1 INFO-VAX	Mon, 05 Aug 2002	Volume 2002 : Issue 428       Contents:C Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS C Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS C Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS C Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS C Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS C Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS C Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS 0 Re: Bad shadowing and SCSI bus error interaction Re: Carl's birthday -- Yea!  Re: Carl's birthday -- Yea!  Re: Carl's birthday -- Yea! $ Clear remains of dead cluster member( Re: Clear remains of dead cluster member Re: DCL Question Dec C & Re: device full: wrong error messages?& Re: device full: wrong error messages? Re: fuser for VMS  Re: fuser for VMS % Re: Help MRU! Robot is not responding % Re: Help MRU! Robot is not responding % Re: Help MRU! Robot is not responding  Re: Hobbyist License Question  Re: Hobbyist License Question  Re: Hobbyist License Question + RE: Itanic2 - the cHumPaq spin continues... + RE: Itanic2 - the cHumPaq spin continues... & OpenVMS Job Opportunities - Boston, MA* Re: OpenVMS Job Opportunities - Boston, MA* Re: OpenVMS Job Opportunities - Boston, MA OpenVMS printers Re: OpenVMS printers Re: OpenVMS printers Re: OpenVMS printers Re: OpenVMS printers* Re: Second KLEZ from ATTBI Today - TROLL ! Re: uaf  Re: uaf  Re: uaf  Re: uaf . VMSINSTAL exiting silently after LINK warnings  F ----------------------------------------------------------------------  # Date: Sun, 04 Aug 2002 18:42:34 GMT  From: system@SendSpamHere.ORG L Subject: Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS0 Message-ID: <00A11F4D.A24B7FF3@SendSpamHere.ORG>  N In article <3D4D5E89.3000700@hp.com>, John Reagan <john.reagan@hp.com> writes: >Paul Repacholi wrote: >  >>  G >> This would be a dance with FP, the CRTL and interupt mode perchance?  >>   >  >The CRTL isn't involved...  > J >By default, the compiler will sometimes use floating divide instructions D >to perform integer division.  It is sometimes faster to convert to J >float, do the floating divide, and convert back to integer than to use a H >pure integer-only subroutine that loops around generating the result a  >few bits at a time. > I >To do this optimization, the compiler needs to use a few floating point  H                                                           ^^^^^^^^^^^^^^H >scratch registers (ie, ones that are not normally saved by the calling   ^^^^^^^^^^^^^^^^^ >standard).       F >However, certain places in OpenVMS (interrupt routines come to mind) H >only save certain user-context for better performance.  In paritcular, J >floating registers (even scratch ones) aren't saved since the assumption H >is that they are not used in those places.  In other words, the system I >can take an interrupt, process it, and return without saving any of the  8 >user's floating point registers since nobody used them.   Huh?  G First, if the calling standard were followed, the FEN (Floating ENable) G would be set to indicate the need for floating point register preserva- F tion.  An interrupt routine, typically, would not need to use floatingF registers -- as you have pointed out -- and therefore, no need to pre-G serve them.  However, it should not be able to corrupt them either.  In G the case of a process context switch, the FEN would/should indicate the H need to preserve floating context.  Even AST delivery -- which preservesG much behind the sceens -- has optional lengths dependent upon the FEN's G status and the AST delivery mechanism will conditionally preserve these 
 registers.    I >To reconcile all of this, the C compiler (and BLISS and Pascal) provide  G >an /INSTRUCTION_SET=NOFLOATING_POINT qualifier.  This qualifier tells  @ >the optimizer not to use the "use floating for integer divide" G >optimization since the generated code will run in a context where the  G >floating registers aren't available (ie, nobody is going to save them  J >before the code would be called).  Instead, it calls the OTS$DIV routine G >which is an integer-only  routine that loops around and generates the   >result a few bits at a time.  > J >At seems that components of TCP were built without using this qualifier. H >  Recent versions of GEM use the "use floating for integer" more often J >than previous versions.  The TCP folks upgraded to the latest C compiler J >which used a newer version of GEM.  This uncovered the bug where the ISR J >would use F0/F1 for integer division.  Since nobody saved/restore F0/F1,   ! F0/F1: From the calling standard:   E F0 : Floating-point function value register. In a standard call that  D      returns a floating-point result in a register, this register isD      used to return the real part of the result. In a standard call,C                                                  ^^^^^^^^^^^^^^^^^^ C      this register may be modified by the called procedure without        being saved and restored.    D F1 : Floating-point function value register. In a standard call thatH      returns a complex floating-point result in registers, this registerF      is used to return the imaginary part of the result. In a standardF                                                          ^^^^^^^^^^^^^H      call, this register may be modified by the called procedure without	      ^^^^       being saved and restored.    J >the interrupted code would see F0 and F1 become magically corrupted.  If J >you are lucky, F0 and F1 weren't being used by the code.  If you weren't $ >lucky, you are calling for the MUP.  H F0, F1 are not defined as "scratch" registers so why should they be used as scratch registers?     J >I believe the MUP is simply a recompile of the affected modules with the 4 >needed /INSTRUCTION_SET=NOFLOATING_POINT qualifier.  H Regardless, there should be better attention paid to the context whereinH these optimizations are utilized and proper programming for said contextH should be implemented.  As far as I can tell -- looking at $INTSTKDEF --I R2 through R7, PC and PS are *still* the only registers preserved by PAL- . code dispatching an interrupt service routine.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM              5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------  % Date: Mon, 05 Aug 2002 07:36:56 +0930 / From: Mark Daniel <Mark.Daniel@wasd.vsm.com.au> L Subject: Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS3 Message-ID: <aik8r6$fjc$1@fang.dsto.defence.gov.au>   . I am a little confused by the ECO description.  6 Friday it looked to me like a 5.3 release ... "V0503".  ? Looking at it again it does say "for OpenVMS V5.1 ECO Summary".   H Yet a bit further down "Corrections for COMPAQ TCP/IP Services V5.3 BFS D Images".  At the bottom of the description has a table of 5.3 image ! identifications, e.g. "V5.3-18D".    Mark Daniel wrote:7 > http://ftp.digital.com.au/pub/ecoinfo/ecoinfo/896.htm  > 9 > http://ftp.digital.com.au/pub/ecoinfo/ecoinfo/chron.htm  >  > Dirk Munk wrote: >  >> Malcolm Dunnett wrote:  >>2 >>> In article <3D4A2542.4010807@wasd.vsm.com.au>,9 >>>     Mark Daniel <Mark.Daniel@wasd.vsm.com.au> writes:  >>>  >>> F >>>> We recently had a *real significant* problem after an upgrade to J >>>> Compaq TCP/IP 5.3.  Not to labour the story too heavily, but as this I >>>> causes data corruption, a *warning* is in order and I couldn't find   >>>> an existing one on c.o.v. >>>>E >>>> If you have this version of TCP/IP installed, are using the NFS  I >>>> server and have any application that does floating point arithmetic  H >>>> :^) apply the just-released AXPVMS-TCPIP_V0503-181-4 *immediately*. >>>> >>>  >>> I >>>    This doesn't seem to be available on ftp.service.digital.com ( at  	 >>> least  >>> I can't find it ). >>>  >>H >> No, it is not there. Only a update for TCPIP 5.1 (I'm looking in the C >> VMS 7.3 patch list). Thta is a bit odd, since this MUP update !  I >> Perhaps full testing hasn't completed yet, and at the moment you will  3 >> only get this patch if you experience problems ?  >> >    ------------------------------  % Date: Mon, 05 Aug 2002 07:27:41 +0930 / From: Mark Daniel <Mark.Daniel@wasd.vsm.com.au> L Subject: Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS3 Message-ID: <aik89r$jdd$1@fang.dsto.defence.gov.au>   5 http://ftp.digital.com.au/pub/ecoinfo/ecoinfo/896.htm   7 http://ftp.digital.com.au/pub/ecoinfo/ecoinfo/chron.htm    Dirk Munk wrote: > Malcolm Dunnett wrote: > 1 >> In article <3D4A2542.4010807@wasd.vsm.com.au>, 8 >>     Mark Daniel <Mark.Daniel@wasd.vsm.com.au> writes: >> >>E >>> We recently had a *real significant* problem after an upgrade to  I >>> Compaq TCP/IP 5.3.  Not to labour the story too heavily, but as this  H >>> causes data corruption, a *warning* is in order and I couldn't find  >>> an existing one on c.o.v.  >>> D >>> If you have this version of TCP/IP installed, are using the NFS H >>> server and have any application that does floating point arithmetic K >>> :^) apply the just-released AXPVMS-TCPIP_MUP-V0503-181-4 *immediately*.  >>>  >> >>H >>    This doesn't seem to be available on ftp.service.digital.com ( at  >> least >> I can't find it ).  >> > K > No, it is not there. Only a update for TCPIP 5.1 (I'm looking in the VMS  K > 7.3 patch list). Thta is a bit odd, since this MUP update ! Perhaps full  I > testing hasn't completed yet, and at the moment you will only get this  $ > patch if you experience problems ? >    ------------------------------  % Date: Mon, 05 Aug 2002 07:47:29 +0930 / From: Mark Daniel <Mark.Daniel@wasd.vsm.com.au> L Subject: Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS3 Message-ID: <aik9f1$8jm$1@fang.dsto.defence.gov.au>   ' John Reagan has posted the in-and-outs.   G    http://groups.google.com/groups?hl=en&selm=3D4D5E89.3000700%40hp.com   I We were the site that originally posted this with the Sydney CSC.  There  K were a lot of exchanges between myself and the specialist that handled the  E call, some of which may have been in-confidence.  It was better that  I someone from HPq post an explanation (probably technically better too ;-)   D *I am still a little concerned* that it appears unrated, at least on  :    http://ftp.digital.com.au/pub/ecoinfo/ecoinfo/chron.htm  L I would have thought that anything with the potential for *generalized data E corruption* and mayhem would be rated *1*.  It cost us ten days lost  3 production plus resources diverted into chasing it.    Stuart Fuller wrote: > Mark Daniel wrote: >  > I >>We recently had a *real significant* problem after an upgrade to Compaq K >>TCP/IP 5.3.  Not to labour the story too heavily, but as this causes data L >>corruption, a *warning* is in order and I couldn't find an existing one on >>c.o.v. >>L >>If you have this version of TCP/IP installed, are using the NFS server andH >>have any application that does floating point arithmetic :^) apply the; >>just-released AXPVMS-TCPIP_MUP-V0503-181-4 *immediately*.  >>7 >>The ECO description seems to rather understate it ...  >> >>= >>>Applications which use floating point find their registers ( >>>modified when the NFS server is used. >>E >>There is an interesting story behind this.  Perhaps someone who has 5 >>first-hand knowlege might be in a position to post.  >  > K > The fact that you know of an interesting story behind this suggests that  4 > you might know some of it.  Why don't you post it? > 
 >         Stu  >    ------------------------------   Date: 5 Aug 2002 10:59:41 -0600 - From: Kilgallen@SpamCop.net (Larry Kilgallen) L Subject: Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS3 Message-ID: <iraex$Uj97Ou@eisner.encompasserve.org>   P In article <00A11FF8.BC9D2B2C@SendSpamHere.ORG>, system@SendSpamHere.ORG writes:  G >>The point is that on Alpha, the use of floating point in an interrupt K >>service or FORK routine is explicitly forbidden.  This is because the ISR L >>could be executing in a process context which does not have floating pointM >>enabled.  The ISR didn't save F0/F1 because the writer had no idea that the G >>compiler was going to use them as scratch registers in a divide.  The M >>compiler didn't know that it couldn't modify the contents of the FP scratch M >>registers.  There is no ISR linkage definition that would have told it that M >>it had to save scratch registers, unless the ISR writer had explicitly told N >>it to save the FP scratch registers in a custom linkage... and in a circularJ >>fashion that you may be detecting now, the writer didn't realize that anJ >>integer operation in C, might be done in floating point behind his back. > @ > So it's not a good idea to do VMS system programming in C?  :)  F Not exactly.  It is not a good idea to do VMS system programming usingF a compiler that defaults to using FP registers for integer divide.  IfC GEM is common between compilers, perhaps other compilers are making # the same mistake in their defaults.    ------------------------------  # Date: Mon, 05 Aug 2002 16:26:50 GMT 5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> L Subject: Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS2 Message-ID: <eHx39.26$B27.525348@news.cpqcorp.net>  * system@SendSpamHere.ORG wrote in message >  ? >So it's not a good idea to do VMS system programming in C?  :)  >   L Quite a bit of new system code is written in C.  But as *should* be the caseJ in any language, you need to know how to use the language for the task andC the environment it's in.  Inner mode, by default should be compiled B /INSTRUCTION_SET=NO_FLOAT.   If you are coding an ISR, you need toC understand what the linkage is that is needed.  If you are touching 6 hardware, you need to know what "volatile" does.  Etc.  G Macro was much more explicit, and therefor to some extent more obvious.   L I've written inner mode code in C since the VAX days (remember the VWS VT200I terminal emulator?  It was loaded into paged pool, and executed in kernel H mode at IPL 2.  So yes, I believe system code can be written in C.  It's0 only recently that it can *all* be written in C.   ------------------------------  # Date: Mon, 05 Aug 2002 17:40:55 GMT & From: John Reagan <john.reagan@hp.com>L Subject: Re: Arithmetic Data Corruption - Compaq TCP/IP Services 5.3 and NFS% Message-ID: <3D4EB6E4.1060300@hp.com>    Larry Kilgallen wrote:  H > Not exactly.  It is not a good idea to do VMS system programming usingH > a compiler that defaults to using FP registers for integer divide.  IfE > GEM is common between compilers, perhaps other compilers are making % > the same mistake in their defaults.   B The only compilers that even provide the switch are BLISS, C, and I Pascal.  The rest default to /INST=FLOAT (except for AMACRO) with no way  E to change it.  So you can't write a device driver in COBOL or BASIC.  G There is no Pascal usage in those areas of OpenVMS (AFAIK, nobody uses  D /INST=NOFLOAT for Pascal, we added it on day-1 in preparation for a H piece of software from the VAX that ended up never being ported).  That ? leaves just BLISS and C.  As Fred mentioned, if you are coding  H inner-mode code, it is upto you to read the manuals and use the correct E compiler settings, the correct header files, adding 'volatile' where   needed.   E AMACRO never uses floating to implement divide.  It always calls the  ? OTS$ routines.  So that makes AMACRO default to /INSTR=NOFLOAT.    --   John Reagan ' Compaq Pascal/{A|I}MACRO Project Leader  Hewlett-Packard Company    ------------------------------    Date: 05 Aug 2002 07:35:40 +0800, From: Paul Repacholi <prep@prep.synonet.com>9 Subject: Re: Bad shadowing and SCSI bus error interaction - Message-ID: <877kj65h9v.fsf@prep.synonet.com>   , "Bill Todd" <billtodd@metrocast.net> writes:  ) <stuff about error staus on a shadow set>   A > That logic would require that the application be informed about B > *every* recoverable error (even read re-tries at the disk) - andD > informed by an error status rather than something like a qualifiedD > success return which is easily ignorable if the application is notC > interested (an RMS write-error return - and for that matter *any*cD > RMS error return - is *not* something you should ignore under most > circumstances).C  & You mean like the `No Retry' modifier?  m > Your logic is flawed.m  F Well, one error you have to handle is EXQUOTA, and for that, you retryF to write. Only the applicaton knows if the data is critical, or can be tossed with little loss.  C It would be nice if there was some way to control and qualify this, D but in the current form, better to know than to die without warning.   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.f@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Sun, 04 Aug 2002 19:42:45 GMTU1 From: "David J. Dachtera" <djesys.nospam@fsi.net>e$ Subject: Re: Carl's birthday -- Yea!' Message-ID: <3D4D886D.15E0F234@fsi.net>n   Didier Morandi wrote:n > ( > Well, the answer is - to me - obvious: > R > The hardware which runs VMS has been produced by the same company which producedF > the operating system. This is (to me) the reason why VMS is so good.  G True. The "goodness" is largely a matter of perception, IMHO, since the G source of the o.s. has near-total control over the underlying hardware.   E The challenge I have with that is that *BSD and Linux can be and have-B been adapted to virtually any current processor not to mention theF accompanying hardware platform. While drivers for various elements areD an issue, the community-at-large has done and admirable job (IMO) of "filling in the gaps".  ? So, if VMS were exposed to the same environment, true, the OVMS.E marketing machine have to get up out of its grave and enlist ISVs and F OEMs to develop drivers and such for the myriad sundy hardware that is  likely to found in such systems.  D In short, to me, it comes down to a matter of the difference between, what folks can do, versus what they will do.   Can it be done? Certainly.  , Will it be done? The jury was not convinced.  Q > When Intel starts developping an OS for their platforms, maybe we will see whato > you are looking for.  H The word "monopoly" looms large there. I don't see that happening, short of BG going "on the grab".  @ > For the moment, I prefer to believe that VMS will run on IA64.  C ...as do I. However, I condition that remark on the appearance of adD viable, ready-for-prime-time IA64 processor and supporting platform.   > So, one day, weM > will have PCs running VMS.  ? In so far as "PCs", taken generically to mean the equivalent of B current-day "billy boxes" that may eventually contain IA64 or some2 variant of 64-bit successor to 80x86, yes I agree.   -- l David J. Dachterad dba DJE Systemsc http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/s   ------------------------------  # Date: Sun, 04 Aug 2002 22:41:03 GMTr0 From: "brad.madison" <brad.madison@mail.tds.net>$ Subject: Re: Carl's birthday -- Yea!, Message-ID: <3D4DAD7F.4991C75C@mail.tds.net>   Mark Tarka wrote:o > A > Just wanted to remind the faithful, that Carl J. Lydick's birthnD > anniversary is coming up and that y'all should celebrate the event > appropriately. >i  E RIP, Carl.  You knew your stuff and shared it.  You were also a royald
 PITA to some.d  A I just got flamed twice, IIRC.  Still, the next time I asked, you 	 answered.  --0 See this:  http://www.corpit.ru/cgi-bin/h0n5yp0t  G Stop relay spam by Labor Day, 2002.  It can be done and is worth doing.h   ------------------------------   Date: 5 Aug 2002 08:45:02 -0700B' From: mark_tarka@yahoo.com (Mark Tarka)I$ Subject: Re: Carl's birthday -- Yea!= Message-ID: <6b70c71c.0208050745.254f957a@posting.google.com>n  Y sasadmin <jec@nospam.net> wrote in message news:<87heie9htv.fsf@Alethion.systasis.net>...e+ > mark_tarka@yahoo.com (Mark Tarka) writes:  > C > > Just wanted to remind the faithful, that Carl J. Lydick's birthaF > > anniversary is coming up and that y'all should celebrate the event > > appropriately. > H > Noted. Without Googling, can you tell me what a kiloGavron is and what > it measures?  > For any given script, the bench-mark for speed/efficiency is a	 "Gavron".t  @ A "kiloGavron" denotes a script performing the same function but taking 1000 times longer.c  E There was a guy claiming to have written a milliGavron script (MurphyrD ?), but it turned out that he was under the influence when he posted :-)o       Mark   ------------------------------   Date: 5 Aug 2002 07:58:43 -0700v$ From: issinoho@slayme.com (issinoho)- Subject: Clear remains of dead cluster membere= Message-ID: <d0141774.0208050658.2d0d64df@posting.google.com>t  A I have recently removed (permanently) a couple of machines from a  cluster. No problems.^  E Remnants of the old machines persist in the remaining nodes though. At: SH DEV D lists the disk from the older machines, albeit inB HostUnavailable state. SH CLUS lists the older machines in BRK_NON state.  C This is not causing any problems as such, but I would like to cleart? things up for the sake of tidiness, hopefully without a cluster  reboot.   
 Any ideas?   ------------------------------  $ Date: Mon, 5 Aug 2002 11:32:38 -0400 From: norm.raphael@metso.com1 Subject: Re: Clear remains of dead cluster membere? Message-ID: <OF52E8C2C4.B25B4B0D-ON85256C0C.00553AED@metso.com>   A AFAIK, the only way to clear the cluster database of such entriesoK is to shut down the entire cluster, then reboot the nodes.  Leaving  up any D node may cause the entire cluster database to retain the stale data.  @ As you stated, the persisting information is unually no problem.        8 issinoho@slayme.com (issinoho) on 08/05/2002 10:58:43 AM  0 Please respond to issinoho@slayme.com (issinoho)   To:    Info-VAX@Mvb.Saic.Com cc:i0 Subject:    Clear remains of dead cluster member    A I have recently removed (permanently) a couple of machines from a$ cluster. No problems.h  E Remnants of the old machines persist in the remaining nodes though. As: SH DEV D lists the disk from the older machines, albeit inB HostUnavailable state. SH CLUS lists the older machines in BRK_NON state.  C This is not causing any problems as such, but I would like to clear-? things up for the sake of tidiness, hopefully without a cluster  reboot.k  
 Any ideas?   ------------------------------   Date: 5 Aug 2002 07:44:38 -0700o, From: evdnep@hotmail.com (Het Kritische Oog) Subject: Re: DCL Questione< Message-ID: <6e6e4dbd.0208050644.3f3c14a@posting.google.com>  | Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> wrote in message news:<01KKEQ5MRZXU9870E1@sysdev.deutsche-boerse.com>...R > > > When you use $On Error Then Continue, and an error happens then the on error > > > setting has gone!sI > > > Everty time an exception happens you have to redefine your on errora > > > definition!  > > > < > > > A nice, quick solution? Begin your DCL procedure with:# > > > $RESET := ON ERROR THEN RESET/ > > > $RESET > > B > > Which destroys $STATUS.  The $ ON ERROR THEN CONTINUE doesn't. > > = > > If your goal is to avoid the default $ ON ERROR THEN EXITw. > > behavior, the clean solution is $ SET NOON > K > Right.  The definition of RESET is a clever hack, but as you say for the 5K > ERROR case, SET NOON (the default being ON ERROR THEN EXIT) is cleaner.  u; > Is there a case where the RESET construction makes sense?    Yes,  9 with MMS the procedure stops at the first compiler error. N I make a DCL file with MMS/NOACTION and add the reset statements at the front.% In this way I get all errors at once.p   XXXXXs   ------------------------------  % Date: Mon, 05 Aug 2002 20:13:55 +0200 & From: Oren Nahum <orenna@zahav.net.il> Subject: Dec C, Message-ID: <aimbo3$7oj$1@news.inter.net.il>   Hi,   D Does anybody know how I can get the size of a certian file in bytes  using DEC C.   Thanks,l Oren   ------------------------------    Date: 05 Aug 2002 07:45:46 +0800, From: Paul Repacholi <prep@prep.synonet.com>/ Subject: Re: device full: wrong error messages?o- Message-ID: <873ctu5gt1.fsf@prep.synonet.com>-  " John Santos <JOHN@egh.com> writes:  C > I mean the pool of free clusters, as listed in BITMAP.SYS (and asAD > cached in memory by VMS so it doesn't have to go out to bitmap.sysF > everytime you allocate a file.)  I think the VMS Cluster members areB > supposed to handshake about this, so if one member wants to grab; > more space than it can get, due to other cluster member'st? > reservations, the others will release some.  Maybe there is a F > problem with this...  I.E. Free space is large enough, and the valueC > that shows in SHOW DEVICE is accurate, but the bits in BITMAP.SYS E > reflect some pre-allocated clusters that have been grabbed by other 6 > systems which are failing to release them on demand.  E That was my first thought, given that you had checked for the obvioushE error source. If an allocation exhausts free space, the XQP dances on F the volume allocation lock to get other nodes to dump their free spaceE cache. From the headers, that area has caused a fair bit of developer> bloodshed already.  eC > I have successfully created files with an allocation equal to thenB > entire free space on a disk mount /cluster (many times), so thisF > generally works right, but maybe there is an infrequent bug or a bug' > in a recent ECO that has broken this.   tB > (I could be confusing the method to allocate disk space with theC > method to allocate file headers, but I think similar methods are\-F > used to preallocate both.  Also, above, I use the word "clusters" toC > mean both "disk clusters" (i.e. sets of blocks used as a unit for ; > disk space allocation), and "VMS clusters".  Oh, well...)@   That is my memory also.    -- R< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.r@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Mon, 05 Aug 2002 12:03:43 -0500b& From: jlsue <jlsuexxxz@screaminet.com>/ Subject: Re: device full: wrong error messages?e8 Message-ID: <pmbtku822kd7prmuv78ag243lcsd6sth0d@4ax.com>  / On Sat, 03 Aug 2002 14:35:12 GMT, "Tom Simpson"t! <simpsont@attbi.com.fubar> wrote:l  M >We had a similar experience last weekend during our weekly file conversions.   >No explanation so far from CSC. >i >The process went like this:8 >1. convert large RMS file ==> to empty drive, no errorsA >2. delete original file (delete/log),  delete status was successP2 >3. copy the converted file back to original drive6 >4. the copy failed because the output drive was full. > C >The converted file is smaller than the original file, so how couldv >the drive be full?a >dJ >No users are on the system while this process occurs.  No Batch processes >running$ >other than the file conversion job.J >The command procedure does a stop/id on any interactive users to be sure. >rF >So either the delete, really didn't or the drive was not really full.C >An analyze/disk/repair on the disk in question indicted no errors.> >> >My $.02 >n  F This is difficult to trouble-shoot without the exact text of the error, message (multiple lines of it, most likely).  B A "disk full"-like message can be generated from different places, e.g.,   ) 	a.	file header full - indexf.sys relatedo, 	b.	very fragmented disk - the copy may fail 		due to too many fragments   C There may be other situations as well, I'm going from my own memory F here on what I've seen over the years.  The exact message is importantB because the interpretation may be that the drive was full, but theA actual message may give more detail as to why it's reporting thatG failure.  ) Not speaking for anyone, certainly not HPo- (get rid of the xxxz in my address to e-mail)    ------------------------------   Date: 5 Aug 2002 01:29:01 -0700e- From: bobmarlow@postmaster.co.uk (Bob Marlow)h Subject: Re: fuser for VMS< Message-ID: <c5d3d5e1.0208050029.21ca39d@posting.google.com>  = I don't know of a VAX fuser, but if you use the output from:-n   $ sho dev/fil disk_namer  @ you could write DCL to do the same thing, would take half a day?   Bobi    N rob.buxton@wcc.spam.govt.nz wrote in message news:<3d49a4b0.249505580@news>... > Hi Chaps,h > 9 > We have a compiled version of fuser running on our VAX. @ > This utility can either list or kill the processes that have aE > particular file open. we use it to remove any users still logged ine  > when the overnight jobs start. > F > It may be that someone here, many moons ago, got the unix source andG > made the changes. Trouble is they're now all long gone and I'm trying  > to get stuff moved to Alpha. > B > Does anyone know if a port of this utility was done and if it is > available anywhere?m >  > TIA  >  > Rob.   ------------------------------  % Date: Mon, 05 Aug 2002 19:06:00 +0010 % From: paddy.o'brien@zzz.tg.nsw.gov.au  Subject: Re: fuser for VMS5 Message-ID: <01KKY6VYHQ82000WXX@tgmail.tg.nsw.gov.au>m  - rob.buxton@wcc.spam.govt.nz wrote in message p! news:<3d49a4b0.249505580@news>...s > Hi Chaps,9 > 9 > We have a compiled version of fuser running on our VAX.t@ > This utility can either list or kill the processes that have aE > particular file open. we use it to remove any users still logged in   > when the overnight jobs start. > F > It may be that someone here, many moons ago, got the unix source andG > made the changes. Trouble is they're now all long gone and I'm tryingC > to get stuff moved to Alpha. > B > Does anyone know if a port of this utility was done and if it is > available anywhere?  >a   Rob,  ) I would not have expected this of you :-)   O For some reason, there have been so many threads recently about killing users. wM It's probably in the FAQ, but if not there are procedures on the FREEWARE CD l and at Hunter's site.e  M The only time I have to do this: $stop/id=?????? is when my users are logged tM in via Reflections and Bill decides to terminate them.  I am often left with i dead processes.-   Regards, Paddy   ------------------------------  * Date: Mon, 5 Aug 2002 15:14:55 +0000 (UTC), From: lewis@mazda.mitre.org (Keith A. Lewis). Subject: Re: Help MRU! Robot is not responding. Message-ID: <aim4pf$6bc$1@newslocal.mitre.org>   "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> writes in article <YEv39.16$e07.416792@news.cpqcorp.net> dated Mon, 05 Aug 2002 14:07:52 GMT:vL >Unless something changed in the recent past, I believe that the only thingsK >the SCSI configuration code will configure is disk and tape class devices.oA >I believe anything that uses GK needs to be manually configured.N >FJ >Of course, it's been a couple years since I looked in that code.  Perhaps >something changed.L  L You are correct.  I added the IO CONNECT command so long ago that I forgot I8 did it.  I was wrong when I said it would autoconfigure.    >Steven Xie wrote in message ...F >>I did IO AUTO things, but it didn't give me back anything. I have to
 >>manually doF >>= >>MCR SYSMAN IO CONNECT GKA501:/DRIVER=SYS$GKDRIVER/NOADAPTER  >> >>Is that created the problem?  K You're telling it that the robot is using SCSI ID 5 with a something (LUN?)eJ of 1.  Since you probably have only the robot and the DLT on that bus, youH could use different SCSI IDs for them.  The important thing is that bothK devices are connected physically and that the ID set for the library at the G console (by that I mean the little console on the front of the library)+A matches the IO CONNECT.  And make sure it's set to "random" mode.d  + --Keith Lewis              klewis$mitre.orgs> The above may not (yet) represent the opinions of my employer.   ------------------------------   Date: 5 Aug 2002 09:11:36 -0700t% From: s.joseph@lph.org (Shawn Joseph)h. Subject: Re: Help MRU! Robot is not responding= Message-ID: <c7ea9784.0208050811.3a174dd2@posting.google.com>d  F I had this same problem for the longest time.  I found this message on@ this same news group and it fixed my problem.  Please review theB message below and let me know if this corrects your problem too...  D You probably fell into the same pit I did.  I just installed a TL892E and they way I interpretted the manual section "DLT Loaders" was that C tape drive one was SCSI id 4, tape drive two is SCSI id 5, and thena@ the robot should be the last id plus 01.  That led me to believe5 GKB501 should be the robot id.  I got the same error.d  B It wasn't until I read the next section of the robot manual "TL8nnD Automated Libraries" that it was made clear what the id of the robotC should be.  In my case it was zero, the factory default, which made)? the proper robot device GKB0.  Once I used that it worked fine.   " Shawn Joseph, Systems Adminstrator    \ "Steven Xie" <r33300@email.mot.com> wrote in message news:<aicmqr$t1k$1@newshost.mot.com>... > Hello there, > J > I have one OpenVMS box (ES40, Vms 7.2-1, Media Robot Utility V1.4-1). OnL > this system I have one DLT891 tape library attached and I'm try to use MRUL > to control the tapes. I have created the device GKA501 (The tape device isG > MKA500), However it doesn't work. The error message shows like below.d > F > ROBOT gka501: is not responding: Robot illegal request. SENSE KEY:5, > ASC:25,ASCQ:0. > M > I have also created a device named gka500, this one shows "device offline".l > / > Dose anybody there has the experience before?  > 	 > Thanks,a > Steven   ------------------------------    Date: 05 Aug 2002 23:09:35 +0800, From: Paul Repacholi <prep@prep.synonet.com>. Subject: Re: Help MRU! Robot is not responding- Message-ID: <871y9dcpg0.fsf@prep.synonet.com>   + "Steven Xie" <r33300@email.mot.com> writes:e  F > I did IO AUTO things, but it didn't give me back anything. I have to
 > manually doo  f= > MCR SYSMAN IO CONNECT GKA501:/DRIVER=SYS$GKDRIVER/NOADAPTERk  n > Is that created the problem?  < The GKDRIVER does not load from an AUTOCONFIGURE, it must be manually loaded.   -- d< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  * Date: Mon, 5 Aug 2002 06:53:08 -0500 (CDT)2 From: Kevin Monceaux <OwnedByDogs@ClearSource.net>& Subject: Re: Hobbyist License QuestionI Message-ID: <Pine.LNX.4.30.0208050651170.13288-100000@Linux.monceaux.com>n  J I found the solution to my own question.  After sending the e-mail I triedG installing other layered products which I had installed before and knew G they worked with the hobbyist license.  There license PAK's were listedeJ but the installations failed saying there was no license.  Rebooting seemsF to have solved the problem.  I found a solution but am not sure of the cause.   Keviny  ) On Sun, 4 Aug 2002, Kevin Monceaux wrote:A   >aE > I'm trying to explore the various layered products available to the-L > OpenVMS hobbyist.  There was a license included in my hobbyist license PAKK > for FMS.  I've tried installing FMS.  The installation failed saying that-F > there wasn't a license registered for FMS.  I've listed my installedJ > license PAKs.  FMS is listed.  I'm running OpenVMS VAX version 7.2.  TheL > layered products set I have is from June 1999(found on eBay).  What is the > FMS license for? >u >t >  > Kevind >.   ------------------------------  % Date: Mon, 05 Aug 2002 14:27:12 +0200 9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>s& Subject: Re: Hobbyist License Question' Message-ID: <3D4E6F20.1A684B7B@aaa.com>I  8 A common error in licensing is now seeing the difference: between "REGISTRED" and "LOADED". A license must be LOADED to be able to be used.  ; Note the the DCL scripts with hobbyist licences from Montag.9 "only" have the LICENSE REGISTER command. You have to addo; a LICENSE LOAD command for each license you actualy want toa
 be active.  > Now, I'm not 100% sure why a reboot solved your problem. Maybe: there are some LICENSE LOAD command in your boot scripts ?   Jan-Erik Sderholm.p   Kevin Monceaux wrote:l > L > I found the solution to my own question.  After sending the e-mail I triedI > installing other layered products which I had installed before and knewhI > they worked with the hobbyist license.  There license PAK's were listedsL > but the installations failed saying there was no license.  Rebooting seemsH > to have solved the problem.  I found a solution but am not sure of the > cause. >  > Kevin  > + > On Sun, 4 Aug 2002, Kevin Monceaux wrote:  >  > > G > > I'm trying to explore the various layered products available to the N > > OpenVMS hobbyist.  There was a license included in my hobbyist license PAKM > > for FMS.  I've tried installing FMS.  The installation failed saying thatlH > > there wasn't a license registered for FMS.  I've listed my installedL > > license PAKs.  FMS is listed.  I'm running OpenVMS VAX version 7.2.  TheN > > layered products set I have is from June 1999(found on eBay).  What is the > > FMS license for? > >m > >e > >B	 > > Kevint > >    ------------------------------  % Date: Mon, 05 Aug 2002 08:37:38 -0400s2 From: rdeininger@mindspring.com (Robert Deininger)& Subject: Re: Hobbyist License QuestionK Message-ID: <rdeininger-0508020837380001@1cust162.tnt2.nashua.nh.da.uu.net>r  0 In article <3D4E6F20.1A684B7B@aaa.com>, Jan-Erik1 =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> wrote:O  9 >A common error in licensing is now seeing the differencer; >between "REGISTRED" and "LOADED". A license must be LOADEDn >to be able to be used.  > < >Note the the DCL scripts with hobbyist licences from Montag: >"only" have the LICENSE REGISTER command. You have to add< >a LICENSE LOAD command for each license you actualy want to >be active.n >)? >Now, I'm not 100% sure why a reboot solved your problem. Maybeo; >there are some LICENSE LOAD command in your boot scripts ?   B By default, something in the startup procedures LOADs all eligible5 licenses.  So yes, rebooting would solve the problem.r   So would the simple command  $ LICENSE LOAD *   ------------------------------  * Date: Mon, 5 Aug 2002 11:09:36 +0000 (UTC) From: david20@alpha2.mdx.ac.uk4 Subject: RE: Itanic2 - the cHumPaq spin continues...+ Message-ID: <ailmdg$juj$2@aquila.mdx.ac.uk>h  W In article <1020805011039.415A-100000@Ives.egh.com>, John Santos <JOHN@egh.com> writes:e! >On 29 Jul 2002, Rob Young wrote:- >- >> In article <BE56C50EA024184DAF48F0B9A47F5CF4026608A2@kaoexc01.americas.cpqcorp.net>, "Main, Kerry" <Kerry.Main@hp.com> writes: 
 >> > Fred, >> > bE >Or is the part of Microsoft that wants to rule the world at war withdB >the part of Microsoft that wants to torment and destroy anyone orB >anything that it deals with?  I vaguely recall a fable about some@ >ferocious creature (a tiger or dragon or something), a boat andC >another creature who was the only one who could keep the boat fromrB >sinking.  At the end, as he drowns, the tiger says "What else did$ >you expect me to do?  I'm a tiger." >   M The version as related on Star Trek Voyager when they do a deal with the BorgeF to cross Borg space (in return for helping the Borg fight an even moreN maleovalent species) involved a Scorpion getting a lift from another creature O (a fox - I can't remember) across a river. The scorpion says it won't sting the5K fox since otherwise they would both drowned. It then proceeds to sting the uH fox halfway across. As they both drown the fox asks why and the scorpion replies "It's my nature".s  
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  $ Date: Mon, 5 Aug 2002 11:25:21 -0400* From: WILLIAM WEBB <WWEBB1@email.usps.gov>4 Subject: RE: Itanic2 - the cHumPaq spin continues...- Message-ID: <0033000075347717000002L072*@MHS>D  , =0AI thought this was one of Aesop's fables.! But it seems that I was mistaken.   2 http://198.64.129.160/critters/malice/scorpion.htm   WWWebb   -----Original Message-----/ From: Info-VAX-Request@Mvb.Saic.Com at INTERNET+% Sent: Monday, August 05, 2002 7:29 AMyB To: Webb, William W Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNET4 Subject: RE: Itanic2 - the cHumPaq spin continues...    H In article <1020805011039.415A-100000@Ives.egh.com>, John Santos <JOHN@= egh.com>) writes: >On 29 Jul 2002, Rob Young wrote:o >'
 >> In articletH <BE56C50EA024184DAF48F0B9A47F5CF4026608A2@kaoexc01.americas.cpqcorp.net=	 >, "Main,e- Kerry" <Kerry.Main@hp.com> writes: >> > Fred,  >> >E >Or is the part of Microsoft that wants to rule the world at war witheB >the part of Microsoft that wants to torment and destroy anyone orB >anything that it deals with?  I vaguely recall a fable about some@ >ferocious creature (a tiger or dragon or something), a boat andC >another creature who was the only one who could keep the boat frombB >sinking.  At the end, as he drowns, the tiger says "What else did$ >you expect me to do?  I'm a tiger." >   H The version as related on Star Trek Voyager when they do a deal with th= e BorgF to cross Borg space (in return for helping the Borg fight an even moreH maleovalent species) involved a Scorpion getting a lift from another cr= eatureH (a fox - I can't remember) across a river. The scorpion says it won't s= ting theH fox since otherwise they would both drowned. It then proceeds to sting = the,H fox halfway across. As they both drown the fox asks why and the scorpio= nb replies "It's my nature".h  
 David Webb VMS and Unix team leader CCSS Middlesex University=d   ------------------------------  $ Date: Mon, 5 Aug 2002 15:28:28 +00002 From: John Eisenschmidt <jweisen@eisenschmidt.org>/ Subject: OpenVMS Job Opportunities - Boston, MAd4 Message-ID: <20020805152828.B19493@eisenschmidt.org>  I was contacted again by Donna Thomas from Roxtel. Apparently her client has decided to move of these contracting positions to Boston, MA where this work is going on. She asked me to report this, although we haven't reached an agreement on a finder's fee yet. =) s  L Please contact her directly at donna.thomas@roxtel.com if you're interested.  P --------------------------------------------------------------------------------   The task requires the skillsets listed below. All persons would be needed full-time for 6-9 months, with strong potential for add-on work, if talent provided is suberb.  Application is in Interstate Transportation Systems (ITS). Location is in Boston, MA.  Work will be on site.  We expect that there will be travel between Boston and Columbia, MD work locations.  No relocation expenses will be reimbursed.  These needs are critical skillsets, so background must match requirements completely.  For immediate co{ nsideration, please forward resume with salary requirements to Donna Thomas at donna.thomas@roxtel.com as soon as possible.   % ODBC server design and implementatione/ - OpenAccess and OpenRDA  experience preferableox - specifically, expand existing ODBC interface between existing custom application and existing custom non-ODBC database - specifically, implementing cross-platform ODBC bridge between existing custom application on OpenVMS and third-party ODBC DB (Sybase) on Windows platform   % OpenVMS programmers (need 6 of these)z - C language requireds - FORTRAN desiredt
 - C++ desired g - specifically, experience upgrading layered products, network, and custom apps from OpenVMS 6.x to 7.xi     --  / John W. Eisenschmidt <jweisen@eisenschmidt.org>e6  Homepage URL    | http://www.eisenschmidt.org/jweisenM  GPG Public Key  | http://www.eisenschmidt.org/jweisen/misc/jeisenschmidt.asc D  GPG Fingerprint | 5F9B F916 5AD1 3295 CF99 BC1E 1F97 E6A3 37E3 BEF2  O This mail is an attachment? Read http://www.jensbenecke.de/misc/outlook.en.htmlx  I "In an Action Mode of 'No Action' You Must Select an Action to Perform." sF                          -- Lawson.Insight Desktop (LID) Error Message   ------------------------------  $ Date: Mon, 5 Aug 2002 16:46:02 +00002 From: John Eisenschmidt <jweisen@eisenschmidt.org>3 Subject: Re: OpenVMS Job Opportunities - Boston, MA04 Message-ID: <20020805164602.E19493@eisenschmidt.org>   Could I have spelled anything else wrong in this? I have been interrupted non stop today. Three calls to GSC, four emails, and it isn't even 1pm yet.n   How are things with you?  S Unless the Voices are Mistaken, John Eisenschmidt (jweisen@eisenschmidt.org) Wrote:u	> I was contacted again by Donna Thomas from Roxtel. Apparently her client has decided to move of these contracting positions to Boston, MA where this work is going on. She asked me to report this, although we haven't reached an agreement on a finder's fee yet. =) s > N > Please contact her directly at donna.thomas@roxtel.com if you're interested. > R > -------------------------------------------------------------------------------- >  > The task requires the skillsets listed below. All persons would be needed full-time for 6-9 months, with strong potential for add-on work, if talent provided is suberb.  Application is in Interstate Transportation Systems (ITS). Location is in Boston, MA.  Work will be on site.  We expect that there will be travel between Boston and Columbia, MD work locations.  No relocation expenses will be reimbursed.  These needs are critical skillsets, so background must match requirements completely.  For immediate } consideration, please forward resume with salary requirements to Donna Thomas at donna.thomas@roxtel.com as soon as possible.e > ' > ODBC server design and implementationh1 > - OpenAccess and OpenRDA  experience preferablesz > - specifically, expand existing ODBC interface between existing custom application and existing custom non-ODBC database > - specifically, implementing cross-platform ODBC bridge between existing custom application on OpenVMS and third-party ODBC DB (Sybase) on Windows platforma > ' > OpenVMS programmers (need 6 of these)o > - C language requiredy > - FORTRAN desiredu > - C++ desiredwi > - specifically, experience upgrading layered products, network, and custom apps from OpenVMS 6.x to 7.x  >  >  > --  1 > John W. Eisenschmidt <jweisen@eisenschmidt.org>r8 >  Homepage URL    | http://www.eisenschmidt.org/jweisenO >  GPG Public Key  | http://www.eisenschmidt.org/jweisen/misc/jeisenschmidt.asc-F >  GPG Fingerprint | 5F9B F916 5AD1 3295 CF99 BC1E 1F97 E6A3 37E3 BEF2 > Q > This mail is an attachment? Read http://www.jensbenecke.de/misc/outlook.en.htmlg > K > "In an Action Mode of 'No Action' You Must Select an Action to Perform." @H >                          -- Lawson.Insight Desktop (LID) Error Message   -- M/ John W. Eisenschmidt <jweisen@eisenschmidt.org> 6  Homepage URL    | http://www.eisenschmidt.org/jweisenM  GPG Public Key  | http://www.eisenschmidt.org/jweisen/misc/jeisenschmidt.ascoD  GPG Fingerprint | 5F9B F916 5AD1 3295 CF99 BC1E 1F97 E6A3 37E3 BEF2  O This mail is an attachment? Read http://www.jensbenecke.de/misc/outlook.en.htmlm  I "In an Action Mode of 'No Action' You Must Select an Action to Perform." aF                          -- Lawson.Insight Desktop (LID) Error Message   ------------------------------  * Date: Mon, 5 Aug 2002 10:25:08 -0700 (PDT). From: Fabio Cardoso <fabiopenvms@yahoo.com.br>3 Subject: Re: OpenVMS Job Opportunities - Boston, MAI@ Message-ID: <20020805172508.91122.qmail@web20204.mail.yahoo.com>   With me ? :-)   & Thanks fot the people who sent me some+ job emails to USA.... but I am intersted to * go to Europe ! USA... may be in the future& after the end of the Osama "stress"...  ) The brazilians dont have a common ethinice/ caracteristic. We are black, white, portuguese,a* german, italian, ukranian, polish, etc ...* My roots are portuguese, so I look like a  jewish/arab ! :-)A   Regardse   FC e  0 --- John Eisenschmidt <jweisen@eisenschmidt.org> wrote:5 > Could I have spelled anything else wrong in this? Is6 > have been interrupted non stop today. Three calls to. > GSC, four emails, and it isn't even 1pm yet. >  > How are things with you? > 3 > Unless the Voices are Mistaken, John Eisenschmidt9# > (jweisen@eisenschmidt.org) Wrote:e6 > > I was contacted again by Donna Thomas from Roxtel.4 > Apparently her client has decided to move of these5 > contracting positions to Boston, MA where this work 4 > is going on. She asked me to report this, although3 > we haven't reached an agreement on a finder's feea
 > yet. =)  > > " > > Please contact her directly at/ > donna.thomas@roxtel.com if you're interested.s > >  > >  >eP -------------------------------------------------------------------------------- > > 5 > > The task requires the skillsets listed below. All 3 > persons would be needed full-time for 6-9 months,o2 > with strong potential for add-on work, if talent3 > provided is suberb.  Application is in Interstated6 > Transportation Systems (ITS). Location is in Boston,2 > MA.  Work will be on site.  We expect that there5 > will be travel between Boston and Columbia, MD workt, > locations.  No relocation expenses will be5 > reimbursed.  These needs are critical skillsets, so 5 > background must match requirements completely.  For  > immediate 2 > consideration, please forward resume with salary! > requirements to Donna Thomas at . > donna.thomas@roxtel.com as soon as possible. > > ) > > ODBC server design and implementationK3 > > - OpenAccess and OpenRDA  experience preferablew2 > > - specifically, expand existing ODBC interface2 > between existing custom application and existing > custom non-ODBC database4 > > - specifically, implementing cross-platform ODBC/ > bridge between existing custom application onc5 > OpenVMS and third-party ODBC DB (Sybase) on Windows-
 > platform > > ) > > OpenVMS programmers (need 6 of these)l > > - C language requirede > > - FORTRAN desiredl > > - C++ desired 0 > > - specifically, experience upgrading layered5 > products, network, and custom apps from OpenVMS 6.xe > to 7.x > >  > >  > > --  3 > > John W. Eisenschmidt <jweisen@eisenschmidt.org>y > >  Homepage URL    |% > http://www.eisenschmidt.org/jweisenO > >  GPG Public Key  | >a: http://www.eisenschmidt.org/jweisen/misc/jeisenschmidt.asc4 > >  GPG Fingerprint | 5F9B F916 5AD1 3295 CF99 BC1E > 1F97 E6A3 37E3 BEF2  > > $ > > This mail is an attachment? Read0 > http://www.jensbenecke.de/misc/outlook.en.html > > 5 > > "In an Action Mode of 'No Action' You Must Select  > an Action to Perform." t6 > >                          -- Lawson.Insight Desktop > (LID) Error Messagey >  > --  1 > John W. Eisenschmidt <jweisen@eisenschmidt.org>g >  Homepage URL    |% > http://www.eisenschmidt.org/jweisen@ >  GPG Public Key  | >:: http://www.eisenschmidt.org/jweisen/misc/jeisenschmidt.asc2 >  GPG Fingerprint | 5F9B F916 5AD1 3295 CF99 BC1E > 1F97 E6A3 37E3 BEF2y > " > This mail is an attachment? Read0 > http://www.jensbenecke.de/misc/outlook.en.html > 6 > "In an Action Mode of 'No Action' You Must Select an > Action to Perform." 4 >                          -- Lawson.Insight Desktop > (LID) Error Messagen     =====i ========================== Fbio dos Santos Cardoso OpenVMS System Manager Rio de Janeiro - Braziln fabiopenvms@yahoo.com.br ==========================  2 __________________________________________________ Do You Yahoo!?( Yahoo! Health - Feel better, live better http://health.yahoo.come   ------------------------------   Date: 5 Aug 2002 02:00:52 -0700 . From: PeterMoreton@hotmail.com (Peter Moreton) Subject: OpenVMS printersi< Message-ID: <c1243f29.0208050100.ecbe4d1@posting.google.com>   Hi,r  B I have to help a customer to specify some new printers for OpenVMSC systems, and looking on the Digital, Compaq, HP website, I can only D find HP laserjet style printers. No 'proper' digital printers like I/ used to use years ago (eg: LN03, LG series etc)   A Can someone update me with the current practise regarding OpenVMS3F printing, is it not possible to buy Digital printers, do we now use HP5 as standard, how are the printer control codes setup?e  C I remember under OpenVMS 5.5 having to write my own printer controlwD libraries to be able to properly manage forms etc. Is this still the# case under current versions of VMS?O   Thanks, Peter Moretons   ------------------------------  % Date: Mon, 05 Aug 2002 19:14:03 +0010o% From: paddy.o'brien@zzz.tg.nsw.gov.aue Subject: Re: OpenVMS printerso5 Message-ID: <01KKY75YAT02000WY3@tgmail.tg.nsw.gov.au>w   Peter Moreton wrote:  C >I have to help a customer to specify some new printers for OpenVMStD >systems, and looking on the Digital, Compaq, HP website, I can onlyE >find HP laserjet style printers. No 'proper' digital printers like Ie0 >used to use years ago (eg: LN03, LG series etc) >oB >Can someone update me with the current practise regarding OpenVMSG >printing, is it not possible to buy Digital printers, do we now use HP 6 >as standard, how are the printer control codes setup? >tD >I remember under OpenVMS 5.5 having to write my own printer controlE >libraries to be able to properly manage forms etc. Is this still the $ >case under current versions of VMS?  O OpenVMS or not, all I remember is that my printer "expert" claimed that HP use .H a non-standard PS.  Ramifications, I know not, except that you might be  interested in this comment.e  E But what the heck, all industry standard (BG) is non-standard anyway.    Regards, Paddy   ------------------------------  % Date: Mon, 05 Aug 2002 10:51:50 +0100- From: Roy Omond <Roy@Omond.net>: Subject: Re: OpenVMS printersm) Message-ID: <3D4E4AB7.B41C8AA6@Omond.net>R  & paddy.o'brien@zzz.tg.nsw.gov.au wrote:   > [...snip...] >7P > OpenVMS or not, all I remember is that my printer "expert" claimed that HP useI > a non-standard PS.  Ramifications, I know not, except that you might beo > interested in this comment.i  K The most glaring non-adherence to the Postscript standard that HP LaserjetsnK used to exhibit was in their response to the control-T sent by e.g. DCPS atcI the beginning of a print job.  If the printer was currently not executing G a print job, then the standard dictated that the printer should return:nI "[ idle ]", instead of which the Laserjets would always return "[ busy ]"eF with the result that print jobs using DCPS would get stuck waiting for@ the "current job" to finish, which of course could never happen.  C HP did produce a firmware fix but I seem to recall that getting themA new firmware was, ahem, rather difficult.  With the fix in place,o> Laserjets were (are) actually quite nice in a VMS environment.  G > But what the heck, all industry standard (BG) is non-standard anyway.d   Indeed !  	 Roy Omondh Blue Bubble Ltd.   ------------------------------  # Date: Mon, 05 Aug 2002 09:53:38 GMTe. From: peter@langstoeger.at (Peter LANGSTOEGER) Subject: Re: OpenVMS printersB5 Message-ID: <CWr39.110862$cU1.2664690@news.chello.at>,  m In article <c1243f29.0208050100.ecbe4d1@posting.google.com>, PeterMoreton@hotmail.com (Peter Moreton) writes:lC >I have to help a customer to specify some new printers for OpenVMSrD >systems, and looking on the Digital, Compaq, HP website, I can onlyE >find HP laserjet style printers. No 'proper' digital printers like It0 >used to use years ago (eg: LN03, LG series etc)   What surprise ;-)o  L DEQ spun the printer business into GENICOM which might be no longer existingH now (eg. DCPS developers are back to HPQ) but their website is still up.   	http://www.genicom.com/  9 I bet, you find more printers there than you expected ;-)B  B >Can someone update me with the current practise regarding OpenVMSG >printing, is it not possible to buy Digital printers, do we now use HPe6 >as standard, how are the printer control codes setup?  L Current practise seems to be using DCPS (from HPQ) and using a (PS!) printerL which is supported by DCPS (or can be made some kind of working with DCPS byL modifying PS code in DCPS's .TLB as an unsupported printer) and is connectedJ via the network (TCPIP and/or LAT and in now rare cases also Appletalk) orM in rare cases another bidirectional (!) communication link (parallel/serial).   M And, yes, a lot of HP Laserjet printers are supported by DCPS. There are much H more of course (refer to http://www.compaq.com/info/SP4415/SP4415PF.PDF)F from other companies and recently DCPS V2.1 added even some more like:   	GENICOM Intelliprint cL160e 	GENICOM Intelliprint mL402  	GENICOM Intelliprint mL450 
 	GENICOM LN45e 	HP LaserJet 2200g 	HP LaserJet 4100. 	HP LaserJet 9000T 	IBM Infoprint 21i 	IBM InfoPrint 32i 	IBM InfoPrint 40 
 	Lexmark C7200
 	Lexmark C9101
 	Lexmark T520c
 	Lexmark T522s
 	Lexmark T620o
 	Lexmark T622 
 	Lexmark W820b 	Xerox DocuPrint N24 	Xerox DocuPrint N32 	Xerox DocuPrint N40 	...  D >I remember under OpenVMS 5.5 having to write my own printer controlE >libraries to be able to properly manage forms etc. Is this still the $ >case under current versions of VMS?  B No. DCPS does it's job very well and is now licensed with OpenVMS.C But you are still able to create your own printer control librariessC if you like/must or insist on using an printer unsupported by DCPS.t   HIH    --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialiste E-mail  peter@langstoeger.atP A-1030 VIENNA  AUSTRIA              I'm looking for (a) Network _and_ VMS Job(s)   ------------------------------  % Date: Mon, 05 Aug 2002 12:05:09 +0200i9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>n Subject: Re: OpenVMS printersg' Message-ID: <3D4E4DD5.AB52BBB6@aaa.com>a  & A printer is a printer is a printer...  ' Why look for a specific "VMS printer" ? / Check what your custumer already have in use byp/ there PC users or others, and try to use those.t  7 Saying that VMS must have it's *own* printers will onlye7 be yet another reason to look for a OS replacement thatd' can use "our normal/standard printers".-  % Some other interating facts could be,A# - Are the VMS environment growing ?s@ - Or are "old" printer replaced ? And what model are they then ?; - Do they have some special apps with hardcoded support forn   DEC-type printers ?o> - Or you might be looking for high-volume printing solutions ?   YMMV, of course...   Jan-Erik Sderholm.i  > > In article <c1243f29.0208050100.ecbe4d1@posting.google.com>,2 > PeterMoreton@hotmail.com (Peter Moreton) writes:E > >I have to help a customer to specify some new printers for OpenVMSoF > >systems, and looking on the Digital, Compaq, HP website, I can onlyG > >find HP laserjet style printers. No 'proper' digital printers like I 2 > >used to use years ago (eg: LN03, LG series etc)   ------------------------------  # Date: Sun, 04 Aug 2002 18:48:58 GMTt> From: andekl_no@saaf_spam.se (=?ISO-8859-1?Q?Anders_Ekl=F6f?=)3 Subject: Re: Second KLEZ from ATTBI Today - TROLL !l; Message-ID: <1fgerua.1qbru98usfy8cN%andekl_no@saaf_spam.se>t  0 Terry C. Shannon <terryshannon@attbi.com> wrote:  M > Thing of it is, ATTBI.COM (who I would recommend no nonody) uses BRIGHTMAILsL > antispam and antivrus filtering. They (ATTBI) are completely unresponsive,N > and the Klez attacks are increasing daily. ATTBI refuses to explain why, butI > if you've very has the misfortune of talking to their tech support, youa > might know why).  D Your Subject lines are very misleading. You don't get the virus from0 ATTBI - you are just blaming the delivery boy ! @ Has ATTBI actually *promised* to filter your mail from viruses ?  gK > This KLEZ stiuff must go. Wouldn't be an issue of DEC negotiated a bettera > deal with Microsoft...  F If you stop using Outlook Express and/or Outlook (don't even *install*C Outlook!), turn off the "Use Microsoft's Viewer" option in Eudora'srG settings, and install a decent antivirus program you'll be fairly safe. A If that's not enough, get a Mac or a VMS box or  switch to Linux.|  F Whare I work - and 50% of the users are still :-) using Windoze - KlezG is virtually a non issue - and always was, except for a few days before51 everyone had updated their NAV virus definitions.    -- tD * Anders Eklf        * Phone: + 46 8581 74712  * "I blame you for *D * Glimmerstigen 46    * ae at radfysdotksdotse  * the moonlit sky" *D * S-196 33 KUNGSNGEN * or andekl at saafdotse  *       ----       *D * SWEDEN              * (sorry - had to garble) *   Tasmin Archer  *   ------------------------------  % Date: Mon, 05 Aug 2002 08:21:25 +0200e- From: Didier Morandi <Didier.Morandi@Free.fr>D Subject: Re: uaf' Message-ID: <3D4E1965.E3A3E63E@Free.fr>s   Howard S Shubs wrote:g >  > ../..a > Are we clear?e  
 we are :-)   D.   ------------------------------  # Date: Sun, 04 Aug 2002 20:41:54 GMTn0 From: John Santos <john.santos@post.harvard.edu> Subject: Re: uaf= Message-ID: <MPG.17b75b9e6124c4e9896b3@news.bellatlantic.net>n  F In article <3D4D8A28.715C0013@Free.fr>, Didier.Morandi@Free.fr says... > Howard S Shubs wrote:B > > M > > Or on node B.  It could be either one, where the one on node B is "good".' > J > No. Original post sez that user on node A gets the error message. So the > offending file is on node A. > O > If it was on node B in sys$specific (i.e. [sysB.sysexe]) user on node A would-@ > not reach it. S:he could only reach [sysA.sysexe]sysuaf.dat orP > [sysA.sys$common.sysexe] which is a synonym for [sysB.sys$common.sysexe] via a > SET FILE/ENTER >  > D.  ? I think Howard is saying there is either a bad file in Node A'so= SYS$SPECIFIC: and a good one in SYS$COMMON:, or a bad file int@ SYS$COMMON: and a good one in Node B's SYS$SPECIFIC:... I agree.? (There could also be a bad file in Node A's SYS$SPECIFIC: and a @ good one in Node B's SYS$SPECIFIC: and a file of unknown quality in SYS$COMMON:.)   --   John   ------------------------------  % Date: Sun, 04 Aug 2002 19:04:16 -0400a' From: Howard S Shubs <howard@shubs.net>n Subject: Re: uaf< Message-ID: <howard-866479.19041104082002@enews.newsguy.com>  ' In article <3D4D8A28.715C0013@Free.fr>,f/  Didier Morandi <Didier.Morandi@Free.fr> wrote:i   > Howard S Shubs wrote:d > > M > > Or on node B.  It could be either one, where the one on node B is "good".C > J > No. Original post sez that user on node A gets the error message. So the > offending file is on node A. > O > If it was on node B in sys$specific (i.e. [sysB.sysexe]) user on node A wouldi@ > not reach it. S:he could only reach [sysA.sysexe]sysuaf.dat orP > [sysA.sys$common.sysexe] which is a synonym for [sysB.sys$common.sysexe] via a > SET FILE/ENTER  E You're missing my point.  Node B works, yes.  Node A doesn't.  Given h; that the SYSUAF.DAT file is in one of the following places:t   SYS$COMMON:[SYSEXE]n SYS$SPECIFIC:[SYSEXE]n< logical name SYSUAF pointing to the real file somewhere else   Then  2 SYS$SPECIFIC:[SYSEXE]SYSUAF.DAT on node A is fried andsI SYS$SPECIFIC:[SYSEXE]SYSUAF.DAT or SYS$COMMON:[SYSEXE]SYSUAF.DAT on node  	 B is fineo   or  I SYS$SPECIFIC:[SYSEXE]SYSUAF.DAT or SYS$COMMON:[SYSEXE]SYSUAF.DAT on node a
 A is fried andW2 SYS$SPECIFIC:[SYSEXE]SYSUAF.DAT on node B is fine.  I Where-ever the SYSUAF.DAT file on node B is located, that file is fine.  .H Meanwhile, the file on node A is bad.  The point is that the files MUST H be different.  You specified one particular possability, and I tried to G clarify your statement by offering a more comprehensive one.  Now I've o= written the most comprehensive statement I can on this topic.   
 Are we clear?    -- a# "Run in circles, scream and shout!"a I hope you have good backups!-* "Is this the right place for an argument?"   ------------------------------  % Date: Sun, 04 Aug 2002 19:05:42 -0400S' From: Howard S Shubs <howard@shubs.net>- Subject: Re: uaf< Message-ID: <howard-F7D6D4.19054204082002@enews.newsguy.com>  = In article <MPG.17b75b9e6124c4e9896b3@news.bellatlantic.net>, 2  John Santos <john.santos@post.harvard.edu> wrote:  A > I think Howard is saying there is either a bad file in Node A'sn? > SYS$SPECIFIC: and a good one in SYS$COMMON:, or a bad file intB > SYS$COMMON: and a good one in Node B's SYS$SPECIFIC:... I agree.A > (There could also be a bad file in Node A's SYS$SPECIFIC: and aeB > good one in Node B's SYS$SPECIFIC: and a file of unknown quality > in SYS$COMMON:.)  I Yup, that too.  Dang, after I posted my long message just now, I thought aH I'd nailed all options, then -you- come along and suggest another one!   <sigh>   --  # "Run in circles, scream and shout!"o I hope you have good backups!s* "Is this the right place for an argument?"   ------------------------------   Date: 5 Aug 2002 04:55:11 -0700e! From: soterro@yahoo.com (Soterro)o7 Subject: VMSINSTAL exiting silently after LINK warningse= Message-ID: <d5440555.0208050355.639074e8@posting.google.com>e   Hello,  F I'm trying to rebuild an installation kit for an in-house product. For? some odd reasons, the installation fails with no specific error-A message. The only problem noticed is: the kitinstal.com links the-> product, and shouts some warnings, then VMSINSTAL abandons theD installation. Is maybe the LINK returning some values (no idea) thatA makes VMSINSTAL barf? Because it never goes further down with theAE KITINSTAL.COM, after the linker messages it simply exits. I tried thei? OPTIONS D with the VMSINSTAL (as suggested in the newsgroup but F missing from the VMSINSTAL manual), but something is not right with itE as it didn't produce any extra output at all. Ok, it did say (WaitingkB for demon to record initial state... ...done) and warned about theC RELEASE_NOTES missing, but that didn't help very much. So I have no F idea why VMSINSTAL is exiting while linking. Are maybe the compilationC warnings seen as bad things? I'm working on the warnings right now,n- but I'm not very optimistic with VMSINSTAL...    OS version:  OpenVMS Alpha 7.2-1i   command in KITINSTAL.COM is: $ LINK/NODEBUG/NOTRACEBACK -2   VMI$KWD:MAINFILE_ALPHA.OLB/LIB/INCL=MAINFILE,		-S   VMI$KWD:XRTTABLE_V201_ALPHA/LIBRARY/INCLUDE=(CLIP,MRMINIT,RENDER,REVERSE_CONV),	-n'   VMI$KWD:XUAFMAESTRO_ALPHA.OPT/OPT			-i"   /EXECUTABLE=VMI$KWD:MAINFILE.EXE! the output as seen on the screen: 7 %VMSINSTAL-I-RESTORE, Restoring product save set B ... +$ %LINK-W-WRNERS, compilation warnings   in module UPDATE_FILES filec< COMPAQ$DKA200:[SYS0.SYSUPD.MAINFILE40]MAINFILE_ALPHA.OLB;407C %VMSINSTAL-E-INSFAIL, The installation of MAINFILE V4.0 has failed.y  >     Adding history entry in VMI$ROOT:[SYSUPD]VMSINSTAL.HISTORY> If I'm missing some important information, please let me know.   Thank you for your help,   Sorin Costea   ------------------------------   End of INFO-VAX 2002.428 ************************