INFO-VAX Sun, 10 Feb 2008 Volume 2008 : Issue 81 Contents: Re: "file locked by another user" mystery Re: "file locked by another user" mystery Re: "file locked by another user" mystery Re: Anyone interested in building a vms-like OS? Re: OMFG. A Usenet post of mine cited in a patent Re: VMSINSTAL Message Re: VMSINSTAL Message Re: Xerox 7760 on DCPS? Re: Xerox 7760 on DCPS? Re: Xerox 7760 on DCPS? ---------------------------------------------------------------------- Date: Sat, 9 Feb 2008 13:09:37 -0800 (PST) From: AEF Subject: Re: "file locked by another user" mystery Message-ID: <77ebedef-58dd-4d76-9eef-323939a9d91d@e10g2000prf.googlegroups.com> On Feb 8, 1:54 pm, chessmaster1...@hotmail.com wrote: > On Jan 31, 10:05 pm, AEF wrote:> On Jan 29, 6:14 pm, Fred Bach wrote: [... some context lost ... top-posting fixed ...] > > > Well, I wouldn't want to STOP on severe errors as I'd probably want to > > do some cleanup. Also, it's somewhat arbitrary what constitutes a > > warning vs. an error vs. a "severe error". Why should editing ADSF: be > > fatal while TYPEing it only a warning?: > > > $ TYPE ASDF:[ASDF]ASDF > > %TYPE-W-SEARCHFAIL, error searching for ASDF:[ASDF]ASDF.LIS; > > -RMS-F-DEV, error in device name or inappropriate device type for > > operation > > $ SH SYM $SEVERITY > > $SEVERITY == "0" > > $ > > $ EDIT/EDT ASDF:[ASDF]ASDF > > %EDT-F-OPENIN, error opening ASDF:[ASDF]ASDF.; as input > > -RMS-F-DEV, error in device name or inappropriate device type for > > operation > > $ SH SYM $SEVERITY > > $SEVERITY == "4" > > $ > The reason the status for TYPing a non-existant file is only a warning > is simple - it does not abort TYPE. If your TYPE command line has a > list of files and one or more of the files in the list do not exist, > TYPE will output the warnings for those files but continue and type > out any files in the list that do exist. The %TYPE-W-SEARCHFAIL > warning is returned as the final status even if the final file in the > list was ok. > > The behaviour is similar if one of the files found can not be opened > by TYPE, for example due to it being locked by another user. You get > a %TYPE-W-OPENIN warning and then TYPE moves on to the next file. > This could happen even if you used a single file specification if it > has wildcards.. > > Strangely though when both of these warnings occur %TYPE-W-SEARCHFAIL > is always returned as the final status even when %TYPE-W-OPENIN occurs > for the final file accessed. > > EDIT, on the other hand, will only process a single file spec (the > default EDIT verb syntax does accept a file list, but as far as I can > tell TPU ignores all but the first file specification in the list). > Since in this case you are only processing one file it makes sense > that any error accessing that file would be a fatal error. > Well, what about the DIRECTORY command, then? It gives warning for no files found, error for directory not found, and severe error for bad device: DCL> DIR ASDF.ASDF %DIRECT-W-NOFILES, no files found DCL> SH SYM $SEVERITY $SEVERITY == "0" DCL> DIR [ASDF] %DIRECT-E-OPENIN, error opening DISK$DATA1:[ASDF]*.*;* as input -RMS-E-DNF, directory not found -SYSTEM-W-NOSUCHFILE, no such file DCL> SH SYM $SEVERITY $SEVERITY == "2" DCL> DIR ASDF: %DIRECT-E-OPENIN, error opening ASDF:[FELDMAN.DCL.PLAY]*.*;* as input -RMS-F-DEV, error in device name or inappropriate device type for operation DCL> SH SYM $SEVERITY $SEVERITY == "4" DCL> And in the last case, the %-severity disagrees with the $SEVERITY severity. OK, when DIRECTORY is given multiple file-specs, the command does continue until it hits an error or severe case; but it could easily conitnue past those also, yet doesn't. So it's consistent with what you wrote when assigning severity levels, but inconsitent with TYPE in that it doesn't attempt to look for more files after hitting. Also, the TYPE command is inconsistent with DIRECTORY in that it gives a warning for all three cases above whereas DIRECTORY doesn't. Furthermore, we're talking about using these commands in a command procedure. There I think what your code is doing matters more than whether you're using TYPE or EDIT (though I admit that the former is probably used far more often). I mean, if you're just running the commands at the DCL prompt, does it really matter much whether the result is warning, error, or severe? But it does matter a lot when the commands are run from a command procedure because of the error handler, of course. And if TYPE can continue on to the remaining file-specs in the list, why doesn't DIRECTORY do so? The reason is that all these commands were written by different programmers at different times and a strict consistency regarding these issues was never enforced. Can you explain why syntax errors in IF statements are only warnings? Do you really want your code to continue after an illegal IF statement, even in testing, as it would with the default error handler setting? I certainly don't. That's one reason I usually start early with ON WARNING THEN GOTO _ERROR or similar, depending on what I'm doing, of course. Then there's the GOTO command. When given a non-existent label, it sets $SEVERITY to 0, but your code immediately exits even if a SET NOON command is in effect! But the portion of an ON WARNING THEN command _is_ executed in such a case. This is yet another reason to start your code with ON WARNING THEN . The book Writing Real Programs in DCL (in which I have no financial interest) does an excellent job in explaining how to deal with these type things. AEF ------------------------------ Date: Sat, 9 Feb 2008 13:49:45 -0800 (PST) From: AEF Subject: Re: "file locked by another user" mystery Message-ID: On Feb 9, 12:45 pm, hel...@astro.multiCLOTHESvax.de (Phillip Helbig--- remove CLOTHES to reply) wrote: > In article > <5c75c7ba-def9-4727-ba0e-b55d8ff5b...@q77g2000hsh.googlegroups.com>, > > chessmaster1...@hotmail.com writes: > > On Jan 31, 10:05=A0pm, AEF wrote: > > > On Jan 29, 6:14 pm, Fred Bach wrote: > > > The reason the status for TYPing a non-existant file is only a warning > > is simple - it does not abort TYPE. > > Makes sense. > > > > Well, I wouldn't want to STOP on severe errors as I'd probably want to > > > do some cleanup. Also, it's somewhat arbitrary what constitutes a > > > warning vs. an error vs. a "severe error". > > Indeed. My pet peeve is MOUNT/CONFIRM/SHADOW. Why would one do this? > Because one wants to make sure that the right members are mounted and, > if a copy is required, that it is in the right direction. If a copy is > required, this brings up -F-. OK, you're interactively building a > shadow set, and you see -F-, even though it is behaving exactly as > expected. Scary. Well, maybe a little. The -F- is there just to keep you alert and encourage you to be extra careful so that you don't overwrite the wrong disk. I'd say that's worth a small scare, no? > > Personally, I think it is clear what success (severity -1), normal (1) > and informational (3) are. Now it gets a bit trickier. I agree that a > warning (0) should be saying "OK, I don't have to abort, but perhaps you > expected something else". This seems more or less standard. Where > things are unclear is the difference between error (2) and fatal error > (4). In both cases, the program needs to abort (except in strange cases > like BACKUP with multiple files). In my own software, an error is > invalid user input which can't be processed, i.e. the user made a > mistake, whereas a fatal error is an internal error in the program. > This is not the paradigm favoured by VMS (if, indeed, there is one), but > at least it is a consistent approach. In other words, you can probably > do something about an error, but not about a fatal error. Well, I'm not aware of any really clear definitions of WARNING vs. ERROR vs. SEVERE. It's really a sort of seat-of-the-pants thing, so different people will have different ideas of what the definitions should be. The BACKUP command, IMO, is pretty good about the way it assigns W, E, and F to its various possible errors. For something like this, %BACKUP-F-POSITERR, error positioning _MKA0:[]SY080201.BCK; you want fatal. You clearly can't continue. Something has gone terribly wrong. OTOH, verification errors, which get the -E- treatment, aren't "fatal" but may be cause for serious concern. %BACKUP-W-ACCONFLICT, SYS$SYSDEVICE:[MONITOR]MONITOR.LOG;9 is open for write by another user OK, some might argue these should also be -E-, but I'm okay with these being -W-. But I find the default error handler, ON ERROR THEN EXIT, to be unsatisfactory, mostly because it won't catch IF errors. AEF ------------------------------ Date: Sat, 9 Feb 2008 17:52:10 -0800 (PST) From: AEF Subject: Re: "file locked by another user" mystery Message-ID: On Feb 9, 7:58 pm, Jan-Erik S=F6derholm wrote: > AEF wrote: > > %BACKUP-W-ACCONFLICT, SYS$SYSDEVICE:[MONITOR]MONITOR.LOG;9 is open for > > write by another user > > > OK, some might argue these should also be -E-, but I'm okay with these > > being -W-. > > They *are* -E- *IF* you do *not* add "/ignore=3Dinterlock" to the BACKUP > command : > > %BACKUP-E-OPENIN, error opening dev:[dir]file as input > -SYSTEM-W-ACCONFLICT, file access conflict > > So if you tell BACKUP to ignore file access conflicts, it > just warns you about them, if not, it gives you an error. > > Rather "logical", I'd say... :-) > > Jan-Erik. You're right! Thanks for reminding me. Good point. I wish I could review all backup error messages with their associated error levels. Does anyone know a way to do that? And for all other error messages? Thanks. AEF ------------------------------ Date: Sat, 09 Feb 2008 13:11:11 -0600 From: David J Dachtera Subject: Re: Anyone interested in building a vms-like OS? Message-ID: <47ADFACF.303B0F8D@spam.comcast.net> Bill Gunshannon wrote: > > In article <47AD22BA.15A1322A@spam.comcast.net>, > David J Dachtera writes: > > Bill Gunshannon wrote: > >> [snip] > >> There are thousands of people out there with the knowledge of linux and unix > >> to do it. There isn't for VMS for the obvious reason that the same level > >> of avaialble information has never existed for VMS. No one has the sources. > >> No one but HP can build versions of it. So, if you want to use VMS you > >> have to go to HP and pay what ever they want for it. That's buysiness. > >> You can't build an equivalent business on software you are required to hand > >> to your competitors. But then, if believe all software must be freei and you > >> still want to feed your family, I understand WallMart and McDonalds are still > >> hiring. > > > > The problem I have with that business model is that it's sort of like a > > light bulb that never fails: you sell it ONCE (think: "everlasting > > gobstopper"). > > Which business model? Two, rather opposite, models are mentioned above. > > > Support, on the other hand, is either a monthly, > > quarterly, yearly or some other recurring charge (read: "cash flow"). > > > > That's known as "residual income" in some circles, rather like rental > > income on an apartment building, royalties of various kinds, etc. > > And that is pretty much where the majority of VMS revenue is coming from > at this point, I would imagine. BUt, the point is, under the GPL business > model you can't even plan on a profit from this (unless you cheat on the > license) because once you have gicen away all the sources to your product > someone with the requisite skills but without the development expenses > to recoup can undercut you on the cost of support leaving you with less > is anything at all. > > I guess what it all comes down to is wether or not people really want to > see VMS remain (and maybe even win back some status) as a viable commercial > operating system or if they are satisfied with it degenerating to the > point of merely another hobbyist toy. > > Having given some thought to the whole thing recently, I have come up with > what might be another interesting concept for those who have a commercial > interest in seeing VMS thrive (or at least survive). The idea of buying > it from HP has been broached but is not realistic as the cost would be > prohibitive and, anyway, it ain't for sale. EVERYthing is "for sale", if the price is right. > That has led to talk of > cloning it and the arguments over licensing. But, how about a third > possible alternative? Anyone ever considered talking to HP about the > possibility of getting a non-exclusive license to develop, sell and > maintain VMS? This was actually done with Unix a lot in the early days. > It is what got us XENIX, Venix, Onyx and any number of other proprietary > Unix clones. The same model might be doable for VMS, especially when > you look at the number of people with serious experience who are no > longer under contract to HP. Comments anyone? I've given it a LOT of thought. Few of us consider "hacking" the system for "fear" of getting into "unsupported" territory; yet, I've already done just that. I updated the V7.1-2 install CD image to apply an important ECO, and even went so far as to repeat that exercise and produce a bootable V7.3-2 CD which includes support for the newer FC HBAs in our GS1280s at work. Out of sheer boredom, I even concocted a replacement for STARTUP.COM based on the "run level" scheme of the UN*X init process and inittab. Never got much farther than that due to VMS's increasing ignominity and obscurity. Otherwise, the world might have seen rebranded Alphas and VMS ala DCC's rebranding of Data General gear and their "flavor" of IRIS known as EOS (late 1970's, early 1980's). Given that, breaking youngsters out of the point-and-click paradigm and into the world of VMS, RMS and ODS seems even more daunting than the actual concept of "swimming upstream" against the current set in motion by HP going out and pushing UX on Itanic instead of VMS. About the only practical action at his point would be port VMS-I64 to x86-64 and try to ride that wave. Database, application, etc. support for it would, of course, remain prime challenges. David J Dachtera DJE Systems ------------------------------ Date: Sat, 9 Feb 2008 17:28:56 -0800 (PST) From: Joseph2k Subject: Re: OMFG. A Usenet post of mine cited in a patent Message-ID: <9f948212-aa94-4846-8563-1b4a5b2ae3de@e25g2000prg.googlegroups.com> On Feb 5, 5:54 am, VAXman- @SendSpamHere.ORG wrote: > In article , "Tim Williams" writes: > >And he didn't even get the date right: > >http://groups.google.com/group/comp.os.vms/browse_frm/thread/bfb510bb... > >28/466fb5c83d9f1f47?lnk=st&q=deleting+files+permanently#466fb5c83d9f1f47 > > Fer shits and giggles, I put my name into that same search and found that > I was cited in a patent application as well. No, even though I am an old > Pennsylvania farm boy, it is not any of the patents for horseshoeing that > turned up. :) It wasn't ausenetpost though. > > Not bad for a VMS ignorant, drooling, cytoplasmic mass prior to '94... or > so it was. > > -- > VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)COM > > "Well my son, life is like a beanstalk, isn't it?" > > http://tmesis.com/drat.html Still, write it down, might be read, nothing's better left unsaid. ------------------------------ Date: Sat, 09 Feb 2008 13:22:10 -0600 From: David J Dachtera Subject: Re: VMSINSTAL Message Message-ID: <47ADFD62.194E030F@spam.comcast.net> Martin Vorlaender wrote: > > Chuck Moore wrote: > > What UNZIP would y'all recommend ? > > Well, the source of all things, http://infozip.sourceforge.net/UnZip.html > says > > VMS binaries for VAX and Alpha/AXP processors can be found at: > ftp://ftp.info-zip.org/pub/infozip/VMS/ (Germany) > http://www.mirrorservice.org/sites/ftp.info-zip.org/pub/infozip/VMS/ (UK) > ftp://sunsite.icm.edu.pl/pub/unix/archiving/info-zip/VMS/ (Poland) > http://vms.process.com/ftp/vms-freeware/fileserv/ (USA) > [also via ftp ftp://ftp.process.com/vms-freeware/fileserv/ ] Indeed. Hunter Goatley does a marvelous job of packaging software, especially the most recent versions. All I did, really, was take his distro. for UNZIP V5.3-2 and modify the LINK procedure so it would work inside a KITINSTAL proc. Could probably do it again for the latest. PCSI is still "black magic" to me: I've yet to see a coherent explanation of how it works or how to deal with it. I think it has some fixed logic ala RPG, but I've never seen a flowchart or anything which explains it as I saw for RPG which essentially was RPG's "rosetta stone" in my case. > http://h71000.www7.hp.com/freeware/freeware70/info-zip/ (USA) The VMS Freeware distro. tens to lag a bit, also, but is still more current on UNZIP than V5.3-2, which is stable but does not handle ODS-5. Steven Schweda ("sms") is the "local" ZIP/UNZIP guru these days. David J Dachtera DJE Systems ------------------------------ Date: Sat, 9 Feb 2008 14:06:50 -0600 (CST) From: sms@antinode.org (Steven M. Schweda) Subject: Re: VMSINSTAL Message Message-ID: <08020914065031_2062A39A@antinode.org> From: David J Dachtera > [...] All I did, really, was take his > distro. for UNZIP V5.3-2 and modify the LINK procedure so it would work > inside a KITINSTAL proc. Could probably do it again for the latest. What's "UNZIP V5.3-2"? The current released version is 5.52, as in: ALP $ unzip -v UnZip 5.52 of 28 February 2005, by Info-ZIP. Maintained by C. Spieler. Send bug reports using http://www.info-zip.org/zip-bug.html; see README for details. [...] What does "UNZIP V5.3-2" call itself? > The VMS Freeware distro. tens to lag a bit, also, but is still more > current on UNZIP than V5.3-2, which is stable but does not handle ODS-5. Once they're released, they're all stable. Or did you mean "works properly" when you said "is stable"? I would try very hard not to use anything older than UnZip 5.52, and unless something gets bungled very badly in unexpected ways, if UnZip 6.0 ever appears, I'd move to that about as quickly as possible (reasons being large-file suport, bzip2 compression, better ODS5 support, and a bunch of miscellaneous bug fixes). ------------------------------------------------------------------------ Steven M. Schweda sms@antinode-org 382 South Warwick Street (+1) 651-699-9818 Saint Paul MN 55105-2547 ------------------------------ Date: Sat, 9 Feb 2008 13:39:03 -0800 (PST) From: ewilts@ewilts.org Subject: Re: Xerox 7760 on DCPS? Message-ID: <05303d5e-1915-42d1-8f04-81f91e0cd64f@e10g2000prf.googlegroups.com> On Feb 9, 12:32 pm, "Richard B. Gilbert" wrote: > How does it benefit HP to support competitor's printers? Sure, it's a > nice and generous thing to do . . . . You may not have noticed that HP supports Linux, an active competitor to HP-UX. They support Windows with printer drivers, camera software, storage, and a ton of other stuff. HP is a large company and the small amount of revenue they'll lose on their printers by supporting them on OpenVMS will be more than made up for by their OpenVMS-based support contracts, both hardware and software. I can't imagine many customers saying that they'll switch operating systems because the printers they want to use aren't supported. You've got hopefully satisfied OpenVMS customers - keep them that way. .../Ed ------------------------------ Date: Sat, 09 Feb 2008 17:54:44 -0500 From: "Richard B. Gilbert" Subject: Re: Xerox 7760 on DCPS? Message-ID: <47AE2F34.3020203@comcast.net> ewilts@ewilts.org wrote: > On Feb 9, 12:32 pm, "Richard B. Gilbert" > wrote: > >>How does it benefit HP to support competitor's printers? Sure, it's a >>nice and generous thing to do . . . . > > > You may not have noticed that HP supports Linux, an active competitor > to HP-UX. They support Windows with printer drivers, camera software, > storage, and a ton of other stuff. > > HP is a large company and the small amount of revenue they'll lose on > their printers by supporting them on OpenVMS will be more than made up > for by their OpenVMS-based support contracts, both hardware and > software. I can't imagine many customers saying that they'll switch > operating systems because the printers they want to use aren't > supported. You've got hopefully satisfied OpenVMS customers - keep > them that way. > > .../Ed Ummm..... I sometimes get the impression that the only people at HP who know or care about VMS are Sue and the remaining developers! ------------------------------ Date: Sat, 09 Feb 2008 20:36:09 -0500 From: JF Mezei Subject: Re: Xerox 7760 on DCPS? Message-ID: <47ae5513$0$30174$c3e8da3@news.astraweb.com> Richard B. Gilbert wrote: > Ummm..... I sometimes get the impression that the only people at HP who > know or care about VMS are Sue and the remaining developers! Inpression ? I though it was FACT ? :-) :-) BTW, you need to step back a number of years when Digital was also a printer company. It made sense back then to have DCPS (and remember that initially, it supported only DEC printers). My guess is that DCPS is not strategic to HP, it is just a leftover that is seen as part of VMS and needs to be maintained. But it is one of the few products still being developped for VAX. And Paul Anderson has historically alwatys provided good communicatiosn with the c.o.v. community. BTW, Mr Anderson, if you read this: The NPMANAGE utility for the mac can be used on OSX (powermac). Runs in the classic environment fine. ------------------------------ End of INFO-VAX 2008.081 ************************