1 INFO-VAX	Thu, 22 Dec 2005	Volume 2005 : Issue 711       Contents: create sub-process fails Re: create sub-process fails Re: create sub-process fails) Re: Deleting alias files (blocks deleted) ) Re: Deleting alias files (blocks deleted) ) Re: Deleting alias files (blocks deleted) 1 Installing DW Motif with PCSI: ovverride checks ? 5 Re: Installing DW Motif with PCSI: ovverride checks ?  Re: Jon Andruszkiewicz Re: Jon Andruszkiewicz Re: Jon AndruszkiewiczC Re: Problem with SYS$CREMBX() writing to a cluster-wide name table? 4 Re: Some people are willing to support their product4 Re: Some people are willing to support their product Re: VMS V8.2 SHOW ERROR Bug? Re: VMS V8.2 SHOW ERROR Bug? Re: Zip question  F ----------------------------------------------------------------------    Date: 21 Dec 2005 23:57:00 -0800 From: thick_guy_9@yahoo.com ! Subject: create sub-process fails C Message-ID: <1135238220.328418.295550@o13g2000cwo.googlegroups.com>    Hi  B I am trying to create a subprocess from my login from DCL (command line).   RUN/DETACH -E /UIC=[513,20503] -    <------------------------------- this is my UIC  /AUTHORIZE - /PRIV=SAME -& /INPUT=dsa10:[test]LOG_something.COM -' /OUTPUT=dsa10:[test]LOG_something.log - & /ERROR=dsa10:[test]LOG_something.err - /PROCESS_NAME = "TEST$PROC" - 	 /PRIO=4 -  SYS$SYSTEM:LOGINOUT.EXE    I get the foll. error:& %RUN-F-CREPRC, process creation failed( -SYSTEM-F-IVLOGNAM, invalid logical name  ' I have the IMPERSONATE and GROUP privs.   * Why do I get this error, and how to solve?   thanks in advance...   ------------------------------  % Date: Thu, 22 Dec 2005 07:24:46 -0500  From: "Jilly" <jilly@hp.com>% Subject: Re: create sub-process fails , Message-ID: <43aa9b0e$1@usenet01.boi.hp.com>  K This issue is that a logical name used in the command is not valid in your  L processes context or in the context of the created process (but this latter E does not appear to be the issue as the detached process does not get  G created).  Check your commands for typos and missing line continuation   hypens.   I Also one of your problems is with the combination of /UIC & /AUTHORIZE .  J What will hapen is that the process will get created with all the quotas, M rights, etc. from the username of the process issuing the RUN command due to  K the /AUTHORIZE and only after it is created will the process have it's UIC  K changed to the one you specified.  The combination of /UIC/AUTH *will not*  E mean that the new process use the quotas, rughts, etc. from whatever  M username might be assocuated with the specified UIC as there can be multiple  L usernames associated with the UIC.  If you want your detached process to be L under the username of the process executing the command then drop the /UIC. G Also I'd drop the /PRIV=SAME as this will not be needed with the /AUTH.   ) <thick_guy_9@yahoo.com> wrote in message  = news:1135238220.328418.295550@o13g2000cwo.googlegroups.com...  > Hi > D > I am trying to create a subprocess from my login from DCL (command > line). >  > RUN/DETACH -G > /UIC=[513,20503] -    <------------------------------- this is my UIC  > /AUTHORIZE - > /PRIV=SAME -( > /INPUT=dsa10:[test]LOG_something.COM -) > /OUTPUT=dsa10:[test]LOG_something.log - ( > /ERROR=dsa10:[test]LOG_something.err - > /PROCESS_NAME = "TEST$PROC" -  > /PRIO=4 -  > SYS$SYSTEM:LOGINOUT.EXE  >  > I get the foll. error:( > %RUN-F-CREPRC, process creation failed* > -SYSTEM-F-IVLOGNAM, invalid logical name > ) > I have the IMPERSONATE and GROUP privs.  > , > Why do I get this error, and how to solve? >  > thanks in advance... >    ------------------------------    Date: 22 Dec 2005 07:16:14 -0600 From: briggs@encompasserve.org% Subject: Re: create sub-process fails 3 Message-ID: <k9lfZqefpElG@eisner.encompasserve.org>   a In article <1135238220.328418.295550@o13g2000cwo.googlegroups.com>, thick_guy_9@yahoo.com writes:  > Hi > D > I am trying to create a subprocess from my login from DCL (command > line).   >  > RUN/DETACH -G > /UIC=[513,20503] -    <------------------------------- this is my UIC  > /AUTHORIZE - > /PRIV=SAME -( > /INPUT=dsa10:[test]LOG_something.COM -) > /OUTPUT=dsa10:[test]LOG_something.log - ( > /ERROR=dsa10:[test]LOG_something.err - > /PROCESS_NAME = "TEST$PROC" -  > /PRIO=4 -  > SYS$SYSTEM:LOGINOUT.EXE   D If you are trying to create a sub-process then both /DETACH and /UIC@ will defeat your attempt.  Both qualifiers force the creation of a detached process instead.   G If you are serious about redirecting sys$error then SYS$SYSTEM:LOGINOUT A will defeat your attempt.  It will redefine SYS$ERROR to the same D record stream used by SYS$OUTPUT -- both logical names will point to
 the same PPF.   F If you are serious about inheriting the privileges of the creator thenF you need to get rid of the /AUTHORIZE qualifier.  That qualifier tells7 LOGINOUT to pick up quotas and privileges from the UAF.    > I get the foll. error:( > %RUN-F-CREPRC, process creation failed* > -SYSTEM-F-IVLOGNAM, invalid logical name  F However, the above quibbles do not explain the reported error message.  @ Your input, output and error values are acceptable.  Even if the@ pointed-to file does not exist, RUN /DETACH doesn't care.  Those? are pure text strings from the point of view of RUN.  It is the @ created process that ends up caring whether they are meaningful.  E Your process name, TEST$PROC is acceptable.  Again, that's pure text.   @ Your image name, SYS$SYSTEM:LOGINOUT.EXE is acceptable.  I don'tA know chapter and verse on whether RUN checks for existence of the A image before creating the process, but SYS$SYSTEM:LOGINOUT.EXE is % quite commonly used and is just fine.    And it works for me...   $ type test.com  $ RUN/DETACH -  /UIC=[BRIGGS] -
  /AUTHORIZE - 
  /PRIV=SAME - <  /INPUT=disk_user:[decuserve_user.briggs]LOG_something.COM -=  /OUTPUT=disk_user:[decuserve_user.briggs]LOG_something.log - <  /ERROR=disk_user:[decuserve_user.briggs]LOG_something.err -  /PROCESS_NAME = "TEST$PROC" -
  /PRIO=4 -  SYS$SYSTEM:LOGINOUT.EXE $ @test.com = %RUN-S-PROC_ID, identification of created process is 0002BE77  $ type log_something.log. $ saved_verify = 0 ! This is log_something.com% $ exit 1 + 0 * f$verify(saved_verify) 8   BRIGGS       job terminated at 22-DEC-2005 07:08:23.57     Accounting information: O   Buffered I/O count:                 28      Peak working set size:       1824 O   Direct I/O count:                   27      Peak virtual size:         169040 O   Page faults:                       133      Mounted volumes:                0 O   Charged CPU time:        0 00:00:00.01      Elapsed time:       0 00:00:00.15   / Could there be something you're not showing us?    	John Briggs   ------------------------------    Date: 22 Dec 2005 08:14:03 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) 2 Subject: Re: Deleting alias files (blocks deleted)3 Message-ID: <4mggdwJAgbGz@eisner.encompasserve.org>    In article <1135209759.547147.96060@g43g2000cwa.googlegroups.com>, "Hein RMS van den Heuvel" <heinvandenheuvel@gmail.com> writes:  >  > Fine example AEF. I > I hope folks realize that one component, which was not nighlighted, and G > which is critical in the behaviour, and that is the fact that A.A and I > B.B live in the same directory. Thus that backlink would be the same no G > matter which file is 'responsible' / the real McCoy. Creating entries + > in other directories will change the game  > * > There is no 'real file' versus aliasses.* > All directory entries are created equal.H > Whether one entry is deemed 'real' or 'just an alias' is by deduction,/ > not through an direct attribute of the entry. F > If the directory entry lives in the directory pointed to by the fileG > header, or if the directory backpointer is invalid, then the entry is F > 'for real'.  So it can even be a matter of timing. Wherever you lookC > first, that will be the real file. The fiel name in teh header is I > immaterial to the file system, but us users may use it to label a given ! > directy entry as real vs alias.   G    The following behaviour under ODS-5 on VMS 7.2-1 disagrees with your 
    assertion:    $ create a.a a.a   Exit  $ set file a.a/enter=b.b
 $ type b.b a.a 
 $ del a.a;
 $ type b.b8 %TYPE-W-OPENIN, error opening USER1:[USER]b.b;1 as input -RMS-E-FNF, file not found $ dir/fu b.b   Directory USER1:[USER]    b.b;1               no such file   Total of 1 file, 0/0 blocks.
 $ del b.b; $ create a.a a.a   Exit  $ set file a.a/enter=b.b
 $ del b.b;
 $ type a.a a.a     C    So the original file (a.a) is treated differently from the alias -    (b.b), even though they have the same DID.    ------------------------------    Date: 22 Dec 2005 06:38:38 -0800< From: "Hein RMS van den Heuvel" <heinvandenheuvel@gmail.com>2 Subject: Re: Deleting alias files (blocks deleted)C Message-ID: <1135262318.237520.132830@g49g2000cwa.googlegroups.com>   E Dang, looks like the name matters after all, when working in the same 
 directory.F Just to be sure I retested Bob's example with the names flipped around2 in case it was being the first file that mattered.D My prior verifications had all been with multiple directories, I had0 not tried teh single directory case exhaustively Thanks for keeping me honest. & May or may not read up some more late.
 Gotta run, Hein.    ------------------------------  % Date: Thu, 22 Dec 2005 10:37:43 -0500 . From: JF Mezei <jfmezei.spamnot@vaxination.ca>2 Subject: Re: Deleting alias files (blocks deleted)- Message-ID: <43AAC80F.387778AD@vaxination.ca>    Bob Koehler wrote: > $ create a.a > a.a  >  Exit  > $ set file a.a/enter=b.b  E HELP SET FILE /ENTER  on VAX VMS 7.2 warns of not creating aliases in  the same directory.   H When you delete B.B, Delete looks up the backlink in indexF.SYS for thatF file ID and finds that it is the directory that holds B.B so it thinks* that B.B is the real McCoy and deletes it.  % And this has interesting corrolaries:   
 Create [X]A.A + The quick brown fox jumps over the lazy dog    SET FILE [X]A.A /ENTER=[X]B.B   M At this point, B.B is "as real" as A.A since they are both in same directory.      RENAME [X]B.B [Y]B.B  E Since B.B sits in the "official" directory for the file, RENAME would F then update INDEXF.SYS to have a backlink to the new [Y] directory. AtB that time [X]A.A becomes an alias because it no longer sits in the  official directory for the file.  < I hope I understood Hein's very good explanations correctly.   ------------------------------  % Date: Thu, 22 Dec 2005 12:38:56 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> : Subject: Installing DW Motif with PCSI: ovverride checks ?, Message-ID: <43AAE470.2F90B652@teksavvy.com>  H PRODUCT INSTALL of the DW 1.25 (for VMS 7.2) insists I reboot my machineD BEFORE it is willing to simply copy files. (due to sysgen parameters: which I have already changed but which are not yet active.  G Is there some simple logical I can define to get that darned PRODUCT to C simply copy the files to where they belong ? Or must I learn how to A unpack the whole .PCSI file, find the offending procedure in PCSI C language that checks those parameters, change it and then  run it ?   F What is the magic incantation to unpack the kit into a directory ? (orH does it require different commands to extract files and the pcsi control files ?   F After that, I gather I can simply use PRODUCT INSTALL/SOURCE=[dev:dir]& DW-MOTIF  /KIT_ATTRIBUTES=REFERENCE  ?    > Also, it wants PQL_MPRCLIM to be 8. I am at 0 (which should beE unlimited, right ?). Is this just some dumb if A < B THEN WRITE ERROR I message, or is there some reason that 0 would be considered truly lower ?   E Alternatively, the kit has some good old VMSINSTALL savesets, but the G doc says to use those for VMS 5.5-2 only. Could this be usable on a 7.2 3 system ? (the last version with display postscript)    ------------------------------    Date: 22 Dec 2005 19:47:13 +01006 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)> Subject: Re: Installing DW Motif with PCSI: ovverride checks ?, Message-ID: <43ab02c1$1@news.langstoeger.at>  \ In article <43AAE470.2F90B652@teksavvy.com>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes:I >PRODUCT INSTALL of the DW 1.25 (for VMS 7.2) insists I reboot my machine E >BEFORE it is willing to simply copy files. (due to sysgen parameters ; >which I have already changed but which are not yet active.   # 1) SET HOST 0/LOG=DWMOTIF-V0105.LOG % 2) Answer YES to allow PCSI to reboot N 3) If SHUTDOWN.COM asks for reboot allowed (which it does with RTA0 terminals)
 answer NO.   --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------    Date: 22 Dec 2005 07:03:10 -0800$ From: "Ed Wilts" <ewilts@ewilts.org> Subject: Re: Jon Andruszkiewicz C Message-ID: <1135263790.050436.166290@z14g2000cwz.googlegroups.com>    john.atoz@gmail.com wrote:I > That is very true, I am not the only person with experience in HBVS ... I > but I was the project leader for several years, until I "retired" on 14  > October 2005 :-)  F Sorry, John, I didn't mean to imply that I questioned your abilities -> just offering a suggestion to the original poster in case yourA retirement meant you really were done (and I'm glad to see you're  not!).  F It always amazes me when I go to my Solaris admins and say "surely theF Veritas Foundation Suite software can do this - we've been doing it onF VMS for a long time!" and they reply "yeah, but that's VMS"...  Things2 like mini-merge, shadow site id with costing, etc.  A I've been using HBVS since 1990 and still remember when the first G shadow set member failed.  I found out about 3 days later when somebody F asked me why the shadow set had only a single member.  That shadow setE contained the sysuaf and the drive dropped out so cleanly that nobody G noticed (except the operator log files but these events happened in the E middle of the night).  It wasn't long after that I implemented shadow A set watchers to make sure we had at least 2 members on all shadow F sets...  The Solaris admins still haven't figured that part out yet...  G It sucks to see a bunch of the DECies take the recent packages.  I fear F that one day I'll have a serious storage problem and all I'll get is a; bunch of young HP dudes that haven't seen the  light yet...   4    .../Ed (proud to have used VMS for over 20 years)   ------------------------------  % Date: Thu, 22 Dec 2005 11:52:25 -0500 - From: William Webb <william.w.webb@gmail.com>  Subject: Re: Jon Andruszkiewicz I Message-ID: <8660a3a10512220852n418a81bck93ee0d080720e82d@mail.gmail.com>   B On 22 Dec 2005 07:03:10 -0800, Ed Wilts <ewilts@ewilts.org> wrote: > john.atoz@gmail.com wrote:K > > That is very true, I am not the only person with experience in HBVS ... K > > but I was the project leader for several years, until I "retired" on 14  > > October 2005 :-) > H > Sorry, John, I didn't mean to imply that I questioned your abilities -@ > just offering a suggestion to the original poster in case yourC > retirement meant you really were done (and I'm glad to see you're  > not!). > H > It always amazes me when I go to my Solaris admins and say "surely theH > Veritas Foundation Suite software can do this - we've been doing it onH > VMS for a long time!" and they reply "yeah, but that's VMS"...  Things4 > like mini-merge, shadow site id with costing, etc. > C > I've been using HBVS since 1990 and still remember when the first I > shadow set member failed.  I found out about 3 days later when somebody H > asked me why the shadow set had only a single member.  That shadow setG > contained the sysuaf and the drive dropped out so cleanly that nobody I > noticed (except the operator log files but these events happened in the G > middle of the night).  It wasn't long after that I implemented shadow C > set watchers to make sure we had at least 2 members on all shadow H > sets...  The Solaris admins still haven't figured that part out yet... > I > It sucks to see a bunch of the DECies take the recent packages.  I fear H > that one day I'll have a serious storage problem and all I'll get is a= > bunch of young HP dudes that haven't seen the  light yet...  > 5 >   .../Ed (proud to have used VMS for over 20 years)  >  >   ; When I worked in banking, the main application we ran was a E MUMPS-based trust accounting product, and, as is frequently the case, D junior staff were assigned to the vendor's after-hours support line.  D It used to aggravate me in the extreme to be paying $125 an hour for; the privilege of teaching somebody *who knew less about the / application than I did* how to fix the problem.    WWWebb   --C NOTE: This email address is only used for noncommerical VMS-related  correspondence. C All unsolicited commercial email will be deemed to be a request for 8 services pursuant to the terms and conditions located at# http://bellsouthpwp.net/w/e/webbww/    ------------------------------    Date: 22 Dec 2005 12:14:31 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)  Subject: Re: Jon Andruszkiewicz 3 Message-ID: <kOsyjae6a+D1@eisner.encompasserve.org>   y In article <8660a3a10512220852n418a81bck93ee0d080720e82d@mail.gmail.com>, William Webb <william.w.webb@gmail.com> writes: D > On 22 Dec 2005 07:03:10 -0800, Ed Wilts <ewilts@ewilts.org> wrote:  J >> It sucks to see a bunch of the DECies take the recent packages.  I fearI >> that one day I'll have a serious storage problem and all I'll get is a > >> bunch of young HP dudes that haven't seen the  light yet... >>6 >>   .../Ed (proud to have used VMS for over 20 years) >> >> > = > When I worked in banking, the main application we ran was a G > MUMPS-based trust accounting product, and, as is frequently the case, F > junior staff were assigned to the vendor's after-hours support line. > F > It used to aggravate me in the extreme to be paying $125 an hour for= > the privilege of teaching somebody *who knew less about the 1 > application than I did* how to fix the problem.   D The last time I called for VMS telephone support I ended up teachingG someone on "The Pascal Team" how Pascal worked.  That was over 20 years  ago.   ------------------------------    Date: 22 Dec 2005 07:52:28 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) L Subject: Re: Problem with SYS$CREMBX() writing to a cluster-wide name table?3 Message-ID: <6Zp61Hv2WwQh@eisner.encompasserve.org>   ` In article <43AA11A6.DF816E3A@comcast.net>, David J Dachtera <djesys.nospam@comcast.net> writes:  F > I haven't played with "node::MBAxxxx:" for many years - I don't evenE > recall if that works across DECnet nodes, assuming suitable proxies  > exist.  H    Oridinary mailboxes were never accessable via DECnet in all the years.    I've used both.  RMD-F-DEV is the response.  K    DECnet mailboxes (akin to IP sockets) provide appropriate functionality.    ------------------------------    Date: 22 Dec 2005 07:50:01 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) = Subject: Re: Some people are willing to support their product 3 Message-ID: <moX4+KP8EkUz@eisner.encompasserve.org>   Z In article <11qkfmbpvkqic59@corp.supernews.com>, Dave Froble <davef@tsoft-inc.com> writes:  H > Would you dispute that VMS has the problem of many people not knowing > > about it, and the majority of the rest think that it's dead?  " Certainly that should be disputed.  A The main problem with VMS is application vendors not choosing it.    ------------------------------  % Date: Thu, 22 Dec 2005 10:30:52 -0500 . From: JF Mezei <jfmezei.spamnot@vaxination.ca>= Subject: Re: Some people are willing to support their product , Message-ID: <43AAC674.83D2EBC@vaxination.ca>   Larry Kilgallen wrote:C > The main problem with VMS is application vendors not choosing it.     E And the main problem with application vendors not choosing VMS is the D total lack of advertising from the owner of VMS. App vendors need to= know that the vendor firmly intends to leverage the product's F capabilities, expands its marketplace (instead of focusing on narrower and narrower niche markets).  A Anyone at HP who agrees that VMS should be thought of  as a niche E product should be sent to the secret overseas US prisons  for torture F until they agree to never again take any actions or say antyhing which would restrict VMS' potential.   ------------------------------  + Date: Thu, 22 Dec 2005 15:32:26 +0000 (UTC) 7 From: moroney@world.std.spaamtrap.com (Michael Moroney) % Subject: Re: VMS V8.2 SHOW ERROR Bug? ( Message-ID: <doegua$hop$1@pcls4.std.com>  4 David J Dachtera <djesys.nospam@comcast.net> writes:   >Guy Peleg wrote: 1 >> I *think* SHOW ERROR ignores template devices.   C >EWA0 is the first ethernet card on many systems (sometimes copper, > >sometimes fibre). I never thought of it as a template device.  G LAN devices are funny.  They are both template/clone devices and actual C physical hardware.  The way they work, xyz0: is a template, then as J users (TCPIP, DECnet or whatever) assign channels, they get clone devices K xyz1: xyz2: etc. that represent the lan protocol being used on the network  6 connected to the physical device represented by xyz0:.   ------------------------------  + Date: Thu, 22 Dec 2005 09:40:09 -0600 (CST) * From: sms@antinode.org (Steven M. Schweda)% Subject: Re: VMS V8.2 SHOW ERROR Bug? 2 Message-ID: <05122209400925_20262AFD@antinode.org>  2 From: David J Dachtera <djesys.nospam@comcast.net>  D > EWA0 is the first ethernet card on many systems (sometimes copper,? > sometimes fibre). I never thought of it as a template device.    alp $ show device /full ewa0  L Device EWA0:, device type DE500, is online, network device, error logging is'     enabled, device is a template only.  [...]   H   Apparently it's the template device for the first Ethernet interface. G (EWA1 is not the second Ethernet interface, for example, but EWB0 might : be the template device for the second Ethernet interface.)  H    Also, on my XP1000 system, EWA0 is not a card.  (EIA0 and EIB0 are on2 a card, however.  That's _a_ card, not two cards.)  H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-org     Saint Paul  MN  55105-2547    ------------------------------  + Date: Thu, 22 Dec 2005 09:49:20 -0600 (CST) * From: sms@antinode.org (Steven M. Schweda) Subject: Re: Zip question 2 Message-ID: <05122209492015_20262AFD@antinode.org>  2 From: David J Dachtera <djesys.nospam@comcast.net>  H > > Is there a version of zip on either VMS or Linux that can zip a file
 > > over 2GB?  > G > My $0.02: I don't actually recommend trying to ZIP anything that big, J > especially a saveset. Since you need to have freedisk space at least theI > size of the saveset in order to be able to restore anything from it, my E > preference would be to simply use ZIP to the do the "backup" in the  > first place. [...]  ?    BACKUP offers some (many?) features which Zip lacks, such as 8 /[NO]ALIAS, more familiar /EXCLUDE operation, and so on.  H    In some cases, the problem is not a lack of disk space on either end,4 but a desire to transmit fewer bits between systems.  ; > A large thank you, Steve, for your work on ZIP and UNZIP! 	     ===== 	     >2GB?   I    Appreciated.  (Of course, if I'd known that _you'd_ be using it, ...)  E The scary thing is that some people now appear to believe that I did  G _all_ the work, while in reality, I tend to try to fix the bugs which I ? find most annoying (which can be plenty of work all by itself).   H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-org     Saint Paul  MN  55105-2547    ------------------------------   End of INFO-VAX 2005.711 ************************