for something else. You might also want to check out DDCNT% = PEEK% (DDB% + 6%). If DDCNT% AND 8192%, it's the owner's console keyboard; if DDCNT% AND 16384%, it's tempo- rarily assigned to FIP; if DDCNT% < 0%, Job JOBNO% explicitly assigned it. SYSTEM PERFORMANCE SIG Message # 229 0 replies Reply to # 217 Left by: Barton F. Bruce On 07-Jun-89 At : (617) 868-1111 Subject: your DHV is hearing itself For whatever reason, it sounds as though that port is talking to itself. The problem is common on long lines with a disconnected terminal (i.e. now NOTHING) at the far end. Something transmitted is coupled into the receive side eliciting some response which is yet another transmission, and this goes on forever. Some call it ear-in-mouth disease, or a what-who loop. 500' is pretty far, and that gives you LOTS of wire with lots of capacitance to couple from transmit to receive. If you are using twisted pair, be sure pins 2 and 3 are each paired with some steady state signal (a ground, or DTR or whatever NON data signals you are passing end to end. I have seen horrible problems when someone paired 2 and 3 together and wondered why it was so prone to chattering to itself. But, runing an LN03, you shouldn't be set up so anything from the printer is going to elicit much of any response. What exactly are the port settings. Also, since the DHV is a DMA device, do you have DMA on? Do you have a scope? VMS SIG Message # 973 0 replies Reply to # 970 Left by: RICK SMITH On 08-Jun-89 At : (419)321-8331 Subject: LOGICAL NAME HASH TABLE Downline load the program in the VAXPRO lib called LMNFILL it will give you the details which you want. DATA/FILE TRANSFER SIG Message # 552 0 replies Reply to # 551 Left by: JOHN S. MIELE On 08-Jun-89 At : (203) 656-5149 Subject: TERMINAL EMULATORS Thankyou for the information. When talking about terminal emulators, it seems we are getting closer and closer to the digital "PCSA" solution for connecting PCs to the Vax. I don't know if you are familiar with PCSA, but Digital has just slashed the price from $500 per PC to $250 per PC. Although it seems that you must also purchase some software to reside on the Vax so the PC software will be functional. I don't know these underlying costs, but if they're negligible, That would make "PCSA" about the same cost as the terminal emulators out there. I will do some more investigation. If you have any pointers, I would appreciate it. Thankyou again VMS SIG Message # 974 0 replies Reply to # 971 Left by: ROBERT G. SCHAFFRATH On 08-Jun-89 At : 914-335-9792 Subject: More information Your code as listed looks fine. You will have to provide some more detail in order to find out what's wrong. One note on security: you are using an INQUIRE statement. If you have any sensitive routines, a user can simply activate verification mode by typing 'F$VERIFY(1)' in response to your prompt. For safety, follow all of your INQUIRE statements with SET NOVERIFY or, better yet, replace your inquire statements with 'READ/PROMPT="Your Choice? " SYS$COMMAND CHOICE'. Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 CLUSTERING/NETWORKING SIG Message # 399 0 replies Reply to # 398 Left by: ROBERT G. SCHAFFRATH On 08-Jun-89 At : 914-335-9792 Subject: DECnet over a Microcom AX/96xx Sorry, I can't give you a specific response to the Microcom AX/9624c. However, I have a Microcom AX/9600 PLUS which has worked DECnet successfully. Robert G. Schaffrath System Manager GeneraL Foods Corporation White Plains, NY 10625 VMS SIG Message # 975 0 replies Reply to # 972 Left by: RUTH MERMELSTEIN On 08-Jun-89 At : HERE Subject: MORE INFORMATION Rober Schaffrath - this is for you, too, since you both asked for more information. Thanks for replying. It's nice to know there are people out there to help when I'm absolutely up against a wall. I got my answer from my boss. I had described the problem to her exactly the way I described it to you, and when she couldn't help me either, I left the message in ARIS. When you said the code looked ok, I got smart and showed my boss the entire procedure, and she spotted the problem. I had the line $ DEFINE/USER SYS$INPUT SYS$COMMAND before the loop instead of in it. Thanks for the tip about READ/PROMPT. I hadn't heard that before, and although my boss had heard that INQUIRE is a security risk, she didn't know why. CLUSTERING/NETWORKING SIG Message # 401 0 replies Reply to # 399 Left by: RICK CURRY On 08-Jun-89 At : (714) 921-7683 Subject: DECNET OVER A MICROCOM AX/96XX Robert, That's good to hear. Is your modem an error correcting type, and if so are you bypassing this feature. Also, what speed is your link and are you using data compression. A few details about your modem configuration would be of great interest to me. Thanks...... FORUM - DEC SIG Message # 66 0 replies Left by: PHILIP ATTNEAVE On 08-Jun-89 At : (206)684-2074 Subject: 8500 VS 3XXX VMS VAX I considering replacing a couple of VAX , one 8200 and one 8530 with three MV3xxx , in a loose cluster. I currently use RA81's, TU81, and a SCSI 8mm Exabyte. There are 4 working groups , 2 Database and 2 mathematical model projects. They do not average many users (3-5 per group), but everybody wants tones of disk space for all the large data sets. There is also a group of about 7 PCSA Dos type machines using files disks. My biggest motivation in going to the micro's is that it appears that ongoing HW and SW support costs will make them cheaper the next 4 years, including purchase. Am I missing something? Anybody got any do's or don'ts in selecting a configuration of 3000 series machines? UTILITIES SIG Message # 372 0 replies Left by: DAVID MOORE On 08-Jun-89 At : 606-878-7900 Subject: LOOKING FOR AWK ON VMS I'm trying to find an implementation of AWK that runs under VMS. AWK is a utility language that is very useful for quick-and-dirty text processing. It was originally developed at AT & T for Unix. It has been ported to MS-DOS, (I've got a version called POLYAWK from Polytron Software), but I need to find it on VMS. Any help would be appreciated. PROGRAMMING LANGUAGES/AI SIG Message # 399 0 replies Left by: DAVID MOORE On 08-Jun-89 At : 606-878-7900 Subject: LOOKING FOR AWK ON VMS I'm looking for AWK running on VMS. AWK is a utility language that is very useful for quick-and-dirty text processing. It was originally developed by AT & T for Unix, but has been ported to other environments. Any assistance would be appreciated. VMS SIG Message # 976 0 replies Reply to # 971 Left by: MARTY JOHNSON On 08-Jun-89 At : (203)666-1737 Subject: SECURITY Since you mentioned this procedure is to be run from a captive account, I hope you are aware that you must take some extra precautions to prevent users from accessing DCL. The little culprit here is the mail utility. Remember that MAIL has a spawn command. It is easy enough to set up the account to prevent spawning. What most people forget is that you can mail messages to yourself such as "$ EXIT" and EXTRACT them as a file called LOGIN.COM. From there, the user can log off and log back on -- to DCL. The easiest solution to prevent this is to set LOGIN.COM to version 32767. RSTS SIG Message # 389 0 replies Reply to # 388 Left by: KELVIN SMITH On 08-Jun-89 At : (203) 357-0504 Subject: GETTING JOB # FOR KB Bruce, I don't know why you couldn't get the SYS call to work. I tried it on my system and it worked with no problems, both with logged-in terminals and with terminals opened by other jobs. I would suggest that you shouldn't need to set byte 21; just A$ = SYS(CHR$(6%) + CHR$(16%) + CHR$(0%) + CHR$(x%)) where x% is the keyboard number, will return in byte 3 the job number times two (as noted, you should AND the result with 126% to clear the 128 and 1 bits, which have other meanings as listed). Are you trying in Basic-Plus, Basic-Plus-2, FORTRAN, or what? It's possible that there's a bug in the language. I tried it in B+. CLUSTERING/NETWORKING SIG Message # 402 0 replies Left by: FRANK INNES On 08-Jun-89 At : 215-448-3461 Subject: DECNET & TCP/IP SUN-VAX I would like to hear about software for networking a VAX running DECnet on Ethernet with a Sun file server running TCP/IP on the same Ethernet. The specifics of my situation are an Ethernet with three VAXes, one PDP, numerous LAT terminal servers, a Sun 4/280 file server and two Sun 3 systems. The traffic with be between a MicroVAX 3500 and the Sun 4/280 and connection to the Sun via terminal servers is also desired. It seems to me that a DECnet implementation for the Sun is probably what I need. Is this so, or is TCP/IP for the MicroVAX preferable? What are the pros & cons, and what LAT software is there for the Sun? VMS SIG Message # 977 0 replies Reply to # 976 Left by: ROBERT G. SCHAFFRATH On 09-Jun-89 At : 914-335-9792 Subject: LOGIN.COM security Definitely a good point. One additional enhancement would be to change the file ownership to SYSTEM and grant EXECUTE access to the user through an ACL. I have two particularly sensitive accounts where this was done. That way if the user does manage to get to DCL, they can't SET FILE/REMOVE, RENAME, or DELETE the LOGIN.COM;32767. However, the ultimate solution would be to store the LOGIN.COM in a special secure directory and change the LGICMD parameter in the UAF to point to that file specification. The user may not even know where the LOGIN.COM is in that case. Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 CLUSTERING/NETWORKING SIG Message # 403 0 replies Reply to # 401 Left by: ROBERT G. SCHAFFRATH On 09-Jun-89 At : 914-335-9792 Subject: AX/9600 modem settings While the modem is basically the same as yours except: a) doesn't support 9600 baud synchronous and b) has an 11000 baud throughput instead of 19200 baud in MNP class 6 service. The modem connects at 2400 baud reliable (MNP Class 4) mode because the other modem doesn't support 9600 baud asynchronous or MNP Class 6. Your modem should be able to run DECnet asynchronously at 9600 baud. You may, however, have to tweek some NCP parameters to allow for possible line delays due to line noise. Sorry, I can't provide a 9600 baud verification. Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 VMS SIG Message # 978 0 replies Left by: On 09-Jun-89 At : Subject: DATA/FILE TRANSFER SIG Message # 553 0 replies Left by: JOHN S. MIELE On 09-Jun-89 At : (203) 656-5149 Subject: EXABYTE DRIVE FORUM - 3RD PARTY SUPPLIERS SIG Message # 103 0 replies Left by: JOHN S. MIELE On 09-Jun-89 At : (203) 656-5149 Subject: EXABYTE DRIVE I am looking for information on the EXABYTE Drive. We decided to go with this drive, as to perform unattended backups. There are many vendors out there. Any comments on the drive itself, and vendor information would be greatly appreciated. We are currently looking at System Industries, Micro Technoligies, and Northwest Peripherals. Thanks, John S. Miele SITE MANAGEMENT SIG Message # 294 0 replies Reply to # 293 Left by: RALPH BRANDT On 10-Jun-89 At : 717-751-5210 Subject: HITMAN We're using a product called hitman on 4.7 and 5.0. Seems to be working since the users are real p.o.d when they get bumped. DATA/FILE TRANSFER SIG Message # 554 0 replies Reply to # 552 Left by: RALPH BRANDT On 10-Jun-89 At : 717-751-5210 Subject: PCSA I have PCSA up on 135 PC's - combination of TANDON, IBM XT AT and about 55 PS/2 It is worthwhile but as usual the DOC stinks. There is a memory problem It uses 202K approx. There are ways to work around this. The terminal emulator is nice but I still have a problem with 132 char mode. Suggest you don't buy the DEC start - it was $4000 down the drain. Send someone to school who can stay awake in class and who can follow diretions. If yoou have PC's with a lot of boards, - call me if you like... EMM and 3270 boards can be real fun. The 3COM microchannel board is nice. Contrary to DEC propoganda, the LK250 keyboard can be placed on the IBM PS/2. You need an adapter DEC doesn't Provide. We bought 164 of the kits just before the price went down. Watch the pricing, There are a lot of quantity package discoounts. 64 copies were cheaper than 50. DATA/FILE TRANSFER SIG Message # 555 0 replies Left by: RALPH BRANDT On 10-Jun-89 At : 717-751-5210 Subject: ETHERNIM Does anyone know why ETHERNIM takes 4 minutes to backup the database - It can be restored in 15 secs. This has been a real pain DATA/FILE TRANSFER SIG Message # 556 0 replies Left by: RALPH BRANDT On 10-Jun-89 At : 717-751-5210 Subject: ETHERNIM Has anyone tried to make reports from ethernim that are more useful than the ones DEC provides. They are either too terse to be useful or so verbose to be workable. This might be a good package if this were improved. DATA/FILE TRANSFER SIG Message # 557 0 replies Left by: RALPH BRANDT On 10-Jun-89 At : 717-751-5210 Subject: PCSA - LN03 Has anyone used PCSA to go to an LN03? I am trying to get a LOTUS spreadsheet there in landscape mode. It takes too many control characters for the lotus setup string. Any ideas? CLUSTERING/NETWORKING SIG Message # 404 0 replies Left by: DICK KRAMER On 10-Jun-89 At : (309) 765-6564 Subject: VAX FILESERVER 6310 We are just getting into workstations, and considering a 63xx FileServer. The network will probably consist of a mixture of VAX and SUN workstations. Obviously, there shouldn't be any problem using a 63xx as a server for VAX, but does anyone out there have any experience with using the FileServer 63xx complete with (VMS/ULTRIX connection), as a server for SUN workstations? We're primarily a VMS shop at this point, and from the stand point of internal support and operations, we'd like to stay that way. However, for certain applications, some of our engineers are convinced that the SUN and UNIX are the only way they can do their job. CLUSTERING/NETWORKING SIG Message # 405 0 replies Reply to # 402 Left by: Philip Gravel On 10-Jun-89 At : HERE Subject: DECNET & TCP/IP I am aware of at least 3 implementations of TCP/IP that run on a VAX. The Wollonogong group has call WIN/TCP, Network Research has one called FUSION, and CMC has one that includes its own Ethernet board which is claimed to provide very high throughputs. A company called KI Research (I think) has an implemention of DECnet that runs on a Sun workstation. (I believe they are located in Nashua, NH.) I don't know of a LAT implementation that runs on a Unix machine although both Emulex and Datability have terminal servers that speak TCP/IP. If I had my druthers, I'd go for the DECnet on a Sun rather than TCP/IP on a VAX. However, the utility of DECnet on a Sun is dependent on how well it is implemented. TCP/IP on a VAX will potentially give you access to any Unix machine on your Ethernet. Hope this helps. Phil VMS SIG Message # 979 0 replies Reply to # 977 Left by: PATRICK MAHAN On 11-Jun-89 At : (714) 720-7868 Subject: CAPTIVE ACCOUNTS One other unsolicited piece of advice. Make sure you turn off control-y capability for the account. We got chewed out at least three different times due to users issuing CTRL/Y's while executing some utility even though NOCONTROL=Y was set in the dcl menu. This especially happened when the user was executing an in-house or third-party software package (they did not handle a graceful exit for CTRL/Y's). We finally settled on a solution that guarantee the user could not perform any mischevious actions by tailoring a a copy of the DCLTABLES for use with captive accounts. This insures that if the user does manage to make it to the DCL prompt, they don't have much to play with! RSTS SIG Message # 390 0 replies Reply to # 385 Left by: Brett Bump On 11-Jun-89 At : (308) 432-2828 or HEA-DUKE Subject: USING LOGICALS WISELY Arthur. If this were my system, I would simply detach my job with a ^D and hang up (although this requires a nonstandard patch to the monitor). I assume that your modem port is already set to /NODIALUP since you say that the program drops to ^C state. Otherwise the job should terminate completely (the recommended norm). I also assume that the reason that you are disconnecting from your system is so you aren't stuck in limbo waiting on a printer (I don't blame you). Obviously the trauma of disconnecting from your system is causing a BREAK signal to be introduced on the line. One way to bypass this problem and keep your program running would be to set your terminal for /NOCONTROL=C. I would much rather see you change your terminal back to /DIALUP if you can (yes, some of us are still stuck using DL lines). Then as an alternative to rewriting anything you could have the output sent to a file, then spool the file to the printer. .example . Assign Sy: Lp ;This assumes that the printer that was . ;opened directly was opened as Run Application ;"Lp:Prtfil.tmp" . $Print Prtfil.tmp Of course there are those who haven't learned the advantages to opening a non-file-structured device with a trailing name. If that is the case, you need to have a NFS disk handy. (Let's use the RAMdisk as an example). .example Nfscpy.bas > 10 On Error Goto 70 . 20 Open "Dv0:" as File 1% Assign Dv0: Lp 30 Open "Prtfil.tmp" as File 2% . 40 Input Line #1%,Prtstf$ Run Application 50 Print #2%,Prtstf$; . 60 Goto 40 Run Nfscpy 70 Close 1%,2% . 80 End $Print Prtfil.tmp As you can see, there are definite advantages to using local logicals and creating programs with the intent of using these logicals wisely. Give me a call if you need any help. CLUSTERING/NETWORKING SIG Message # 406 0 replies Reply to # 405 Left by: RICHARD B. GILBERT On 11-Jun-89 At : (609) 452-5130 Subject: DECNET & TCP/IP Philip, There are at least six TCP/IP implementations for the VAX. In addition to the three you mentioned there is a package from Excelan which requires their board to run, one from SRI called Multinet and one from Carnegie-Mellon University. The Carnegie-Mellon package is "cheapware" (License media and doc are less than $200) and is not supported by CMU. The commercial packages all cost "big bucks" but they are supported. DECNET integrates seamlessly on the VAX end of things. Except for having to specify a remote nodename the user need not even know it's there. TCP/IP implementations require running FTP to transfer files while DECNET lets the user use the familiar COPY, TYPE and DIRECTORY commands at DCL level. TCP/IP is the "native mode" networking for Unix systems. I suspect that it might look better on a Unix system than DECnet would. (I've never used DECNET under Unix). It is possible to run both DECNET and TCP/IP simultaneously on the same Ethernet controller. I use DECNET to communicate between my VAXen and IBM PCs and TCP/IP to talk to the rest of the world. EDITORS SIG Message # 216 0 replies Left by: KEITH PINNEY On 12-Jun-89 At : (612) 778-5191 Subject: CALLABLE WINDOW EDITOR Does anyone know of a callable EDT-like editor for VMS that can accept limits for maximum lines, maximum columns, displayed lines, displayed columns and a starting row and column on the screen. I would like to be able to call an editing routine like this and use it for field entry much the same as FMS or TDMS, but for larger fields of text. These fields have a known, limited number of columns and lines. I have investigated writing this with SMG calls, but I'd rather not re- invent the wheel if it exists somewhere. Thanks! MISCELLANEOUS SIG Message # 513 2 replies Left by: JOHN V. J. FINKELL On 12-Jun-89 At : (404) 980-1190 Subject: VAX/PC NETWORKING PRO&CON We have a VAXstation 2000 and 10 PC's. We are using PCSA software to run the PC's off the VAX. Currently we have certain programs running as disk services with their associated data on a file service. I would like to keep this scenario, but I have a hard time convincing the powers that be that this is the way to go. They think the data and programs should be on the individual machine, IN CASE THE VAX GOES DOWN. They also don't know why we need pc's if the data and programs are on the VAX. If any can supply me with convincing arguments for or against this I would like to here from you. Is this the ultimate use of the VAX and PC?! MASS STORAGE SIG Message # 260 0 replies Left by: FINKELL On 12-Jun-89 At : 404-980-1190 Subject: TK-50'S ANOTHER PROBLEM We have a VAXstation 2000 running VMS v4.7 and a TK-50 tape drive, as well as a RD-54 hard drive. When I want to back up the entire hard drive which takes two tapes I must be present when the tapes switch. The only help I got from DEC is a patch which extends the timeout period to 30 minutes, but this still does not allow me to run it at night as a batch process. And do to the high amount of I/O the whole system slows down considerably, so I am forced to give up 3 hours on saturday to come in and do the backup, and with plans of getting another disk drive I don't relish the idea of giving up 6 hours, due to DEC's irresponsibilty with the TK-50 tapes. Is there a way around this timeout problem or a way to tweak the system so the backup does not slow down our users, setting the priorty doesn't help since it is the I/O that's the killer? DEC said VMS V5.0 solved the timeout problem, but there is still problems with the TVDRIVER, is this true? All these problems are only problems for the TVDRIVER present in VAXSTATION 2000 & 3000 and the MICROVAX 2000.b Please Send Help !!!!!! SITE MANAGEMENT SIG Message # 296 0 replies Reply to # 293 Left by: MARTY KUHRT On 12-Jun-89 At : (818) 249-4707 Subject: WATCHDOG MISCELLANEOUS SIG Message # 514 0 replies Reply to # 513 Left by: RICHARD B. GILBERT On 12-Jun-89 At : (609) 452-5130 Subject: VAX & PC John, Well? Why do you do it with a dozen or so PCs and a VAX? You doubtless have your reasons for doing it that way and some time spent setting down those reasons for your bosses might pay dividends. Off hand, I can think of a couple of issues that might be relevant: 1. The data must be on a file server (the VAX) because it must be shared by several users. 2. The processing is done on PCs because the software is available for PCs. Writing it for a mini (VAX) would cost $xxxxx. Of course neither of these points may have anything to do with the facts. FORUM - DEC SIG Message # 68 0 replies Reply to # 66 PROGRAMMING LANGUAGES/AI SIG Message # 400 1 replies Reply to # 397 Left by: PHILLIP A CREWS On 12-Jun-89 At : AMEX HSG 704 333 1276 Subject: ATEXIT () Here are some hints for using ATEXIT() as I understand it: Be sure to point to the handler block;i.e. atexit(&exit_handler_block); Make sure the exit handler block is static, writable, and defined for the entire length. There is a longword in the exit handler block which must be initialized with the address of the exit handler routine. There may be some limitation to the compiliation of the exit handler itself; C must not set up its own exit handler for the exit handler. I believe there is a CC qualifier that avoids normal run-time initialization. (I know that in BASIC it is /NOSETUP, but I can't remember about C.) Try a very simple procedure to make sure that something in your code is not causing an access violation; just do a printf() or something. PROGRAMMING LANGUAGES/AI SIG Message # 401 0 replies Reply to # 400 Left by: PHILLIP A CREWS On 12-Jun-89 At : AMEX HSG 704 333 1276 Subject: ATEXIT () I hope I didn't confuse you too much with my reply in #400 - much of the description I gave was for SYS$DCLEXH rather than ATEXIT. ATEXIT takes the address of the function to run at exit ( &exit_handler ). Sorry for my mistake. Still attempt to use a simple procedure (such as simply containing a printf() ) to tell if it's a C problem or not. MISCELLANEOUS SIG Message # 515 0 replies Reply to # 513 Left by: Phil Anthony On 12-Jun-89 At : Here Subject: VAX/PC NETWORKS John, I'd like to add a few more reasons for hooking PCs to a VAX to Richard Gilbert's excellent summary: 1. Storing information on the VAX ensures that backups of the data used by the PCs gets done. This is a major hassle on standalone PCs, as well as on networked PCs where the data are stored locally. 2. If executable code for the PCs is stored on the VAX, it's a relatively easy job to upgrade to new releases - there's only one copy of the program to re- place. (This is not, of course, an invitation to evade licensing requirements by purchasing one copy of a standalone program and then using it on multiple PCs. In fact, trying to run single-user software in a multiuser environment usually ends up costing more in aggravation and garbled data than buying a networked version to start with.) 3. Since your work is being done by a PC's CPU instead of the VAX's, more work gets accomplished in less time, with less strain on the VAX's resources, than if everybody used terminals connected into the VAX. This assumes that you could find VAX software that would do the same job you're doing on the PCs. 4. Related to the last point is the fact that while a single PC with no hard disk costs quite a bit more than a terminal does, adding 10 or 15 PC worksta- tions with no hard drives to your network costs less than adding the same num- ber of workstations and then having to upgrade to a more powerful VAX to get adequate response time. What neither Richard's nor my replies speak to is the question why you have the VAX to start with. All the benefits we've outlined - plus full fault tolerance in case the server has problems - could be obtained at lower cost by running a strictly PC network off a powerful system such as a Compaq 386/25. (Two-Gbyte tape systems are available for PCs, too, which would make unattended backup possible.) But I'd guess that you have other programs that are solely VAX-based, where the PCs simply serve as dumb terminals. Speaking philosophically, yes, I do think this is the ultimate use of minis. The trend first became apparent to me, at least, in high-CPU situations such as engineering; the mini server provides an excellent means of supplying shared data without losing centralized control and security, while the workstations' CPUs bring the load on the mini down to manageable proportions. Commercial applications took quite a while to follow suit, but as users discov- ered how much more sophisticated PC programs are than most mini programs - and systems specialists learned how resource-intensive it is to provide equivalent features in mini software - commercial PC/mini networks started springing up. CPU-intensive graphics interfaces and use of color are becoming more prevalent in commercial software, and I expect to see more and more use of intelligent terminals - workstations - and fewer and fewer terminals of the VTxxx variety. Eventually I think the software companies will embrace the trend fully and pro- vide true distributed processing: VAX (or other mini) programs that communi- cate directly with PC programs to make the best use of both types of computer. Oracle, I understand, is already doing this in the database arena. MISCELLANEOUS SIG Message # 516 1 replies Reply to # 16 Left by: JOHN V. J. FINKELL On 13-Jun-89 At : 404-980-1190 Subject: NEW AREA CODE FOR 617 PLEASE NOTE ANY REFERENCES TO AREA CODE (617) FOR MASS. THE AREA CODE HAS BEEN CHANGED. THE NEW AREA CODE IS (508). MISCELLANEOUS SIG Message # 517 0 replies Left by: JOHN V. J. FINKELL On 13-Jun-89 At : (404) 980-1190 Subject: NEW AREA CODE FOR 617 PLEASE BE ADVISED THERE IS A NEW AREA CODE FOR (617) MASS. IT IS NOW (508). FORUM - DEC SIG Message # 69 1 replies Reply to # 65 Left by: Phil Anthony On 13-Jun-89 At : Here Subject: CDA/DDIF In the absence of any specific information (Hey! Any of you guys in the Graphics SIG listening?), I'd suggest that you try DEC again. Get in touch with your local DEC people and tell them how disappointed you are that their DECwindows class didn't live up to its advertising. Ask them - nicely, of course - what DEC is going to do to honor its commitment. If the local people don't respond, kick the question upstairs, all the way to Maynard if necessary, till you get some satisfaction. FORUM - DEC SIG Message # 70 0 replies Reply to # 66 Left by: Phil Anthony On 13-Jun-89 At : Here Subject: 8530 VS 3XXX From what I've heard, the 3000 series are the hottest items in the VAX mar- ket right now, for a lot of good reasons. Three caveats come to mind, though: 1. You're basing your cost analysis on DEC's current prices and policies. DEC has a history of "adjusting" their charge for hardware and software sup- port according to their current marketing strategy; would anybody here like to venture a guess about where DEC will want to take the 3xxx's in the next five years? 2. The smaller machines, whether they're VAX 3xxx's or PDP-11/23s, work best with a lean, mean operating system - smaller percentage of the system's resources taken up with O/S overhead, as well as less O/S overhead per user process. Larger systems tend to be served best by more feature-rich, resource-intensive operating systems; the bigger guys can afford the over- head, and the added functionality provides more flexible user support. The trend in VMS development has been toward ever larger operating systems, to the point that a couple of years back, users with small VAXes (11/750s and 11/780s) had to trade up just to support new (V4.x) versions of VMS. Will DEC continue in this direction, and have you allowed enough slack in your new configuration to run VMS V6.x and V7.x when they come out over the next four years? 3. My understanding is that the bus structure of the 8530 will continue to give you better throughput than the 3xxx's for your disk-intensive database application. Check out your current I/O utilization statistics (MONITOR should give you sufficient information) and extrapolate from that using the published benchmarks to see if the new guys will do the job for you without degrading your performance. (Any comments from the System Performance SIG?) Depending on your answers to these questions, you might want to consider re- placing the 8200 with a single 3000 for the sake of upgrading while saving money. Then you could expand your storage capacity, which seems to be your main resource deficit at present, over the new two-system cluster. You might want to keep the database groups on the 8530 and migrate the math- ematical model groups over to your 3000; the DOS folks, who probably don't have heavy resource requirements, could be attached to whichever VAX can provide them the storage capacity they need. MISCELLANEOUS SIG Message # 518 0 replies Reply to # 512 Left by: Phil Anthony On 13-Jun-89 At : Here Subject: DECUS SESSION NOTES If nobody here has gotten in touch with you in the past week, Jerry, the notes may not be available through the ARIS community (RSX SIG, do you copy?). Your best bet may be to send out an SOS - the call, not the editor - through your DECUS LUG. CLUSTERING/NETWORKING SIG Message # 407 0 replies Reply to # 397 Left by: RALPH BRANDT On 13-Jun-89 At : 717-751-5210 Subject: LAT SERVERS I have been reading the comments on LAT and the server 200. They are somewhat disturbing. On the 9600/4800 baud issue. Can soomeone tell me how to determine if the servers I have are slowing to 4800? This wuld be a real problem in soome future appls where devices cannot autotbaud. Also - what about the server 550? I have 3 - two fully populated with DEC423 boards. I do have user complaints but had written them off to too many ALLIN1 users (approx 100) on an 8530. LASER PRINTERS SIG Message # 71 0 replies Left by: RALPH BRANDT On 13-Jun-89 At : 717-751-5210 Subject: LN03 / SERVER 550 We tried to run a LN03 at 19200 on a DEC 423 port on a server 550. Never got it to run. Went back to 9600. Has anyone done this successfully or is there a known DEC "feature" that prohibits this? MISCELLANEOUS SIG Message # 519 0 replies Reply to # 516 Left by: ROBERT G. SCHAFFRATH On 13-Jun-89 At : Subject: 508 AREA CODE That 508 area code only applies to those areas in Eastern Massachusetts outside of Boston/Cambridge! 617 is still in effect for Boston and Cambridge. FORUM - 3RD PARTY SUPPLIERS SIG Message # 104 0 replies Reply to # 103 Left by: Phil Anthony On 13-Jun-89 At : Here Subject: EXABYTE 8-MM TAPE DRIVES John, two people who left information in the 3rd-Party Suppliers SIG earlier this year may be able to help you out. They are Daniel Servranckx, at (514) 990-0805, who left Message #90 concerning bad reports he had heard about the Exabyte on 06-Jan-89; and Gerald Vezzosi, at (201) 802-6249, who gave his experience with the GT88 in Message #91 on 01-Mar-89. DATA/FILE TRANSFER SIG Message # 558 0 replies Reply to # 557 Left by: Phil Anthony On 13-Jun-89 At : Here Subject: LOTUS AND LN03s Ralph, my best suggestion would be to use a different spreadsheet. If you can find one that supports the LN03 directly, you'd be best off; I think Microsoft Excel does (via Microsoft Windows?), but I'm not sure. You might also check out Twin - see almost any PC magazine for their ads. Lacking ex- plicit LN03 support, you might investigate AsEasyAs, from Trius, Inc., 15 Atkinson St., Lynn, Mass. 01905. It permits upward of 80 characters for a printer initialization string. It's shareware and costs $30 a copy. I'd say that Lotus printer support is archaic and antedeluvian except for one thing - I can't find any Lotus printer support to say it about. As near as I can tell, the only reason anybody uses the overpriced and underfeatured 1-2-3 any more is the huge investment many firms have in Lotus add-ons for functionality 1-2-3 should've given them to start with. The question of templates the users have already created is largely a red herring, since most other spreadsheet packages today will read in 1-2-3 spreadsheets and translate the macros with no or minimal modification. CLUSTERING/NETWORKING SIG Message # 408 0 replies Reply to # 403 Left by: MARTY KUHRT On 13-Jun-89 At : (818) 249-4707 Subject: DECNET BY MODEM This is in response to the first message about using a 9600 baud modem for DECnet. I am currently running DECnet across 17,000 feet of four wire DC metallic phone line (a Pac Bell lease line) with two Virtual Cable Modems. The modems are rated at 9600 at 3-3.5 miles, but unfortunately, do to assertion problems with the MicroVAX II's DZ board, it will only run 4800 clearly (on this virtual modem it needs a pin 20 assertion level of 8-15v and the board provides 6v). For economy though you can't beat this set up, the lease line is the cheapest thing Pac Bell provides, and the modems cost around $100. - Not really a direct response to the question, but proof that you don't have to spend lotsa bucks on a high speed modem and 'special' lease line. - Some of the problems and caveats I ran into... - DEC supports DDCMP upto 9600, but they say they have people running it at 19200 without problems. - DEC is pretty adamant about NOT having error correction or compression on the modems when running DDCMP, they say it has something to do with packet framing. Since the modems I have a just basically line drivers, error correction is not a factor. ^re (maybe I should invest in an error correcting BBS typer) Right now, I am only running at 4800, but if DECnet is used wisely, it is not too much of a problem. I'm still playing around with it, and will report back if I find any handy tips for performance enhancement. Once in a while I lose line sync for 5 seconds, so I'll check into Roberts suggestion about NCP fiddling. p.s. any ncp/ddcmp tips would be appreciated. Marty Kuhrt VAX Mgr Executive Software (818) 249-4707 APPLICATIONS SIG Message # 334 0 replies Left by: PETER MYO KHIN On 13-Jun-89 At : (202) 268-2411 Subject: BARCODES APPLICATIONS SIG Message # 335 0 replies Left by: PETER MYO KHIN On 13-Jun-89 At : (202) 268-2411 Subject: BARCODE PRINTING CAPABILITY I am in search of a VAX 8xxx application package that will permit me to print address barcodes to USPS specifications. This package must be accessible from the All-In-1 platform and be printed on LN03s. Any information on this is welcome. Thank you! UTILITIES SIG Message # 373 0 replies Left by: TOM PALL On 13-Jun-89 At : (302)695-1938 Subject: CMS FOR NON-SOURCE FILES There are more and more "configurable" software packages out in the process control/data acquisition arena, ala GENESIS, THE FIX, LABVIEW that after "programming" in them, result not in source code but binary files containing graphics, pointers, hooks being produced. The runtime packages "runs" thru these files during "program" execution. I am searching for a "code" management system for these files. Polytron and DEC/VMS/VAXSET's CMS appear to only handle source code (card image files of ASCII text). I don't really care about running DIFFERENCES. But I do like having revisions, releases, generations, all the things that DEC's CMS offers, plus annotations for any revision, extraction. Try doing that with a READ.ME file or a label attached to the floppy. Transfer to/from MS-DOS, MPW and VMS no problem IF I can find a product. Any leads? P.S. I'm with a large chemical company headquartered in Wilmington, Delaware. RSTS SIG Message # 391 0 replies Reply to # 389 Left by: BRUCE A. GREMBOWSKI On 13-Jun-89 At : 408/291-3364 Subject: JOB NO. FROM KB NO. Phil and Kelvin, Thank you for your (as always) prompt replies. Phil, I tried the monitor table stuff, and I'm not getting the right stuff. I must've gotten some- thing wrong somewhere, but I'm still looking. Kelvin, I'm using RST V9.2 and Basic-Plus V7.1 (I think). All other SYS() calls are working fine. Is there any other way besides polling all the jobs? Maybe I'll just do that anyway. I was using the 128 to return /PERM characteristics for a terminal, as we have NEC spinwriters hooked up to ports with /NOBROADCAST normally set, but sometimes set to /BROADCAST. I was going to use the result of the SYS(6+16+0) call to find out if the port was just a printer. Anyway, I'll keep trying! FORUM - DEC SIG Message # 71 0 replies Reply to # 69 Left by: PATRICK MAHAN On 14-Jun-89 At : (714) 720-7868 Subject: CDA/DDIF Thanks for the suggestion. We have already started a protest movement starting with our sales rep and with Digital educational division. However, this the date this project goes to pilot mode is only 2-1/2 months away and I don't have time to play letter and phone tag with Digital for support. Oh well, this is not the first time my back's been up againist the wall. Thanks for the help. CLUSTERING/NETWORKING SIG Message # 409 0 replies Reply to # 405 Left by: PATRICK On 14-Jun-89 At : (714) 720-7868 Subject: DECNET & TCP/IP I don't know how much you have been told but Sun provides its one DECnet package. It allows for the Sun to behave as a DECnet end node. However, as soon as Phase V DECnet is out with support for NFS (this is supposely the connection link for the VMS/Ultrix connection) Sun has said that they will drop support for their package. We have decided to implement TCP/IP on our main VAX router. We use a package from EXLAN that is a combination of hardware and software support. This gives us full TCP/IP functionality such as file transfers: FTP, remote login: TELNET, and the ability to send VAX/VMS Mail over the ethernet to our Unix machines. We find this is much easier than support dozens of third party DECnet products on all our Unix machines (Sun's, Apollo's, HP's, SGI's, etc). EDITORS SIG Message # 217 0 replies Reply to # 216 Left by: PATRICK MAHAN On 14-Jun-89 At : (714) 720-7868 Subject: CALLABLE EDT-EDITOR I don't know if you know it but the EDT editor has a callable interface to it. However, I don't believe you can tailor it to your needs. Have looked at using TPU and building your own custom editor? The TPU editor has a callable interface that you can access from inside your program and TPU is very tailorable. For more information see the TPU programmers guide. APPLICATIONS SIG Message # 336 0 replies Reply to # 333 Left by: BRIAN DESANTIS On 15-Jun-89 At : (201)670-0286 Subject: USERWARE Rich Nardiello - I had to laugh when reading your message abour Userware and Userbase. Not 5 minutes before, I was thinking of putting a message on Aris about the very same thing - I just received my package two days ago and installed it yesterday. I am running version 3.4-11 and would love to talk to you about it and vice versa... I am a little miffed though - I was playing around with all the programs and then the license was installed today, and now I learn I cant play with them anymore, because we dont have a license for them. Oh well.... Look forward to hearing from you. VMS SIG Message # 980 0 replies Reply to # 976 Left by: BRIAN DESANTIS On 15-Jun-89 At : (201)670-0286 Subject: SECURITY I had a problem with some 'mischievous' users and one of the ways I kept control was to modify their command language tables - it is very hard to perform taks in DCL when all they get is: %DCL-W-IVVERB, unrecognized command verb - check validity and spelling As for using READ instead if INQUIRE - it is a much safer solution, and provides some nice qualifiers that handle errors (/ERROR=branch), control/z's (/END_OF_FILE=branch) and allow the program to branch if the user doesnt input something within a certain amount of time (/TIME_OUT=seconds). Under VMS V4.7 and below, there was an equation that would return the total number of interactive users, from DCL. It was: users = f$cvsi(0,16,f$fao("!AD",4,%X80002BFC)) Does anyone have the V5.0 and abover 'version' of this statement??? I found it very useful to have.... SITE MANAGEMENT SIG Message # 297 0 replies Left by: KEVIN VERBLE On 15-Jun-89 At : 1-714-998-5500 Subject: NODE LICENSING I'd like to know if a PCSA member license is tied to the PC or to the node name. I have two PC's that could take advantage of the PCSA facilities but I only have one license. I'd like to boot each of them as needed onto the network using the same boot floppy, which would mean the same node name. They don't have to be on the network at the same time. If the licensing is by node name, this should abide by the license. Thanks for any information. MICROVAX SIG Message # 207 0 replies Left by: DR. R On 15-Jun-89 At : AT THE GOLF COURSE Subject: GNU MICROVAXES Gentle Readers: Yes, it's true. Uncle Ken and his crew are planning to spring a whole slew of new MicroVAXes on the DEC community! First we'll see the MicroVAX 3100, an upscale replacement for the obsolescent MicroVAX 2000. Later in the year (fall seems reasonable to me) we can expect DEC to transmute the Firefox workstation into a family of multiprocessor MicroVAXes! Alas, we'll have to wait until next year before DEC unleashes a 7VUPS Rigel-based system . . . facts not rumours, DR. R UNIX/ULTRIX SIG Message # 40 0 replies Left by: DR. R On 15-Jun-89 At : RECOVERY ROOM Subject: OFF TO SEE THE WIZARD Yes, the good doctor is off to see the wizard--the Wizard of OZIX, of course. According to sources close to the DECwest lab where this alluring new O/S is being crafted even as we speak (or type), OZIX is a next-generation VMS followon that will run on an as-yet unimplemented 64-bit version of the MIPS RISC architecture. Can you say "Son of MICA?" Sure you can! facts not rumours, Dr. R. TERMINALS/PLOTTERS/PRINTERS SIG Message # 303 0 replies Left by: DR. R On 15-Jun-89 At : INTENSIVE CARE Subject: NAME THAT TERMINAL DEC has a problem and they need your help. Seems that DEC is all set to program-announce their much-vaunted (and even more heavily rumoured) DECwindows terminal and they need a name for this $2,500 gem. "DT300" is the current internal name for the product, but this sobriquet lacks zest appeal! So c'mon and give DEC a hand--why, maybe they'll send the winner of the name-that-terminal contest to KAPRI! facts not rumours, Dr. R CLUSTERING/NETWORKING SIG Message # 410 0 replies Left by: DR. R On 15-Jun-89 At : EVERYWHERE Subject: VAXCLUSTER II: THE SEQUEL Coming soon to a cluster near you--an active star coupler, an HSC100 and parallel CI-buses. Also an increase in maximum number of nodes (perhaps as many as 64) and other goodies too secret to broach in a forum such as this. Sorry, Charlie: no fibre optic CI just yet, so you'll have to sit on that inventory of optical backplane polish for a while longer . . . facts not rumours, Dr. R CPU'S SIG Message # 51 0 replies Left by: DR. R On 15-Jun-89 At : MR04 Subject: THE AGE OF AQUARIUS Admiral Olsen sez full speed ahead with the successor to Nautilus! Ken's latest and greatest is the air-cooled Aquarius, a hot new VAX CPU that will be available in uni through quadratic configurations that will deliver 30 through 108VUPS scalar performance and 55 through 200 debit/credit transactions respectively! Angling for vectors? Great, 'cuz Aquarius' built-in vector accelerator will belt out 18MFLOPS (100 x 100 Linpack) and 125MFLOPS (peak) on a per-CPU basis! That's substantially better than the performance of the vector option that will be offered--at some point in the future--on the VAX 6400 and Firefox workstation. Don't expect to see Aquarius in the Electronic Store until early next year. One and two-CPU systems may ship in winter/spring 1900; you'll have to wait a bit longer for three and four processor versions and Ultrix support. facts not rumours, Dr. R RSTS SIG Message # 392 0 replies Reply to # 391 Left by: Phil Anthony On 15-Jun-89 At : Here Subject: BLUSH! Whoops! Bruce, I just looked over that code again, and I found the problem immediately: The "32" in Mike Maysfield's wonderful book is octal, not decimal ... and BASIC+ doesn't speak octal. (Darn!) Try substituting . DEVTBL% = PEEK (DEVPTR% + 26%) And if that doesn't work, try 24% - the explanatory text doesn't match the numbers given very well. Mike writes about Offsets 0-16: "These twelve words [in DEVNAM, the device name table, whose offsets correspond to the en- tries in DEVTBL] contain the names of all the disks supported by RSTS/E, in the order: DV, DF, DS, DK, DL, DM, DP, DR, DB, DZ, DW, DU. All disk names are included, even if the device is not configured in this system." Now, my octal counting would put 12 word-size offsets at 0-26 octal, or 0-22 decimal, so the "1" in "16" is probably a typo. Then Mike says that Offset 32 (octal) in DEVNAM contains the device name "KB". He doesn't say what, if anything, is at Offset 30 (octal) = 24 (decimal), and I don't have a V9.6 system handy to try it out on. But he talks about Offset 32 (octal) in sev- eral other tables as being the position of entries for KB-type devices, so I suspect that 26% is correct. EDITORS SIG Message # 218 0 replies Reply to # 206 Left by: CARL HOUSEMAN On 15-Jun-89 At : 703-949-1323 Subject: EDT WORKALIKE FOR MS-DOS For the price of a phone call you can download SEDT from various bulletin boards. This is an editor written by a DEC employee on his own time, which improves on EDT with two-windows, overstrike mode, block cut/paste, etc. It is also included in the PCSA client distribution, I believe. Kevin Barkes (of this magazine) should have it on his Opus bulletin board. (412-854-0511) UTILITIES SIG Message # 374 0 replies Reply to # 373 Left by: CARL HOUSEMAN On 16-Jun-89 At : 703-949-1323 Subject: CMS HANDLES NON-SOURCE FILES Since version 3.0 CMS has had the capability to store binary files. Unfortunately, each successive generation will take about as much storage as keeping a separate version of the file, because a single-byte difference in a binary file tends to make the rest of the records different (and CMS is still record-oriented). CPU'S SIG Message # 52 0 replies Left by: DR. R On 16-Jun-89 At : ON CALL Subject: I WANT A NEW SIG Yes. Methinks the time has come for a new ARIS SIG, one dedicated to Digital's soon-to-be expanding RISC product family. Soon-to-be expanding? Read on . . . Reliable sources have informed me that DEC will deploy a plethora of pyrotechnics one week after our nation's natal day. In addition to the VAX 6400 that we've all been waiting for, DEC will roll out its first DECsystem data center servers. These mysterious machines are likely to include R2000 and R3000 uniprocessors in MicroVAX cabinets as well as R2000 and R3000 dual-processor ASMP systems packaged in commodious Calypso cabinetry. Prefer workstations? No sweat, Chet--DEC will ring your chimes with an entry-level Sparcstation-slayer called the DECstation 2100. Also on tap are R3000-based DECstations with 3D graphics and perhaps an entry-level, bounded PVAX based on the KA640 CPU board. facts not rumours, Dr. R SITE MANAGEMENT SIG Message # 298 0 replies Left by: DR. R On 16-Jun-89 At : AT THE FREE CLINIC Subject: AT YOUR SERVICE Yes. DEC's Field Circus ringmasters soon will announce Integrated Services Delivery System, a new service offering that includes Site Service Processor hardware and gee-whiz AI-based predictive diagnostic software. But if an Aquarius mainframe VAX is on your shopping list, you need not make a separate trip to buy your SSP--a MicroVAX-based SSP will be concealed in the cabinet of each and every commercial Aquarius system! facts not rumours (accept no substitutes) Dr. R CPU'S SIG Message # 52 0 replies Left by: DR. R On 16-Jun-89 At : ON CALL Subject: I WANT A NEW SIG Yes. Methinks the time has come for a new ARIS SIG, one dedicated to Digital's soon-to-be expanding RISC product family. Soon-to-be expanding? Read on . . . Reliable sources have informed me that DEC will deploy a plethora of pyrotechnics one week after our nation's natal day. In addition to the VAX 6400 that we've all been waiting for, DEC will roll out its first DECsystem data center servers. These mysterious machines are likely to include R2000 and R3000 uniprocessors in MicroVAX cabinets as well as R2000 and R3000 dual-processor ASMP systems packaged in commodious Calypso cabinetry. Prefer workstations? No sweat, Chet--DEC will ring your chimes with an entry-level Sparcstation-slayer called the DECstation 2100. Also on tap are R3000-based DECstations with 3D graphics and perhaps an entry-level, bounded PVAX based on the KA640 CPU board. facts not rumours, Dr. R SITE MANAGEMENT SIG Message # 298 0 replies Left by: DR. R On 16-Jun-89 At : AT THE FREE CLINIC Subject: AT YOUR SERVICE Yes. DEC's Field Circus ringmasters soon will announce Integrated Services Delivery System, a new service offering that includes Site Service Processor hardware and gee-whiz AI-based predictive diagnostic software. But if an Aquarius mainframe VAX is on your shopping list, you need not make a separate trip to buy your SSP--a MicroVAX-based SSP will be concealed in the cabinet of each and every commercial Aquarius system! facts not rumours (accept no substitutes) Dr. R RSTS SIG Message # 393 1 replies Left by: JOHN HEINZ On 16-Jun-89 At : 513 866-3300 Subject: V8.0 ON PDP11/84 DOES ANYBODY OUT THERE USE A VERSION OF RSTS BELOW 9 ON A PDP 11/84?? WE ARE PLANNING ON UPGRADING OUR HARDWARE BUT WANT TO USE V8.0 . I KNOW THIS ISN'T OFFICIALLY SUPPORTED BY DEC BUT WILL IT WORK OK??? I'D APPRECIATE ANY INFO AT ALL.... THANKS!!!! RSTS SIG Message # 394 0 replies Reply to # 393 Left by: DON ROBERTS On 16-Jun-89 At : HERE Subject: V8.0 ON 11/84 I don't use it, but some years ago, when I was invovled in RSTS, I encountered a few 84's that were on 8.0, so I would presume it would still work. SITE MANAGEMENT SIG Message # 299 0 replies Reply to # 277 Left by: TOM CACHAT On 16-Jun-89 At : ARIS Subject: VAX SHUTDOWN IF POWER FAIL We've come up with two plans to automatically shut down our VAX if we have a power outage: 1. Our electricians our installing some 110V outlets off of the commercial feed to the UPS. We will plug in our VAX modems to these outlets, set the modems to always have carrier detect on and have the VAX ports set /MODEM. If the modem is on, the VAX port will have the /REMOTE characteristic; if the commercial power goes out, the port will have /NOREMOTE due to the loss of the carrier detect signal from the modem. A DCL command procedure can then monitor these modem ports for the the status of the /REMOTE characteristic. 2. The UPS interface to the AS400 will be connected to the AS400. If there is a power problem, the AS400 will be informed via this inter- face and can be programmed to tell the VAX, either directly or via a PC. VMS SIG Message # 981 0 replies Left by: TOM CACHAT On 16-Jun-89 At : ARIS Subject: BYPASSING DEVICE CONTROL LIBR The following is how we print to our LN03 using condensed print, wihtout using the device control library. SH SYM PRLE16 PRLE16 = "@DUA0:[SYS0.SYSMGR.UTILITIES]LN03_16" USAGE: PRLE16 DUA0:[SYS0.SYSMGR.UTILITIES]LN03_16.COM --------------------------------------- $define sys$output nl: $print/que=letter/noflag/noburst/noheader utilities:ln03_16. $deassign sys$output $print/que=letter/noflag/noburst/noheader 'p1' $define sys$output nl: $print/que=letter/noflag/noburst/noheader utilities:ln03_default. $deassign sys$output SHQ LETTER/FULL Terminal queue LETTER, on TTA6: /BASE_PRIORITY=4 /NOENABLE_GENERIC /FORM=DEFAULT Lowercase /OWNER=[1,4] /PROTECTION=(S:E,O:D,G:R,W:RW) /SCHEDULE=(NOSIZE) (*** need this ***) UTILITIES:LN03_16. ----------------------- 90 C4w UTILITIES:LN03_DEFAULT. (changes back to default pitch) ----------------------- c VMS SIG Message # 982 0 replies Reply to # 980 Left by: Robert G. Schaffrath On 16-Jun-89 At : 914-335-9792 Subject: Interactive User Information Brian, The following DCL code will work for VMS V5.0 and V5.1-1: users = f$cvsi(0,16,f$fao("!AD",4,%X800044c0)) ! SYS$GW_IJOBCNT Also, if you want to know the interactive limit: limit = f$cvsi(0,16,f$fao("!AD",4,%X800082de)) ! SYS$GW_IJOBLIM The two memory locations you are looking at are known as SYS$GW_IJOBCNT and SYS$GW_IJOBLIM respectively. In the future, if you wish to determine there locations in a new release of VMS, all you need do is type, in DCL: $ SEARCH SYS$SYSTEM:SYS.MAP SYS$GW_IJOBCNT $ SEARCH SYS$SYSTEM:SYS.MAP SYS$GW_IJOBLIM The '8000nnnn' number returned is the memory address you reference. Also note that you can examine these locations with the DCL 'EXAMINE' command. All you need type is: $ EXAMINE/WORD/DECIMAL %X800044C0 ! Interactive count $ EXAMINE/WORD/DECIMAL %X800082DE ! Interactive limit The method you use, however, allows you to assign the value to a symbol for later use. Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 TERMINALS/PLOTTERS/PRINTERS SIG Message # 304 0 replies Reply to # 303 Left by: RALPH BRANDT On 16-Jun-89 At : 7177515210 Subject: ANSWER I'd suggest they call it a 3290 - That's what Ibm called a similar white elephant. (DEC not only has it now, DEC builds them too, white elephants, I mean) Dr R has arrested social and mental development - he can't give his correct name and location. (He probably has bad breath too) TERMINALS/PLOTTERS/PRINTERS SIG Message # 305 0 replies Left by: RALPH BRANDT On 16-Jun-89 At : 717-751-5210 Subject: DEC BILLING Does anyone have a good billing system they could sell to Ken Olsen. I haven't seen a bill that is right for six months. Being a blue box retread who was accustomed to getting bills that were right month after month, this was a real shock. Any ideas? TERMINALS/PLOTTERS/PRINTERS SIG Message # 306 1 replies Reply to # 304 Left by: DR. R On 17-Jun-89 At : NETWORK-WIDE Subject: DECWINDOWS TERMINAL A *white elephant*? Surely you jest, Ralph. A DECwindows terminal is just what DEC needs to lower the entry price to the wonderful world of DECwindows (try convincing management to purchase an $8K PVAX for everyone on staff!). Incidentally, as a fully distributed process I do not feature attributes such as halitosis or spatial locality. facts not rumours, Dr. R TERMINALS/PLOTTERS/PRINTERS SIG Message # 307 0 replies Reply to # 306 Left by: RALPH BRANDT On 17-Jun-89 At : Subject: MANAGEMENT Management will do what is recommended if you have the proper justification When people say management doesn't listen, they arent talking the right language to them... That's the usual techie copout... There's no substitute for having your ducks in a row, asking for what you need and no more. Most techies would rather spend their time arguing the merits of unibus, Bibus and vanagons rather than talking improved productivity of personnel. After all, the manager only wants a few things. 1) Keep his job 2) Look good to his boss. Give him that and you get what you want. Taking the system down in the middle of the day to test a pet thoery isn't the way. PS. I'm not a manager... Never got that desparate for a job. Just found out the hard way how to survive... (In an IBM shop no less...) This DEC environment is a piece of cake. You guys have had it easy all these years.. MISCELLANEOUS SIG Message # 520 1 replies Left by: DR. R On 18-Jun-89 At : MAKING HOUSE CALLS Subject: DECSALES SUMMER SCHOOL Yes, it's true: Ken Olsen finally has responded to recurring complaints about the quality and competence of his sales force! DEC has rented a major college campus (a "colorful" one in Rhode Island) and is gearing up to send 8,000 salesfolk to a two-week-long cram course on the DEC product line and how to sell it . . . facts not rumours, Dr. R TERMINALS/PLOTTERS/PRINTERS SIG Message # 308 1 replies Reply to # 307 Left by: Phil Anthony On 19-Jun-89 At : Here Subject: MANAGEMENT REDUX Ah, Ralph, if only that were true! In the ideal operation, with an ideal manager, it is; but the world is hard on ideals. A lot of managers have subscribed to a theory of cost containment which says that if you can keep your budget under last year's budget, you'll get brownie points. Forget the question whether you might get more productivity out of your decisions by spending some money, especially if the productivity comes out of another department's hide. It might be more true to say that part of the techie's job is salesmanship. First, you have to identify what the manager's goals are. Then, you justify what you want to accomplish in terms of his or her goals. Don't ask whether your manager's goals are in the company's best interests, sensible, or even realistic; chances are they're not, and your justification will be eyewash of the purest ray supreme. But if you can persuade your manager that what you want to do falls in with how he or she wants to be perceived, you have a reasonable chance of getting it. Not always, of course. A few years back I was involved in a situation where the best justification I could make for expanding the company's VAX presence fit the manager's goals reasonably well. He was overruled, however, at the highest executive levels, where the feeling was that if a computer system wasn't Blue, it was an aberration within the company to be removed at the first possible opportunity. My recommendation unfortunately involved spending money (upgrade to a ma- chine big enough to do the job they were requiring of it). It gave the top executives *their* justification to get rid of DEC completely and move the job to a mainframe - costing initially four times as much as the VAX upgrade would have. That didn't count, however, since they officially made their decision in two parts: first, junk the VAXes citing the additional cost, and second, determine what it would take to migrate the application, with a cou- ple of months between the two. At that point my only possible response was simple: understand what "Mene, mene, tekel, upharsin" meant, tell them they'd made the correct choice, help them as best I could to do the mainframe conversion, shake hands, and find another client. Unfortunately, other folks don't have the options I did. I agree completely, though, that any techie or system manager who doesn't understand that production comes before play is just asking for trouble. And speaking personally, I prefer the Vanagon. (DEC's decision to switch to VAXBI, necessitating total system replacement rather than just CPU upgrade, played a prominent role in the executives' official decision.) MISCELLANEOUS SIG Message # 521 0 replies Reply to # 520 Left by: RALPH BRANDT On 19-Jun-89 At : Subject: DEC BILLING Now he needs to get the people who write the bills trained to write one right... Those bozos could screw up a one car funeral.. TERMINALS/PLOTTERS/PRINTERS SIG Message # 309 0 replies Reply to # 308 Left by: RALPH BRANDT On 19-Jun-89 At : Subject: PHIL You missed part of my thesis, ie. that you have to know what the manager wants, then decide how you can best serve within that framework. Sounds like you understand the situation. For every one like you there are a host that still blame the manager for their own inadequacies PROGRAMMING LANGUAGES/AI SIG Message # 402 0 replies Reply to # 399 Left by: V. JAMES KRAMMES On 19-Jun-89 At : (614) 228-2001 X 341 Subject: FREE SOFTWARE The Free Software Foundation has an AWK-like utility (called GAWK). Never having used AWK, I cannot vouch for it's quality, but I've been using their C compiler for several months now without a single problem. A collection of their stuff was on the (I think) VAX88A1 tape from DECUS. You can contact the FSF directly on the internet at: gnu@prep.ai.mit.edu I don't have a US-Snail address for them. If you need more info (this is sketchy at best) please call me at the above number [(614) 228-2001 x 341]. RSX SIG Message # 277 0 replies Reply to # 274 Left by: Barton F. Bruce On 20-Jun-89 At : (617) 868-1111 or HERE Subject: online BRU of LB: is OK First off, you should be able to do what you are trying, but you have a 3rd party controller, and BRU has a long history of weird behavior... Anyway, I regularly BRU from the system disk I'm booted on. No problem. Try to have as little running as possible, and if you are verifying, expect [1,6] errorlog related file differences between passes. We are running M+ exclusively, now, but when we were in transition, we always (against DEC's advice) used the LATEST M+ BRU on M. Our disks and tapes are all non DEC. I only mentioned your 3rd party controller because you didn't identify it. Newer controllers tend to be MSCP emulations and so the disks are DUx:. Yours must be older. CLUSTERING/NETWORKING SIG Message # 411 0 replies Reply to # 405 Left by: Barton F. Bruce On 20-Jun-89 At : bruce@ccavax.camb.com Subject: another source for VMS TCP/IP In ATLANTA there was one vendor that really stuck out as being worth looking at. They are: TGV (Two Guys and a Vax). They gave a Sunday PSS, and in other sessions easily answered questions that stumped other vendors. I think they are the 'guys' that wrote what some others are now selling, and they are the ones that at least one of the other vendors was trying to hire (all rumors at Decus, but I bet true). Anyway they are now developing/selling TCP/IP s/w on their own, although I think one of the previously mentioned sources (SRI) is also peddling their s/w. Someone who was checking prices said that he was ordering from TGV 1) for price, and 2) to be dealing with the source. CLUSTERING/NETWORKING SIG Message # 412 0 replies Reply to # 395 Left by: Barton F. Bruce On 20-Jun-89 At : Decuserve (EISNER::BRUCE) Subject: NOVELL supports VMS Novell has s/w to let you run on Ethernet and (I think) even use a VAX as a file server. I doubt they seriously expect anyone to buy their product if the user already had PCs using DEC's PCSA, the VAX end of which is licensed FREE to all VMS machines. But they do intend users to buy it if they are already heavily into a NOVELL environment, but user are screaming for better connectivity to their VMS world. With DEC's new pricing that puts the PC end PCSA license at < $100 (at normal discounts) over the price of the PC's DEPCA Ethernet controller card (when they are bought together as a package) I also doubt Novell is seriously trying to catch even new PC users at a VMS site. But for the company with well established Netware users, and equally well established VMS users, their product lets them hookup to each other using the connectivity of Ethernet. And that may fit your situation. Your 2 seperate sites both have Ethernet, now you only have to bridge them together. There are many many Ethernet bridges that would work. You can spend from well under $5k per end to well over $20k per end. You should have as much bandwidth as you can afford between them. 56kb is a good minimun, but some folks do it at 9600 (ouch). If you have MANY users and will be file serving across it, look at T1 type speeds. The latest rage is 'fractional' T1. The 24 DS0 time slots (64kb each) of a T1 are sold at much more attractive prices than the traditional 56kb DDS service. You must rent a T1 access line at each end to the carrier's POP (Point Of Presence) but then you only buy the DS0s you need. For example: Boston to NYC is $211./month POP to POP for each 64kb from one of the carriers. Your T1 access line might also carry additional circuits to other sites as well as in or out WATTS traffic, all effectively lowering your portion of the cost of the T1 'pipe' into the long haul carrier. If you are INTRA-LATA, pray the local TELCO has filed similar tarrifs. Long haul C&W has it, and recently even AT&T has it under some marketing name. VMS SIG Message # 983 1 replies Left by: BARRY YOUNG On 20-Jun-89 At : HERE Subject: BROADCASTING TO SERVERS Does anyone know of a method to broadcast messages to DECserver ports that are not logged in?? We need to be able to send messages to disconnected terminals to inform them of systems being available for use. Any help would be appreciated. Regards, RSTS SIG Message # 395 1 replies Reply to # 392 Left by: BRUCE A. GREMBOWSKI On 20-Jun-89 At : 408/291-3364 Subject: JOB NO. FROM KB NO. Phil, I reread that section you mention and tried looking at DEVPTR% + 2%/4% etc. and found that DEVTBL% = PEEK(DEVPTR% + 20%), at least in RSTS V9.2 (which is what we are running). The routine works great now! Thanks for pointing me in the right direction initially. Bruce RSTS SIG Message # 396 1 replies Reply to # 396 Left by: BRUCE A. GREMBOWSKI On 20-Jun-89 At : 408/291-3364 Subject: RSTS V8.0 ON 11/84 John, At my former job we ran RSTS V8.0 on PDP-11/70s, a PDP-11/44, a PDP-11/24, and PDP-11/23s. The only problem we ran into was with the RA-80 and RA-81 disk drives. For some reason, when they were genned on the system as the main disk, an use of the RT11 run-time system would cause a system crash. As PIP uses RT11, this was a major problem. DEC finally had to send some teen-aged guru from Colorado to fix our problem. He came up with a bizarre solution: we re-genned the system, telling it that we had one of the RP06/RP07 type devices on the system. This solved the problem. Other than that, everything runs great. This fix was necessary on the 11/70s, but not on the 11/23s, so you may or may not need it. In any case, good luck! Bruce PROGRAMMING LANGUAGES/AI SIG Message # 403 0 replies Reply to # 397 Left by: DAVID ROBISON On 20-Jun-89 At : 702-364-8633 Subject: ATEXIT() I tryed a very short program but still got the access violation. I will check for any appropreate compiler switches. LASER PRINTERS SIG Message # 72 1 replies Left by: RAGU RAO On 20-Jun-89 At : (703)264-1080 Subject: LASER PRINTER SETUP Is there a (simple !) way of changing the standard VMS print symbiont, or the terminal port setup to get messages from the printer back on the terminal screen of the user who initiated the print request. Thanks. MASS STORAGE SIG Message # 261 1 replies Left by: DR. R On 20-Jun-89 At : AT THE GOLF COURSE Subject: A SCUZZY RUMOUR Rumour? (Sorry, I must have confused myself with that phony rumourmonger in the tabloid!) No, the *FACT* for the day is that DEC will announce a 3-1/2" SCSI floppy and a 632MB SCSI hard disk for the DECstation 3100 on July 11. Possibly an 8mm tape, too . . . FACTS not rumours, Dr. R RSTS SIG Message # 397 0 replies Reply to # 395 Left by: Phil Anthony On 21-Jun-89 At : Here Subject: JOB NO. FROM KB NO. Bruce, the official way to find the offset into the DEVPTR table is to first get hold of DEVNAM. The device name table contains a series of two-letter device names, starting with disks and then moving on to other devices. One reads the device name table until the letters "KB" appear. The offset into the device name table is the same offset you use into the device pointer table to get the address of the DDB for KB0:. Then you add keyboard number times 2 to the value in the device pointer table to locate the address of the DDB for the one you want. If you do it this way, you'll be reasonably safe for any release of RSTS you come across. I tried to shortcut the procedure since the good Mr. Mayfield seemed so sure about the hardwired Offset 26. Probably he's a hundred per cent correct - for V9.6 of RSTS only. CLUSTERING/NETWORKING SIG Message # 413 1 replies Reply to # 406 Left by: FRANK INNES On 21-Jun-89 At : 215-448-3461 Subject: DECNET & TCP/IP FROM CMU Having begun this discussion, I thank all of you for your very informative replies. Can you provide me with an address &/or phone number for the Carnegie-Mellon "cheapware" TCP/IP implementation for the VAX? Thanks again for your help. UTILITIES SIG Message # 375 1 replies Left by: ROGER BALDWIN On 21-Jun-89 At : 312-766-2800X558 Subject: LINK ERR PSECT/TRUNC When linking a basic program and several macro routines the linker generates a truncation error in one of the macro routines. I have tried tracing it based on the error offset in the message but it doesn't seem to match up. I would appreciate any suggestions on how to track this type of error since the manuals tend to give it about 3 lines. MASS STORAGE SIG Message # 262 1 replies Reply to # 261 Left by: KIT NOLSEN On 21-Jun-89 At : DEC Subject: DR. R Dr. R is a DEC employee who's sloe job is to make announcements of DEC products without DEC seeming to be responsible for the announcement. There's a switch, DEC being responsible. VMS SIG Message # 984 0 replies Reply to # 983 Left by: ROBERT G. SCHAFFRATH On 21-Jun-89 At : 914-335-9792 Subject: BROADCASTING TO SERVERS It may be possible to broadcast to a server by creating an application port through LATCP, mapping it to the desired DECserver port, and then 'writing' a message to it. Afterwards, delete the port and move on to the next one. This would definitely be messy but should work. The process could be likened to making each port a line printer port temporarily and print a one line message to the port. Under VMS V5.0, there are $QIO services which will let you map a port. You could use LATCP to create a port such as LTA123, and then run a program which changes the port mapping and each time writes a one line message to the server. Hope this helps. Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 PROGRAMMING LANGUAGES/AI SIG Message # 404 0 replies Left by: STEPHANIE LYON On 21-Jun-89 At : (609) 737-0425 Subject: UNIX SYSV COMPATABILITY LIB FOR VMS I am looking for a UNIX System V (or ULTRIX) compatability library for VMS. In particular, we are interested in IPC. We are porting some software from UNIX to VMS (and maybe ULTRIX) and would like to get away easily. It matters not from where it comes. VMS SIG Message # 985 0 replies Left by: STEPHANIE LYON On 21-Jun-89 At : (609) 737-0425 Subject: UNIX SYSV COMPATABILITY LIB FOR VMS We are looking for UNIX System V (or ULTRIX) Compatability library for VMS. In particular, we are interested in IPC. We are porting some software from UNIX to VMS and would like to get away easily. It matters not from where it comes. UTILITIES SIG Message # 376 0 replies Reply to # 375 Left by: ANDREW DUGGAN On 21-Jun-89 At : Subject: LINK ERR PSEC/TRUNC Roger, You you need at most 2 things to track this error. 1) A linker map 2) A Macro Listing The linker map will tell you about each .psect and each object module and where they start. Subtract the starting offset from the linker map for each object module that contains that .psect. The smallest positive difference will tell you which is the offending object mod- ule. The macro listing will give the offsets from the .psect directive which will take you straight to the offending code. To speed the process along, you might try putting different .psect directives at the begining of each macro routine, if you haven't already done that, which will make locating the code from the map/ listing easier. Hope this helps. MASS STORAGE SIG Message # 263 0 replies Reply to # 262 Left by: DR. R On 21-Jun-89 At : CERTAINLY *NOT* AT DEC! Subject: A MASS OF STORAGE TRIVIA No, the good doctor is not the proud possessor of a Digital employee badge. But, the doc *does* (from time to time) possess high quality information about unannounced DEC goodies. Rather than ratholing all the lurid details in the Dr. R Upcoming Products File (UPF), I would prefer to share the wealth through Doctor Dave's Distributed VAX Emporium! What wealth? Thought you'd never ask! How's about removable RF30 disk drives? Not unexpectedly, DEC refers to these as RSEs, or Removable Storage Elements. A 300MB two-pack will set you back about $14K, which is equal to one month's malpractice insurance. Want even more capacity? Well, it just so happens that the Digital Storage Architects are hard at work crafting a 4.48GB array of removable RA70 drives. This 16-spindle aggregation is slated for availability sometime before the debut of an even more commodious RA70 32-pack. Also on tap in the near future will be DSSI versions of the aforementioned RA70 as well as the capacious and capable RA90. Does this mean that dual-porting is coming to the DSSI family? And is this a portent of the fabled Local Area System? Stay tuned for details! facts not rumours, Dr. R SITE MANAGEMENT SIG Message # 300 0 replies Reply to # 294 Left by: HOWARD DIAMOND On 22-Jun-89 At : 301-427-7436 Subject: HITMAN PROCESS KILLER I AM VERY INTERESTED IN A PROCESS KILLER LIKE THE HITMAN PACKAGE WHICH YOU MENTIONED. CAN YOU PROVIDE ME WITH ANY INFORMATION ABOUT HOW TO GET IT, AND WHETHER IT MAY AVAILABLE ON A SHAREWARE BASIS. ANY INFORMATION WOULD BE APPRECIATED. THANK YOU. VMS SIG Message # 986 1 replies Left by: JOHN P. FLYNN On 22-Jun-89 At : (615)269-2266 Subject: SET TERMINAL/INTERACTIVE Hi! Can anyone give me some information about the /INTERACTIVE terminal characteristic? All I know is that /NOINTERACTIVE is the same as /PASSALL, but these characteristics don't seem to be documented anywhere. Also, is there a reason why it's not documented? Thanks in advance. John. DATA/FILE TRANSFER SIG Message # 559 0 replies Left by: Phil Anthony On 22-Jun-89 At : Here Subject: ARIS MENUS Well, I suppose they had to do it (sigh) ... the old menus did make it dif- ficult to locate some of the options (especially "Dear Sysop"). But every time the prompts and menus change, it takes me one to two hours to rewrite my automatic message download scripts. This time I also had to regen, recompile, and relink the program that gets rid of inter-message prompts and stray s as well. Thank Heaven for Mirror's LEARN mode and lex on my XENIX system! Now that the work is done, anybody else who uses Mirror on their PCs to get hold of ARIS messages (Dave Kaplan, take note) would be welcome to a copy of my scripts or the prompt stripper program. Give me a call at (215) 884-9128 to arrange a download. 'Course, there'll be yet another version next week after the login sequence changes .... EDITORS SIG Message # 219 1 replies Reply to # 217 Left by: KEITH PINNEY On 22-Jun-89 At : (612) 778-5191 Subject: CALLING A CUSTOMIZED TPU I have looked at calling TPU also, but it isn't tailorable enough. I'm looking for the ability to basically edit in a window on the screen and TPU takes over too much. I'm really looking for something that handle a screen definition like FMS, but with a more powerful in-field editing capability. PC'S SIG Message # 222 0 replies Reply to # 216 Left by: JEFF CLEARY On 22-Jun-89 At : HERE Subject: VT220 EMULATION Another choice aside from Poly-comm is the Rainbow connection software which can still be ordered from Digital. I suggest however that you download a program called LCTERM from Kevin Barkes bulletin board 412-854-0511 and use the terminal emulation contained with it. Save some money too. SYSTEM PERFORMANCE SIG Message # 230 0 replies Left by: DR. R On 22-Jun-89 At : IN THE PSYCH WARD Subject: A FULL ACCOUNTING Yes. After years of providing an operating system which is bereft of robust project-based accounting facilities, Digital is hard at work hacking up just such an offering for VMS! Could this be related to a massive VMS system software development program that was sparked by the forthcoming AQUARIUS system's need for mainframe-class "data center management" software? Could be; stay tuned for all the lurid details as they become available. facts not rumours, T is more... to continue, or enter S to skip to end of message? Dr. R MISCELLANEOUS SIG Message # 522 3 replies Reply to # 515 Left by: FINKELL On 23-Jun-89 At : 404-980-1190 Subject: VAX UPTIME PHIL, DO YOU KNOW WHERE I MIGHT BE ABLE TO GET SOME STAITISTICS ON VAX UPTIME, ESPECIALLY ON THE VAXSTATION 2000. JOHN F. MISCELLANEOUS SIG Message # 523 0 replies Reply to # 522 Left by: Phil Anthony On 23-Jun-89 At : Here Subject: VAX UPTIME John, the specifics I have about VAX CPUs are hopelessly outdated. (In my defense, I'll simply mention the DEC field engineer who told me he couldn't keep track of all the new VAXen himself and referred to them generically as the "VAX 8995 plus tax." No, Mr. Olsen, the name is confidential.) But DEC should be willing to provide you that information; or if Richard Gilbert or Robert G. Schaffrath is listening in, one of them might be able to help. MASS STORAGE SIG Message # 264 1 replies Left by: JOHN HENLEY On 23-Jun-89 At : ARIS Subject: UVAX 2000 & 9-TRACKS This probably sounds totally crazy, but does anyone out there have any ideas for getting a uVAX 2000 to write 9-track tapes. We have a TK50, and short of using one of our clients machines (which is a hassle), we have no way of sending a client a 9-track. I realize we could change machines, but we're only 1-2 users, and the expense doesn't seem justified. We are willing to spend money to solve this, so any hints are greatly appreciated. John Henley VMS SIG Message # 987 0 replies Reply to # 986 Left by: PATRICK MAHAN On 23-Jun-89 At : (714) 720-7868 Subject: /INTERACTIVE If you will look in the I/O Users guide, Part I under the chapter on the Terminal driver you will find explanations on the characteristics of the terminal port. EDITORS SIG Message # 220 0 replies Reply to # 219 Left by: PATRICK MAHAN On 23-Jun-89 At : (714) 720-7868 Subject: CUSTOMIZED EDITORS Keith, You may write your own tpu editor. The tpu editor is not just what you get when you enter EDIT/TPU or use the callable TPU interface. You may write you own editor in tpu from the ground up. Digital's vast doc set supplies a manual on how to write tpu code, but the name eludes me. Good hunting. LASER PRINTERS SIG Message # 73 0 replies Reply to # 72 Left by: PATRICK MAHAN On 23-Jun-89 At : (714) 720-7868 Subject: It depends on what kind of laser printer are you running? If you are using the Printserver 40 or Scriptprinter then you are utilizing the LPS software (Laser Printer Software?). The newest release v2.0 supports an accounting feature that allows you to track what user started the job and how many pages did he print. There is a article in Digitial Review, June 19, 1989, page 39 on the Scriptprinter that gives a good overview of the LPS software. DATA/FILE TRANSFER SIG Message # 560 0 replies Reply to # 496 Left by: CLAUDE BAUDOIN On 23-Jun-89 At : (512) 356-3108 Subject: VMSMAIL-PROFS GATEWAY Sorry that this is so late, but... There is a VMSmail-PROFS gateway made by Interlink. I use it all the time, it connects the half of National Semiconductor which uses VMSmail to the half that uses PROFS. It translates the addresses, the text of the message, and the format correctly, as long as the VMSmail users don't have a username longer than 8 characters (that's the PROFS limitation). From the VAX side, you address to GATWAY::IBMNODE::USERID, as if it was a non-routing DECnet intermediary node. From the IBM side, you send directly to VAXNODE(USERNAME), and all sysops on the IBM side have to declare the VAXNODE in their RSCS tables, indicating that the route to it passes through the gateway. Replying to messages, forwarding, etc., is handled well in both directions. PROFS acknowledgment requests (".ak" line) are lost, however, for the obvious reason that there is no such concept in VMSmail. CLUSTERING/NETWORKING SIG Message # 415 0 replies Reply to # 413 Left by: RICHARD B. GILBERT On 23-Jun-89 At : (609) 452-5130 Subject: DECNET & TCP/IP FROM CMU Frank, CMU-TEK TCP/IP may be ordered from: Carnegie Mellon University CMU-TEK Software 4910 Forbes Avenue Pittsburgh, PA 15213-3890 Attn: Lynne Stonis You may also send E-mail to Lynne Stonis; Lynne.Stonis@VB.CC.CMU.EDU MISCELLANEOUS SIG Message # 524 0 replies Reply to # 522 Left by: RICHARD B. GILBERT On 23-Jun-89 At : (609) 452-5130 Subject: VAX UPTIME John, Here is a random sampling of VAX uptimes. VAXStation 2000 has been up for 92 days 6 hours. Last reboot was because the owner tried to run some kernal mode code for VMS V4 under V5 and it crashed! VAXStation 2000 has been up for 33 days. Reason for last reboot unknown, probably power failure. VAX 8200 has been up for 10 days. Last reboot was due to shutdown for PM. This system has run for more than 60 days without a reboot. VAX 11/750 runs for seven days unless there is a power failure. It is rebooted every Monday morning following disk backup and disk compression using Mike Levine's JUICER package. If I didn't need to compress the disk on a regular basis, there is no reason why this machine couldn 't run 30 to sixty days between reboots. In short, a VAX can be expected to run damn near forever if you can supply uninterrupted electric power and cooling (for the bigger VAXen). Most downtime is schedule downtime for disk backups, PM, software installation, system tuning, etc. MASS STORAGE SIG Message # 265 0 replies Reply to # 264 Left by: RICHARD B. GILBERT On 23-Jun-89 At : (609) 452-5130 Subject: UVAX 2000 & 9-TRACKS John, Attaching a nine track tape to a uVAX 2000 would be a custom job from beginning to end. Expensive would be an understatement. There is nothing on the uVAX 2000 to attach a tape drive to. If you must have nine track tape, buying a VAX with a bus of some sort would be by far the cheapest route. It might be possible to attach a tape drive to a VAXStation 3100 but don't believe it until you see it. I would suggest a used uVAX II as the cheapest solution. MISCELLANEOUS SIG Message # 525 0 replies Reply to # 522 Left by: PHILIP GRAVEL On 24-Jun-89 At : HERE Subject: VAX UPTIME Let me add my two cents worth. I have several MicroVAX II's that are doing real time data acquisition. One system (the one that is most busy) has been up for 143 days. (Watch, it be down when I go in on Monday!) The last time it was down was to add an additional disk drive. I agree with everything Phil Anthony and Richard Gilbert have said. The only reason a VAX might to down other than a hardware failure (machine check) or software failure (bugcheck) would be if the pagefile were too small and it became so fragemented inter- nally, that VMS could no longer work with it. Phil VMS SIG Message # 988 0 replies Left by: PHILIP GRAVEL On 24-Jun-89 At : HERE Subject: READING MEMORY LOCATIONS Is there a way I can query the contents of specific memory location from a higher level language such as Fortran? For example, I'd like to be able to find the value stored in %X8000443C (PSM$GL_USER). I can get this value using the DCL command EXAMINE, but I can't figure out how to do it from Fortran. Using SYS$GETSYI I can get the boot time, but not the value of time in user mode. Phil RSTS SIG Message # 398 0 replies Reply to # 387 Left by: Brett Bump On 25-Jun-89 At : (308) 432-2828 or Hea-Duke Subject: Try It Again Bruce. I'm sorry that I didn't catch your message sooner but my Arismail is usually a minimum of 7 days behind, and my Arisanswers another 7 days behind that. Either solution that Phil or Kelvin have given will provide you with the job number (although uu.trm is recommended). You're obviously describing byte 4 (or FQSIZM), which as you have discovered contains the keyboard number. Back up 1 byte. Grab the one on the right. In the directives manual it looks like this. (3-477 V9) (3-322 V8) . +-------------------------------------------+ 3 |/////////////////////| current job no. * 2 | 2 (your bytes 2 and 1) . +-------------------------------------------+ 5 |255(current) or KB no| KB status | 4 (your bytes 4 and 3) . +-------------------------------------------+ The only change that has been made over the years is the addition of the status bits 0&7 (in KB status) in V9.x. Otherwise the information here has not changed. I take that back. As I recall byte 2 (///) is undocumented info for the total Kb's genned into your system (probably invalid for 9.6). VMS SIG Message # 989 0 replies Reply to # 984 Left by: BARRY YOUNG On 25-Jun-89 At : HERE Subject: Robert, Thanks for your help once again. I was hoping for an simple undocumented method, but DEC have obviously not left any loopholes here (better luck next time). I'd better get back to our system and start knocking up a macro routine to do this before our users complain that no-one tells them anything (mind you, I have to work like that so why shouldn't they???) Thanks again, rega > CPU'S SIG Message # 53 0 replies Reply to # 51 MISCELLANEOUS SIG Message # 526 0 replies Reply to # 523 Left by: ROBERT G. SCHAFFRATH On 26-Jun-89 At : 914-335-9792 Subject: VAX UPTIME VAX uptime, the final frontier? It is an interesting topic. I can't recall a study being done on the subject. I know DEC pushes VAXclusters so that you can keep a system running 24 hours a day seven days a week but you still have to shutdown the individual CPU's occasionally. Just to add to the pool of information already supplied, our 6330 has been up 60 days and our MicroVAX II has been up 63 days. This is the longest any of the systems has ever been up (I am thinking of getting one of those 'plant safety' signs which says 'xxx days without an accident, the previous record was yyy' and changing it to read 'crash'!). Most of our down time lately has been due to either a bugcheck from some flakey third-party software (it uses $CMKRNL naturally!), preventive maintenance work, SYSGEN parameter adjustments, software upgrades, and disk drive repair. Still, we have managed to stay up for 2 months now. In fact, the formation date of our cluster is March 15, 1989. I hope someone is making note of all these uptime statistics being presented by various people. Looks like a paper for DECUS. Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 VMS SIG Message # 990 0 replies Reply to # 988 Left by: ROBERT G. SCHAFFRATH On 26-Jun-89 At : 914-335-9792 Subject: READING MEMORY LOCATIONS IN A HLL It is quite easy to read memory locations in a High Level Language. Even easier if those locations are protected to allow user mode read. One note, on your message, the memory location is PMS$GL_USER not PSM$GL_USER. To read this location using BASIC, the following little code will do it: 1 external long PMS$GL_USER print PMS$GL_USER end Compile the code and then link it using the following command (assuming the program is called PMS): $LINK PMS,SYS$SYSTEM:SYS.STB/SELECTIVE_SEARCH This will then print the contents of the PMS$GL_USER memory location. Also, just for information if you change the line 'external long PMS$GL_USER' to 'external long constant PMS$GL_USER', the variable PMS$GL_USER will be the memory location itself, not the contents of that memory location which is what you want. Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 CLUSTERING/NETWORKING SIG Message # 416 0 replies Reply to # 386 Left by: Barton F. Bruce On 26-Jun-89 At : (617) 868-1111 Subject: CRC speedup due soon The emulated CRC inefficiency has become a very sore point. The soon to be released (actually it will be a staged release with succesive changes to allow for backwards compatibility issues) new version of BACKUP will have a DRAMATICALLY faster CRC calculation. It is done with a modestly larger in memory table than whatever is currently used, but the performance is supposed to be dramatic. I believe that the CRC change is NOT at the BACKUP level, but at the VMS level and so other users of the offending instruction will also benefit. Of course all the other super benefits of the new BACKUP are implemented in BACKUP. MIGRATION ISSUES SIG Message # 50 1 replies Reply to # 48 Left by: Barton F. Bruce On 26-Jun-89 At : (617) 868-1111 Subject: The easiest transitions from RSX to VMS (or the reverse!) are often done over a period of time. If you have DECNET on Ethernet this is trivial. Your files can exist on EITHER machine and programs on EITHER machine can access them (within reason). You can migrate applications one by one. Another trick is to USE the dual ports on your disk drives. A cable from each machine goes into each drive. Note WELL that SIMULTANEOUS access is not possible, but any disk can be mounted on either machine. RSX doesn't understand ODS-2 disks, so you can't easily use your regular VMS disks under RSX, but RSX's ODS-1 disks (or ones INIted for ODS-1 on the vax - see INI /STRUcture=...) can be used on either system. If you have an RA60 for various occasionally used stuff, both systems can share the drive, and VMS can use it for ODS-2 packs. Even if the A/B button isn't pushed in to allow access from a port, the drive NUMBER is still visible to the controller connected to that non enabled port. Any controller (e.g. UDA50) NUST NOT SEE the same drive number twice on 2 different cables. To coexist with a VAX, it is often desirable to NOT be limited to unit numbers of simply 0-3. M+ sysgen makes this a royal PAIN, as they LIMIT you to 4 UCBs being generated per controller, and since there is NO functionality such as: CON SET PORT DU3: /PHYS_UNIT_NUMB=7 you are effectively stuck with 4 numbers, generally 0-3. If you need more, it is simple to fix! Extract module SGNMAS from SYSGEN.CLB and there are two places you will have to edit. One is an explicit limit of possible unit numbers of (4 x Controllers). The second is a limit of 4 units for each controller. These tests are not too near each other - this is a BIG command file. Keep a copy of the original .CLB, and stuff the fixed module back into a new copy of the .clb and do a component mode sysgen to just rebuild DUDRV. I often build for 0-17, but remember that each UCG so generated means that much less primary pool, so do what you need, but don't get too piggy. If you are not too comfortable with command files, I could mail you the CAMPARE listing of the modified vs original SGNMAS, as this would only show a few lines of DEC's copyrighted module and is 'legal' to do. Please please SPR this, as DEC needs to KEEP hearing this is &(K{v%[ about this at DECUS symposia. MIGRATION ISSUES SIG Message # 51 0 replies Reply to # 50 Left by: Barton F. Bruce On 26-Jun-89 At : (617) 868-1111 Subject: ...more line noise chopped that previous message off and before I had entered a 'Subject', sorry. and unlike with VAXNOTES , I don't know of a way to fix it... Anyway BUG DEC about fixing SYSGEN to allow > 4 units per controller. To answer the 'can VMS read BRU taoes' question, there have been such tools on VMS SIG tapes. Try in [VAX000] or maybe in [vax000.tools]. There used to be something called VAXBRURDR there. I have NEVER needed it because DECNET over Ethernet is FAST, and, if I needed to make a VMS readable tape for someone, I would either use BACKUP from the files copied over or from an ODS-1 disk mounted on the VAX, or simply use PIP to get them to ANSI mag tape. Some people swear BY (or AT) using RMS CONVERT and related utilities to get stuff between RSX and other DEC op syses. I haven't tried it. VMS SIG Message # 991 0 replies Reply to # 987 Left by: JOHN P. FLYNN On 26-Jun-89 At : (615)269-2266 Subject: SET TERMINAL /INTERACTIVE Sorry, the reason I asked about /INTERACTIVE (and /PASSALL) is because they do not appear to be documented in the Terminal Driver section of the I/O manual (V4 or V5) or anywhere else that I can find. So, again, can anyone tell me what this terminal characteristic means? Thanks. John. VMS SIG Message # 992 0 replies Left by: PHILIP GRAVEL On 26-Jun-89 At : HERE Subject: DIRECTORY SIZE I'm sure I have seen this question before, but I can't seem to locate the conversation. At any rate, what is the limitation on how large a directory can be? As I recall, it is limited by the size of the directory file and the requirement that the file be contiguous. I have one directory that was causing a problem and after deleting some files the directory file had a size of 997/1026. Thanks, Phil VMS SIG Message # 993 0 replies Reply to # 947 Left by: MATT QUIGLEY On 27-Jun-89 At : PACT (215)662-5500 UNTIL 7/15/89 Subject: BACKUP SAVESET NAMES ON TAPE >17 CHARS Hi folks! This msg should explain partly why BACKUP has problems finding those long-named savesets on tape. I ran into the same problem a while ago and, after re-reading my VAX/VMS Guide to DISK & Mag. Tape operations, i found out why. Altho BACKUP reads/writes a tape as a non-labelled ASCII tape, it does write file header & trailer labels on the tape for each saveset. This lets you mount the tape later as a file-structured (VMS Standard Label) tape and copy the savesets to disk (as files). The ANSI specs for writing file header labels specify that if the filename exceeds 17 chars (VMS allows up to a 39 char file + 39 char file type for a BIG filename), that the remaining chars be written in the HDR3/HDR4 labels. BACKUP deals with the tape as a NL (unlabelled) tape, and apparently only looks at the HDR1/HDR2 set of file header labels (compatibility with earlier versions of VMS?). It doesn't truncate the saveset name on the command line because you might just be writing that saveset to disk (where the big-name is OK). Suggestions/Work-Arounds: i) Limit your saveset names to 17 chars or less. Seventeen chars should be plenty to describe a saveset. ii) If you should need more than 17 chars, only specify the first 17 chars when restoring the saveset. iii) Or, use the "SET MAGTAPE/SKIP=FILE:n" command to position the tape at the beginning of the saveset and use "BACKUP tape: disk: ..." without specifying the saveset name. BACKUP will default to using the next saveset it finds on the tape, without bothering to check the saveset name. mq/VAX MISCELLANEOUS SIG Message # 527 2 replies Reply to # 522 Left by: Phil Anthony On 27-Jun-89 At : Here Subject: VAX UPTIME Sorry. I should've added Philip Gravel to my list, and probably one or two others as well. Just to stir up a hornet's nest (gotta keep things hopping around here!), I was a little surprised that the uptime was so short - 93 and 63 days, for instance, being considered good. Yes, I noted the 143-day record, but that appeared to be the exception rather than the rule. That's pretty well com- parable to my experience on the older VAXes about five years ago, and I'd kind of expected an improvement. It's particularly disturbing because the PDPs I worked on ten years ago generally gave uptime approaching the sum of mean time to failure for the individual hardware components, or about six to eight months. Sorry, I don't have exact figures any more. I wonder if the quoted figures would change significantly if the question were modified to ask about uptime excluding planned shutdowns for PM, BACKUP, and power failure. That might provide a more meaningful figure by eliminating causes of downtime which aren't inherent in the VAX hardware or the VMS operating system, and which can be avoided by foregoing PM (I gather that's the case with the 143-day record), not doing system-wide back- ups (what's an operating system among friends if the data are safe?), and installing an UPS. All of these are reasonable if uninterrupted uptime is the primary concern. SITE MANAGEMENT SIG Message # 301 0 replies Reply to # 300 Left by: RALPH BRANDT On 27-Jun-89 At : 717-751-5120 Subject: HITMAN Phone # is wrong - 5210 HITMAN is only one of many process killers. We chose it because we didn't want to support one ourselves and the price was reasonable. It went on 5.0 and 4.5 and 4.7 and is behaving nicely - unless you listen to the users who get tagged! Just make sure you exclude all processes you don't want hit. Suggest you give Manager's secretaries and Managers (if they have accounts) more time. They create too much hassle if they get hit too often. It is available from SIAGA Systems Suite 215 801 Sixth St Cagery Alberta T2P3V8 403-263-1151 Prices vary by CPU - so I'll let you get it from them. TERMINALS/PLOTTERS/PRINTERS SIG Message # 310 0 replies Left by: ABE SULEIMAN On 27-Jun-89 At : United Cerebral Palsy (212) 947-5770 Subject: DOWNLOADING SOFT FONTS TO VT320 I have a DCL Command procedure that downloads some fonts to my VT220 and sets the prompt to a cute graphics figure like snoopy, Mickey etc ... When I tried to use it on my VT320, the terminal "froze" and the prompt was set to a weird string; I tried to modify it to work on the VT320 but, alas, unsuccessfully Does anyone out there have any experience in downloading fonts to VT320. I used the VT320 manual but it's very cryptic and I couldn't get anywhere. In a following message I will include the DCL procedure. CPU'S SIG Message # 54 1 replies Left by: ABE SULEIMAN On 27-Jun-89 At : UCPA (212) 947-5770 Subject: UPGRADING A VAX 8200 Does anyone out there know if the VAX 8200 computer is upgradable to 8350? My thought is that it would be cheaper than getting a MicroVax 3xxx and additional software licenses. I read that the 8350 provides 2.3 VUP vs. 1.0 VUP on the 8200. How true is this statement? CPU'S SIG Message # 55 0 replies Left by: DR. R On 28-Jun-89 At : THE BULLDOG COFFEE SHOP Subject: THE VAUNTED VECTORIZED VAX Yes. On 11 July, those ever-so-clever Maynard Marketing Mavens will program announce VECTOR SUPPORT for the VAX 6000 family of fine processors! No, this is not to be confused with the so-called Pegasus nonsense that a certain tabloid has been babbling about, this is the REAL THING. But wait, there's more! Well, actually, there's less. You see, DEC won't have VAX FORTRAN V6.0 and its automatic vectorizing and parallelizing slicing-and-dicing capabilities on the street until well after the next iteration of Ground Hog Day. What's more, early versions of the compiler are likely to contain more bugs than a roach motel--my fee-splitting pals in neurosurgery tell me that it takes a year or so after FCS to provide a really robust and solid vectorizing compiler. I must be off--my flight to Amsterdam (and world-renowned Bulldog Coffee Shop) await. I'll be back on July 5 with a full report unless I get too confused whilst perusing the menu at the Bulldog! Facts not Rumours Dr. R MISCELLANEOUS SIG Message # 528 0 replies Reply to # 512 Left by: Barton F. Bruce On 28-Jun-89 At : 617.868.1111 or on DECUSERVE "BRUCE" or BRUCE@CCAVAX.CAMB.COM or on PAGESWAPPER "US159353" I just dug out the RSX SIG's Session Notes - a VERY thin thing indead. I am pretty sure what you want isn't there. There are a couple of things about CPRSX given by Spit Brook Road troops (Brian IS on that project, but these seem to be by other people), but not what you want. If all else fails, try mailing him a nice letter. Spit Brook Road is in Nashua NH. If you call me I will give you his phone #, but won't post it here. Actually you can get any DECies # by simply calling MAYNARD and asking how to reach the person. The RSX groups phone #s seem to be answered by a machine most of the day, so mail is simpler. Brian used to be active on Decuserve, but gave up because of too much DEC bashing. What you are looking for is the sort of thing that is reasonable to ask someone to upload. Try to convince him to post it somewhere. There is a much better chance he would post it on DECUSERVE, or PAGESWAPPER than here. TERMINALS/PLOTTERS/PRINTERS SIG Message # 311 0 replies Reply to # 191 Left by: DAVE JOHNSON On 28-Jun-89 At : (415) 620-3023 Subject: VIDEO OUTPUT FROM VT340 We bought an adaptor for a VT340 from COVID (602-966-2221). There are two parts: a COVID 700 analog interface ($505) and a COVID 70-340-00 cable ($85). It's easy to install. and it works great RSTS SIG Message # 399 1 replies Left by: BRUCE A. GREMBOWSKI On 10-Jul-89 At : 408/291-3364 Subject: IS THERE ANYBODY OUT THERE? Is ARIS still working? No new messages between 29 June and now? Well, welcome back from vacation, everyone! PC'S SIG Message # 223 0 replies Left by: GEORGE HERSON On 10-Jul-89 At : 212 241-9363 Subject: DIGITAL MODEM My institution uses a digital PBX telephone system that makes it impossible to use a regular modem without some sort of line change and adapter. Is there a modem that makes this (and the associated month wait) unnecessary? RSTS SIG Message # 400 0 replies Reply to # 399 Left by: JOHN BRIGGS On 10-Jul-89 At : (301)231-3061 Subject: IT'S BACK Glad to see someone else couldn't get on. APPLICATIONS SIG Message # 339 0 replies Reply to # 336 Left by: RICH NARDIELLO On 10-Jul-89 At : (609)764-0100 Subject: USERWARE Brian, Thanks for the reply. Good luck in using the Package. I'll think you will find the report writer utilities very good. Do me as favor. Please give me the name of the developer that you acquired Userbase from? We are looking for a local developer. Thanks, Rich VMS SIG Message # 994 2 replies Left by: HOWARD HENNING On 10-Jul-89 At : (702)364-8633 Subject: SPAWN VS MONOLITH We are a C shop, and have implemented an application using 30-50 images which spawn each other as subprocesses. For example, the main menu level is simply a series of image path names which are spawned upon the request of the user. This type of structure is advantageous for maintenance, and allowing the user access to any image from anywhere in the structure. Normally, there are no more than three processes running per user. Also such a structure is standard for UNIX or ULTRIX. However, several of our clients who have programming staff have questioned this method over monoliths where all the code is linked into one giant program. In our case the monolith would be some 6 Mb plus. Has anyone benchmarked this? or done any kind of study on the resources required? Other than the 6-9 seconds that the user waits, when starting a process, and having to reopen some files does this method waste system resources? Or is this a unfounded concern? Thanx in advance. WORD PROCESSING SIG Message # 127 0 replies Left by: SONJA STEIS On 11-Jul-89 At : 313 651 5300 X324 Subject: DOCUMENT SHARING QUESTION I would appreciate any suggestions of software that will accomodate my users needs of Document Sharing on a per document per individual basis. I am already aware of the Advanced Ideas product and was wondering if there were any others that I might be able to consider. Best Regards SYSTEM PERFORMANCE SIG Message # 231 0 replies Reply to # 203 Left by: KEVIN DULANEY On 11-Jul-89 At : 813-587-6746 Subject: DE-FRAGMENTTATION WE USE RABBIT-7 DISK OPTIMIZER WITH MUCH SUCCESS. YES, IT HAS THE FLEXIBILITY OF MANY OPTIONS, HOWEVER, ONCE YOU HAVE ESTABLISHED YOUR SITE SPECIFIC PARAMETERS ... GO AT IT! WE EXECUTE RABBIT-7 EVERY NIGHT ON A PARTICULAR DRIVE, ALL SET UP AUTOMATICALLY TO EXECUTE THROUGH A COMMAND FILE IN THE BATCH QUEUE. KEEP IN MIND THAT IF YOU EVER HAVE DRIVE PROBLEMS ... MAINTENANCE COMPANIES LIKE TO POINT THE FINGER AT DISK OPTIMIZERS. WE HAVE NEVER HAD ANY FILES CORRUPTED OR DESTROYED BY USING RABBIT-7. APPLICATIONS SIG Message # 340 1 replies Left by: Rod Falanga On 11-Jul-89 At : 505-892-7393 Subject: help with DTC, PLEASE!! We recently acquired the program DTC from one of the DECUS tapes (VAXSP84 I think). DTC stands for Desk Top Calendar. It was written a while ago and has had,apparently some steady updates to it since. One of the features that DTC supposedly offers is the ability to write a appointment to another schedulke file in another account (or at least to another schedule file.) What I would like to know is how we can do this?? I have the documentation and the source code, but after having spent most of yesterday morning looking into it, I just cannot figure out how it is done. I know that it involves declaring the year to be 99, it should the year be 1999, 9999 or just 99?? Also, what I don't understand is the format of the rest of the record which will contain the pointer to the schedule file to be written to. Does it look like this: filename=(filespec) or filespec=(filespec) or (filespec)= which is it?? If anyone out there knowsthe format for that record, I would appreciate it if you would give the format AND please also give an example!! Thanks Rod Falanga RSX SIG Message # 278 0 replies Left by: BRUCE PRENDERGAST On 11-Jul-89 At : 1-213-940-7177 Subject: RSX11M 4.1 UPGRADE TO RSX11M 4.4 I HAVE SYSGENED 4.4 AND THE SYSTEM BOOTS (SOFT) HOWEVER WHEN WRITING THE BOOT BLOCK TO DISK, THE HOME BLOCK IS DESTROYED. THE CONFIGURATION IS PDP 11/23+ WITH 512K AND EMULEX SC02 CONTROLLER USING FUJI DISK M2312 CONFIGURED AS 1-RK06 AND 2 RK07. THE LATEST ROMS HAVE BEEN INSTALLED IN THE CONTROLLER. EMULEX HAS STATED THAT THEY DO NOT SUPPORT BEYOND RSX11M 4.2 WITH THE SC02. EMULEX PROVIDED PATCHES TO DMDRV.MAC AND SYSCM.MAC FOR 4.1 THESE SAME PATCHES WERE USED FOR 4.4 AND THE RESULTANT SOURCE CODE APPEARED TO BE IDENTICAL EXCEPT FOR COMMENT LINES. THE GOAL IS TO UPGRADE TO 4.4 SO AN 11/73 MAY BE INSTALLED. ANY HELP IS APPRECIATED IN SOLVING THIS PROBLEM. ANOTHER APPROACH WOULD BE GETTING AN 11/73 TO WORK UNDER RSX11M 4.1 THANK YOU VMS SIG Message # 995 2 replies Left by: ED BASS On 11-Jul-89 At : (713) 220-4534 Subject: TELLING IF DCL .COM IS RUNNING I need to implement something along the lines of the IBM System/36 "If procedures is active (running)" test. In other words, I need to know if a certain DCL procedure is currently running in either ineractive or batch mode. More than one users at a time will be accessing these procedures as well, which complicates matters. I need a way to just test if a DCL procedure [line noise] tly running anywhere on my cluster. Testing if the file is open would work but I don't really want to do it that way.... WORD PROCESSING SIG Message # 128 0 replies Reply to # 12 Left by: SONJA STEIS On 11-Jul-89 At : ITT AUTOMOTIVE 313 651-5300 X324 Subject: DOC SHARE PROGRAM I am interested in your document sharing program. Best Regards, Sonja Steis EDITORS SIG Message # 221 3 replies Left by: BOB ZISEK On 11-Jul-89 At : (203)798-5240 Subject: SPECIAL CHAR ATTRIBUTES I would like to know how to include imbedded escape codes to control character attributes within EDT. We are running VMS 4.7. This is mainly in regard to using bolding, underlining, flashing etc. in mail messages. I know about the escape character definition for 7 or 8 bit codes, however, is there a way to do this on the fly directly within EDT? Thanks in advance. Bob Zisek VAX system manager Boehringer Ingelheim LTD 90 East Ridge PO 368 Ridgefield, Ct. 06877 VMS SIG Message # 996 0 replies Reply to # 995 Left by: PHIL GRAVEL On 11-Jul-89 At : HERE Subject: IS A COMMAND PROCEDURE RUNNING? One quick and dirty way to do this is to do a SHOW DEVICE/FILE on the device on which the command procedure resides. If the command procedure is in use, it will be listed (SHOW DEVICE/FILE lists all files which are open on a device) along with the process name(s) of the processes that are accessing it. Phil VMS SIG Message # 997 0 replies Reply to # 991 Left by: RICHARD B. GILBERT On 11-Jul-89 At : (609) 452-5130 Subject: SET TERMINAL/INTERACTIVE John, PASSALL and INTERACTIVE are holdovers from VMS V3. PASSALL has been replaced by PASTHRU. The difference is that PASSALL passed everything through to the application program. PASTHRU allows the driver to handle XON/XOFF flow control and passes everything else through to the application. MISCELLANEOUS SIG Message # 532 0 replies Reply to # 527 Left by: PHIL GRAVEL On 11-Jul-89 At : HERE Subject: VAX UPTIME Well, it had to come to an end as ungraceful as it was. My system went down after 157 days on-line. The janitorial service was cleaning the room where the MicroVAX resides and were moving this around and apparently moved something against the power plug enough to cause a momentary disruption in power. Somewhat disconcerting was a MicroVAX that I had upgraded to VMS 5.1 that appeared to deadlock after 60 days on-line. Terminals could not get a response from the system and attempting to set host to left a message about insufficient resources at the remote node to make the connection. After rebooting the system, I could find no information in the error log or accounting files or the system dumpfile that would indicate what caused the problem. The operator logfile was inadvertently purged during the rebooting process (there were several reboots). I am left to wonder if the deadlock which occurred was due to the know problem (feature??) in the VMS V5 Modified Page Writer and pagefile fragmentation. * Phil PS. Yes I know an UPS would have kept my system on-line, but..... VMS SIG Message # 998 0 replies Reply to # 994 Left by: RICHARD B. GILBERT On 11-Jul-89 At : (609) 452-5130 Subject: SPAWN VS MONOLITH Howard, Process creation and image activation are both fairly expensive things to do on VMS. Since VMS is a virtual memory system you don't really care how big the image is. What counts is how much the user needs to have in memory at one time. I'd say that 6mb is not an unreasonable size for an image. If the image is INSTALLed /SHARE, then multiple users can use the same copy of the code in memory and require their own private copies of the data only. Unless you have some other reasons for having a system of 30-50 images, I'd say go monolithic! MISCELLANEOUS SIG Message # 533 0 replies Reply to # 527 Left by: RICHARD B. GILBERT On 11-Jul-89 At : (609) 452-5130 Subject: VAX UPTIME Phil, I think that the numbers would go through the roof if you excluded power failures and scheduled down time. DEC hardware seems to be extremely reliable. If it makes it through the first 90 days, you can't kill it with an axe! EDITORS SIG Message # 222 0 replies Reply to # 221 Left by: RICHARD B. GILBERT On 11-Jul-89 At : (609) 452-5130 Subject: SPECIAL CHARACTER ATTRIBUTES Bob, I'm afraid you're going to be disappointed on this one. MAIL transmits special characters faithfully but it will not send them to your terminal. There is a security problem. Many terminals can be programmed with escape sequences to load strings into function keys or store an auto-answerback string. People have been known to abuse this feature by programming an answerback string like DELETE /NOCONFIRM *.*;*. You can get the full text of a message, including escape sequences by EXTRACTING it. EXTRACT TT: will display it on the screen for you. The escape sequence for getting special character attributes is of the form [Pn;Pn;Pn;...m. The Pn designates a numeric parameter which is 1 for bold, 4 for underline, 5 for blink and 7 for reverse video. 0 calls for all attributes off. You would set bold, blinking, reverse video by sending the string '[1;5;7m'. must be replace by a real escape character, decimal 27, hex 1B. CPU'S SIG Message # 57 0 replies Reply to # 54 Left by: ROBERT G. SCHAFFRATH On 12-Jul-89 At : 914-335-9792 Subject: UPGRADING A VAX 8200 Yes you can upgrade a VAX 8200 to a VAX 8350. However, DON'T DO IT!!!!!!!! The price for the upgrade is outrageous and the performance of the machine (even under V5.0) is unimpressive when compared to a MicroVAX 3XXX series. Another division here at GF did in fact spend $35000 to upgrade their VAX 8200 to an 8350 and was VERY disappointed (in fact they expect their new VAX 6310 shortly to give them the performance they were expecting). However, if you are short on cash, you might want to do the upgrade yourself by purchasing a used KA-820 cpu board. They are selling for around $10K. It simply plugs into the BI backplane and you are off. One note though, if you are running VMS prior to V5.0, ASMP was licensed separately and you will not be able to start the CPU. Under V5.0, it will start automatically. Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 MICROVAX SIG Message # 208 1 replies Left by: FRED DUTCHER On 12-Jul-89 At : 213-477-1421 Subject: 3800 VS 8350 PERFORMANCE We are considering a VAX purchase and have looked at purchasing a used 8350 or a new Microvax 3800. There is some concern over I-O bottlenecks due to the Q-Bus bandwidth. We would like to use RA90's or RA82's on the 8350 or RF71 drives on a 3800. I was wondering what sort of performance levels people were getting with these two machines. The limit of 64 meg of memory has some of us worried too. We are a COBOL software development company. UTILITIES SIG Message # 377 1 replies Left by: FRED DUTCHER On 12-Jul-89 At : 213-477-1421 Subject: CODE MANAGEMENT SYSTEMS We are currently looking at code and module management systems. I'd like the names of ones other than the DEC/CMS/MMS that might be available on the market. A number of the companies would be appreciated too... RSX SIG Message # 279 0 replies Reply to # 273 Left by: MICHAEL CHRISTIAN On 12-Jul-89 At : 317-267-6946 Subject: LOADABLE RSX DIRECTIVES I attended that session and have the session notes. It was presented by Brian McCarthy of DEC in Nashua, NH. I beleive that he is one of the RSX managers at DEC. APPLICATIONS SIG Message # 342 2 replies Reply to # 340 Left by: MICHAEL CHRISTIAN On 12-Jul-89 At : 317-267-6946 Subject: LATEST AND GREATEST DTC I don't know the answer to your specific problem. We are using a version of DTC from the Fall '88 tape. It has the features that you are describing, but they are supported by some new commands. There is some reference to making an entry in the DTC file of the form "999999080[dir]filename.typ=" to point to another file. No one here has ever tried that option. My recommendation is to get a copy of the Fall '88 tape if possible. APPLICATIONS SIG Message # 343 0 replies Reply to # 342 Left by: ROD FALANGA On 12-Jul-89 At : 505-822-7393 Subject: DTC HELP Thanks for your help, I'll give it a try!! Rod PROGRAMMING LANGUAGES/AI SIG Message # 406 0 replies Reply to # 404 Left by: DAVID R ROBISON On 12-Jul-89 At : 702-364-8633 Subject: UNIX SYSV COMPATABLE LIBRARY I have just finished writing a C callable library which emulates IPC message queues and shared memory under VMS. The library is linked as a shared image. All the routines run in user mode, therefore, if a message que is a system wide message queue then all the processes using it must have WORLD privileges and if it is a group message queue, they will need GROUP privileges. This allows message writers to wake process waiting to read. My company is willing to release both the shared image and source code to the public domain in order to encourage expantion and development of the library. If you would send me a blank TK50 tape and a stamped, self-addressed return envelope or box, I would be happy to send you a copy. (p.s. I would guess that the library is about 90-95% compatible). David. PROGRAMMING LANGUAGES/AI SIG Message # 407 0 replies Reply to # 404 Left by: DAVID ROBISON On 12-Jul-89 At : 703-364-8633 Subject: UNIX SYSV COMPATIBLE LIBRARY I just realized that I forgot to give you my address. The address is: David R Robison Health Care Computer Works Suite 205 Las Vegas, NV 89102-8316 702-364-8633 David. MICROVAX SIG Message # 209 0 replies Reply to # 208 Left by: RICHARD B. GILBERT On 12-Jul-89 At : (609) 452-5130 Subject: 3800 VS 8350 PERFORMANCE Fred, Unless you are getting a real deal on an 8350, I'd say don't bother. The 82xx and 83xx series machines are obsolete. AT best, the 8350 would give you 2.5 VAX Units of Processing (VUPS) which means 2.5 times the power of an 11/780. The 8350 is a dual processor model which is fully supported only under VMS V5. VMS V4 supports it only in a master CPU/slave CPU mode in which all I/O, indeed all kernel mode code, must be executed on the master CPU. If what you are looking for is I/O bandwidth rather than CPU power, the machine's BI bus has a 13Mbyte/second bandwidth which would allow you to support mulitple disk contollers and a great deal of simultaneous disk I/O. The level of I/O performance available would leave any QBUS machine far behind. Most applications, however, do are not limited by the maximum disk transfer rate but rather by the disk access time. I believe that a uVAX 3800 would give you 3+ VUPs. Consult a salesman for the exact figure. The QBUS wins no prizes for I/O bandwidth but it should be adequate unless you know that you will be making extremem demands on it. UTILITIES SIG Message # 378 0 replies Reply to # 377 Left by: RICHARD B. GILBERT On 12-Jul-89 At : (609) 452-5130 Subject: CODE MANAGEMENT SYSTEMS Fred, PVCS from Polytron is available in a VMS version as well as an MS-DOS version. I have not used either version but I understand that the MS-DOS version is well thought of. The package includes PolyMake. Polytron's phone is 1-800-547-4000. If you buy this product, let us know what you think of it. VMS SIG Message # 999 0 replies Reply to # 994 Left by: MARTY JOHNSON On 13-Jul-89 At : (203)666-1737 Subject: SPAWN VS MONOLITH On a resource tight machine, we found that spawning can take 30 - 45 seconds - sometimes longer. Also with spawning, we had to limit the number of users as spawned processes use up balance slots. We were not able to make a Monolith as several of the programs used are from Digital and third party vendors. Our solution was to consolidate as many programs as possible and chain. "Chaining" back from third party programs was no problem as everything is executed from the command procedure: $ FLOW:=LOOP $LOOP: $ RUN MAIN $ GOTO 'FLOW' $END: $ EXIT The program MAIN will chain to the appropriate program using the LIB$DO_COMMAND. To exit the procedure, MAIN sets the symbol FLOW to END and exits. As far as using a Monolith program, INSTALLing the program shared will reduce image activation and memory overhead. However, you should check user's WSQUOTA and WSEXTENT to prevent excessive paging. In fact, you should check your current programs for excessive paging. Excessive paging programs should be redesigned to reduce paging as the Monolith will be paging as the users move between menu options and you don't really need any more paging than necessary. You might also want to consider multiple pagefiles residing on different disks. Bottom line: Chaining is better than spawning, paging is better than chaining, and no paging is best of all! :-) Marty EDITORS SIG Message # 223 0 replies Reply to # 221 Left by: KITTY GEORGE On 13-Jul-89 At : HERE (ARIS) Subject: IN EDT Bob, Mr. Schlaffrath (or RiffRaff or something like that) is right re: not using escape sequences within MAIL for highlighting messages. If you want to degrade your VAX back to version 2.5 or thereabouts, you can get the fancy highlights in your messages. If you still want to crank out Escape sequences, Mr. Schlaffrath gives the details, except for how to create the ESC character. If you have a DEC terminal, hitting the ESC key twice does the trick. On any terminal with a key and the "[" character, hitting [ twice in succession gives you . Have fun. Kitty :-) EDITORS SIG Message # 224 0 replies Reply to # 221 Left by: BRUCE A. GREMBOWSKI On 13-Jul-89 At : 408/291-3364 Subject: ESC IN EDT Bob, It is fairly simply to embed escape sequences in EDT; you simply type [0;1;4;5;7m, selecting which attributes you want. 0 is for normal, 1 for bold, 4 for underline, 5 for blink, and 7 for reverse video. The key is just to type twice to get the character in your text. Hope this helps, Bruce DATA/FILE TRANSFER SIG Message # 561 1 replies Left by: CHARLES PRITT On 13-Jul-89 At : 214-740-8701 Subject: TAR ON THE VAX I will very shortly have a need to be able to read TAR formatted tapes on my VAX. I believe their is a tar utility available on a DECUS tape. A.) I need to know which DECUS tape has the TAR utility on it and B.) I need to know how to get access to this TAPE. Thanks to any help which can be provided. VMS SIG Message # 1001 2 replies Reply to # 995 Left by: DAVID DESROCHES On 14-Jul-89 At : (508) 875-7511 Subject: TELLING IF DCL .COM IS RUNNING Excuse the previous burp, I hit control-z and it exited. I had a need to do the same thing and wrote a DCL procedure using the technique mentioned by Phil Gravel. I defined SYS$OUTPUT to be a temporary file and then opened the file and scanned it for the information I was looking for. Alternately, you could use the DCL SEARCH command to see if the command procedure was in the temporary file. CONTROLLERS/MEMORY SIG Message # 113 0 replies Reply to # 111 Left by: DOUGLAS TAYLOR On 14-Jul-89 At : 202-767-6893 Subject: Q-BUS Thanks for the info, I don't know why the third party RX02 drive wouldn't work in the Micro PDP-11. Maybe there was a conflict with the RQdx1 controller. Sorry for the late reply, I didn't get back to this BB for some time. VMS SIG Message # 1002 1 replies Left by: DAVE MARLIN On 14-Jul-89 At : Litel Telecommunications 614-785-6270 Subject: CLUSTER WIDE SYSUAF.DAT LOCKUP Two weeks ago I upgraded from VMS 4.7 to A5.1 then last weekend to VMS 5.1-1 Cluster wide (8650, (3)8550, 6330,6320). Since then there have been several instances of interactive & Batch logins hanging between the USERNAME prompt and PASSWORD. SHOW DEVICE/FILES shows many processes cluster wide with the SYSUAF.DAT file open. I repair this situation by using BACKUP/IGNORE=INTERLOCK SYSUAF.DAT []*.*; to get an unlocked copy to enable logins and batch work again. 4 systems are on one system disk, with separate disks for the other 2. 5 of the systems access the common SYSUAF.DAT, RIGHTSLIST.DAT and netproxy.dat . This Custer supports about 600 concurrent interactives with some having multiple sessions. I have spent several frustrating days in SDA (analyse/system) trying to trace the problem. Today with DEC CSCs help I traced the process with the Granted locks but found no resource wait states or cross lock conditions. Comments/fixes/birds-of-a-feather ??? I answer all calls. I'm tied to this chair untill this one is fixed. VMS SIG Message # 1003 0 replies Left by: PAUL LUCERO On 14-Jul-89 At : (805) 388-5700 Subject: PRINTING FROM DETACHED PROCESS his question must have been asked before. 8888 VMS SIG Message # 1004 1 replies Left by: PAUL LUCERO On 14-Jul-89 At : (805) 388-5700 Subject: PRINTING IN A DETACHED PROCESS Is there a way to print within a detached process. This must have been asked before. I have a program that I want to run as a detached process with the function LIB$SPAWN issuing the DCL PRINT command, but unfortunately, CLI is not present in a detached process. Is there anyway I can beat this limitation? The program has to run as detached. Help!!! APPLICATIONS SIG Message # 345 0 replies Reply to # 335 Left by: JIM MCGERIGLE On 14-Jul-89 At : 408/439-1153 Subject: BARCODE S/W DEAR SIR; I AM CURRENTLY EVALUATING A PRODUCT CALLED ON-TAP VMS FROM INTEGRATED SOFTWARE DESIGN 508/339-4928 IN MASSACHUSETTS. THIS PROD WILL DO MANY TYPES OF BARCODES AND WILL INFACT PRINT ON AN LN03. ASK FOR MR. MARK R. LANIA. GOOD LUCK; JIM MCGERIGLE PROGRAMMER ANALYST BORLAND INTERNATIONAL, INC. SCOTTS VALLEY, CA. 95066 APPLICATIONS SIG Message # 346 0 replies Left by: JIM MCGERIGLE On 14-Jul-89 At : 408/439-1153 Subject: CONFERENCE ROOM SCHEDULING S/W DEAR FELLOW VAX USERS; DOES ANYONE KNOW OF ANY FREE BULLETIN BOARD S/W THAT DOES CONFERENCE ROOM SCHEDULING. THIS IS A KIND OF OFFICE AUTOMATION TYPE OF APPLICATION. MY MARKETING PEOPLE ARE IN PANIC MODE TO GET SOMETHING HERE TO PREPARE THEMSELVES FOR THE NEXT COMDEX. ANY ASSISTANCE IS APPRECIATED; JIM MCGERIGLE PROGRAMMER/ANALYST 1800 GREEN HILLS ROAD SCOTTS VALLEY, CA. 95066 C/O BORLAND INTERNATIONAL, INC. MASS STORAGE SIG Message # 266 0 replies Left by: DR. R On 15-Jul-89 At : AT THE FREE CLINIC Subject: GNU DEC DISKS Yes. Look for DEC to unveil a double-density version of the RZ55 in the near future. The Maynard Mass Storage Mavens already have cut a deal with Micropolis, so the new SCSI disk should be debuting in very short order. Also, there's a reasonably good chance that DEC will take an equity position in Micropolis. As for the rumoured 8mm Exabyte cartridge tape backup system, it would appear that a certain weekly tabloid is once again auffering from premature speculation. A DEC version of the Exabyte drive is coming, but it sure didn't arrive on 11 July. Sorry, Charlie--you blew it again! facts not rumours, Dr. R CPU'S SIG Message # 58 0 replies Reply to # 52 Left by: DR. R On 15-Jul-89 At : BACK FROM THE BULLDOG Subject: RISCY VAX Yes. Once again Digital has seen fit to verify my prognostications by formally announcing the products that I have preannounced in this electronic forum. Based on Digital's 11 July announcement, I'm confident that the Maynard Microprocessor Mavens are crafting a next- generation computer architecture that will be based on a 64-bit implementation of the MIPS Rxxxx RISC platform. You don't have to possess the clairvoyant skills of the Wizard of OZIX or Merlin the Magician to conclude that DEC's Vax Architecture eXtended (or VAX II) platform will incorporate vector processing, dynamic reconfiguration of multiple processors, instruction-level failover, crossbar memory and new communications registers that will support hardware-based parallelism. No wonder Digital is claiming that the VAX architecture will undergo some radical changes during the next decade. And remember, you read it here first, because Dr. R Has It Now! facts not rumours, Dr. R CLUSTERING/NETWORKING SIG Message # 418 0 replies Reply to # 408 Left by: LEE K. GLEASON On 16-Jul-89 At : 713/528-1859 Subject: COST EFFECTIVE PERFORMANCE INCREASE About $10 worth of transistors, resistors, and capacitors would be enough to build a level shifter for that pin 20 signal... CLUSTERING/NETWORKING SIG Message # 419 0 replies Left by: DR. R On 16-Jul-89 At : ON CALL Subject: SUPERCLUSTERS Yes. DEC this fall will unveil Superclusters, second-generation VAXclusters that will support twice as many processor nodes, eight times the I/O bandwidth and much greater node-to-node distances than existing CI-clusters. Also slated to debut are the much-vaunted HSC100 controller and an active star coupler. Superclusters are likely to be announced in conjunction with the VAX 9000 series "Aquarius" processors. Don't get mugged by tabloid-resident rumourmongers--when it comes to factual DEC futures, Dr. R has 'em now! facts not rumours, Dr. R CLUSTERING/NETWORKING SIG Message # 420 1 replies Left by: DR. R On 16-Jul-89 At : AT THE MILL Subject: AT YOUR SERVICE Yes. Tomorrow DEC will unveil the DECserver 300, a 16-line version of the venerable DECserver 200 terminal server. facts not rumours, Dr. R MISCELLANEOUS SIG Message # 534 0 replies Left by: A FORMER COWORKER On 16-Jul-89 At : BOSTON Subject: FALSE PROPHETS The original Charlie Matco of Rumor Roundup infamy resigned his position a year and a half ago and is now employed as a DEC analyst at IDC. VMS SIG Message # 1005 1 replies Reply to # 1004 Left by: RICHARD B. GILBERT On 16-Jul-89 At : (609) 452-5130 Subject: PRINTING IN A DETACHED PROCESS Paul, You can print from a detached process by using the $SNDJBC system service to queue a print job. I've never needed to do it so I can't give you details but you'll find the necessaries in the "System Services Reference Manual". VMS SIG Message # 1006 0 replies Reply to # 1005 Left by: PAUL LUCERO On 17-Jul-89 At : (805) 388-5700 Subject: DETACHED PROCESS AND PRINTING Thank you for the info. It helps beat the task of searching the answer from the manuals and don't know where to start. - Paul PC'S SIG Message # 225 0 replies Left by: STEVEN G. BERMAN On 17-Jul-89 At : (800) 832-4362 Subject: PCSA 2.2 Has anyone gotten PCSA 2.2 installed and running yet? I would be interested to know if anyone has performed an upgrade from 2.1...especially with PCs remote booting to the host. Thanks. Steve Berman Datamedia Customer Support PC'S SIG Message # 226 0 replies Left by: CAM TAMBLYN On 17-Jul-89 At : (403)429-5834 Subject: PRINTING FROM PC TO DS200 I am looking for a way to spool information from a PC connected to a DecServer 200 port out to one of my Vax printers connected to another DS200 on the same ethernet. I realize that there are several products available which will do this via the host Vax, but you should theoretically be able to do it directly. It would require a program running on the PC which would intercept printed output, spool it onto the local PC hard disk, and connect to a printer on the ethernet to print it. This has the considerable advantage of not requiring participation of the Vax host. Dec makes a vague reference to this capability in section 1.8.6 of the DecServer 200 Management Guide: "With an appropriate applications program, a personal computer can also use a printer offered as a service by the server" What I need is this "appropriate applications program"!!!! Can anyone help? CLUSTERING/NETWORKING SIG Message # 421 0 replies Reply to # 420 Left by: Dr. Gonzo On 17-Jul-89 At : Subject: DECserver 300 To my colleague Dr. R., the DECserver 300 was announced on July 13th along with the LAT Protocol Technology Licensing program! Pricing? Well a DSRVF-BA will cost you $4400 at a price-per-port of $275! As for the Licensing program, now that Digital has obtained its beloved patents for LAT it is licensing the protocol for three different types of use: LAT/HOST Technology - Writing drivers for alien operating systems to use LAT. (ie. writing your own LTDRIVER for UNIX, AOS/VS, GCOS etc.) cost: $10,000 LAT/SERVER (SINGLE PORT) - Writing LAT implementations to allow PC's to act as LAT terminals (ie. the VT240 program on VAXmates). cost: $10,000 LAT/SERVER (MULTI PORT) - Building your own competitive terminal server. (ie. a faster more cost effective DECserver) cost: $10,000 + $10,000 up t royalty fee. Subsequent royalties due based on the number of LAT ports sold by the licensee! Also, the DECserver 250 program was announced. This little device will provide 2 parallel and 4 serial ports exclusively for the purpose of driving line printers via LAT. VMS SIG Message # 1007 0 replies Reply to # 1002 Left by: ROBERT G. SCHAFFRATH On 17-Jul-89 At : 914-335-9792 Subject: MORE INFORMATION Perhaps some more information on your configuration. I gather you have: 6 systems 4 using a common system disk 2 using their own system disks 1 of the two with their own system disks (call it #5) is accessing the common\ system disk for the purpose of reading SYSUAF.DAT. correct so far? If so, is system #5 referencing SYSUAF.DAT through the cluster of via DECnet? Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 SITE MANAGEMENT SIG Message # 302 1 replies Left by: CAM TAMBLYN On 17-Jul-89 At : (403)429-5834 Subject: ACCESS DS200 VIA DECNET We centrally support a number of Vax 3600 sites. They are interconnected via DECNET using DecRouter 200's and 9600 baud lines. In order to configure terminal servers at the remote sites, we first set host to that site, then run NCP and CONNECT to the appropriate DS200 node. This routine frequently causes us 2 problems. The main one is that we often get a hung screen and eventually a "Target not responding" message when we try to display a full screen of server info. The second problem is that connecting this way seems to generate a large volume of Decnet traffic, even if you are not doing anything. Does anyone have a better way of doing this? I realize that Dec has a Terminal Server Manager product, but I think that this is overkill. A related capability that we would like is the ability to connect to a service offered by a DS200 on a remote node. We currently do this by getting the Local> prompt via the above method. This causes similiar problems. Thanks, Cam DATA/FILE TRANSFER SIG Message # 563 0 replies Reply to # 561 Left by: V. JAMES KRAMMES On 17-Jul-89 At : (614) 228-2001X341 Subject: DECUS TAR STUFF I have a tape with a collection of decus stuff on it, and the TAR stuff is on the VAX88A0 part of it; however, I don't think that means anything. If you would send me a TK50 or "reel" (real?) tape at the address listed below, I'll send you the DECUS TAR stuff that I have, though I'm not sure that it the complete library (i.e., sources, etc.) it worked to unpack the Free Software Foundation stuff in the VAX88A1 part of my tape. The address is: V. James Krammes Systems Programmer The Midland Mutual Insurance Company 250 E. Broad Street, Route 39 Columbus, OH, 43215 Jim SITE MANAGEMENT SIG Message # 303 0 replies Reply to # 302 Left by: V. JAMES KRAMMES On 17-Jul-89 At : (614) 228-2001 X 341 Subject: SET HOST / CONNECT NODE I experienced similar problems managing multiple cpu's and their associated terminal servers. The problem only appeared when doing a $ set host to a remote node and then doing a connect node from within NCP. The displays (longer than a few lines) would pause every so often. Hitting will allow the display to continue. Eventually, however, the connection would fail. Sorry, I can't supply any "fix" information, unless you are willing to do what we did: Get rid of all but one of your cpu's. Jim APPLICATIONS SIG Message # 347 0 replies Reply to # 342 Left by: Rod Falanga On 17-Jul-89 At : 505-822-7393 Subject: DTC revisited Michael, I tried using the 9999[dir]filname= for the record in the DTC.DAT file, and it didn't seem to work. Any other ideas?? Rod Falanga CLUSTERING/NETWORKING SIG Message # 422 0 replies Reply to # 385 Left by: BILL WEISSBORN On 17-Jul-89 At : (214) 980-7924 Subject: FTSV WORKS GREAT! Someone in our organization got a copy of FTSV and as I understand it, licensed it for all our sites. We have a very wide-area network and the start/restart capabilities are great time-savers. I can highly recommend it. As I understand it, the program is available thru the DEC Assets program, whatever that is. Check with your local sales-person. Don't have any idea as to cost, fees etc. CLUSTERING/NETWORKING SIG Message # 423 0 replies Reply to # 402 Left by: BILL WEISSBORN On 17-Jul-89 At : (214) 980-7924 Subject: DECNET ON SUN I can speak for the use/implimentation of DECNET on Suns. We have two flavors currently running. Sun 3/260's running SunOs 3.5 which uses SunDNA V5.1 and Sun 4/110's and 386i/250's running SunOS 4.0.x which requires SunDNI 6.0(note the name change!). SunDNI 6.0 seems to be a bit more robust than the 5.1 version. 5.1 has a habit of "losing" the permanent database. 6.0 has not developed this problem-yet. 6.0 also loads in much easier than the earlier versions. You can do almost anything on the Sun except send mail(although there is some software out in the Unix world that will allow you to send mail to your VAXen, I have not had the chance to try it for SunOs 4.x but it worked ok on 3.5). I have not found a way to get mail from the VAX to the Sun, however. Bill Weissborn VMS SIG Message # 1008 0 replies Reply to # 1007 Left by: DAVE MARLIN On 17-Jul-89 At : LITEL TELECOMMUNICATION Subject: REPLY TO MORE INFORMATION Bob, Correct to the last clarification question. The "Cluster" systems access the same SYSUAF.DAT but the last one has it's own SYSUAF.DAT. (development system with no interaction except common queues (JBCSYSQUE.DAT).J*e DEC has put an area guy on site to observe. Oh well. Clusters are such a joy. Should we talk about LMF ? Dave Marlin RSTS SIG Message # 403 0 replies Reply to # 399 Left by: Phil Anthony On 17-Jul-89 At : Here Subject: NOBODY HERE BUT US CHICKENS Bruce, yours was the first message since ARIS switched to Xyplex terminal servers - you're right, it wasn't working. In fact, I didn't find out it was working again till today, when I called Kevin and the Bonnie One up to find out if they had any news. Seems that differences in the Xyplex changed the timing involved in logging on completely .... For those (like me) who rely on PCs and script-driven communications soft- ware, I'm now waiting six seconds after making a connection, then issuing two carriage returns with a 0.1-second pause in between them. If I haven't gotten the "." prompt that means there's a VAX listening on the other end, I wait six more seconds and repeat. If I do that 100 times and there's still no response (i.e., after almost 11 minutes), I assume the VAX is out to lunch and quit. The delays are probably over-generous, and for people who are calling long distance, I'd suggest experimenting with the initial delay cut back to two or three seconds and giving up after a minute or two. The key step seems to be sending a pair of s with a wee bit of delay in between after allowing a little droppie of time to let VMS ask the Xyplex what speed it's supposed to set itself at: A single return and no delay will fail every time. P.S. Dave, I have the new Mirror script for ARIS ready for you. It'll be up Thursday night. MISCELLANEOUS SIG Message # 537 0 replies Reply to # 533 Left by: Phil Anthony On 17-Jul-89 At : Here Subject: VAX UPTIME Richard, that wasn't exactly what I said. *All* computers I've ever worked on require downtime for preventive maintenance, and *all* computers will go down if you pull the plug - at least if you pull it long enough (an UPS, be- ing battery-powered, will only keep things running for a finite amount of time). What I was looking for was a figure that would show the difference between VAXes running VMS and other machines/operating systems. Scheduled downtime for software upgrades or tuning should definitely be included; there are other systems where I've genned and installed all system software (excluding the operating system) and performed all required system tuning on line, with no interruption to production. Even downtime for a new O/S varies a great deal from one system to the next: How often does a new release come out, and how necessary is it to have the latest version up? Also, how much downtime is required when you do install new system software? Many operating systems and other system software can be genned while the system is up and running, and require only five or ten minutes to install them after genning, if they require downtime at all. This would be in addi- tion to unscheduled downtime, such as hardware failures or software bugs (the infamous interrupt above ASTDEL, or whatever it is, or the recent uVax/V5.1 lockup that Phil Gravel reported). CPU'S SIG Message # 59 1 replies Left by: DR. R On 17-Jul-89 At : IN CLOSE PROXIMITY TO MR02 Subject: THE AGE OF AQUARIUS Yes. One of the most intriguing features of the VAX 9000 Aquarius processor is its incorporation of proprietary high density signal carrier (HDSC) technology that Digital acquired from the abortive Trilogy project. With HDSC, ECL macrocell arrays are mounted on "cold plates" whose thermal conductivity has helped eliminate the need for liquid cooling. ("Plumbing? We don' need no stinking plumbing!," say my pals in the High Performance Systems Group). HDSC modules are being assembled at this very moment at Digital's supersecret--well, *formerly* supersecret--Cupertino, CA facility. Look for these clandestine components to play an important role in both standard and "tall-cabinet" Aquarius configurations, as well as in the 50VUPS Aquarius II and the 150VUPS gallium arsenide-based Centaurus followon products that will debut in 1991 and 1993 respectively! Facts not rumours, Dr. R CPU'S SIG Message # 60 0 replies Reply to # 59 Left by: DR. R On 17-Jul-89 At : IN QUARANTINE Subject: IT'S ALL COMING BACK NOW... Yes. The good doctor estimates Aquarius performance to be as follows: SCALAR SPECS (Uniprocessor and Quad flavors) VUPS 30 - 108 TPS 55 - 200 VECTOR VELOCITY (Per vector accelerator) LINPACK MFLOPS 18 PEAK MFLOPS 125 Facts not rumours, Dr. R CLUSTERING/NETWORKING SIG Message # 424 0 replies Reply to # 421 Left by: DR. R On 17-Jul-89 At : ON ARIS Subject: FAITH RESTORED! Yes. Digital has once again restored my faith by formally announcing yet another product that has been the subject of one of my fearless prognostications. Now that the DECserver 300 is reality, I guess I can begin wondering how many sheckels Digital will try to extract from Emulex and Datability for each copy of a third-party LAT license! Facts not rumours, Dr. R MISCELLANEOUS SIG Message # 538 2 replies Reply to # 523 Left by: FINKELL On 18-Jul-89 At : (404) 980-1190 Subject: VAX UPTIME CONT. PHIL, IS THERE ANY INFORMATION AVAILABLE AS TO WHEN THE VAXen DO GO DOWN, APPROX., HOW LONG ARE THEY DOWN FOR? SPECIFICALLY FOR THE VAXSTATION 2000, AND SPECIFIC AREA SUCH AS UPGRADING, DISK DRIVE REPAIRS, PM. WE HAVE HAD OUR VAXSTATION SINCE SEPT. '88 AND IT HAS COME DOWN ONLY DUE TO POWER FAILURE, AND JUST RECENTLY TO INSTALL A NEW DISK DRIVE. IS DEC RESPONSIVE ENOUGH THAT IT WOULD BE SAFE TO SAY THAT FOR A SMALL OPERATION SUCH AS OUR'S THAT THE ACTUALLY DOWN TIME WOULD BE LESS THAN 1 DAY?!!! VMS SIG Message # 1009 0 replies Reply to # 1008 Left by: ROBERT G. SCHAFFRATH On 18-Jul-89 At : 914-335-9792 Subject: Well the DEC engineer on site would certainly be much more experienced in cluster problems than myself so it looks like you are in good hands. As for LMF, there are many wonders yet to be seen... VMS SIG Message # 1010 0 replies Left by: MITCHELL KATZ On 18-Jul-89 At : 212-232-9878 Subject: DISK TIMEOUTS Hello one and all, The problem I have is concerned with mount verification timeout. With 36 disks in a homogeneous vaxcluster of 7 vaxs. What I have done is set the sysgen parameter mv_timeout to 64000 (the max) so that I have an hour to repair the disk before it times out (which means rebooting the world). Obviously, an hour may not be long enough if I need parts. So what I am working on is a macro program to reset the timer associated with that device so that I can get enough time to repair the disk without necesitating a reboot. Any comments would be appreciated. mak GRAPHICS SIG Message # 89 0 replies Left by: FRAN SHIELDS On 18-Jul-89 At : 215-293-0410 Subject: OUR GRAPHIC NEEDS WE ARE IN THE PROCESS OF REVIEWING GRAPHIC PACKAGES AVAILABLE ON THE VAX. IN ADDITION, WE ARE LOOKING AT THE POSSIBILITY OF USING A MAC OR AN IBM PC CONNECTED TO THE VAX TO ACCOMPLISH OUR GRAPHIC NEEDS. I HAVE NOT BEEN ABLE TO FIND A 'SIMPLE' PACKAGE THAT RUNS ON THE VAX, BUT HAVE SEEN SEVERAL NICE LITTLE PACKAGES ON THE PC'S.... OUR REQUIRMENTS ARE 'SIMPLE' BUSINESS GRAPHICS... I PREFER TO LOOK AT THE MAC TO VAX APPROACH BUT WOULD LIKE TO HEAR SOME CONSTRUCTIVE CRITICISM? THANKS SITE MANAGEMENT SIG Message # 304 0 replies Left by: FRAN SHIELDS On 18-Jul-89 At : 215-293-0410 Subject: UPS SYSTEMS I AM STARTING A REVIEW OF UNINTERRUPTIBLE POWER SYSTEMS AND WOULD LIKE COMMENTS FROM ANYONE... I HAVE INFO FROM 1 COMPANY (TO START) .... ISOGUARD. I KNOW THAT DEC PROVIDES SYSTEMS AND I'M SURE THERE ARE QUITE A FEW OTHERS. WE HAVE 2 MICRO VAX'S, 4 RA81'S AND ABOUT 15 MODEMS/MUX'S. ANY HELP? THANKS FRAN VMS SIG Message # 1011 0 replies Left by: MITCHELL KATZ On 18-Jul-89 At : 212-232-9878\ Subject: DISK TIMEOUTS I FOUND THE ANSWER TO MY PREVIOUS MESSAGE! It seems that although the minimum value in sysgen for mv_timeout is 1, you can still totally disable mount verification timeout bvy setting mv_timneout to 0 (you must disable checks to do this). Therefore, no program is necessary. mak PC'S SIG Message # 227 0 replies Reply to # 204 Left by: SCOTT TAYLOR On 18-Jul-89 At : (303)930-7264 Subject: MEMORY UTILITIES Your problems with PCSA hogging memory have plagued us since we installed PCSA 2.0. Our original solution (R:Base was the program we had trouble with) was to re-boot the PC using another disk every time the user needed to use R:Base. This other disk contained only enough drivers to start the file server, as R:Base's directory must be writable. This opened up about 445Kb on the PC (we generally use Compaq 286 & 386 machines with DEPCA's), and allowed the user to use R:Base. However, since the users wanted to have an IS lynching party, we sought some alternatives. There are a number of good utilities available that work well on 286 and 386 machines. These include 386max, Abovedisc, etc. Contact any good software store for more information on these products. Generally, they are fairly inexpensive, and easy to install. 386max, I know, moves any drivers, etc. from conventional memory (the initial 640K) to higher memory (either expanded or extended). This leaves nearly all the base memory free for your programs, without affecting PCSA performance. MASS STORAGE SIG Message # 267 0 replies Reply to # 264 Left by: TOM MYNAR On 18-Jul-89 At : 714-956-9159 Subject: I'VE GOT A SYSTEM U COULD USE I've got a UV I I'm trying to sell, it includes VMS and a Cipher 1600 BPI tape drive (includeing controller). With a DZ, you could DECNET between them and transfer the files. Call me (I hope this is an acceptable message concerning the non-advertising clause of using the system). Tom PC'S SIG Message # 228 0 replies Left by: SCOTT TAYLOR On 18-Jul-89 At : (303) 930-7264 Subject: SECURITY WITH PCSA V2.1 We have an interesting problem with PCSA file security that has plagued us since we installed PCSA V2.0. Much of this is due to the unusual nature of our network, and the structure we need to have in place for directories. We have a number of different groups (determined by UIC's), that don't use the default [360,*] that PCSA likes. We are able to set the UIC ourselves after editing PCSA_ADMIN.COM a little bit. This structure effectively keeps users from different groups out of each other's files. The problem occurs in users personal directories, and our Word Processing directories. PCSA sets default file protection to (S:rwd,O:rwd,G:r,w:) on all newly created files within a user's directory. Unless we make a different group UIC for each user on the network (we have over 150 currently), other users can connect to their account using the command NET USE d: \\NODE\USER1%USER2 USER2PASSWORD. This way, user2 can read user1's file (although he can't delete them) using his own password to connect to user1's service. I originally tried setting access using the NET ATTRIB/DEFAULT command in the user's AUTOUSER.BAT file, however this sets the file protection only for that session, and involved so much maintenance that it wasn't worth it. It would be ideal if the /DEFAULT on the ATTRIB command worked for more than the current session, however, maybe I'm too close to the problem to see an easy solution. Additionally, I tried using ACL's, however, maintenance here is not worth the trouble. The second problem is similar, however, a bit more complex. There are a number of subdirectories underneath our WP directory that need to be accessed by the Word Processing department. The structure is as follows: BUGS$DUA0:[PCCOMMON.WPC] BUGS$DUA0:[PCCOMMON.WPC.INVEST] BUGS$DUA0:[PCCOMMON.WPC.LEGAL] ETC. WPC (the group) needs full access to the INVEST and LEGAL (as well as 10-15 others) directories, therefore, needs to be the owner of these directories. However, LEGAL should not have ANY access to WPC or INVEST. If the user is good at DOS, or realizes that WPC is above them in directory structure, they can have complete access to that directory, or any other underneath the WPC directory. One last problem - Through VMS we change access to user's AUTOUSER.BAT to (S:RWD,O:,G:W:), to prevent users from changing the file or accidentally deleting it. However, one "power user" discovered that if he copied AUTOUSER.BAT to another name he could edit it. Then he was able to rename AUTOUSER.BAT to another name, and replace it with the new (edited AUTOUSER.BAT). What a way around VMS security! Please let me know if you have any solutions. VMS SIG Message # 1012 0 replies Left by: ED BASS On 18-Jul-89 At : 713-220-4534 Subject: SHARED FILE COPY/SORT I need to be able to COPY and SORT files without locking the source (input) file for exclusive access, thus locking out users of these files who may want to update them with online maintenance programs. I could use CONVERT/SHARE for the COPY situation but it is SLOW...I could write a program to take care of the COPY situation as well, but I'm just curious if there's an easier way. As for the SORT problem, I'm lost. Any way to sort a file without locking the input file? I'd be interested in any third-party software or tips on SORT features that I may be unaware of. SITE MANAGEMENT SIG Message # 305 0 replies Reply to # 302 Left by: Phil Gravel On 18-Jul-89 At : HERE Subject: TERMINAL SERVER Connecting to a terminal server via the NCP CONNECT NODE command causes considerable network traffic even if nothing is being done. Who knows why... Apparently, the dropping of a connection to a terminal server in NCP from a node that you have SET HOST to is a known feature. DEC's CSC was helping me set up printers connected to a terminal server. When the connection to the terminal server from NCP was lost on a node that was accessed via SET HOST, the response from the support person was "Oh yeah, you can't do that. That happens..." Cie la vie. Phil RSTS SIG Message # 404 1 replies Reply to # 403 Left by: Phil Gravel On 18-Jul-89 At : HERE Subject: ARIS CONNECT TIME Since the changes that were made to ARIS earlier this month, it routinely takes me about 2 minutes to a response from the ARIS VAX once the modems have completed their handshaking. Awhile back, users were complaining about the 15-30 second response time before the recent changes. I wonder how happy they'll be with the longer time?! I hope people don't get frustrated and quit logging into ARIS because of this long time delay. SYSOP: Isn't their anything that can be done to improve the response time?? Two minutes for every long distance call begins to add up to big $$$. Phil RSTS SIG Message # 405 1 replies Reply to # 404 Left by: Phil Gravel On 19-Jul-89 At : HERE Subject: ARIS CONNECT TIME Phil Anthony revealed the secret to fast connect times -- a couple of quick carriage returns causes the connection. I guess the single carriage return that I was issuing every 5-10 seconds wasn't doing the trick. Phil RSTS SIG Message # 406 0 replies Reply to # 405 Left by: BRIAN DESANTIS On 19-Jul-89 At : ARIS Subject: CONNECTION TO ARIS Forgive me, but I dont seem to find that long a wait when trying to connect to Aris. During the day, frommy VAX I will use VAXNET on a 1200 baud line and after getting the connection established msg, I will hit return three or four times, will get the "." that was mentioned before, it will delay about 4-5 seconds, then display the first screen, etc. Am I doing something 'wrong'??? It is 2:58a.m. and I am home and using 2400 so the word delay is pretty much unknown at this time... Sidenote- regarding ARIS' upcoming text retrieval section... Will those articles be available for actual download,not just a 'text dump' intoa log file?? Since I assume these will be text files and with the cost of long distance charges what they are, it would seem logical to be able to d/l to host system and then read/print/etc.. Keep fingers crossed.... RSTS SIG Message # 407 1 replies Reply to # 406 Left by: ROBERT G. SCHAFFRATH On 19-Jul-89 At : 914-335-9792 Subject: ARIS CONNECT TIME How did this wind up under the RSTS SIG? I haven't noticed an unusual response time for logging into ARIS. I usually sit here and pound on my carriage return key about 20 times before I get that little "." prompt. But that is just force-of-habit from doing a 'SET TERMINAL/INQUIRE' on DECserver ports for sooooooo long! If you don't hit carriage return a couple of times, it takes about 10 seconds for the SET TERMINAL/INQUIRE to function. RSTS SIG Message # 408 0 replies Reply to # 405 Left by: Phil Anthony On 19-Jul-89 At : Here Subject: ARIS CONNECT TIME Phil, I only discovered the "secret" after a chat with Kevin at ARIS. The credit really belongs to him; I just implemented my script after he gave me the technical details. My thanks to him, and to all the staff at Profes- sional Press, for their speedy response to questions and problems I've had with ARIS over the years. By the way, Kevin tells me that this machine is now dedicated to ARIS. We can expect even better response times except during ARIS maintenance and backups. RSTS SIG Message # 409 0 replies Reply to # 407 Left by: Phil Anthony On 19-Jul-89 At : Here Subject: ARIS CONNECT TIME Well, Robert, now you know who the *really* devoted ARIS users are! MISCELLANEOUS SIG Message # 539 1 replies Reply to # 538 Left by: Phil Anthony On 19-Jul-89 At : Here Subject: MORE VAX UPTIME That's a big question! I've never known a PM on any single DEC system to keep it down any longer than a couple of hours unless a hardware problem turned up in the course of diagnostics. System software installation doesn't take very long either if you do your homework up front and determine what else may have to be changed after you install the new version (custom mods to SYSGEN parameters, new features you may want to implement, changes to file structures, for instance). The worst I ever experienced under VMS was 14 hours, and that involved upgrading major versions of VMS (V3.5 to V4.1) plus several other software packages. Usually upgrades only take 1 to 4 hours on a single system, depending on the complexity of the software. When it comes to hardware glitches, the situation is more nebulous. Once a problem determination is made, it generally doesn't take very long to re- place a faulty part; typically a few minutes to a few hours, with an hour or two of diagnostics afterward to make sure the new guy is working right. (You may also have a few hours of work afterward, such as restoring your data after a disk replacement. Make sure you have good backups!) DEC does a pretty good job of keeping parts handy, and the turnaround time before a field engineer shows up depends on what level of field service you've paid for. But the debugging process for certain failures (e.g., intermittent problems, especially ones involving two or more components) can take a long time - weeks, even months. That's true on any computer, with any operating system. Anybody else have comments? PC'S SIG Message # 229 0 replies Reply to # 228 Left by: Phil Anthony On 19-Jul-89 At : Here Subject: PCSA SECURITY Scott, I've never used PCSA, but your problems are fairly common on PC net- works in legal and accounting offices. It seems to me you ought to be able to do some of what you want, such as keeping users from changing AUTOEXEC.BAT, by implementing a common set of initialization files in a di- rectory with (S:RWED,O:,G:,W:R) protection. Those scripts would determine who the user is, and on that basis, set his/her default to the appropriate directory where she or he would have full (RWED) access. Within VAX structure, I can't think of any way to handle a word processing group that can work in two or more directories while another group (LEGAL, INVEST, etc.) can only access one of them - only one group can own a VMS di- rectory. In a Novell network, though, the problem is trivial. Multiple groups can have trustee rights to a single directory, and you can control access on a directory-by-directory basis; different groups can also have different rights in the same directory. Under Novell, the right to read or search in a parent directory doesn't nec- essarily mean you can do anything in a subdirectory. File modification (such as renaming or resetting the READ-ONLY flag) is a separate right that need not be granted to those who don't have to have it. By the way, the Novell overhead is so small that workstations have no problems with large PC programs such as R:Base. APPLICATIONS SIG Message # 348 0 replies Reply to # 347 Left by: MICHAEL CHRISTIAN On 19-Jul-89 At : 317-267-6946 Subject: DTC WORKS FOR ME. I did some experimenting, and had successful results. Remember that I am probably using a later revision of DTC off the Fall '88 tape. 1.) In the DTC.DAT file, I added the line: 99999999080USR$A:[CHRISTIAN.ARIS]SECDTC.DAT= 2.) Inside the DTC program, there are two commands: DTC>S 072089 8:00 Time to be at work. This command files the scheduled event into the alternate file, but not the primary. DTC>G 072089 5:00 Go home now. This command files the scheduled event into both the primary and alternate file. I only tried one alternate file, but it claims that you can use more than one. If this doesn't work for you, feel free to call me. If you need the latest version of DTC, you can try to get the Fall '88 DECUS tapes, or we could probably make arrangements for me to send you the new DTC. Michael Christian Thomson Consumer Electronics 317-267-6946 PROGRAMMING LANGUAGES/AI SIG Message # 408 4 replies Left by: LEE GOSS On 19-Jul-89 At : 913-573-7941 Subject: VAXC How can one use the getchar feature of c on the vax. I am trying to get one character of screen input at a time with a limit of 16 characters. The way it is working now is, the input is buffered until a CR is hit. PROGRAMMING LANGUAGES/AI SIG Message # 409 0 replies Reply to # 408 Left by: V. JAMES KRAMMES On 19-Jul-89 At : (614) 228-2001 X 341 Subject: EMULATE C'S GETCH() Lee, I, too wondered how to do this though I wanted to do it from MACRO. The answer I came up with is to call the SYS$QIO(W) system service. You can tell RMS to that your buffer is only a single character, but it will read until a carriage return or other terminator is hit. If you tell $QIO to only read 1 character, that's all it will read. First, open a channel (you can use a file name of "TT:") with the SYS$ASSIGN system service as shown: $iodef $qiodef ttchan: .word 0 tty: .ascid /TT:/ buf: .blkb 133 buflen: .blkl 1 . . $assign_s devnam=tty,chan=ttchan . . $qiow_s chan=ttchan,func=#io$_readlblk!io$m_noecho,- p1=buf,p2=#1 If you want the character to echo as it is typed, remove the "!io$m_noecho" part from . Of course, you need to check error codes, too. To use such a routine from a high-level language, return the put the character into register 0 and declare it (in the HLL) as integer or long. If you would like a copy of the code, let me know. Hope this helps! --Jim MISCELLANEOUS SIG Message # 540 0 replies Reply to # 539 Left by: BRIAN DESANTIS On 19-Jul-89 At : (201)670-0286 Subject: DOWNTIME I too, have never experienced a long downtime for PM's, if the system came down at all. And in the case of problematic hardware, my local DEC field office is only 15-20 mintues away and parts can get here very quickly. I have also never had a problem in getting a part - I have never been in a situation (knock wood, or computer floor) where I have had to wait for a part. If they didnt have one, they went and found one, even if it was at their cost, and they had to go and fly it in.... On the whole, the service that my VAX has provided and my field service as well has been top-notch. I have heard many horror stories from people who have had problems and my sympathies are with them... LASER PRINTERS SIG Message # 74 0 replies Reply to # 33 Left by: BRIAN DESANTIS On 19-Jul-89 At : (201)670-0286 Subject: CUSTOM APPLICATIONS, INC, Regarding the software from Customs Applications that prints Postscript on an HP LJII- has anyone used this? Is it for IBM, Mac, Vax??? CLUSTERING/NETWORKING SIG Message # 425 0 replies Reply to # 375 Left by: JESS GOODMAN On 19-Jul-89 At : 814 238-2811 Subject: RECORD LOCKING OVERHEAD We had the exact problem you described when we first brought up our cluster. I eventually found a work around. If you are still having performance problems please give me a calll and I may be able to help you. CLUSTERING/NETWORKING SIG Message # 426 0 replies Reply to # 394 Left by: JESS GOODMAN On 19-Jul-89 At : 814 237-0309 Subject: SI CLUSTER We have installed an SI cluster. If you are still looking for feedback on this, give me a call. MISCELLANEOUS SIG Message # 541 1 replies Reply to # 523 Left by: JOHN FINKELL On 19-Jul-89 At : 404-980-1190 Subject: CAN'T READ REPLIES PHIL, MESSAGE # 538 (VAX UPTIME CONT.) KSAYS THERE ARE 3 REPLIES BUT WHEN I TRY TO READ THEM IT SAYS IT CAN'T. I NOTICED THE SYSTEM WAS DOWN, IS IT POSSIBLE THE RESPONSES WERE LOST?! E PROGRAMMING LANGUAGES/AI SIG Message # 410 0 replies Reply to # 408 Left by: Phil Anthony On 19-Jul-89 At : Here Subject: VAX C and getchar () Lee, that's the way getchar () has always worked. It's intended as a means of retrieving one character at a time *from buffered I/O* (which could be a file - getchar () is nothing but a macro for getc (stdin)), not a way to get hold of a character as soon as it's typed at the keyboard. The previous re- sponse about writing a getch () function in Macro-32 may be the best way for you to go. You might also look at the setbuf () function. Normally this is used to as- sociate a user-declared buffer with a FILE stream (note that it has to be called before any I/O is done on the stream). However, if you declare the buffer address as NULL, the file becomes unbuffered. I don't know if VAX C will hand you the character immediately when you declare stdin unbuffered - the boys and girls in Maynard (or was it Nashua, N.H.?) had to play some pretty messy games to emulate ASCII stream file structure under VMS/RMS, and the terminal driver does its own buffering as well unless a QIO tells it not to - but it wouldn't take more than 15 minutes to write a quick-and-dirty test program to try it out. Give it a shot both with stdin and with another stream for the terminal - i.e., ttptr = fopen ("TT:", "r"); /* Open the terminal on a new stream */ setbuf (ttptr, NULL); /* Make the new stream unbuffered */ Good luck! MISCELLANEOUS SIG Message # 542 0 replies Reply to # 541 Left by: Phil Anthony On 19-Jul-89 At : Here Subject: BUGBUGBUGBUGBUG John, the ARIS downtime had to do with a new Xyplex terminal server; this problem seems to involve a bug in the ARIS code. When I left my message, your message header already said there were two replies - but somehow they got lost. Another message I left correctly says it's a reply to a previous message, but the previous message's header says there have been no replies. A third message shows up twice when you follow that particu- lar thread. Bonnie! Oh, Bonnie .... TERMINALS/PLOTTERS/PRINTERS SIG Message # 312 0 replies Reply to # 5 Left by: MITCHELL KATZ On 19-Jul-89 At : 212-232-9878 Subject: LCP01 WE USE IT- the lcp01 that is. We have an HP plotter as back up. That is an HP pen plotter. Give a call. mak PROGRAMMING LANGUAGES/AI SIG Message # 411 0 replies Reply to # 408 Left by: DR. R On 19-Jul-89 At : SOMEWHERE IN TAXACHUSETTS Subject: C++ Yes. Soon we'll have to deal with another C-related issue, 'cuz Roger Heinen's DECwest OS/SB crew is hard at work hacking a Digital C++ compiler. When the compiler will be announced I can't reveal, but it's almost certain to hit the streets before OZIX, the OS/SB's new Unix implementation for OLTP applications. facts not rumours, Dr. R FORUM - 3RD PARTY SUPPLIERS SIG Message # 105 0 replies Reply to # 95 Left by: TOM MYNAR On 19-Jul-89 At : 714-956-9159 Subject: ANY SHOULD WORK I got it from a DEC employee (so it's a reasonably good rumor) that any SCSI tape drive will work. Now, I've got a 60 megger coming in the mail that I would love to try. Maybe we could work something out ? Tom EDITORS SIG Message # 225 0 replies Reply to # 204 Left by: JOHN RICE On 19-Jul-89 At : 312-940-9000 Subject: UTILITIES SIG Message # 379 0 replies Left by: MICHAEL M. FRAZIER On 19-Jul-89 At : (714) 788-7490 Subject: WHAT I am looking for a utility known as WHAT.EXE. This is a System Management Utility that I ran into at the last place I worked. Its banner stated it was an unsuported VMS utility. I am curious if anyone out there knows anymore about it I have the EXE for 4.X, but when we upgraded to 5.1 it needed to be relinked. If you have any info about this utility, please let me know. I feel lost without WHAT! later... Michael M. Frazier UTILITIES SIG Message # 380 1 replies Left by: HARRY MOSKOWITZ On 20-Jul-89 At : 818-8884329 Subject: VMS BACKUP WITH TU81-PLUS I am currently using VMS Version 5.0-2 on a VAX 6210. When I use BACKUP to copy files from RA82 drives to my TU81-PLUS tape drive, I get increased performance by specifying /NOCRC. I would like to know if there are any drawbacks to using /NOCRC, since the TU81 manual says you can gain perfor- mance with the NOCRC option if operating in GCR mode. (I am using 6250 density tapes and therefore should be operating in this mode.) In my telepone conversations with Digital, they still recommend CRC checking (the default). It is unclear to me why this is necessary if the hardware is doing the parity checking, etc. Any ideas or opinions would be appreciated! UTILITIES SIG Message # 381 0 replies Reply to # 380 Left by: Phil Anthony On 20-Jul-89 At : Here Subject: VMS BACKUP AND CRC Some time back there was extensive discussion on this SIG involving the need for software CRC checking in VMS BACKUP. Read the thread starting with Mes- sage No. 144 - it covers your questions pretty thoroughly. VMS SIG Message # 1013 1 replies Reply to # 1012 Left by: JOHN P. FLYNN On 20-Jul-89 At : (615)269-2266 Subject: SHARED FILE COPY I use BACKUP/IGNORE=INTERLOCK for a lockout-free copy operation. As for sorting, I am no help. John. VMS SIG Message # 1014 0 replies Reply to # 1013 Left by: Phil Anthony On 20-Jul-89 At : Here Subject: SORT /IGNORE=INTERLOCK Short of writing your own sort program, the kludgy workaround might be to copy the file you want to sort (maybe using BACKUP /IGNORE=INTERLOCK) and then sort the copy of the file. This wouldn't be too useful for large files because of disk space requirements, but it might work until some- body submits an SPR on the subject and DEC gets around to doing something about it "in a future release." One of the third-party developers (EG&H, maybe?) might get around to it more quickly if you were willing to pay for a custom version of their sorting software. PC'S SIG Message # 230 0 replies Left by: DAN DONOHUE On 20-Jul-89 At : (215)439-3276 Subject: PC-PLOT IV TERM. EMUL. Is there anyone out there who has experience running a Tektronics Terminal Emulator called PC-Plot IV, from Microplot systems? We have been having trouble running it over PCSA. Works like a charm via dial-up. Now they have added a network interface option, but the details are sketchy and their support is sketchier... If anyone has used this new feature of PC-Plot, over any network, please contact me or leave details where you may be contacted for a few basic questions. Thanks so much. DATA/FILE TRANSFER SIG Message # 564 0 replies Left by: DAN DONOHUE On 20-Jul-89 At : (215)439-3276 Subject: PC-PLOT TERM. EMUL. If anyone is familiar with tektroniks terminal emulator PC-Plot IV, please see message 230, 20-Jul-89, in PCs SIG. Thanks! UTILITIES SIG Message # 382 1 replies Reply to # 379 Left by: ROBERT G. SCHAFFRATH On 20-Jul-89 At : 914-335-9792 Subject: WHAT.EXE? UTILITIES SIG Message # 383 0 replies Reply to # 382 Left by: ROBERT G. SCHAFFRATH On 20-Jul-89 At : 914-335-9792 Subject: CONTINUATION OF LAST MESSAGE (something funny is going on with my terminal) I have never heard of a utility from DEC (unsupported or supported) called WHAT.EXE. Are you sure it is a DEC creation? More than likely it is a piece of DECUS software that slipped into your system. What does it (or should I say DID) it do? Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 CLUSTERING/NETWORKING SIG Message # 427 0 replies Reply to # 395 Left by: VERNON L. PALANGO On 20-Jul-89 At : (201) 679-0200 Subject: NOVELL PC/LAN - DECNET Novell has a product called Netware for VMS, it provides transparent file sharing between a VAX host and a Netware LAN. The set up for it is very easy and it runs without hitches. However, if you are looking for the same file server performance as your LAN file server gives you on the VAX you will be disipointed. The product is intended to give transparent communication between your VAX and your LAN, it does that well. Netware for VMS also tosses in a very good LAT type package, that allows any PC on your network to access your VAX as a terminal. Good Luck!! Vernon L. Palango Systems Manager JWP - Information Systems VMS SIG Message # 1015 0 replies Reply to # 983 Left by: VERNON L. PALANGO On 20-Jul-89 At : (201) 679-0460 Subject: BROADCASTING TO DECSERVERS Have you tried assigning all the ports on your decservers predefined logical terminal ID's using the LATCP program??? I don't know if it will work, but it will make VMS recognize the terminals when they are not logged in. V. Palango Systems Manager JWP - Information Systems UTILITIES SIG Message # 384 0 replies Reply to # 377 Left by: PATRICK MAHAN On 20-Jul-89 At : (714)720-7868 Subject: CODE MANAGEMENT We use a code management system called CCC developed by a company called Softtool Corporation in Santa Barabra. It works great! EDITORS SIG Message # 226 0 replies Reply to # 221 Left by: PATRICK MAHAN On 20-Jul-89 At : (714) 720-7868 Subject: EMBEDDING ESCAPE SEQUENCES IN EDT Bob, To embed escape sequences under EDT use the 7-bit mode (ie CSI = [) and insert the escape character () by striking the CTRL-[ key twice. VMS SIG Message # 1016 0 replies Reply to # 1012 Left by: PATRICK MAHAN On 20-Jul-89 At : (714) 720-7868 Subject: SHARED FILE COPY/SORT Ed, For shared file copying use the BACKUP utility with the qualifier /IGNORE=interlock. I don't know if this will allow a user to open the file while you are accessing it (it will work if the user already has opened the file.) For shared Sort, could you try using the callable SORT utility and perform a record sort? For information on the SORT utility see the VAX/VMS utilities manual. LASER PRINTERS SIG Message # 75 0 replies Reply to # 72 Left by: VERNON L. PALANGO On 20-Jul-89 At : (201) 679-0200 Subject: PRINT REPLYS I don't know exactly what you are looking for, but if your are trying to get a message on your terminal when a print job is completed, try this: $ PRINT/NOTIFY You will get a message at your terminal when the job is completed. MICROVAX SIG Message # 210 0 replies Left by: RON SALAZAR On 20-Jul-89 At : 818-967-9521X425 Subject: VMSKITBLD I have a need to distribute VMS 4.7 to a VAXStation 2000. Our configuration is an LAVC with a VSII/GPX as a boot node with an RA81 as our common disk. Digital's documentation gives a very description of VMSKITBLD at best. I have spoken to the CSC about it, they recommended using the BUILD function with SYS$SYSDEVICE as the source disk & [SYS0.SYSCOMMON] for the top directory. The command file executes properly, but when I go to boot the VS2000 standalone, I receive "UNABLE TO LOCATE SYS.EXE" which sends the machine into the power-on test procedure. I have done this once before on another machine, but unfortunately MY documentation was lacking!! (Something about working at 2:00 AM does that to me!!) If anyone has done this and has managed to keep better notes than I did, I'd appreciate the help. Ron Salazar EG&G Power Systems Covina, Ca. APPLICATIONS SIG Message # 349 1 replies Left by: SCOTT TAYLOR On 20-Jul-89 At : (303)930-7264 Subject: SEARCH FOR A CORPORATE DATABASE! Over the past nine months, we have been searching for a corporate database. Of course, we've heard every claim in the world, and are very confused! I think we've narrowed the choices down to two. Either Rdb with Smartstar (and maybe Focus's report writer), or Oracle. I personally don't like Smartstar due to their Report writer, and the fact that we're owned by a predominantly IBM company. This means if we ever switch to IBM (God forbid!) with Rdb we have to re-write all our applications. In addition, Smartstar's references were not as good as I would expect. Oracle is my first choice due to their portability across multiple platforms, and their wonderful reporting capabilities and user-friendliness. Because of our inexperience with databases, we'd love some input from those of you out there who are currently using either of these, or something else you prefer. Our considerations are: 1) primary use will be for user inquiry, 2) database will be installed on a 6320 also running PCSA with 200-250 total users. Concurrent use approx. 100-150 users. 3) Query must be quick. 4) Must not overburden the 6320. 5) Report writing capabilities must be very flexible (for MIS), and user interface easily understood (for the end- users). We'd appreciate any input. Thanks for your help. MISCELLANEOUS SIG Message # 543 0 replies Reply to # 539 Left by: VERNON L. PALANGO On 20-Jul-89 At : (201) 679=0460 Subject: VAX DOWNTIME I recently had a VERY bad experience with system down time at one of my clients in New York. They are slowly going through the process of upgrading all of thier in house applications for use under V5.0 (they have completed this task now by the way), but they had just installed a 6220 and were running a mixed version cluster (VMS 4.7 & VMS A5.1) with an 8350 and 750. They started having problems keeping the 750 (which is their development system) up, it kept crashing with a disturbing variety of bugchecks. After an eternity of calls to Collorado support, field service was put on the case. A week long process involving 5 (count 'em five) engineers working on the machine was needed to figure out that the system had a faulty power supply. Oh, I didn't mention that in the process they: Wiped out the system disk Bent the ethernet cable under the floor bringing the whole network down It took a total of about 3 weeks to resolve this (simple) problem, and an unknown amount of cost in terms of system down time. I think it should be noted that the quality of service engineers (board swappers) that DEC has put out lately has gone down a great deal in recent years. There are still some very good engineers out there, but DEC should spend more time training these people to DIAGNOSE problems before swapping parts. Vernon L. Palango Systems Manager JWP - Information Systems SITE MANAGEMENT SIG Message # 306 0 replies Reply to # 304 Left by: CAM TAMBLYN On 20-Jul-89 At : (403)429-5834 Subject: UPS'S The May 29 issue of Digital Review has an article about UPS's and lists about 100 different models along with features and prices. APPLICATIONS SIG Message # 350 0 replies Reply to # 349 Left by: DR. R On 20-Jul-89 At : EVERYWHERE Subject: ORACLE VS RDB Oracle is a great marketing company. facts not rumours, Dr. R MISCELLANEOUS SIG Message # 544 0 replies Reply to # 534 Left by: A FORMER SSC MEMBER On 20-Jul-89 At : ARIS Subject: ORIGIN OF PSEUDONYM No Virginia, there is not--and never was--a "Charlie Matco." The name that launched a thousand rumors (and one editorial embargo) was in actuality the name of the virtual system manager on the "DCS VAX," an ALL-IN-1 system used by DECUS leaders to conduct Society business. The DECUS Support Services Committee (SSC), a group tasked with oversight responsibilities for the DCS system, decided to create a pseudonymous account to which questions and complaints about DCS could be directed. Just to make things more interesting (and to seed the DECUS membership list just as magazines seed their mailing lists), the SSC decided to assign a name to their new account. But what name to choose? After discovering that the Marlboro Public Library lacked a copy of "What to Name Your Bogus Computer Account," the resourceful SSC members convened at DECUS US headquarters on Boston Post Road for a good old fashioned brainstorming session. During a lull in the cerebral storm, one of the meeting attendees glanced out the window and spotted the sign of a nearby automobile dealership, Charles Chevrolet. A spirited discussion ensued: "Charles?" "Nah, too formal..." "Charlie?" "Charlie... yeah, Charlie! That's the ticket!" And with that, the SSC selected the first name of DECUS' virtual system manager! Choosing a last name, however, proved to be a far more difficult task. The SSC "name-that-sysmunger" brainstorm mounted in intensity until it reached 12 on the Beaufort scale, but no surname emerged from the tempestuous session. Until another committee member looked out the window and began to gaze at the passing traffic on Boston Post Road. Out of the corner of his eye he spotted a large, gleaming white panel truck. Emblazoned on the side of the vehicle (in letters at least three feet tall) was the logo: MATCO TOOLS. And the rest is history! PROGRAMMING LANGUAGES/AI SIG Message # 412 0 replies Left by: MICHAEL JESSE CHONOLES On 21-Jul-89 At : GENERAL ELECTRIC Subject: ADA UTILITY PACKAGES? I'm interested in anyone who has experience with the available Ada utiltiy packages, such as Math Advantage NAG BOOCH GRACE ASR We need to purchase one in the VERY near future and we are interested in any evaluation you can share. Thanks Michael Jesse Chonoles General Electic Bldg 25, RM 13S25 P.O. Box 8048 Phila., PA 19101 215/354-3185 APPLICATIONS SIG Message # 351 0 replies Reply to # 349 Left by: V JAMES KRAMMES On 21-Jul-89 At : (614) 228-2001 X 341 Subject: ORACLE We currently have three Oracle databases running, two large and one small. Some hints: Installation is (to put it mildly) non-trivial. Don't expect to find VMSINSTAL. I've had to re-link all of my Oracle images after upgrading VMS from 5.0 to 5.1. Hopefully I won't have to do it again when we go to 5.1-1. It is a real pain. Execution is fast for simple queries, with increasing time as the queries become more complex (as you would expect). With the various tools available, like SQL*CALC, SQL*PLUS, EasySQL, SQL*ReportWriter, etc, the Oracle system is fairly complete - But extremely expensive. Each tool mentioned above is a cost item. Key mapping is a nightmare. The same key in different Oracle applications or even in different parts of the same application can mean different things. This is complicated by the fact that here we use PC's running WRQ's Reflection 1 and 2, each of which has a different key map. Augh! All my experience is from the systems side of the equation; I'm not an applications programmer. If you'd like, I'll see if my applications people can give you some info from their perspective. Call me. Jim Krammes The Midland Mutual Life Insurance Co. (614) 228-2001 x 341 APPLICATIONS SIG Message # 352 0 replies Reply to # 351 Left by: ROBERT G. SCHAFFRATH On 21-Jul-89 At : 914-335-9792 Subject: DATABASES Here at GF (at least my division), ORACLE is our database. However, as already noted, it is VERY VERY expensive. We upgraded an 8350 to a 6330 and paid $120,000 in tribute for the ORACLE upgrade (almost 1/3 the cost of the hardware!). ORACLE does however port well to other systems. It is supported on (to name a few) AOS/VS, MS-DOS, UNIX, and MVS (so it will run on IBM). In fact we are about to receive a decision from our corporate parent that will eliminate DEC equipment from GF altogether and move everything to a pure IBM environment (3090's and AS/400's). So it looks as though ORACLE was probably the right choice. If you are in danger of the same thing happening then you probably should use ORACLE. The portability should be your main concern. Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 UTILITIES SIG Message # 385 0 replies Reply to # 383 Left by: MIKE MIKOWSKI On 21-Jul-89 At : (313)524-8678 Subject: WHAT.EXE WHAT.EXE is an unsupported utility from Digital that is typically brought in by Digital employees who are writing custom software for customers. That is the way it was brought into our shop. It is a useful monitoring utility for processes, mailboxes, and other system parameters. It will show current process quotas, number of mailbox messages currently in a mailbox, all sort of wonderful things. You may be able to get it from your salesman although it may cost some money. VMS SIG Message # 1017 0 replies Left by: JEFF ALLISON On 21-Jul-89 At : BREA Subject: NOISY RS-232 LINES Several of the terminal lines on our 750 always show up as login processes. Allocating the terminals keeps them from trying to log in, but it also keeps the owner from logging in at his/her terminal. Could this be due to "noisy" or "floating" RS-232 lines? If so, does anyone have any suggestions as to what can be done to quiet them? Or perhaps this is solvable via VMS. I would appreciate any feedback, as I'm tired of LOGINOUT.EXE being the most frequently invoked image on the system. Thanks -- Jeff Allison APPLICATIONS SIG Message # 353 0 replies Reply to # 348 Left by: ROD FALANGA On 21-Jul-89 At : 505-822-7393 Subject: DTC HELP Michael, Thanks again, I'll give your suggestions another try. By the way, even though the tape was named VAXS84P (or something like that), the date on the files are all from 1987 or 1988. So I think that they're more recent. Rod Falanga APPLICATIONS SIG Message # 354 0 replies Reply to # 349 Left by: ROD FALANGA On 21-Jul-89 At : 505-822-7393 Subject: WHICH DB TO USE Scott, We have FOCUS here, and at least in my opinion, I don't care for it. It is a BEAR to work with (from the programmers point of view.) One of the other programmers here calls it "HOCUS POCUS FOCUS", but that, again, is his opinion. It seems to me that you would be better off with Oracle. Rod Falanga PROGRAMMING LANGUAGES/AI SIG Message # 413 0 replies Left by: ROD FALANGA On 21-Jul-89 At : 505-822-7393 Subject: HELP W/ FORTRAN We have got FORTRAN version 5.2-33. Whenever I INCLUDE a system module from a text library and then do a compile with the /LIST option, I get page after page of listings of all of the parameters that are defined in the text libraries that are included. What I would like to know is, is there any way that I can make the /LIST option NOT give the position in the listing of parameters that are NOT used in my code?? Rod Falanga VMS SIG Message # 1018 0 replies Reply to # 1017 Left by: Phil Anthony On 22-Jul-89 At : Here Subject: NOISY RS232 LINES Jeff, it seems to me the first thing is to determine why the lines are so noisy. Are you connected via RS232 null modem cables? If so, you might want to think about replacing them with twisted pair between short-haul mo- dems - maybe you have a distance problem, or too great a voltage difference between your computer and terminal grounds. This would also help if your offices are located near a source of RFE, such as hospital equipment or a radio or TV antenna. Where are your cables? Be careful that they don't run next to power lines, fluorescent lights, copiers, elevators, or other machines with heavy-duty motors - even vacuum cleaners can introduce noise. Are your cables making good connections? A loose connection, or an intermittent short in a DB25 connector, will really do a number on serial lines. Is the problem the ter- minals you're using? Some inexpensive terminals, especially older ones that aren't shielded too well, can become an excellent source of line noise. Are the power lines to the terminals properly grounded? And it sounds as if DEC still hasn't done anything about LOGINOUT's hyper- sensitivity. I sent them an SPR under VMS V3.x concerning the fact that LOGINOUT shouldn't respond to any stray control character coming down the line, only to a carriage return - this after a streaming VT102 managed to extend ACCOUNTNG.DAT enough to fill up the system disk one weekend. They promised me they'd look into it "in a future release" (their favorite phrase) and then forgot about it completely. I wonder what would happen if every ARIS user running VMS sent them an SPR about it. Maybe they'd get around to cleaning up their act at long last? PROGRAMMING LANGUAGES/AI SIG Message # 414 0 replies Reply to # 413 Left by: RICHARD B. GILBERT On 22-Jul-89 At : (609) 452-5130 Subject: PARAMETER LISTS Rod, Yes and no. You can elimate all those PARAMETER definitions from your cross reference with a qualifier on the Fortran command. It's something like /NOSINGLE as I recall. The only way to eliminate the unused stuff from your listing is to not list any of it by putting a NOLIST qualifier on the include statement. What some people do is simply copy the definition of SS$_WHATSIT and hope that DEC won't change it anytime soon. SITE MANAGEMENT SIG Message # 307 0 replies Reply to # 304 Left by: RICHARD B. GILBERT On 22-Jul-89 At : (609) 452-5130 Subject: UPS SYSTEMS Fran, I have purchased UPS systems from Best Power Technology (2KVA) and from RTE-DELTEC(7.5KVA). Both systems have performed flawlessly for the last two years or so. As someone else noted, there are hundreds of manufacturers selling UPS systems. VMS SIG Message # 1019 0 replies Reply to # 1011 Left by: LEE K. GLEASON On 26-Jul-89 At : 713/528-1859 Subject: FYI... 64000 seconds (max value for mvtimeout) is approximately 17.78 hours...usually, my field circus types can get a disk problem sorted out in this much time... CPU'S SIG Message # 61 0 replies Reply to # 54 Left by: Brett Bump On 26-Jul-89 At : Here Subject: NOT UPGRADING A VAX 8200 Abe. I would have to agree with Robert. Don't upgrade to an 8350. Our college decided to purchase "the more powerful 8350" after one of the other colleges had received its 8200. At the time, Vms 4.6 was in beta so we were running 4.5. I was quite disappointed that the machine couldn't beat my desktop PC in number-crunching ability. If V5 will give the 8350 an honest 2.5Vups (as Richard states), at least it's closer to my 4Vups PC, but I wouldn't call that much of an improvement. I'd opt to leave the wide-bussed 8200 as it is. After all, my PC will only handle me! MISCELLANEOUS SIG Message # 546 0 replies Reply to # 539 Left by: Brett Bump On 26-Jul-89 At : Here Subject: DEC UPTIME Phil. As far as a hardware issue goes, I would have to agree with Richard as to some DEC hardware running forever. The cheap (cheap,cheap,cheap) PDP system that we had for students of programming classes ran forever. Since this system never had PM (no $) it ran one year from September to late April before a hardware failure (the air conditioner died) caused us to bring the system down. Student file backups (Backups? Well, once in a while!) were done online and software updates came in June (maybe, after much begging). As to the software, the only bug (maybe you could call it a bug) that was an annoyance to live with, was the RSTS system status telling you the system had been up since ???:??:?. This happens after the system has been up over 1000 hours. In contrast, the administrative VAXes were brought down every night for security reasons (a holdover from a previous computer with no security), so I can't give you any figures over 9 hours for Vms 4.5-4.7. But I always thought that the VAXes would run as long as the PDPs if they were given a chance. Maybe DEC changed that with V5. RSTS SIG Message # 410 0 replies Reply to # 409 Left by: Brett Bump On 26-Jul-89 At : Here Subject: Shhh! Careful, Phil, you'll give away our secret. Peter Dick already gave away the real abilities of the P90 in his article "RSTS 80th Birthday Celebration" in the July quarterly of Decuscope. DATA/FILE TRANSFER SIG Message # 565 1 replies Left by: Phil Anthony On 26-Jul-89 At : Here Subject: XMODEM FOR VMS Help! A new ARIS user (his management finally permitted him to hook a modem to his VAX) needs a copy of XModem for VMS. As I've mentioned in the past, ARIS could also use a working XModem, though the good folk at ARIS are a little more particular - they'd also like source code. Anybody who has a copy they'd be willing to share, please leave a message here, or for the new ARISite you could call Andy Yett directly at (203) 677-1341, ext. 315. Thanks in advance to whoever can help out. UTILITIES SIG Message # 386 0 replies Reply to # 379 Left by: BRIAN DESANTIS On 26-Jul-89 At : (201)670-0286 Subject: WHAT.EXE Just in case nobody asks, what is what.exe? and what does it do?? SYSTEM PERFORMANCE SIG Message # 232 0 replies Left by: MATT HALL On 26-Jul-89 At : GIANT INDUSTRIES (602) 274-3584 Subject: DISK THROUGHPUT - BI BUS We are upgrading from an 8600 to a 6320 with RA90's over this weekend and would be curious as to the saturation point of a RA90 (and or RA81) on a BI bus (as shown on MONITOR DISK). Any comments would be appreciated. MISCELLANEOUS SIG Message # 547 0 replies Left by: JOHN R. O'HARA On 26-Jul-89 At : 312-385-3468 Subject: ONLINE JOBS DATABASE I would like to announce to all users of this system the opening of a new online system dedicated strictly to the listing of job opportunities. Here's the info: ************************************* The Online RECRUITER ************************************* 2400 baud ===> (312) 385-4447 1200 baud ===> (312) 385-4475 8 bit, no parity, one stop VT100 emulation required The system is available FREE of charge and is confidential. The database is filling now, so please be a bit patient with us while we get things going, but the application is running clean. We are not a recruiting agency or placement firm and do not make money off the users who log in. It's simply a place where jobs are listed. All we ask is that you use it and spread the word. Thanks, -JRO DATA/FILE TRANSFER SIG Message # 566 0 replies Reply to # 565 Left by: RICHARD B. GILBERT On 26-Jul-89 At : (609) 452-5130 Subject: XMODEM Phil, I'm taking care of Andy Yett. He is going to send me a tape and return postage if he can't get it from my system with Kermit. Where should I upload a copy? I have a .TLB file containing the Fortran source, documentation, help files, etc, etc. Its 132 blocks. The same program, though perhaps a more recent version, is available from the DECUS library as V00095 (Double check the order number in the catalog). It was also on the Fall 1985 DECUS VAXSIG Symposium tape. It is also available from the VAX Forum on CompuServe. UTILITIES SIG Message # 387 0 replies Reply to # 381 Left by: Phil Gravel On 26-Jul-89 At : HERE Subject: BACKUP & CRC Rumor has it that Digital is working to improve the speed of the CRC function which will, in turn, improve the performance of utilities such as BACKUP. Phil DATA/FILE TRANSFER SIG Message # 567 0 replies Reply to # 566 Left by: Phil Anthony On 27-Jul-89 At : Here Subject: XMODEM FOR ARIS Thanks a lot, Richard, for taking care of Mr. Yett - I've often been im- pressed by how much help there is on ARIS for people who ask for it. I also left a message for Bonnie Auclair, suggesting that she get in touch with you directly to arrange an upload/download/tape swap so that we can have a work- ing XModem on ARIS; she should be calling you within a couple of days. DATA/FILE TRANSFER SIG Message # 568 0 replies Reply to # 555 Left by: Phil Anthony On 27-Jul-89 At : Here Subject: ETHERNIM Sorry, Ralph, but I don't think anybody here knows anything about ETHERNIM; the people I asked about it had never even heard of it. We'd be interested in learning about it, though - is it a DEC package? What does it do? After using it, what's your assessment of it? DATA/FILE TRANSFER SIG Message # 571 0 replies Reply to # 568 Left by: ROBERT G. SCHAFFRATH On 28-Jul-89 At : 914-335-9792 Subject: ETHERnim NMCC/VAX ETHERnim is DEC's Ethernet Network Integrity Monitor (which I believe was developed by HP). According to the Silver book, it is: ...a network management tool that aids in fault isolation and configuration management of extended local area networks. ETHERnim utilizes Phase IV DECnet software on a system configured as either a full-function DECnet node or a DECnet end node. It runs as a layered product on the VMS operating system. Full functionality of ETHERnim is achieved by having DECnet software and remote test command files in target nodes. ETHERnim is designed to test the communications path through Ethernet protocol and DECnet. It also has the ability to gather system information for its database from remote VMS and RSX nodes by invoking test probes (remote command files). ETHERnim is a multivendor product that recognizes non-DECnet local area network addresses, and gives the network manager the capability of maintaining a large amount of information about each node via a database that can be edited. It can be operatied from local terminals using ReGIS graphics for graphic representations of the netowrk under test.... PROGRAMMING LANGUAGES/AI SIG Message # 415 0 replies Reply to # 406 Left by: STEPHANIE LYON On 28-Jul-89 At : (609) 737-0425 Subject: UNIX COMPAT LIB Dave, I gave up hope!! I will be trying to reach you by phone for more info. Stephanie UTILITIES SIG Message # 388 0 replies Left by: ROBERTSON G. DAVENPORT On 28-Jul-89 At : HERE (ARIS) OR 216-646-3602 Subject: NEWMAIL COUNT IN VMS 5.1 I used to have a simple .com file that would check VMSMAIL.DAT for a person's newmail count for VMS 4.x, but now that we've upgraded to VMS 5.1-1 it doesn't work. Can anyone tell me what's changed in the format of the VMS MAIL .DAT file (I forget what it's called in 5.x)? (i.e. how do I get my mail count in 5.1) Much Thanks in Advance, Rob Davenport Allen-Bradley Company 747 Alpha Drive Highland Heights, Ohio 44143 (216) 646-3602 UTILITIES SIG Message # 389 0 replies Left by: AL GOCHTOVTT On 28-Jul-89 At : SMS Subject: PF KEY DEFINITIONS What is the correct syntax for layering function key definitions? Currently, I have the following code in my login file: DEFINE/KEY "" /SET_STATE=GOLD/NOLOG UTILITIES SIG Message # 390 0 replies Left by: AL GOCHTOVTT On 28-Jul-89 At : SMS Subject: FUNCTION KEYS Sorry about message 389, I had keyboard problems. My question concerns the layering of function keys using the DEFINE/KEY command. I have the following commands in my login file: DEFINE/KEY PF1 "" /SET_STATE=GOLD/NOLOG . . DEFINE/KEY PF1 "" /SET_STATE=SILVER/IF_STATE=GOLD/NOLOG DEFINE/KEY PF2 "MAIL" /IF_STATE=GOLD/TERMINATE/NOLOG . . DEFINE/KEY PF1 "LO/FULL" /IF_STATE=SILVER/TERMINATE/NOECHO/NOLOG The problem is that PF1 works in the silver state but PF2 doesn't get defined in the GOLD state. Am I missing something? Sincerely, Alex Gochtovtt Message # 391 1 replies Reply to # 388 Left by: ROBERT G. SCHAFFRATH On 31-Jul-89 At : 914-335-9792 Subject: VMSMAIL_PROFILE.DATA The new file is called VMSMAIL_PROFILE.DATA. The format of the file is rather involved to display here. However, the new mail count is located as follows: bytes data ----- ---- 1-31 username 32-33 0 (new mail count identification) 34-35 new mail count If you want the complete layout of the file, drop me a line and I can FAX it to you. Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 UTILITIES SIG Message # 392 0 replies Reply to # 391 Left by: ROBERT G. SCHAFFRATH On 31-Jul-89 At : 914-335-9792 Subject: CORRECTION After I saved the message, I realized that I had made a mistake. The format should be: +--------------------------------+ | 0 | 1 | 0 | 2 | new mail count | +--------------------------------+ 31 32 33 34 35 37 the 1 in byte 32 identifies the information as new mail count. the 2 in byte 34 indicates the information is 2 bytes long (one word). The username occupies the space from 0 to 30. VMS SIG Message # 1022 0 replies Reply to # 918 Left by: STEVE LOOS On 31-Jul-89 At : (414) 447-4451 Subject: VAXSTATION 3100 CONSOLE There is a method of using an alternate console for the Vaxstation 3100. On the back of the 3100 you will notice a printer port that you can use to plug a VT terminal into. Right beside this port is a SMALL Dip switch that toggles between the normal display terminal (VR290) and an alternate console. Flip the dip switch to use the VT console. Make sure to set the baud rate at 9600 on the VT console. See the Vaxstation installation guide for more info. PROGRAMMING LANGUAGES/AI SIG Message # 416 0 replies Left by: FRANSICO VILLAO On 31-Jul-89 At : (718)330-5151 Subject: EXPERT SYSTEM INTERVIEWS Dear Whomever- I'm in the market for text on Expert system interviewing techniques. (IE picking someone's brain in a structured maner. Anyone ever hear of such a anamal ? Fransico Villao System Analysist NCYTA 370 JAY ST. BROOKLYN NY 11201 MISCELLANEOUS SIG Message # 555 0 replies Reply to # 539 Left by: BRUCE A. GREMBOWSKI On 31-Jul-89 At : 408/291-3364 Subject: DEC SERVICE Phil, Apparently, the service DEC provides on the East coast is superior to that we get here on the West coast. The site I'm responsible for is in San Jose, in the heart of silicon valley, yet very often DEC field service does not have a replacement RA-80 disk drive, UDA-50 controller, or power supply for our PDP-11/44s. It is very frustrating to know the problem is with the controller and have to wait for the DEC field service rep to call the office to have his boss walk him through the diagnostics. I try telling him that the ERRLOG seems to indicate the controller, but he can't read it, and wants to try everything else first, since they don't have a controller in stock. Then, finally admitting that it must be the controller, they have to have one shipped in from SoCal, so I have to stay with these guys for over 24 hours, when I knew immediately what the problem was. The first DEC rep we had knew what he was doing; I presume they made him head of the Santa Clara office, as none of the other reps seems to know anything about PDPs (and I'm not so sure they know too much about VAXes, either). COMMUNICATIONS SIG Message # 284 0 replies Left by: IRA SOLOMON On 01-Aug-89 At : (617) 736-4576 Subject: HST MODEMS Can anyone tell me the correct settings for an HST di COMMUNICATIONS SIG Message # 285 1 replies Left by: IRA SOLOMON On 01-Aug-89 At : (617) 736-4576 Subject: HST MODEMS Can anyone tell me the proper settings for HST dual standard modems on a VAX Host (VMS 5)? I connect OK, but get very poor results in EVE. Lots of invalid escape sequence errors. Thanks UTILITIES SIG Message # 393 0 replies Reply to # 390 Left by: DAVID DESROCHES On 01-Aug-89 At : (508) 875-7511 Subject: DEFINE KEYS I can not figure out where your problem is as I typed in the key definitions that you indicated and it worked on my machine. Are there other PF keys defined in your command procedure? Could they be changing the definition of the "Gold" PF2 key? If there are no other keys defined, I would simply define PF2 without indicating a "Gold" state. APPLICATIONS SIG Message # 355 0 replies Reply to # 348 Left by: ROD FALANGA On 01-Aug-89 At : 505-822-7393 Subject: DTC revisited again Michael, I tried what you suggested with the DTC.DAT file (to make it point to another DTC.DAT file) and I too have now got better success. Thank you very much!! One thing though, what I want to be able to do is write to another User's DTC.DAT file (file protection is not an issue here, as that is what my User's are requesting, i.e.: the ability to write to other Users' DTC.DAT files.) When I do do that, however, DTC writes to my own DTC.DAT file twice?! Why is that?! For example, if I give it the command: G 090189 08:00 Power breakfast at McDonalds ;-) It will write the string "Power breakfast at McDonalds ;-)" for 8AM and 8:30 AM on Sept. 1st. Why is that?? Rod PROGRAMMING LANGUAGES/AI SIG Message # 417 0 replies Reply to # 414 Left by: ROD FALANGA On 01-Aug-89 At : 505-822-7393 Subject: UNUSED FORTRAN PARAMETER Richard, Thank you for the advice of /NOSINGLE. I'll give it a shot. And, of course, as you've pointed out, if I REALLY am getting tired of this, then I can just not get a listing of the file. Rod Falanga CLUSTERING/NETWORKING SIG Message # 428 0 replies Left by: ALAN ANESHANSLEY On 01-Aug-89 At : 405-323-2332 Subject: DECNET REMOTE PRINTING I HAVE A MICROVAX II RUNNING VMS V5.0 WITH A VAXMATE RUNNING OVER DECNET. THE MICROVAX HAS SEVERAL QUEUED TERMINAL PRINTERS. I AM HAVING DIFFICULTY GETTING THE REMOTE PRINTING FUNCTION OF PCSA TO WORK OVER THE NETWORK. SO FAR I HAVE RESEARCHED THE SOFTWARE DOCUMENTATION AND CONFIGURED THE MICROVAX AND VAXMATE AS FOLLOWS, WITHOUT SUCCESS: IN SYS$MANAGER, WITH THE PCSA_ADMIN UTILITY I HAVE ADDED AND REGISTERED A PRINTER SERVICE NAMED HPLJ (THIS IS ALSO THE QUEUE NAME FOR THE TERMINAL PORT QUEUE ON THE MICROVAX FOR PORT TXA1). ON THE VAXMATE BOOTED OVER THE NETWORK, TO REDIRECT THE LOGICAL TO THE REMOTE QUEUE I HAVE ENTERED....... F:\>NET USE LPT1 \\CDS\HPLJ IN THE FORMAT "NET USE logical \\nodename\queue" THE VAXMATE RESPONDS "COMMAND COMPLETED SUCCESSFULLY" WHEN I ATTEMPT TO PRINT A FILE OVER THE NETWORK EITHER USING SOFTWARE CONTROL OR BY ISSUING THE "NET PRINT filename logical" COMMAND FROM THE DOS PROMPT OF THE VAXMATE, I GET THE FOLLOWING ERROR... NET817:NETWORK DEVICE TYPE INCORRECT AND THE FILE DOES NOT PRINT ON THE REMOTE QUEUE ON THE MICROVAX. ON THE MICROVAX THERE IS CREATED IN THE DEFAULT SPOOL DIRECTORY A FILE NAMED SYS$SYSDEVICE:[PCFS_SPOOL.HPLJ]ECHOPCFSnnnnnn.prt. IF I ISSUE THE DCL COMMAND PRINT/QUEUE=HPLJ FOR THESE FILES, THEY WILL QUEUE TO THE PRINTER NORMALLY. I HAVE COMPLETED ALL THE NETWORK TROUBLESHOOTING CHECKLISTS IN THE DEC DOCUMENTATION, BUT TO NO AVAIL.... I HAVE THE FOLLOWING VERSION OF PCSA/MSDOS SERVICES FOR VMS: LAD$KERNEL V0.32 LADDRIVER V0.68 PCFS_SERVER T2.0-5E PCSA_MANAGER T2.0-5E ALL OTHER FUNCTIONS OF PCSA SEEN TO BE OPERATING NORMALLY, I.E. NETWORK FILE SERVICES, NETWORK FILE TRANSFERS (NFT), VIRTUAL DISKS, ETC..... IF ANYONE HAS ANY SUGGESTIONS, NEEDLESS TO SAY I'D BE EXTREMELY GRATEFUL!!! COMMUNICATIONS SIG Message # 286 0 replies Reply to # 285 Left by: Phil Anthony On 01-Aug-89 At : Here Subject: MODEM CONFIGURATION? Ira, the information you've given doesn't suggest a modem configuration problem to me: You don't say anything about text (straight ASCII printable characters) being messed up - just escape sequences. Might your terminals not be telling EVE correctly what kind of terminal they are, so that EVE is expecting and sending one set of escape sequences while the terminal is us- ing another set? Do you have a $ SET TERM/DEV= in LOGIN.COM, or in your system startup file, that doesn't match the terminals you're actual- ly using? Or do you have PCs coming in on the line, using software that may do too good a job of emulating whatever terminal you want it to be? COMMUNICATIONS SIG Message # 287 0 replies Reply to # 286 Left by: Phil Anthony On 02-Aug-89 At : Here Subject: MODEM CONFIGURATION? Sorry - a word got left out of my previous message. The last line should read: " ... that may not do too good a job of emulating .... " SYSTEM PERFORMANCE SIG Message # 233 1 replies Left by: TIM MANVILLE On 02-Aug-89 At : (202)653-0164 Subject: CPU EXHAUSTED/MEMORY AVAILABLE I have a 750 with 14mb memory. We dedicate the system to a POWERHOUSE application. When the application is running, the CPU is exhausted but very little memory is being used. Can I increase performance by taking advantage of the memory which is not being used? VMS SIG Message # 1023 0 replies Reply to # 992 Left by: BILL GUTSCHOW On 02-Aug-89 At : 716/862-4085 Subject: DIRECTORY SIZE 1026 IS THE MAJIC NUMBER THAT THE DIRECTORY WILL SHUTDOWN.. WE HAD SIMILAR PROBLEMS.. SINCE DIRECTORY FILES MUST BE CONTIGUOUS, DISK FRAGMENTATION WILL ADD TO THE PROBLEM. EVEN THOUGH YOU SHOW 997/1026 BLOCKS, IF THE DIRECTORY FILE CANNOT BE EXTENDED CONTIGUOUSLY, NO NEW FILES CAN BE CREATED.. SUCH LARGE DIRECTORIES TAKE A TERRIBLE PERFORMANCE HIT ALSO. VMS WILL CACHE DIRECTORIES, BUT WITH THEM SO LARGE, THE CACHE IS ALWAYS GETTING FLUSHED ANYWAY... INCREASING SYSGEN PARAMETERS ACP_DIRCACHE, AND ACP_HDRCACHE MAY HELP A BIT, BUT IT CERTAINLY WON'T BE ANYTHING DRASTIC... REGARDS... BILL GUTSCHOW ITT STANDARD BUFFALO NY MISCELLANEOUS SIG Message # 556 0 replies Reply to # 508 Left by: BONNIE On 02-Aug-89 At : HERE Subject: RAINBOWS We aren't using them here anymore... in fact, last I heard we have a dead rainbow collection. I still use one at home, though. So... Does that answer your question? What do you need. bsa SYSTEM PERFORMANCE SIG Message # 234 0 replies Reply to # 233 Left by: ROBERT G. SCHAFFRATH On 02-Aug-89 At : 914-335-9792 Subject: Poor Lil' 11/750 You are suffering from a classic case of being CPU bound! All the memory in the world will not make a slow CPU go faster (unless of course your processes are swapping all over the place). You might try a MONITOR MODES command and see what percentage of time you are spending in user mode. If it is way up, and you have a lot of processes in the COM state, you are going to have to get a faster CPU. You may not be able to increase memory usage because your processes probably don't want any more (they are having a tough time as is with what they have!). If your processes are page faulting excessively, you may be able to correct the situation by increasing the working set sizes (resulting in better memory usage). Note also that excessive faulting will slow your CPU down. Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 UNIX/ULTRIX SIG Message # 41 0 replies Reply to # 40 Left by: KEN OLSON On 02-Aug-89 At : Subject: RUMORS RARELY HAVE SO MANY PEOPLE BEEN SO MISLED BY SO FEW... THE COMBINED VMS/UNIX OPERATING SYSTEM IS A PIE-IN-THE-SKY DATA/FILE TRANSFER SIG Message # 572 0 replies Reply to # 559 Left by: DAVE KAPLAN On 02-Aug-89 At : 919-878-0204 Subject: MIRROR ARIS SCRIPT Phil - Thanks for logging on to our system and uploading your latest Mirror III ARISSIGS.* stuff. After some minor twiddling for my needs, I used it successfully (1st try!) last night to pickup all ARIS activity since 15-Jun. That's the last time I recall being on ARIS. Been very busy lately and I haven't had time to come up for air! I'm using Mirror III with the MNP addin. We have Microcom and Multitech MNP level 5 modems on our PDP-11/83 and our MicroVAX II. Since I'm using a NEC Multispeed laptop with an internal 2400 baud modem, software MNP is my only option. However, I can say I'm thoroughly pleased with the MNP addin. FYI (you probably know this already) the MNP addin was not written by SoftKlone. It's from a company called MagiSoft (I think that's their name) - they market their own communications package called MTE. I bought a copy for $35 about 5 months ago - unfortunately it's quite crude; no script capability, no assortment of terminal emulations, etc. Another MNP tidbit - Some older MultiTech 2400 baud modems with MNP level 5 have a "bug" in their code which prevents the MNP addin of Mirror III (and MTE for that matter) from establishing a "reliable" link. It seems that the MultiTech starts negotiating at MNP level 2, which is not supported by the software implementation of MNP. You can call MultiTech and get upgraded ROM's to fix this problem. I'll try to get onto your BBS tomorrow night with further comments about your RSTS SIG ideas. Thanks again... DATA/FILE TRANSFER SIG Message # 573 1 replies Left by: DAVE KAPLAN On 02-Aug-89 At : 919-878-0204 Subject: ARC/DEARC FOR VMS OR RSTS? Just fantasizing here... Does anyone know if any programs exist which will do PC-ARC-compatible file ARChiving and de-ARChiving under VMS or RSTS? I'm assuming that if such software was out there, I would probably have heard about it, but just thought I'd ask. VMS SIG Message # 1024 1 replies Left by: DAVE KAPLAN On 02-Aug-89 At : 919-878-0204 Subject: VAX BASIC MAGTAPE HANDLING We use VAX Basic extensively on our MicroVAX II, and do alot of data inter- change on 1/2" tape. We've written "generic" utilities to read and write tapes with just about any recordsize/blocksize/character-code specifications, as well as several dedicated programs for particular tape applications. With all this experience, there remains one thing which we can't seem to figure out how to do: When writing large amounts of data to tape (we're talking non-file-structured mode here), we'd like to be able to detect the EOT marker, write EOF marks, and proceed to an additional reel of tape. However, we can't seem to figure out how to trap for EOT? We do this on our RSTS system, but the Basic+2 "trapable" error (I forget just what it is at this moment) you get when the tape drive detects the EOT marker doesn't exist in VAX Basic V3.0. Can anyone help me out on this? Dave Kaplan; Independent Data Processing; Raleigh, NC 919-878-0204 (or ARIS) DATA/FILE TRANSFER SIG Message # 574 0 replies Left by: DAVE KAPLAN On 02-Aug-89 At : 919-878-0204 Subject: KERMIT LONG PKT, SLIDING WINDOWS We have "fairly recent" versions of VMS and PDP (for RSTS) Kermit. Our PDP Kermit supports long packets (up to 600 bytes I believe). Our VMS Kermit does not. Can anyone tell me if there are more recent implementations for either operating system which support long packets and/or sliding windows? Dave Kaplan; Independent Data Processing; Raleigh, NC 919-878-0204 (or ARIS) VMS SIG Message # 1025 0 replies Reply to # 1024 Left by: RICHARD B. GILBERT On 02-Aug-89 At : (609) 452-5130 Subject: "TRAPPING" EOV Dave, There is a status code returned in word one of the I/O Status block when a QIO operation causes the tape to reach the EOT marker. Check the I/O User's Guide, chapter on the tape driver for details. As usual I don't have a copy handy but I think that the code is SS$_EOV or SS$_EOT or something like that. If you can't find it, give me a call and I think I can turn it up for you. If you are using BASIC's I/O statements to do the I/O you may have to substitute QIO calls if BASIC does not return the I/O status code somehow. MISCELLANEOUS SIG Message # 557 0 replies Reply to # 555 Left by: RICHARD B. GILBERT On 02-Aug-89 At : (609) 452-5130 Subject: DEC SERVICE Bruce, I too have had problems (on the East Coast) with DEC not having the repair parts they need to support my systems. I have experienced things like calling to report a smoked power supply on a Micro 11/23, waiting 24 hours for somebody to show up only to have him tell me, "Yup. It's a smoked power supply, we'll have to order one!". Another one of my favorites is the broken cabinet latch (the hex key things) that took seven or eight months to get replaced. I finally did something about it when my CE showed up with a printer specialist to work on an annoying glitch in my printer. The part they wanted to replace was not in the kit that they brought. I blew a fuse and called the field service manager. She wasn't at her desk so I called her manager and dumped two and a half years of frustration in his lap. This caused a bit of a flap. The field service manager visited my site and apologized. The parts situation seems to have improved a little. I also make it clear to all concerned that it's death to show up at my site without parts. I complain. And I remember when the customer satisfactions surveys come around in the spring. Be sure to fill out those surveys (goodies are handed out to your local office, or not, depending on the results of the survey.) It helps to keep your own little log of when each call was placed and with whom. Who showed up and when, what was done and noting if parts were on hand or had to be ordered. We pay DEC big bucks for our service contracts and as far as I'm concerned, they'll stock the parts needed to support our systems or else! DATA/FILE TRANSFER SIG Message # 575 0 replies Reply to # 573 Left by: RICHARD B. GILBERT On 02-Aug-89 At : (609) 452-5130 Subject: ARC FOR VMS Dave, There are at least three programs which provide some capability to manipulate .ARC files under VMS. None of them provides total compatiblity but they do make a start. There is one called something like VAXARC or VMSARC and another called something like VMSSWEEP. Both of these programs and, I believe, one other are available on CompuServe's VAX Forum. There is also a text file comparing the facilities and compatibility of the three programs. Some of these programs may be available through DECUS as well. DATA/FILE TRANSFER SIG Message # 576 0 replies Reply to # 575 Left by: Phil Anthony On 03-Aug-89 At : Here Subject: ARC FOR VMS For the record, System Enhancement Associates does have a version of ARC for VMS. However, it's compatible only with old PC versions of ARC (V4.x - the latest PC ARC release is V6.02). If anybody wants this, it's available from SEA's bulletin board at (201) 473-1991. I may still have a copy of ARC V4 kicking around on floppy somewhere, if I haven't reused the floppy, and will try to track it down if anybody's interested. So far as ARC for RSTS goes, Brett Bump and I have talked about porting it but haven't actually done it yet. DATA/FILE TRANSFER SIG Message # 577 0 replies Reply to # 572 Left by: Phil Anthony On 03-Aug-89 At : Here Subject: MIRROR AND MNP Good seeing you here again, Dave! I'm glad the scripts worked out for you without any trouble. The MNP addin to Mirror is an interesting piece of software, but apparently it suffers from the same problem that MNP modems do: With some remote MNP modems it works perfectly, while with others it simply doesn't establish a reliable connection. For example, the Mirror addin syncs with the ATIs SoftKlone uses on their BBS and whatever System Enhancement Associates uses, but it won't make the connection with a genuine Microcom 2400 on another board I call. On the other hand, my Everex Evercom 24e (MNP Level 3) works with the Microcom but not with SEA's modem .... I haven't tried the addin with the Evercom yet. Incidentally, on my board MNP is disabled. Too many non-MNP modems have trouble handling the MNP handshake. UNIX/ULTRIX SIG Message # 42 0 replies Reply to # 41 Left by: SPELLING POLICE On 03-Aug-89 At : Subject: GEE KEN, WHEN DID YOU CHANGE THE SPELLING OF YOUR NAME? LAST TIME I LOOKED IT WAS STILL 'OLSEN'. MISCELLANEOUS SIG Message # 558 0 replies Reply to # 557 Left by: Robert G. Schaffrath On 03-Aug-89 At : 914-335-9792 Subject: Another tale of trouble Rich, sounds like you've had similar experiences. I figured I'd add a little anecdote to the collection. About three months ago our system disk, an RA82 (need I say more?), died. A very strange problem. The motor would try to spin up but would disengage after about 3 seconds. The drive would fail with a tachometer error. DEC came in with an RA82 spares kit and proceded to replace the motor, all logic boards, belts, wiring harness, power supply, and HDA (not necessarily in that order). In fact they replace all the components in the drive except for one. The motor start capacitor. Seems this part never goes bad and for this reason is not included in the spares kit. Well, after two motors, a re-install of the old HDA, and 6 hours of down time, DEC gave up. We happened to have a brand new SA482 which was waiting to be returned to DEC (we never opened it and traded it in on an SA600). Well, we opened it up, took out a new RA82 and installed it as our system disk. We never got the old RA82 to work and DEC came by two days later and took it away still in pieces. An inquiry two weeks later revealed that no further action was taken on the drive and that it was returned to the central depot (where ever that is) as-is. COMMUNICATIONS SIG Message # 289 0 replies Reply to # 286 Left by: IRA SOLOMON On 03-Aug-89 At : Subject: HST MODEMS I am using a terminal emulator which works fine when connected at 2400 and also worked fine with a different, quite obscure , psuedo-9600 baud modem.w3 *Urq{{?<#E01G$A0ss?6}~R=v~[{ COMMUNICATIONS SIG Message # 290 0 replies Reply to # 287 Left by: IRA SOLOMON On 03-Aug-89 At : (617)7364576 Subject: HST Sorry about garbled message. Am using terminal emulator that works fine at 2400 with this modem, fine direct connect at 9600, and worked fine with a different "pseudo-9600" modem. I did try a different emulator and the results were the same. I have wondered if the type ahead buffer in the VAX is big enough, bu~rzVf*t{O}9$X?B~o8)]5Z612BY@ $ohOeHVzi&[|?R $ohOeHVzi&[|?R5Gz(~#k{~r I hate to get someone to screw around without some clue that that is the right approach. I hope to find someone who has done it already and knows that that is the problem. COMMUNICATIONS SIG Message # 291 0 replies Left by: DANNY HO On 03-Aug-89 At : 18005432379 Subject: VAXSTATION 3520 I just received my first VS3520 and it comes with FireFox Q-bus Adapter Module. But this Q-bus backplane can only support one Q-bus module which is TQK70. Any- thing that I put in the Q-bus would prevent the system to boot. I was wondering anybody knows how to get around that and make the Q-bus to function as standard Q-bus. I only need the Q-bus to accept DRV11-J which is 64 line parallal inter- face card. Please help. MICROVAX SIG Message # 211 0 replies Reply to # 89 Left by: MICKEY STEIN On 03-Aug-89 At : 619-587-2272 Subject: MVAX II CRASH WELL YOU'RE NOT THE ONLY ONE ON EARTH (OR WHEREEVER YOU ARE) TO HAVE SEEN THIS. IT'S BEEN HAPPENING HERE FOR SEVERAL MONTHS. IN THIS CASE, IT SEEMS TO BE A CASE OF A SECOND-SOURCE DISK CONTROLLER BOARD SET TO THE WRONG INTERRUPT LEVEL. ALL THE DEC STUFF COMES SETUP AT BR 4, BUT OCCASIONALLY, SOMEONE SETS ONE UP TO INTERRUPT AT A HIGHER PRIORITY (5,6 OR 7). IF THIS IS THE CASE, THEN ACCORDING TO THE DEC DEVICE SUPPORT MANUAL, THE HIGHER PRIORITY BOARD MUST BE PHYSICALLY PLACED IN THE CLOSEST AVAILABLE Q-BUS SLOT NEAR THE CPU BOARD. I'M NOT SURE I REALLY UNDERSTAND ALL THIS BUT IT HAS TO DO WITH DECS INTERRUPT PRIORITIES AND HARDWARE-INTERRUPT DAISY-CHAINING, ETC. IN ANY CASE, WE'RE NOT OUT OF THE WOODS YET, BECAUSE THE SERVICE CALL HASN'T BEEN COMPLETED. IF I FIND OUT MORE, I'LL LET YOU KNOW. GOOD LUCK. VMS SIG Message # 1026 0 replies Left by: NORA MILLER On 03-Aug-89 At : 503-222-5161 Subject: SECTBLFUL ERROR I have two users of a very large FORTRAN model, still in development, who have suddenly started getting SECTBFUL errors when they fire up the debugger to run the model. This almost never happens the first time they run the model, and occasionally never happens, even after repeated runs. But it almost always happens after they have run the model, made changes, relinked and tried to rerun. The error message book tells them to inform the system manager that not enough space was allocated to global symbol tables when the system was generated. Unfortunately, I am the system manager, and I don't know what to change to help this problem. I am guessing that maybe it is PROCSECTCNT, but mine is set higher than that parameter on a networked 8650 running the same model without these problems. Any suggestions?? I hope this is one of those, "oh yeah I should have thought of that!" I am open to ridicule, even, if you can make this go away! VMS SIG Message # 1027 0 replies Left by: On 03-Aug-89 At : Subject: SYSTEM PERFORMANCE SIG Message # 235 0 replies Left by: NORA MILLER On 03-Aug-89 At : 503-222-5161 Subject: SECTBLFUL ERROR Please take a look at the message I left under SIG 37 (VMS) concerning my SECTBLFUL error. I need all the help I can get! Thanks! VMS SIG Message # 1028 0 replies Reply to # 924 Left by: JOHN CHIVIAN On 03-Aug-89 At : 214 690-3333 Subject: MV 3300 We have been experiencing sporadic problems along these same lines on our MicroVax 3300. It occasionally locks itself (and unfortunatly the rest of the cluster) up with the console message.. PAGE FILE BADLY FRAGMENTED - SYSTEM CONTINUING If anyone knows if this is the same problem discussed in #924-926 I would be very interested in hearing from you. Our solution is of cource to shutdown the 3300 (which releases all the cluster locks freeing up everyone else) and bring it right back up. What I really want to know, though, is why DEC failed to mention this 'LOST PAGE PATCH' the last time I called them with the same problem. It's always something... VMS SIG Message # 1029 1 replies Reply to # 1024 Left by: ANDREW P DUGGAN On 03-Aug-89 At : (215) 925-5400 Subject: VAX BASIC MAGTAPE HANDLING Dave: You must use $QIO to get status about end_of_tape from VMS. I have a set of similar routines that were originally written in Basic+ under RSTS. RSTS is much friendlier about this than VMS. I ran into the same EXACT problem you described and solved it. If you do $QIO Sys_Status = Sys$QioW(,T1Lun(T1_Channel) By Value, & Integer(IO$_WriteVBlk,Word) By Value, & IO_Status::Blk By Ref,,, & T1_M_Buf_Loc By Value, & T1_M_Buf_Len By Value,,,,) If (IO_Status::QIO_Status <> SS$_Normal) then If IO_Status::QIO_Status = SS$_ENDOFTAPE then Sys_Status = Sys$QioW(,T1Lun(T1_Channel) By Value, & IO$_WritEOF By Value, & IO_Status::Blk By Ref,,,,,,,,) Sys_Status = Sys$QioW(,T1Lun(T1_Channel) By Value, & IO$_WritEOF By Value, & IO_Status::Blk By Ref,,,,,,,,) Sys_Status = Sys$Dismou("PHRED__TAPE:" By Desc, & Dism_Flags by Value) End If End If $ VMS SIG Message # 1030 0 replies Reply to # 1029 Left by: ANDREW P DUGGAN On 03-Aug-89 At : (215) 925-5400 Subject: MAGTAPE Sorry about that last message (Please do not publish either)... Dave: I did not mean to upload that code. (I just wanted to look at IO function codes and statuses). What you need to do is use $QIO(W). In your IOSB you should check for SS$_ENDOFTAPE. If you get that status then you may do some $QIO with a function of IO$_WRTVBLK with a modifier IO$M_WRTEOF. I do two of them at end of tape. Then you can call SYS$DISMOU then SYS$MOUNT for the next tape. If you would like a copy of my MAGTAPE code library (VAX BASIC source code). Give me a call and I send you a tape. APD SYSTEM PERFORMANCE SIG Message # 236 0 replies Left by: GRANT MACDONALD On 03-Aug-89 At : PRATT & WHITNEY CANADA Subject: DECWINDO SYSTEM PERFORMANCE SIG Message # 237 0 replies Left by: GRANT MACDONALD On 03-Aug-89 At : (902)873-4241 Subject: DECWINDOWS I have just installed DECwindows on a microvax II which serves diskless VAXstation 2000's. After following all the hype from DEC and the user community, I told our users that this is going to be the greatest thing since sliced bread. What a PIG! Creating windows usually takes over 30 seconds! We only have 4 MB per station (DEC recommends 6), and it does quite a bit of swapping at startup. I did a little bit of tuning, but it doesn't look good.The users are screaming for their Workstation software to be installed again. Have any of you other folks installed DECwindows, and if so how do you feel about it? Hows response? Being a fairly new product, I haven't heard too much yet from the user community. Thanks. Grant MacDonald Pratt & Whitney Canada Halifax, Nova Scotia VMS SIG Message # 1031 1 replies Reply to # 1026 Left by: ROBERT G. SCHAFFRATH On 04-Aug-89 At : 914-335-9792 Subject: Bloated Section Table Your problem is that the GBLSECTIONS parameter in SYSGEN is set too low. However, you mention that this problem does not occurr on the first run of the program but on subsequent runs. This would seem to indicate that you are creating process private sections and not deleting them when you are through. So, you basically have two choices, modify the program to cleanup after itself or modify the GBLSECTIONS parameter. If you are going to modify GBLSECTIONS, do it through AUTOGEN. Edit MODPARAMS.DAT and insert the following line: ADD_GBLSECTIONS = 50 and see if that helps. You don't want to set GBLSECTIONS too high because that will waste physical memory. Still, if I had my choice, I would dig into the application first and play with SYSGEN last. Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 SYSTEM PERFORMANCE SIG Message # 238 0 replies Reply to # 237 Left by: ROBERT G. SCHAFFRATH On 04-Aug-89 At : 914-335-9792 Subject: DECwindows and it relation to a certain farm animal The problem you are describing is not new and has been described in various publications for some time now. DECwindows is a pig. DEC recommends a minimum configuration of 6MB but you would probably be better off with 12MB. Also, a major problem is the fact the CVAX chip is just too slow to handle the amount of processing required for the windows environment. Don't forget that besides running DECwindows, you are running VMS. That also take a fair amount of processing power. My personal opinion on DECwindows is that I will not deal with it until DEC releases an equivalent to the VAXstation 3100,35xx series based on the new Rigel CPU (the VAX 6400 series) chip. You will need all 7 of those VUPS to get any kind of decent response. Finally, DECwindows is most likely the future direction VMS. Even though it is slow now, doing development in it is probably a smart move. Going back to the old workstation software will only leave you working with an obsolete environment. Become familiar with DECwindows now and avoid the Christmas rush! Robert G. Schaffrath System Manager General Foods Corporation White Plains, NY 10625 VMS SIG Message # 1032 0 replies Reply to # 1031 Left by: NORA MILLER On 04-Aug-89 At : 503-222-5161 Subject: I THOUGHT I HAD ENOUGH? I have looked at INSTALL LI/GL/SUM and there are 109 GBLSECTIONS used. My sysgen setting is 190. Is it possible for the difference to be used up by private process sections? I thought that number you get from install is really how much is used--if it is not, is there somewhere else I can get the actual info from--how many total sections are being used? I will try making the change tonight. The other approach might be more reasonable except that I don't have the time to spend on investigation! Thanks for the tip and I will let you know if it works. N Miller DATA/FILE TRANSFER SIG Message # 578 0 replies Left by: DOUG TRAVIS On 04-Aug-89 At : 615-453-6940 Subject: XMODEM - NEEDED FOR RSTS NEED XMODEM FOR RSTS 8.2 ... ONLY HAVE RL02 FOR REMOVABLE MEDIA ... NO TAPE COULD ALSO USE KERMIT .. PLEASE HELP ... DO NOT KNOW WHERE TO START RSTS SIG Message # 411 0 replies Left by: DOUG TRAVIS On 04-Aug-89 At : 615-453-6940 Subject: XMODEM NEEDED FOR RSTS XMODEM NEEDED FOR RSTS 8.2 ... NO TAPE ... ONLY RL02 USED FOR REMOVABLE MEDIA ... CAN GET TAPE CONVERTED, HOWEVER. THANKS ... NEED HELP CLUSTERING/NETWORKING SIG Message # 429 0 replies Reply to # 428 Left by: ERIC S. DUNGAN On 04-Aug-89 At : 612-334-3633 Subject: PCSA PRINTERS Can you check the protection on the queue ? I do know that version 2.2 of PCSA requires queues associated with print services to have W:W world:write access. FORUM - 3RD PARTY SUPPLIERS SIG Message # 106 0 replies Left by: ANDI DURBIN On 04-Aug-89 At : (303)866-0406 Subject: PARADYNE MUX/MODEMS Does anyone using Paradyne equipment use someone other than Paradyne for maintenance? I'm real tired of the high prices and slow response times. Thanks. RSTS SIG Message # 412 0 replies Reply to # 411 Left by: Phil Anthony On 05-Aug-89 At : Here Subject: XMODEM (AND KERMIT) FOR RSTS Doug - got you covered. I'll call you Monday with XModem source code for RSTS in BASIC+, which we can PIP over (checksum only - I've never gotten around to adding CRC), and we can use that to send you Kermit. VMS SIG Message # 1033 0 replies Reply to # 823 Left by: VICTOR F. PITTMAN On 05-Aug-89 At : (404)296-7979 Subject: SYSTEM DUMP FILE n VinIn V5.0 it is possible to have you dump file ( or page and swap files ) VMS SIG Message # 1034 0 replies Reply to # 823 Left by: VICTOR F. PITTMAN On 05-Aug-89 At : (404) 296-7979 Subject: SYSTEM DUMP FILES (AGAIN) Sorry about that last reply , I'm using a PC and I,m not sure how the keys are mapped. Anyway, in v5.0 it is possible to have your dump files (and page and swap files for that matter) on a disk other than then boot disk. If your running out of disk space and are 'n not on V5 you cacan use your page file to hold the dump heres how : rename the old dump file to SYSDUMP.old . Don't delete it yet !!! Set the SYSGGEN parameter SAVEDUMP to 1 Reboot Now you can delete the old dump file To release a dump file from the page file : $ ANALYZE/CRASH_DUMP SYS$SYSTEM:PAGEFILE.SYS SDA> COPY disk:[directory]newdump.file SDA> EXIT you can put this in SYSTARTUP.COM com> UNIX/ULTRIX SIG Message # 43 0 replies Reply to # 41 Left by: DR. R On 05-Aug-89 At : HERE AND THERE Subject: OZIX Yes. OZIX is a forthcoming operating system for OLTP servers, not the successor to Mica. OZIX is being developed at the much-vaunted DECwest facility in Bellevue, Washington. DATA/FILE TRANSFER SIG Message # 579 0 replies Left by: Dick Kramer On 05-Aug-89 At : John Deere Harvester Works 1100 - 13th Ave. East Moline, IL 61244 (309) 765-6564 Subject: VMS/ULTRIX CONNECTION Does anyone out there have any experience with DEC's VMS/Ultrix Connection on a VAX serving non-DEC Unix workstations? We have been using VAX/VMS for several years now and have been very happy with both the hardware and software. However, we are beginning to see some non-DEC workstations running UNIX appear in our environment. We'd like to use a VMS based VAX as a file server for both VMS and UNIX workstations and DEC is recommending a product called the VMS/Ultrix Connection. I'd appreciate hearing from anyone hving experience with it. DATA/FILE TRANSFER SIG Message # 580 0 replies Reply to # 568 Left by: RALPH BRANDT On 05-Aug-89 At : 7177515210 Subject: ETHERNIM Our purchase of ethernim was the result of a network study design and install done by a vendor, DEC. They recommended we install certain networking software... ETHERNIM is an EtTHERNET monitor and it's really helpful, however, keeping the data base updated is a real job - we've gone from 7 nodes and 200M of cable to 165 nodes (including PC's) and over 4000M of thick cable, 11000 FT thin wire and nearly 2km of fiber. I've spent days keeping it up ot date and cant produce the reports because of software errors. They closed the problem while I was on vacation - because they couldn't reach me... in spite of an answering machine that told them when I would be back and gave an alternate they never called... (Oh how I long for the efficiency of IBM software support. At least they call back...) Now, I have another abend problem and no fix. Ethernim could be a good product because it provides a base of information about your network, alerts you when a new node comes on line and allows you to test links and devices... If it just worked... PS - the tarriff is high. We also installed Lan Traffic Monitor, (I wouldn't spend the money again). Likewise Remote Bridge Management system. (RBMS) I've written DCL to look at the bridges on the network since we have some redundancies. If a goes active that shouldn't the sys mgr and I get messages. We have a problem somewhere. That is expensive - but works. I still have gobs of questions abut it. Likewise Terminal Server Manager. The sys mgr likes it, I don't have occasion to use it. MISCELLANEOUS SIG Message # 559 0 replies Reply to # 558 Left by: RALPH BRANDT On 05-Aug-89 At : Subject: ONE BETTER I keep seeing stories about the board swappers.... A few years ago - I can give date and place and engineer's name - I know of a full CPU swap being done to fix a problem. The repair ticket on it when it went back was sh.. in the processor The unit had set under a leaky sewer pipe!!! Guess I should tell you the box was blue.. That happened while Ken O was a college student.. IBM was 20 years ahead even then. MISCELLANEOUS SIG Message # 560 0 replies Left by: RALPH BRANDT On 05-Aug-89 At : 717-751-5210 Subject: MODEMS-OUTBOUND I've read soewhere about attaching a modem to a terminal server and use it for aoutbound traffic. I can't find it and I was sure it was in DEC PRO... Anyone know where?