1 INFO-VAX	Mon, 17 Oct 2005	Volume 2005 : Issue 579       Contents:: Re: Announcing the closing of the VMSGateway Experts Group: Re: Announcing the closing of the VMSGateway Experts Group Re: ES40 MODEL NUMBER F$GETSYII Re: fork() (Was: Re: Will "COM for OpenVMS" be ported to Itanium ?) ?) ?) I Re: fork() (Was: Re: Will "COM for OpenVMS" be ported to Itanium ?) ?) ?)  JOnAS for OpenVMS ' Re: pws600au vms startup console output  Re: VMS support strategy?  Re: VMS support strategy?  Re: VMS support strategy?  Re: VMS support strategy?  Re: VMS support strategy? 1 RE: Will "COM for OpenVMS" be ported to Itanium ? 1 Re: Will "COM for OpenVMS" be ported to Itanium ? 1 Re: Will "COM for OpenVMS" be ported to Itanium ? ! Re: [Fwd: File marked for delete]   F ----------------------------------------------------------------------    Date: 16 Oct 2005 12:20:41 -0700# From: "H Vlems" <hvlems@freenet.de> C Subject: Re: Announcing the closing of the VMSGateway Experts Group C Message-ID: <1129490441.227864.127450@g49g2000cwa.googlegroups.com>   6 Didier, congratulations and good luck in your new job.E Does that mean you've got to move? IIRC you were in Toulouse and this   new job is in the Vendee, right?   Hans   (au Lavandou cette semaine :-)   ------------------------------  % Date: Mon, 17 Oct 2005 06:10:01 +0200 - From: Didier Morandi <prenom.nom@freesurf.fr> C Subject: Re: Announcing the closing of the VMSGateway Experts Group 4 Message-ID: <43532419$0$28138$626a14ce@news.free.fr>   H Vlems wrote:8 > Didier, congratulations and good luck in your new job.G > Does that mean you've got to move? IIRC you were in Toulouse and this " > new job is in the Vendee, right? >  > Hans >   > (au Lavandou cette semaine :-)  Q right. I drive 550 Km every week. I forgot to say that Edouard is in the College  6 this year. Antoine, his brother, is still in Toulouse.   D.   ------------------------------  # Date: Sun, 16 Oct 2005 20:05:39 GMT   From: John Santos <john@egh.com>' Subject: Re: ES40 MODEL NUMBER F$GETSYI * Message-ID: <noy4f.5275$W32.4977@trnddc06>   BRANDON, JOHN M wrote:< > Anyone out there with an ES40-2 with 4 each 833-MHz CPU's? > 4 > If so, could you please post the following result? > ) > $ WRITE SYS$OUTPUT F$GETSYI("HW_MODEL")  > 	 > Thanks!  >  >  >  > John "REBOOT" Brandon  > VMS Systems Administrator , > firstname.lastname.spam.me.not@dalsemi.com   1987  6 (Definitely an ES40 w/ 4 833MHz's.  Not sure if it's a> model 1 or model 2.  It has lots of PCI slots, if that helps.)     --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Sun, 16 Oct 2005 19:39:23 +0100 - From: John Laird <nospam@laird-towers.org.uk> R Subject: Re: fork() (Was: Re: Will "COM for OpenVMS" be ported to Itanium ?) ?) ?)8 Message-ID: <h475l1dk9v999f9prbioc0lrr165qtuq4a@4ax.com>  7 On Thu, 13 Oct 2005 13:11:02 -0400, "Stanley F. Quayle"  <squayle@insight.rr.com> wrote:   + >On 13 Oct 2005 at 9:44, Bob Koehler wrote: @ >>    But mostly people used single processor systems, where all# >>    processes see the same thing.  > D >Not true.  I had a set of processes communicating through a global C >section.  Worked perfectly on VAX, but totally messed up on Alpha  E >until I put explicit memory barrier instructions.  Single processor.  > 6 >Hoff told me I'd have to do that, and he was right...  L You didn't have unaligned data in there, did you ?  When we ported to Alpha,H we re-aligned global sections so that the compiler could generate atomicL instructions to read and write values.  If it sees unaligned data, it has toF generate multiple fetch/update sequences, between any of which anotherJ process can be scheduled.  It would probably be worse still if the sectionE was passed as an argument in which case the compiler would (iirc, and L talking Fortran, but there are common back-ends) generate instructions whichE assumed alignment, but which caused PALcode routines to kick in if at H run-time the target memory was not aligned.  (Equivalent of a Vax trap.)  ) It was described to me as "word-tearing".    --  B Star Trek Friday the 13th Part XXXVI: Jason Stalks the Enterprise.   ------------------------------  % Date: Sun, 16 Oct 2005 17:25:20 -0400 2 From: "Stanley F. Quayle" <squayle@insight.rr.com>R Subject: Re: fork() (Was: Re: Will "COM for OpenVMS" be ported to Itanium ?) ?) ?)6 Message-ID: <200510162127.j9GLR0827982@disk.stanq.com>  * On 16 Oct 2005 at 19:39, John Laird wrote:4 > You didn't have unaligned data in there, did you ?  D Almost certainly -- arrays of 2-byte items come immediately to mind.  D Also, even longword (4 byte) data is subject to problems, since the @ Itanium works in quadword.  I compiled everything with longword " granularity, but that didn't help.   ------------------------------    Date: 16 Oct 2005 10:43:45 -0700 From: thierry.uso@wanadoo.fr Subject: JOnAS for OpenVMSC Message-ID: <1129484625.416941.284590@g47g2000cwa.googlegroups.com>   G JOnAS is a J2EE applications server. It is developed under the guidance 9 of the ObjectWeb consortium (http://jonas.objectweb.org).   @ Jean-Yves Bourles and I have ported JOnAS 4.3.4 on OpenVMS. This@ release is Sun J2EE 1.4 certified. Porting means testing all theF functions, writing startup, shutdown and admin procedures and creating a PCSI kit.   G JOnAS is a free software under an LGPL license. JOnAS for OpenVMS needs E Java 1.4.1-2 or later. It MUST be installed on ODS5 volume. JOnAS has , been tested on Alpha but not yet on Itanium.  5 The PCSI kit (70 Mo) is based on the official archive 4 jonas4.3.4-jetty5.1.2.tgz and can be downloaded at :  1 http://perso.wanadoo.fr/thierry.uso/jonas-en.html   F The kit has been tested only on a few sites. Remarks and appreciations are welcome.    Some performance consideration :  @ 1. During the first startup, JOnAS deploys its applications in aD directory called work. 127 subdirectories and 1333 files are createdA during this deployment. It takes a lonnnnng time (7 minutes on my G AXP900 workstation) on OpenVMS (the usual RMS bottleneck with Java...). G This work directory is persistent. So the following startups are faster $ (1 minute on my AXP900 workstation).  8 2. The files created on-the-fly by the applications (jspG compilation...) are put in sy$scratch. I redirect sys$scratch towards a F directory called scratch under the JOnAS root directory in the startup@ procedure. If you have DECram, redirecting sys$scratch towards a+ RAMdisk greatly improves the response time.    ------------------------------  % Date: Sun, 16 Oct 2005 13:57:43 -0500 ) From: Wayne  Sewell <wayne@tachysoft.com> 0 Subject: Re: pws600au vms startup console outputB Message-ID: <1129489063.ef92f0bfb0a2a4f00a58a8f346ebde47@teranews>  K On 2005-10-16 09:19:25 -0500, Kilgallen@SpamCop.net (Larry Kilgallen) said:   E > In article <1129435170.3db6f352d7e1cbc01a5d7a499217b25f@teranews>,  - > Wayne  Sewell <wayne@tachysoft.com> writes:  > I >> Yep, turns out it's TTA0, not OPA0.  See my response to David Sneddon.  >>  @ >> The question is why you are allowed to boot from TTA0 at all. > 3 > Perhaps as insurance against the day a UART dies. B > Of course one has to hope it is an isolated incident rather than3 > a case where both UARTs die in the "wrong" order.  > F > I suppose the term UART may no longer be accurate, but the principle
 > remains.  F Yes, it's good that SRM allows you to boot even if the console serial E port dies.  The problem is that even though SRM recognizes tta0 as a  A console, VMS does not.  So you cannot see the boot sequence (vms  F starting, vms version, waiting to join cluster, now a cluster member, D etc.).  If some sort of boot error occurs, you have no idea what is G wrong.  Nor can you do a conversational boot.  Nor can you install vms  E or console firmware.  In all of these cases, without the true serial  G console you have to drag over a graphics terminal, assuming you have a  
 spare one.  + So this "feature" is of limited usefulness.   D As I said in another message, a better solution would be for SRM to E automatically assign opa0 to whichever port sends commands to it and  E make the other port tta0.  That would give full functionality in all  F scenarios, including my admittedly dumbass mistake of using the wrong . port.  Basically either port would be correct.     --   Wayne Sewell   ------------------------------    Date: 16 Oct 2005 11:08:33 -0700 From: mark_hpq@yahoo.com" Subject: Re: VMS support strategy?C Message-ID: <1129486112.965613.120950@f14g2000cwb.googlegroups.com>    .  >  > But just gripingG > here does nothing but annoy those of us trying to engage in technical  > communication.  @ I am always surprised by this kind of remarks. There are already= tons of information or garbage on the net. If you find a post / is not interesting, just avoid answering to it.    ------------------------------    Date: 16 Oct 2005 13:36:59 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen) " Subject: Re: VMS support strategy?3 Message-ID: <r+BDIWsEuz14@eisner.encompasserve.org>   ^ In article <1129486112.965613.120950@f14g2000cwb.googlegroups.com>, mark_hpq@yahoo.com writes: > .  >> >> But just griping H >> here does nothing but annoy those of us trying to engage in technical >> communication.  > B > I am always surprised by this kind of remarks. There are already? > tons of information or garbage on the net. If you find a post 1 > is not interesting, just avoid answering to it.   B Apparently you do not understand the reasoning behind the decisionC to divide posts into separate newsgroups.  Those who are interested D in both corporate politics and VMS technical details can participateE in multiple forums, just like those of us interested in VMS technical A details, the Ada language, TECO, and Motorola cellphones have the @ courtesy to put our various posts into newsgroups where they are	 on-topic.    ------------------------------    Date: 16 Oct 2005 21:22:58 -02006 From: eplan@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)" Subject: Re: VMS support strategy?, Message-ID: <4352c4b2$1@news.langstoeger.at>  c In article <r+BDIWsEuz14@eisner.encompasserve.org>, Kilgallen@SpamCop.net (Larry Kilgallen) writes: C >Apparently you do not understand the reasoning behind the decision * >to divide posts into separate newsgroups.  ; Dividing is always worse. eg. See what happened to VMSNET.*   D >                                           Those who are interestedE >in both corporate politics and VMS technical details can participate F >in multiple forums, just like those of us interested in VMS technicalB >details, the Ada language, TECO, and Motorola cellphones have theA >courtesy to put our various posts into newsgroups where they are 
 >on-topic.  J They are On-Topic in COMP.OS.VMS and they are also On-Topic in ITRC.HP.COM= No need for yet another group with a very limited audience...    --   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: Sun, 16 Oct 2005 17:45:54 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> " Subject: Re: VMS support strategy?+ Message-ID: <4352CA08.BAC1798@teksavvy.com>    Larry Kilgallen wrote:E > to divide posts into separate newsgroups.  Those who are interested F > in both corporate politics and VMS technical details can participate > in multiple forums,     E When was the last time the charter for thew newsgroup comp.os.vms was 8 posted here ?  Does it restrict it to technical issues ?    D When you have a dwindling interest in a platform, you do not wish toH split newsgroups because you then lose critical mass to make a newsgroup8 worthwhile since none of the groups have enough posters.    > The PSION newsgroups are a good example which follow VMS in an accelerated way:  F there used to be comp.sys.psion to include all discussions about psionF PDAs. Traffic grew to a point where it was decided to split it betweenH different newsgroups. A few naysayers started to see problems with PSION> (lack of marketing for one) and they/we were scalded for beingG naysayers. Then PSION decided to give up on PDAs, proving the naysayers  were right all along.   E The naysayers proved to be the most loyal are are still around in one O remaining active newsgroup where what is left of the psion community "gathers".     F Like it or not, VMS is not in its heydays. The community has dwindled.G You blame it on the naysayers. The naysayers are not to blame, they are G only pointing out the real cause of the downsizing of the VMS community H and showing how VMS's current owner could begin to fix the problems with* simple solutions which HP refuses to take.   ------------------------------    Date: 16 Oct 2005 16:07:11 -0700' From: "Pete" <peter.giza@spitbrook.com> " Subject: Re: VMS support strategy?C Message-ID: <1129504031.150711.236730@g43g2000cwa.googlegroups.com>   F I realized when I posted to this group that it is a technical forum to= discuss software engineering issues.  My quest was to find an C appropriate group to discuss the Future and Strategy for Supporting ? VMS.  However, it seems that no such forum exists outside of HP 7 corporate forums. As I state I have created a new group B http://groups.google.com/group/systems-support-vms to separate theG management and marketing issues from the core technical issues involved G in supporting VMS such as are discussed here.  If you are interested in D the marketing and management issues please join.  More importantly IC invite any and all VMS users and decision makers that have concerns E regarding the future of their VMS investment to join in.  The goal is  to make our voices heard.    -pete    ------------------------------  % Date: Sun, 16 Oct 2005 17:06:28 -0400 ' From: "Main, Kerry" <Kerry.Main@hp.com> : Subject: RE: Will "COM for OpenVMS" be ported to Itanium ?R Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB70C45B@tayexc19.americas.cpqcorp.net>   > -----Original Message-----= > From: Richard Maher [mailto:maher_rj@hotspamnotmail.com]=20   > Sent: October 15, 2005 8:16 AM > To: Info-VAX@Mvb.Saic.Com < > Subject: Re: Will "COM for OpenVMS" be ported to Itanium ? >=20	 > Hi Don,  >=204 > "Main, Kerry" <Kerry.Main@hp.com> wrote in messageH > >My understanding (hence, not official) is that since Microsoft COM isG > >ancient history i.e. legacy, (Microsoft statement), it does not make 5 > >much sense to port those legacy pieces to Itanium.  >=20H > Unfortnately, it has been my painful experience that absolutely no oneH > involved with VMS albeit DEC/Compag/HP has ever had even the slightest4 > inkling of what Microsoft is thinking/planning :-( >=20= > Although newsgroups are newsgroups, can I suggest a vist to G > microsoft.public.platformsdk.complus_mts (at the msnews.microsoft.com @ > server) for at least some examples of what people are doing=20 > with COM+ and G > MTS in general. It's a very low-noise newsgroup with a lot of helpful 	 > people.  >=20  D Richard - what I stated was correct i.e. COM is ancient history, butH Arne corrected me in that COM+ is the follow-on technology which is part of .Net.  E COM+ was never available on OpenVMS (not even Alpha), so it was not a , case of not porting COM+ to OpenVMS Itanium.    F > >Instead, OpenVMS is focussing on improved J2EE and .Net integration > >technologies. >=20@ > Oh really? And how is that MTS/DTC support manifesting itself? >=20 > Regards Richard Maher   G I have no idea, but I would also have to ask - I know its possible, but F how many Cust's are actually doing 2PC in real production environments today with Windows servers?   * [Not trying be nasty, just really curious]   [snip..]    
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  4 OpenVMS - the secure, multi-site OS that just works.   ------------------------------    Date: 16 Oct 2005 16:04:10 -0700 From: bob@instantwhip.com : Subject: Re: Will "COM for OpenVMS" be ported to Itanium ?C Message-ID: <1129503850.689444.120310@g44g2000cwa.googlegroups.com>    none if you don't do the port!   ------------------------------  % Date: Mon, 17 Oct 2005 07:32:49 +0100 3 From: "Richard Maher" <maher_rj@hotspamnotmail.com> : Subject: Re: Will "COM for OpenVMS" be ported to Itanium ?1 Message-ID: <diuo22$sgu$1@news-02.connect.com.au>   	 Hi Kerry,   H >I have no idea, but I would also have to ask - I know its possible, butG >how many Cust's are actually doing 2PC in real production environments  >today with Windows servers?  I I don't know the numbers exactly, but send me a list of all VMS customers K (especially in Oz) with contact names and I'll be happy to ring them up and  find out :-)  D But seriously, the sad truth is when a customer has an architecturalC requirement for functionality A on platform platform X and it isn't K available what do you think that customer will do? Anyway I've had a bit of J interest in hotTIP (my Transaction Internet Prorocol compliant Transaction, Manager) and I'll let you know how I get on.  B BTW. What's the difference between COM and COM+ apart from the 2PC capability?    Cheers Richard  2 "Main, Kerry" <Kerry.Main@hp.com> wrote in messageL news:FD827B33AB0D9C4E92EACEEFEE2BA2FB70C45B@tayexc19.americas.cpqcorp.net...   > -----Original Message-----: > From: Richard Maher [mailto:maher_rj@hotspamnotmail.com]  > Sent: October 15, 2005 8:16 AM > To: Info-VAX@Mvb.Saic.Com < > Subject: Re: Will "COM for OpenVMS" be ported to Itanium ? > 	 > Hi Don,  > 4 > "Main, Kerry" <Kerry.Main@hp.com> wrote in messageH > >My understanding (hence, not official) is that since Microsoft COM isG > >ancient history i.e. legacy, (Microsoft statement), it does not make 5 > >much sense to port those legacy pieces to Itanium.  > H > Unfortnately, it has been my painful experience that absolutely no oneH > involved with VMS albeit DEC/Compag/HP has ever had even the slightest4 > inkling of what Microsoft is thinking/planning :-( > = > Although newsgroups are newsgroups, can I suggest a vist to G > microsoft.public.platformsdk.complus_mts (at the msnews.microsoft.com = > server) for at least some examples of what people are doing  > with COM+ and G > MTS in general. It's a very low-noise newsgroup with a lot of helpful 	 > people.  >   D Richard - what I stated was correct i.e. COM is ancient history, butH Arne corrected me in that COM+ is the follow-on technology which is part of .Net.  E COM+ was never available on OpenVMS (not even Alpha), so it was not a , case of not porting COM+ to OpenVMS Itanium.    F > >Instead, OpenVMS is focussing on improved J2EE and .Net integration > >technologies. > @ > Oh really? And how is that MTS/DTC support manifesting itself? >  > Regards Richard Maher   G I have no idea, but I would also have to ask - I know its possible, but F how many Cust's are actually doing 2PC in real production environments today with Windows servers?   * [Not trying be nasty, just really curious]   [snip..]    
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  kerryDOTmainAThpDOTcom (remove the DOT's and AT)   4 OpenVMS - the secure, multi-site OS that just works.   ------------------------------  % Date: Mon, 17 Oct 2005 04:41:18 +0800  From: prep@prep.synonet.com * Subject: Re: [Fwd: File marked for delete]- Message-ID: <87wtkdry8h.fsf@prep.synonet.com>   ' BRAD <bradhamilton@comcast.net> writes:   D > A disk is mounted on 3 cluster nodes - 2 Vax, 1 Alpha. Versions ofE > VMS are VAX 6.2, Alpha 7.2-1. Disk is at zero free blocks due to an F > application error. I delete a file around 300000 blocks which is notB > open (or even likely to have been open for 5 years) according toA > SHOW DEV/FILE on all nodes. After the DELETE free disk space is C > still 0 blocks. I do an ANAL/DISK and it finds the file as marked : > for delete (only warning it finds on the disk). I now doF > ANAL/DISK/REPAIR/CONFIRM and respond Y to fix this error. AfterwardsE > disk is still at zero free blocks and anal/disk still shows file as E > marked for delete. Try this a couple of times with same result. Ten E > minutes later disk free space is around 300000 blocks and anal/disk 0 > no longer finds this file. What happened here?  B When a file is deleted, it is truncated back to 0 blocks first. To@ keep the disk consistant, it is done extent at a time, unless itF crosses a bit map block boundry, in which case it will treat each sideF of the boundrary as a seperate extent. This can result in thousands ofE BUFFERED IOs, not Direct IOs. The IOs are done by the XQP, not by the  process.   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------   End of INFO-VAX 2005.579 ************************