1 INFO-VAX	Tue, 29 Oct 2002	Volume 2002 : Issue 597       Contents:3 Re: 6 free RA81 disks near philly (for pickup only) 3 Re: 6 free RA81 disks near philly (for pickup only) 3 Re: 6 free RA81 disks near philly (for pickup only)  <None>
 Re: <None>
 Re: <None>
 RE: <None>
 Re: <None>
 Re: <None>
 RE: <None>% Re: ??== OVMS 7.3-1:.PCSI_INSTALL_MIN ' ??== OVMS 7.3-1:.PCSI_INSTALL_MIN hangs + Re: ??== OVMS 7.3-1:.PCSI_INSTALL_MIN hangs  Re: Cluster  Re: Cluster  Re: File Attributes  Re: File Attributes  Re: File Attributes  Re: File Attributes  Free VAX Re: Immutable laws of the PC Re: Immutable laws of the PC Re: Immutable laws of the PC Re: Immutable laws of the PC RE: Immutable laws of the PC Re: Immutable laws of the PC Re: Immutable laws of the PC Re: Immutable laws of the PC Re: Immutable laws of the PC Re: Immutable laws of the PC Re: Immutable laws of the PC RE: Immutable laws of the PC Re: Immutable laws of the PC Re: Immutable laws of the PC Re: Immutable laws of the PC Re: Immutable laws of the PC7 Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek 7 Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek 0 Re: Now where's the OpenVMS OpenOffice software? RE: OpenVMS 7.3-1 and CSWING Re: OpenVMS 7.3-1 and CSWING Re: OpenVMS 7.3-1 and CSWING Re: OpenVMS 7.3-1 and CSWING Question on XFC  Re: Question on XFC ( Re: Quorum disk appears to not be voting( Re: Quorum disk appears to not be voting Re: Rejuvenating OpenVMS? Re: Rejuvenating OpenVMS [was Re: So I went to the HP IT forum] ? Re: Rejuvenating OpenVMS [was Re: So I went to the HP IT forum] ' Re: SAP? still a market for consultants ' Re: SAP? still a market for consultants ' Re: SAP? still a market for consultants ' Re: SAP? still a market for consultants & SMTP_CFGETERRORs with TCPIP V5.3 ECO 1* Re: SMTP_CFGETERRORs with TCPIP V5.3 ECO 1* Re: SMTP_CFGETERRORs with TCPIP V5.3 ECO 1* Re: SMTP_CFGETERRORs with TCPIP V5.3 ECO 1  Re: So I went to the HP IT forum  Re: So I went to the HP IT forum  Re: So I went to the HP IT forum$ Re: TCPIP 5.3 and SMTP authorization$ Re: TCPIP 5.3 and SMTP authorization9 Re: The VMSNET group of forums (was: Re: Acrobat Reader?)   Re: Timezone-change observations Re: TSM on the Freeware disk* Unique problem involving virtual terminals3 [TCPware V5.6-2] Default Route gets lost (on Alpha) 7 Re: [TCPware V5.6-2] Default Route gets lost (on Alpha)   F ----------------------------------------------------------------------  # Date: Mon, 28 Oct 2002 14:10:23 GMT * From: "Mark E. Levy" <levy@sysman-inc.com>< Subject: Re: 6 free RA81 disks near philly (for pickup only)? Message-ID: <jzbv9.142565$zE6.492902@rwcrnsc51.ops.asp.att.net>   5 >> From what I've been told, DEC knew this but it was B > > cheaper to deal with the problem than to fix it at the source.  L That doesn't even remotely make sense. DEC *did* fix the problem once it wasJ discovered, but by that time, many hundreds of thousands of the things hadL been manufactured. Those had to be replaced, and when they were refurbished, a different sealant was used.   L It makes no sense that DEC would keep using the same bad sealant on new HDAsG while they were actively replacing failed HDAs in the field. The bad PR  alone would be quite expensive.    Mark   ------------------------------  % Date: Mon, 28 Oct 2002 13:17:46 -0700 % From: Dan O'Reilly <dano@process.com> < Subject: Re: 6 free RA81 disks near philly (for pickup only)7 Message-ID: <1035836267.1298.1.camel@ntbsod.psccos.com>   0 On Mon, 2002-10-28 at 07:10, Mark E. Levy wrote:7 > >> From what I've been told, DEC knew this but it was D > > > cheaper to deal with the problem than to fix it at the source. > N > That doesn't even remotely make sense. DEC *did* fix the problem once it wasL > discovered, but by that time, many hundreds of thousands of the things hadN > been manufactured. Those had to be replaced, and when they were refurbished, > a different sealant was used.  > N > It makes no sense that DEC would keep using the same bad sealant on new HDAsI > while they were actively replacing failed HDAs in the field. The bad PR ! > alone would be quite expensive.   F The RA81 HDA replacement program was HUGELY expensive for DEC.  At theD time they were doing this, I worked for the Advanced Technology DiskF Business at DEC in the Colorado Springs plant.  We processed literallyH thousands of the bad HDA's through the plant daily.  The did replace the bad sealants on them.     --   ------J +-------------------------------+----------------------------------------+H | Dan O'Reilly                  |  "There are 10 types of people in this | H | Principal Engineer            |   world: those who understand binary   | H | Process Software              |   and those who don't."                | H | http://www.process.com        |                                        | J +-------------------------------+----------------------------------------+   ------------------------------   Date: 29 Oct 2002 02:03:06 GMT+ From: "Joe Heimann" <heimann@ecs.umass.edu> < Subject: Re: 6 free RA81 disks near philly (for pickup only), Message-ID: <apkq8q$au0$2@odo.ecs.umass.edu>  3 Atlant Schmidt <atlantnospam@mindspring.com> wrote:  > Howard S Shubs wrote:   8 >> In article <apffb9$130n1$1@ID-135708.news.dfncis.de>,- >>  bill@cs.uofs.edu (Bill Gunshannon) wrote:  >>J >> > Actually, I have had relatively few failures and when you figure thatL >> > all of these disks bounced their way here in the back of station wagonsN >> > and trucks it goes a long way to proving the quality of computer products >> > from that era.  >>J >> Not R81 disks.  They used a sealant that broke down after about a year,H >> in my experience.  From what I've been told, DEC knew this but it wasA >> cheaper to deal with the problem than to fix it at the source.   , > You're referring to the famous "Glue" ECO?. > I thought DEC *DID* deal with it, eventually1 > replacing all the failing HDAs. Did they not do 3 > so to your liking? (Hint: We had a *LOT* of RA81s ) > in-house at DEC as well. We were *VERY*  > aware of the failure rates.)  H They did replace the HDA's for the ones with the problem glue.  But onlyF if the drive was on hardware maintenance or failed within the one yearJ warranty.  Few did fail in that one year that I heard of.  But even thoughI it was a problem with the materials used in manufacturing, it cost if the I drives failed and they were not on maintenance.  Our drives were on self- J maintenance, it ended up costing us a lot to fix them as they failed.  TheF core refund on the bad HDA's was just a fraction what it cost to get aH refurbished one.  Digital marked up one of its first big negative points1 here due to that problem and how they handled it.   3 > By the way, don't the rotary positioners in these . > drive have some wonderfully-powerful magnets2 > that you could carefully scrap out of them? (But6 > try not to *SNAP!* lose any fingers in the process.)   > Atlant  H Have one of those positioners up on a cabinet right now.  Or it might beH out of an RA82 instead.  Definitely very powerful magnets.  Used to haveF a motor or two around, 1/3 HP and the size of a washing machine motor.    Joe Heimann   ------------------------------   Date: 28 Oct 02 23:28:55 +0100) From: p_sture@elias.decus.ch (Paul Sture)  Subject: <None> ) Message-ID: <hhWHLvaKjuNt@elias.decus.ch>   G This is a difficult one, I'll admit, but I have just been asked to come E up with a figure to compare apps on VMS/Alpha versus how many MIPS on ? an IBM mainframe (apparently the IBM folks are talking MIPS, so ' that is the measure we have to go by)..   C The question is roughly "How many MIPS does your Alpha app need, so ) that we can price that up in IBM terms?".   ? I know this question is as long as a piece of string (and would > you prefer that in inches or centimetres, or with fries, Sir?)  @ I can use the VUP.COM piece of DCL to come up with some ballpark. figure, but really don't feel happy with that.   Any advice considered.   TIA.   --  
 Paul Sture Switzerland    ------------------------------    Date: 28 Oct 2002 16:21:54 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)  Subject: Re: <None> 3 Message-ID: <4EZe17M1reAU@eisner.encompasserve.org>   U In article <hhWHLvaKjuNt@elias.decus.ch>, p_sture@elias.decus.ch (Paul Sture) writes: I > This is a difficult one, I'll admit, but I have just been asked to come G > up with a figure to compare apps on VMS/Alpha versus how many MIPS on A > an IBM mainframe (apparently the IBM folks are talking MIPS, so ) > that is the measure we have to go by)..   C MIPS works with System/390 only because the machines being compared  have the same instruction set.  3 Are you saying IBM issues MIPS figures for AS/400 ?   E > The question is roughly "How many MIPS does your Alpha app need, so + > that we can price that up in IBM terms?".  > A > I know this question is as long as a piece of string (and would @ > you prefer that in inches or centimetres, or with fries, Sir?) > B > I can use the VUP.COM piece of DCL to come up with some ballpark0 > figure, but really don't feel happy with that. >  > Any advice considered.  F Ask them for the SpecInt numbers for a 1 MIPS machine and do the math.D That is horribly flawed, but no more flawed than much more expensive approaches.    ------------------------------  # Date: Mon, 28 Oct 2002 23:25:02 GMT # From: "John Smith" <a@nonymous.com>  Subject: Re: <None> I Message-ID: <iHjv9.87210$mxk1.10116@news04.bloor.is.net.cable.rogers.com>   6 "Paul Sture" <p_sture@elias.decus.ch> wrote in message# news:hhWHLvaKjuNt@elias.decus.ch... I > This is a difficult one, I'll admit, but I have just been asked to come G > up with a figure to compare apps on VMS/Alpha versus how many MIPS on A > an IBM mainframe (apparently the IBM folks are talking MIPS, so ) > that is the measure we have to go by)..  > E > The question is roughly "How many MIPS does your Alpha app need, so + > that we can price that up in IBM terms?".  > A > I know this question is as long as a piece of string (and would @ > you prefer that in inches or centimetres, or with fries, Sir?) > B > I can use the VUP.COM piece of DCL to come up with some ballpark0 > figure, but really don't feel happy with that.    H This is one of those situations where the best comparison is the logicalE unit of work and how many each can do per minute. You'd probably have K several different metrics to analyze - I/O operations, transactions against  a db, etc...  J Since each of those is also dependent on the subsystems, there can be manyK variations in how high the throughput is on each platform, depending on the F installed options. You wind up comparing apples and oranges - DB2 on aJ mainframe vs. Oracle on Alpha, a current version of MQ Series vs. an older/ version of MQ on VMS, CICS vs. ACMS, and so on.   * Is this for a new or existing application?   ------------------------------  % Date: Mon, 28 Oct 2002 19:28:20 -0500 ' From: "Main, Kerry" <Kerry.Main@hp.com>  Subject: RE: <None> T Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF402660AE6@kaoexc01.americas.cpqcorp.net>   Paul,   . Re: IBM MIPS vs. Alpha performance comparisons  = This article might be of interest: (one url will likely wrap)   4 "Performance Comparisons: Why Must They Be So Hard?"4 http://www.iseriesnetwork.com/Article.cfm?ID=3D12036  C Key point is that IBM uses different benchmarks for different class < systems and hence, in their view, are not valid cross server comparisons.   Quote :   H "Being able to generate a number does not make it valid for cross-serverH performance comparisons. I stated this several times in the article, butE others are still misusing my words to justify some possibly erroneous G performance claims. Therefore, I felt it was best to remove the article ? from this Web site before any further misinterpretations occur.   A Perhaps sometime in the future if I ever get the urge to write an G article on performance comparisons, I will remember the words of wisdom F on a plaque that hung over my desk many years ago. It said, "For everyE complex question there is a simple answer. And, that simple answer is  wrong." end quote.   Regards   
 Kerry Main Solutions Architect  Hewlett-Packard (Canada) Co.! Consulting & Integration Services  Voice: 613-592-4660  Fax   : 613-591-4477 Email: Kerry.Main@hp.com     -----Original Message-----3 From: Paul Sture [mailto:p_sture@elias.decus.ch]=20  Sent: October 28, 2002 5:29 PM To: Info-VAX@Mvb.Saic.Com  Subject: <None>     G This is a difficult one, I'll admit, but I have just been asked to come H up with a figure to compare apps on VMS/Alpha versus how many MIPS on anH IBM mainframe (apparently the IBM folks are talking MIPS, so that is the measure we have to go by)..   H The question is roughly "How many MIPS does your Alpha app need, so that$ we can price that up in IBM terms?".  C I know this question is as long as a piece of string (and would you : prefer that in inches or centimetres, or with fries, Sir?)  H I can use the VUP.COM piece of DCL to come up with some ballpark figure,& but really don't feel happy with that.   Any advice considered.   TIA.   --=20 
 Paul Sture Switzerland    ------------------------------   Date: 29 Oct 2002 02:13:19 GMT( From: bill@cs.uofs.edu (Bill Gunshannon) Subject: Re: <None> 5 Message-ID: <apkqru$2hf4v$1@ID-135708.news.dfncis.de>   ) In article <hhWHLvaKjuNt@elias.decus.ch>, , 	p_sture@elias.decus.ch (Paul Sture) writes:I > This is a difficult one, I'll admit, but I have just been asked to come G > up with a figure to compare apps on VMS/Alpha versus how many MIPS on A > an IBM mainframe (apparently the IBM folks are talking MIPS, so ) > that is the measure we have to go by)..  >   * MIPS??  Mythical Instructions Per Second??   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  % Date: Mon, 28 Oct 2002 20:52:56 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca>  Subject: Re: <None> , Message-ID: <3DBDE9F7.C8FF803F@videotron.ca>   Paul Sture wrote: E > The question is roughly "How many MIPS does your Alpha app need, so + > that we can price that up in IBM terms?".   D You need to find one commeon metric between *an* IBM system and you.  J If for instance, your Alpha system does 15 tps, and IBM has a model with aL published tps value of 25, and IBM also published MIPS numbers for that sameN model, then your could faily easily deduct how many "IBM MIPS" your Alpha has.  M Another possibility is to find some old VAX-ALPHA performance comparisons, at J which point you may be able to get a VUP number for your current alpha and> since VUPS are essentially a MIPS, you could give that to IBM.    I However, one has to be careful about the application itself. Will its IBM L equivalent perform as efficiently or more efficeintly than the VMS version ?   ------------------------------  % Date: Mon, 28 Oct 2002 18:45:14 -0800 # From: "Tom Linden" <tom@kednos.com>  Subject: RE: <None> 9 Message-ID: <CIEJLCMNHNNDLLOOGNJIIELLFPAA.tom@kednos.com>   9 FWIW, some years ago we used to figure that 2 VUP = 1 MIP   9 But if you are really going to do it right, run a typical > application on both machines and compare.  Specmarks are in my' view more hype than reality.  TPS  too.    >-----Original Message----- 5 >From: JF Mezei [mailto:jfmezei.spamnot@videotron.ca] ' >Sent: Monday, October 28, 2002 5:53 PM  >To: Info-VAX@Mvb.Saic.Com >Subject: Re: <None> >  >  >Paul Sture wrote:F >> The question is roughly "How many MIPS does your Alpha app need, so, >> that we can price that up in IBM terms?". > E >You need to find one commeon metric between *an* IBM system and you.  > K >If for instance, your Alpha system does 15 tps, and IBM has a model with a D >published tps value of 25, and IBM also published MIPS numbers for 
 >that same@ >model, then your could faily easily deduct how many "IBM MIPS"  >your Alpha has. > ? >Another possibility is to find some old VAX-ALPHA performance   >comparisons, atK >which point you may be able to get a VUP number for your current alpha and ? >since VUPS are essentially a MIPS, you could give that to IBM.  >  > J >However, one has to be careful about the application itself. Will its IBMD >equivalent perform as efficiently or more efficeintly than the VMS 
 >version ? >  >---' >Incoming mail is certified Virus Free. ; >Checked by AVG anti-virus system (http://www.grisoft.com). B >Version: 6.0.404 / Virus Database: 228 - Release Date: 10/15/2002 >  --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.404 / Virus Database: 228 - Release Date: 10/15/2002    ------------------------------   Date: 28 Oct 2002 09:49:59 GMT7 From: sy18889@rabmbit.famrp.cosm (Bradford J. Hamilton) . Subject: Re: ??== OVMS 7.3-1:.PCSI_INSTALL_MIN! Message-ID: <YJuQ94eLGxeX@rabbit>   J I have just successfully installed a minimum root on one of my disks under V7.3-1:    XXXXXX::SY18889 $ sh sys/noproc K OpenVMS V7.3-1  on node XXXXXX  28-OCT-2002 09:48:51.06  Uptime  6 20:24:00      Directory DATA3:[SYSE]  J DIA$TOOLS.DIR;1            1/25       28-OCT-2002 09:45:10.10  28-OCT-2002 09:45:10.10 J MOM$SYSTEM.DIR;1           1/25       28-OCT-2002 09:45:10.10  28-OCT-2002 09:45:10.22 J SYS$I18N.DIR;1             1/25       28-OCT-2002 09:45:10.11  28-OCT-2002 09:45:10.11 J SYS$LDR.DIR;1              1/25       28-OCT-2002 09:45:10.11  28-OCT-2002 09:45:10.11 J SYS$STARTUP.DIR;1          1/25       28-OCT-2002 09:45:10.11  28-OCT-2002 09:45:10.11 J SYSCBI.DIR;1               1/25       28-OCT-2002 09:45:10.12  28-OCT-2002 09:45:10.12 J SYSCOMMON.DIR;1            2/25       28-OCT-2002 09:44:52.81  28-OCT-2002 09:45:09.75 J SYSERR.DIR;1               1/25       28-OCT-2002 09:45:10.12  28-OCT-2002 09:45:10.12 J SYSEXE.DIR;1               1/25       28-OCT-2002 09:45:10.12  28-OCT-2002 09:45:10.12 J SYSHLP.DIR;1               1/25       28-OCT-2002 09:45:10.12  28-OCT-2002 09:45:10.12 J SYSLIB.DIR;1               1/25       28-OCT-2002 09:45:10.14  28-OCT-2002 09:45:10.14 J SYSMAINT.DIR;1             1/25       28-OCT-2002 09:45:10.14  28-OCT-2002 09:45:10.14 J SYSMGR.DIR;1               1/25       28-OCT-2002 09:45:10.14  28-OCT-2002 09:45:10.14 J SYSMSG.DIR;1               1/25       28-OCT-2002 09:45:10.14  28-OCT-2002 09:45:10.14 J SYSTEST.DIR;1              1/25       28-OCT-2002 09:45:10.14  28-OCT-2002 09:45:10.14 J SYSUPD.DIR;1               1/25       28-OCT-2002 09:45:10.15  28-OCT-2002 09:45:10.15 J TNT.DIR;1                  1/25       28-OCT-2002 09:45:10.15  28-OCT-2002 09:45:10.15     s In article <aus-18533E.15102828102002@wrzx08.rz.uni-wuerzburg.de>, "Hans M. Aus" <aus@vim.uni-wuerzburg.de> writes: & > Can't install min VMS in VMS 7.3-1.  > I > After selecting the disk, SYS$SYSTEM:AXPVMS$PCSI_INSTALL_MIN.COM hangs   > and nothing happens. > J > I can mount all disks on my Alpha and PCSI_INSTALL_MIN worked fine with  > OVMS 7.3.  > # > Have "they" changed things again?  >  > --  D > Cheers, Hans M. Aus, Wuerzburg, Germany,  aus@vim.uni-wuerzburg.de --   Bradford J. Hamilton& braMdhamAilPtoSn@aMtAtPbi.cSom		(home)& sMy1A88P89S@rabMbit.fAmPr.coSm		(work)  ; "All opinions that I express are my own, not my employer's"  "Lose the MAPS"    ------------------------------  % Date: Mon, 28 Oct 2002 18:45:48 +0100 3 From: "Aus, Hans Magnus" <aus@vim.uni-wuerzburg.de> 0 Subject: ??== OVMS 7.3-1:.PCSI_INSTALL_MIN hangsB Message-ID: <aus-D7037C.18454828102002@wrzx08.rz.uni-wuerzburg.de>  $ Can't install min VMS in VMS 7.3-1.   G After selecting the disk, SYS$SYSTEM:AXPVMS$PCSI_INSTALL_MIN.COM hangs   and nothing happens.  H I can mount all disks on my Alpha and PCSI_INSTALL_MIN worked fine with 	 OVMS 7.3.   ! Have "they" changed things again?    -    --  4 Hans Magnus Aus, Wuerzburg, aus@vim.uni-wuerzburg.de   ------------------------------  # Date: Tue, 29 Oct 2002 02:22:34 GMT , From: "Ken Randell" <ken.randell@fortel.com>4 Subject: Re: ??== OVMS 7.3-1:.PCSI_INSTALL_MIN hangs7 Message-ID: <Khmv9.29640$iV1.8581@nwrddc02.gnilink.net>   I I have used this command procedure successfully in the past week.  Simply F invoked it, gave it the name of my target disk (which I had previouslyF init'd and mounted privately), and it ran properly.  The 'min' disk it7 created works great for standalone backup and the like.   " Sorry if this is of not much help.  G Perhaps you could run it with VERIFY and see what exactly is 'hanging'?    ------------------------------  % Date: Mon, 28 Oct 2002 21:41:46 +0100 " From: "Hans Vlems" <hvlems@iae.nl> Subject: Re: Cluster5 Message-ID: <apk7es$2b3ej$1@ID-143435.news.dfncis.de>   0 "Vivek Soni" <visoni@bmc.com> schreef in bericht) news:urqdalq1d8iv0f@corp.supernews.com... 	 > Hi All,  >  > Please through some light. > I > How can I finf out the collecters and the interconnects connected to my   > VAX/Alpha system in a cluster. >  $ show cluster/contin  add circuits  I As soon as you start typing the "add circ" command you will get a command  prompt.   . And of course there is $ help show cluster :-)   Hans   ------------------------------  % Date: Tue, 29 Oct 2002 10:52:54 +0530 # From: "Vivek Soni" <visoni@bmc.com>  Subject: Re: Cluster/ Message-ID: <urs6m8odpp4j88@corp.supernews.com>    Hans,    This info is of help. Thanks.   ; Working with what else we can do see with the show/cluster.    thanks Vivek     + Hans Vlems <hvlems@iae.nl> wrote in message / news:apk7es$2b3ej$1@ID-143435.news.dfncis.de...  > 2 > "Vivek Soni" <visoni@bmc.com> schreef in bericht+ > news:urqdalq1d8iv0f@corp.supernews.com...  > > Hi All,  > >  > > Please through some light. > > K > > How can I finf out the collecters and the interconnects connected to my " > > VAX/Alpha system in a cluster. > >  > $ show cluster/contin  > add circuits > K > As soon as you start typing the "add circ" command you will get a command 	 > prompt.  > 0 > And of course there is $ help show cluster :-) >  > Hans >    ------------------------------  % Date: Mon, 28 Oct 2002 08:39:38 -0500 2 From: Atlant Schmidt <atlantnospam@mindspring.com> Subject: Re: File Attributes. Message-ID: <3DBD3E1A.BB60AE4D@mindspring.com>   Carl Perkins wrote:   P > [wrt "popen"] Have you ever considered checking these things out for yourself?  - I have no interest in doing so. The operating . system *I* currently develop upon provides all/ these things and more with no doubt about their 
 availability.    Atlant   ------------------------------  % Date: Mon, 28 Oct 2002 08:43:09 -0500 2 From: Atlant Schmidt <atlantnospam@mindspring.com> Subject: Re: File Attributes. Message-ID: <3DBD3EED.6767AF91@mindspring.com>   Carl Perkins wrote:   G > RMS can get complicated, but not in this case. All you do for this is D > define the data structures, initialize them, and then do somethingH > that loads them like a Sys$Open( fab) (followed by a Sys$Close( fab)).G > What could be easier? Certainly not spawning a subprocess and parsing  > it's text output.   1 With regard to your last statement, have you ever 3 used Perl? Between `backticks`, split(), and reqexp 1 matching, Perl makes these things about as simple  as possible.  . But as the Perl motto goes, "There's more than one way to do it!".    Atlant   ------------------------------  % Date: Mon, 28 Oct 2002 23:19:09 -0600 7 From: "Craig A. Berry" <craigberry@nospam.telocity.com>  Subject: Re: File AttributesG Message-ID: <craigberry-C912E4.23190928102002@news.directvinternet.com>   . In article <3DBD3EED.6767AF91@mindspring.com>,4  Atlant Schmidt <atlantnospam@mindspring.com> wrote:   > have you ever 5 > used Perl? Between `backticks`, split(), and reqexp 3 > matching, Perl makes these things about as simple  > as possible. > 0 > But as the Perl motto goes, "There's more than > one way to do it!".   D True.  Here are two of the ways to get the RDT on VMS; I'm not sure E spawning a subprocess is really that much easier.  (excuse wrappage).   ? $ perl -e "use POSIX qw(strftime); @s=stat('login.com'); print  4 strftime('%a %b %e %H:%M:%S %Y', localtime($s[9]));" Sat Mar 30 22:30:04 2002  $ $ perl -e "print `write sys\$output , f\$file_attributes(""login.com"",""RDT"")`;" 30-MAR-2002 22:30:04.73    ------------------------------   Date: 28 Oct 2002 23:58 CST ' From: carl@gerg.tamu.edu (Carl Perkins)  Subject: Re: File Attributes- Message-ID: <28OCT200223582204@gerg.tamu.edu>   6 Atlant Schmidt <atlantnospam@mindspring.com> writes... }Carl Perkins wrote: } Q }> [wrt "popen"] Have you ever considered checking these things out for yourself?  } . }I have no interest in doing so. The operating/ }system *I* currently develop upon provides all 0 }these things and more with no doubt about their }availability. }  }Atlant   < So you you prefer to come here and waste our time by posting, dumb-ass comments that are incorrect. Great.  9 Stop wasting our time. If you want to know something, try = spending the 2 minutes to find out for yourself. Just because : you are an ignorant twerp doesn't mean you should show off
 that fact.  + There are two things that are certain here: % 1) VMS is better than the OS you use.  2) You are an ass.   --- Carl   ------------------------------  # Date: Tue, 29 Oct 2002 00:20:52 GMT ) From: "dneprcomp" <dneprcomp@rmortho.com>  Subject: Free VAX A Message-ID: <Evkv9.1808$H_3.115205373@newssvr13.news.prodigy.com>    The following items are free: 9 2 MicroVax II model DH-63003-E2 TK70 OS VAX IVMS ver. 1,7  Terminal VT 220  Tape Backup 8 mm Plotter Bruning model ZETA 8A. Location: Los Angeles / If interested please call Yakov at 818-704-8198    ------------------------------   Date: 28 Oct 2002 19:47:20 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)% Subject: Re: Immutable laws of the PC 5 Message-ID: <apk488$29drd$1@ID-135708.news.dfncis.de>   3 In article <5cfv9.27$lT7.1196154@news.cpqcorp.net>, 8 	"Fred Kleinsorge" <kleinsorge@star.zko.dec.com> writes: > I > The 64-bit capabilities may take years for anyone to seriously exploit. " > Probably database vendors first.   I would make that:J  "The 64-bit capabilities may take weeks for anyone to seriously exploit." And:  "Probably gamers first."    bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  % Date: Mon, 28 Oct 2002 13:27:20 +0000 ' From: Andrew Harrison SUNUK Consultancy % Subject: Re: Immutable laws of the PC . Message-ID: <3DBD3B38.4070801@nospamn.sun.com>   Rob Young wrote:Z > In article <3DB97DB9.4010409@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy writes: >  >>Rob Young wrote: >>e >>>In article <20021025081223.1b64f794.mathog@caltech.edu>, David Mathog <mathog@caltech.edu> writes:  >>E >>>	AMD doesn't have to run out of cash (even though that is likely), F >>>	Opteron won't have *nearly* the TPC marks of Monticeto/Madison norA >>>	the SpecFp.  Opteron's niche is narrow compared to Itanium's.  >>>  >>9 >>What makes you think that Itanium will out TPC Opteron, 7 >>concise technical response please not too much fluff.  >> >  > 9 > 	Cache size, cache latency.   Latency and L2 effects is E > 	a tricky find.  However, as it was/is a definitive work or Richard B > 	Sites, of course a quick search is an indirect reference to it: > ^ > http://www-2.cs.cmu.edu/afs/cs.cmu.edu/academic/class/15740-f97/public/doc/21264-article.pdf >   7 This article appears to have no relevance whatsoever to 4 a discussion about why IA-64 should have a TPC-C tpm2 advantage over Opteron. Its also worth noting that1 despite the rosy glow that the article casts over 4 the 21264 processor it never had a per CPU advantage5 over HP-PA or SPARC for TPC-C TPM quite the opposite.     9 > 	Trying groups.google.com yields indirect results also:  > Y > http://groups.google.com/groups?selm=3prbid%249jr%40crl7.crl.com&oe=UTF-8&output=gplain  > + > 	But we do know better latency helps tpm.  >   2 Contains a lot of speculation about the effects of5 latency and cache size on TPC-A and TPC-B but nothing  else. Fluff   + > 	Regarding size... that is even trickier:  > A > http://research.compaq.com/wrl/projects/Database/isca99_rac.pdf  >  > 	But a better result.  > 	   > Humm, not a good example either, the RAC is neither a L3 cache8 or main memory in that it appears to be worse than both.  6 > 	Itanium 2 with 1 cycle L1, 5 cycle L2, 12 cycle L3. > A > 	Montecito?  Maybe they keep the same cycles and associativity, @ > 	don't know.  But I suspect L2 goes to 1.5 MBytes leaving 8 or> > 	so for L3... all this to say we can see from slide 3 in theG > 	isca99_rac.pdf showing a 50% increase in tpm when going from 4 MByte @ > 	to 8 Mbyte.  I do admit I may be comparing apples to oranges, > 	but I made a stab.  >   > Given that a OLTP type workload can cover a very wide spectrum< of useage spculating that 4-8 MB will show a 50% increase in0 TPC-C TPM based on this paper is pretty dubious.  5 And since this white paper was published by the Alpha 8 team at Compaq lets look at conclusions you have infered4 from the paper in the light of Compaqs own TPC-C TPM results.  - GS320 32 x 731 Mhz 4MB ecache 155,179 TPC-TPM , GS320 32 x 1 Ghz 8 MB ecache 230,533 TPC-TPM  ) This clearly shows that in the real world + moving from 4-8MB of cache at best accounts + for a 10% performance hike and nowhere near  the 50% you are suggesting.   4 Remember the memory and I/O subsystem hasn't changed. for these two benchmark results only the cache size and the clock rate.  . People also expect better throughput for newer+ versions of Oracle/OS on TPC-C particularly - when you are pushing the scalability envelope , as you do, while this isn't easy to quantify0 it does leave precious little room to illustrate0 the claimed advantages of having a larger cache.  , So nice try but no cigar. Your response is a- perfect example of the dangers of reading too / many white papers while studiously ignoring the  real world.    regards  Andrew Harrison    ------------------------------  # Date: Mon, 28 Oct 2002 20:09:17 GMT 5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> % Subject: Re: Immutable laws of the PC 2 Message-ID: <NPgv9.39$F8.1783936@news.cpqcorp.net>  J Yeah, sure.  Cool Quake-64 for Windows.  How much "extra" will you pay forI it?  How many people buy PC's just as gaming stations (as opposed to that ? just being one additional thing you do) rather than, say a PS2?   G You really think that shrink wrap games will show up "in weeks" without  32-bit versions?      $ Bill Gunshannon wrote in message ...4 >In article <5cfv9.27$lT7.1196154@news.cpqcorp.net>,9 > "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> writes:  >>J >> The 64-bit capabilities may take years for anyone to seriously exploit.# >> Probably database vendors first.  >  >I would make that: K > "The 64-bit capabilities may take weeks for anyone to seriously exploit."  >And:  > "Probably gamers first." >  >bill  >  >-- K >Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolves E >bill@cs.scranton.edu     |  and a sheep voting on what's for dinner.  >University of Scranton   | ? >Scranton, Pennsylvania   |         #include <std.disclaimer.h>    ------------------------------  % Date: Mon, 28 Oct 2002 12:29:01 -0800 ' From: David Mathog <mathog@caltech.edu> % Subject: Re: Immutable laws of the PC 8 Message-ID: <20021028122901.550d4e33.mathog@caltech.edu>    On Mon, 28 Oct 2002 18:18:41 GMT6 "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote:  J > And note - the only real business for "shrink wrap software" (outside ofL > perhaps Apple) is Windows.  So that is really the only place the IA32-nessN > is an advantage.  Linux (despite Linus's dislike if IPF) will happily run on, > IPF, or IA32, or one would imagine x86-64.  C Well, yes, and no.  I had a bunch of Alpha Linux boxes.  Yes it was = Linux.  No, it wasn't always trivial to get a given piece of  G code from the x86 space to run properly on it.  For instance, there was H an incredibly odd bug in PVM which apparently only showed up on my linuxA DS10s [:-( ] which caused an unpredictable 5 second delay for PVM H jobs to start.  It was some sort of access violation apparently, becauseC throwing in a ridiculous set of print statements on the slave nodesiD in a nearby section of the code "cured" it.  Yes, I did stumble onto: this by debugging via printf.  That's the sort of nonsenseE you have to deal with when you're running minority hardware even on aLH platform like linux.  It isn't good for productivity to be on a platformD which triggers all the latent bugs in software.  (Which, by the way,8 was what happened all the time when we used to run VMS.)     > J > Time will tell.  I would not be suprised to see a Sun/AMD merger and theI > abandonment of Sparc, and see Opteron turned into the Solaris platform.o  @ Egads, what a nightmare!  The only thing keeping CPU prices downB is AMD plugging away at Intel.  Sun would want to keep Opteron outD of everybody else's hands, hence driving the price up, and making it? a much less desirable chip.   Just look what Digital did to the B Alpha with a similar "us only" strategy.   Sun buys AMD and we can< expect higher prices for servers and PCs - across the board.   Regards,   -- B David Mathog mathog@caltech.edu> Manager, Sequence Analysis Facility, Biology Division, Caltech   ------------------------------  % Date: Mon, 28 Oct 2002 12:39:34 -0800 $ From: Shane Smith <ssmith@icius.com>% Subject: RE: Immutable laws of the PCr0 Message-ID: <01C27E7F.21DB1EF0@sulfer.icius.com>  G An awful lot of people buy PCs for gaming first. I know I do. There areSE PC manufacturers catering purely to the gaming community. Sure it's a G smaller market than business computing, but there's enough to support at2 surprisingly large number of specialist companies.   Shanea   -----Original Message-----: From: Fred Kleinsorge [mailto:kleinsorge@star.zko.dec.com]' Sent: Monday, October 28, 2002 12:09 PMt To: Info-VAX@Mvb.Saic.ComP% Subject: Re: Immutable laws of the PC     F Yeah, sure.  Cool Quake-64 for Windows.  How much "extra" will you pay forID it?  How many people buy PC's just as gaming stations (as opposed to that? just being one additional thing you do) rather than, say a PS2?h  G You really think that shrink wrap games will show up "in weeks" withoutb 32-bit versions?      $ Bill Gunshannon wrote in message ...4 >In article <5cfv9.27$lT7.1196154@news.cpqcorp.net>,9 > "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> writes:g >>J >> The 64-bit capabilities may take years for anyone to seriously exploit.# >> Probably database vendors first.0 >1 >I would make that:MK > "The 64-bit capabilities may take weeks for anyone to seriously exploit."y >And:i > "Probably gamers first." >1 >billE >9 >--rK >Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesiE >bill@cs.scranton.edu     |  and a sheep voting on what's for dinner.c >University of Scranton   |e? >Scranton, Pennsylvania   |         #include <std.disclaimer.h>    ------------------------------  # Date: Mon, 28 Oct 2002 17:43:31 GMTf# From: "John Smith" <a@nonymous.com>d% Subject: Re: Immutable laws of the PCnI Message-ID: <7Hev9.116365$%h2.46436@news02.bloor.is.net.cable.rogers.com>e  4 "David Mathog" <mathog@caltech.edu> wrote in message2 news:20021028091705.3fff90df.mathog@caltech.edu... > C > This is exactly why Opteron is going to drop depth charges on theeK > already submerged Itanic.  Joe Gamer or Mary Officeuser can upgrade theiruF > machine from a Pentium or Athlon to an Opteron machine and still runL > all the same software.  Sure, it won't take advantage of the 64 bit stuff,G > but it will run faster than what they had.  And when Windows XP64 (oraC > whatever they call it) ships they will still be able to run theircB > existing applications and then add any new 64bit ones they need. >lJ > In other words, for at least 95% of the computer market it's a darn nearH > painless migration to 64bit land with an increase in 32bit performance thrownI > in to sweeten the pot.  Or turning that argument over, Opteron is goinghD > to have a lot of sales that have nothing whatsoever to do with its > 64bit capabilities.g >iL > With Itanic, assuming any actually ever appear for sale, you're looking atH > all new software from the get go.  And the only reason to go to Itanic? > is to get a 64 bit machine. It's a jarring transition and oneuD > that there's no compelling reason to make outside the data center.= > Not that the case there is all that compelling either.  Oh,pF > it's compelling enough for two companies: HP (who burned their Alpha bridgeD > at leat 3 years prematurely) and Intel (who needs somebody to footA > the bill for Itanic).  But it sure doesn't look very appetizing  > for their customers. >lD > The primary beneficiary of the HP/Intel Itanic axis is going to be > IBM and Sun (in that order).    J Maybe there ought to be a Senate investigation of HP before it melts down.J Rosy predictions to accomplish a merger, leading statements that investorsH relied upon when voting. I don't think 'Safe Harbor' provisions apply in this instance.   ------------------------------  # Date: Mon, 28 Oct 2002 18:18:41 GMTh5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>c% Subject: Re: Immutable laws of the PCo3 Message-ID: <5cfv9.27$lT7.1196154@news.cpqcorp.net>    David Mathog wrote in message@/ <20021028091705.3fff90df.mathog@caltech.edu>...( >lB >This is exactly why Opteron is going to drop depth charges on theJ >already submerged Itanic.  Joe Gamer or Mary Officeuser can upgrade theirE >machine from a Pentium or Athlon to an Opteron machine and still runtK >all the same software.  Sure, it won't take advantage of the 64 bit stuff,RF >but it will run faster than what they had.  And when Windows XP64 (orB >whatever they call it) ships they will still be able to run theirA >existing applications and then add any new 64bit ones they need.  > I >In other words, for at least 95% of the computer market it's a darn neartG >painless migration to 64bit land with an increase in 32bit performanceo thrownH >in to sweeten the pot.  Or turning that argument over, Opteron is goingC >to have a lot of sales that have nothing whatsoever to do with itsA >64bit capabilities. >p  J This is where it gets interesting.  The assumption is that Opteron will beH able to compete on price/performance with whatever generation of IA32 isK then current when it finally appears.  As you imply, there is no compelling,G reason at this point for 64-bits on the desktop.  We can construct someiK examples where 64-bits might potentially be used, but the mass-market ISV'stJ are not going to retarget their applications for the IA64x and exclude theL IA32 market.  So the question really gets down to:  Can Opteron IA32 competeL with Intels IA32.  The big problem that AMD has is that Intel can underpriceL them, and has a significant market clout both with consumers and with systemG builders.  So it's unlikely that Opteron will be able to compete with aoI better price, and we'll get to find out just how much that "Intel Inside" " sticker really means to consumers.  J That leaves performance for Opteron to compete on, and if it's performanceK will be significant enough for people to pay a premium for it...  how largenG is the potential base of super-high-end desktop PC users?  If the pricetF point today for an entry PC is about $600, and a high-end is say $1500E (ignoring options) - will the user pay more for "64-bit ready OpteroncL Inside"?  How much more?  Or is it more likely that the system builder needsJ to hit a price point - and needs to eat the incremental cost for the chip?  H So, I believe that the only logical target for Opteron is the higher-endI IA32 server space (say up to 8-way SMP boxes) - where people will pay forsK the performance - NOT the desktop which is far more price sensitive.  WhichlK if I am right, reduces that "95%..." you speak of to 95% of a small segments of the IA32 space.  G The 64-bit capabilities may take years for anyone to seriously exploit.u  Probably database vendors first.  H And note - the only real business for "shrink wrap software" (outside ofJ perhaps Apple) is Windows.  So that is really the only place the IA32-nessL is an advantage.  Linux (despite Linus's dislike if IPF) will happily run on* IPF, or IA32, or one would imagine x86-64.  K >With Itanic, assuming any actually ever appear for sale, you're looking atr" >all new software from the get go.  G Not really.  You need new software for the applications you really carerL about performance for, and others can use the IA32 capabilities.  So, if youK are buying a 8-way Windows Server to run Oracle - you really only care that K Oracle is native.  Not that Solitare doesn't run faster.  So Intels job haseI to be to convince the major server ISV's to produce IPF versions of their/ software for Windows64.i  & >  And the only reason to go to Itanic> >is to get a 64 bit machine. It's a jarring transition and oneC >that there's no compelling reason to make outside the data center.   L There is no jarring transition for most people.  Yes, it may be work for theI ISV, but fewer and fewer customers are writing their own code these days,yK instead of using off-the-shelf software, and applets written in things likee Java.t  < >Not that the case there is all that compelling either.  Oh,L >it's compelling enough for two companies: HP (who burned their Alpha bridgeC >at leat 3 years prematurely) and Intel (who needs somebody to foot @ >the bill for Itanic).  But it sure doesn't look very appetizing >for their customers.@ >s  F You are projecting a bit much here.  HP wanted IPF to replace PA-RISC.$ Compaq made the decsion about Alpha.  K IPF *does* look very appealing to customers.  It competes head-to-head with H the fastest CPU's out there, including Alpha.  Perhaps it's not bleedingJ edge performance, but it outperforms SPARC.  The Opteron numbers are vaporF until there is a real chip and system.  And by that time there will beI faster versions of IPF.  What it doesn't look appealing to is the typical 2 non-server IA32 user **AND NEITHER DOES OPTERON**.  C >The primary beneficiary of the HP/Intel Itanic axis is going to bee >IBM and Sun (in that order)., >r  H Time will tell.  I would not be suprised to see a Sun/AMD merger and theG abandonment of Sparc, and see Opteron turned into the Solaris platform.e   ------------------------------  # Date: Mon, 28 Oct 2002 20:16:31 GMTw5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>e% Subject: Re: Immutable laws of the PCo2 Message-ID: <zWgv9.40$b9.1806442@news.cpqcorp.net>  E Atlant Schmidt wrote in message <3DBD8598.245A9CA5@mindspring.com>...  >Fred Kleinsorge wrote:e >rJ >> This is where it gets interesting.  The assumption is that Opteron will beK >> able to compete on price/performance with whatever generation of IA32 iseC >> then current when it finally appears.  As you imply, there is noA
 compellingJ >> reason at this point for 64-bits on the desktop.  We can construct someH >> examples where 64-bits might potentially be used, but the mass-market ISV'siI >> are not going to retarget their applications for the IA64x and exclude5 the6 >> IA32 market.e >a2 >Fred, I'm sorry, but this sounds naive. Nowadays,- >I won't buy a computer for our SOHO that has 0 >less than a Gigabyte of main memory. Now, IIRC,4 >main memory usage is growing at a rate of something9 >like 1.2 bits/year, so in just 3+ years, I can expect to   >be bumping the 32-bit VM limit. >   G Really?  I can easily see running into a 32-bit *physical* limit, but I K wanna know what single application virtual address space requires that 33rd I bit on a typical home computer today, or even in 3 years.  We can come uprK with a list of things that potentially *could* use it... but that isn't theoH same as requiring it.  Nor do I believe that any ISV in their right mindH would target the 64-bit space at the exclusion of the 32-bit space.  TooJ many IA32 systems in the world.   So far, I haven't heard what the "killer app" for desktop 64 bits is.    3 >You *BET* I'll stretch to buy those retread 64-bita- >Power4 chips when IBM starts selling them toI >Apple to fit into Macintoshes!i >t  8 Yeah, but Atlant I *know* you are the lunatic fringe ;-)  ) >I wouldn't expect the PC world to be anyi >different.  >e4 >And there's really no need to wait for applications6 >to port; it'd be an entirely obvious thing to provide. >a 32-bit environment within the larger 64-bit+ >environment just like MacOS/X provides thes1 >"Classic" (MacOS/9) operating environment within . >a much-richer MacOS/X environment. And within3 >Classic is a full MC68K emulation so programs laste2 >compiled for the pre-RISC architecture still run! >M  E You don't need a 64-bit machine, to add more than 32 bits of physicalkF addressing, as long as the virtual addressing remains limited within a single VA to 32-bits.w   ------------------------------  % Date: Mon, 28 Oct 2002 13:44:40 -0500e2 From: Atlant Schmidt <atlantnospam@mindspring.com>% Subject: Re: Immutable laws of the PC . Message-ID: <3DBD8598.245A9CA5@mindspring.com>   Fred Kleinsorge wrote:  L > This is where it gets interesting.  The assumption is that Opteron will beJ > able to compete on price/performance with whatever generation of IA32 isM > then current when it finally appears.  As you imply, there is no compellingiI > reason at this point for 64-bits on the desktop.  We can construct some$M > examples where 64-bits might potentially be used, but the mass-market ISV's L > are not going to retarget their applications for the IA64x and exclude the > IA32 market.  1 Fred, I'm sorry, but this sounds naive. Nowadays,o, I won't buy a computer for our SOHO that has/ less than a Gigabyte of main memory. Now, IIRC,a3 main memory usage is growing at a rate of somethingo8 like 1.2 bits/year, so in just 3+ years, I can expect to be bumping the 32-bit VM limit.o  2 You *BET* I'll stretch to buy those retread 64-bit, Power4 chips when IBM starts selling them to Apple to fit into Macintoshes!  ( I wouldn't expect the PC world to be any
 different.  3 And there's really no need to wait for applications-5 to port; it'd be an entirely obvious thing to provide0- a 32-bit environment within the larger 64-bita* environment just like MacOS/X provides the0 "Classic" (MacOS/9) operating environment within- a much-richer MacOS/X environment. And withino2 Classic is a full MC68K emulation so programs last1 compiled for the pre-RISC architecture still run!t   Atlant   ------------------------------  % Date: Mon, 28 Oct 2002 16:52:49 -0500s- From: JF Mezei <jfmezei.spamnot@videotron.ca>u% Subject: Re: Immutable laws of the PCm, Message-ID: <3DBDB1AF.E5B84694@videotron.ca>   Fred Kleinsorge wrote:I > You really think that shrink wrap games will show up "in weeks" withoutL > 32-bit versions?  L Do you have teenagers ? If the newest version of Quake comes in 64 bit only,E you can bet that you'll have hundreds of thousands of young teenagers.N blackmailing their parents into buying a new 643 bit PC for christmas or their, birthday (or any other day for that matter).  N You only one guy in a class room to get 64 bit Quake and the rest will follow.  M Do not underestimate the power of gaming in the home PC market. Have you seenoJ those games lately ? They have live 3d rendering of objects that is bloodyC impressive.  given 64 bit, and they will get evben more impressive.r  K What used to take days on a VAX to generate is now done live/real time on amJ home PC. And what used to take months to generatye (such as some scenes onM titanic) will soon be incorporated into some video game. And it won't be long > before you start to get 3D games that you use with 3D glasses.   ------------------------------  % Date: Mon, 28 Oct 2002 16:45:53 -0500s- From: JF Mezei <jfmezei.spamnot@videotron.ca>d% Subject: Re: Immutable laws of the PCB, Message-ID: <3DBDB010.97C92FD9@videotron.ca>   Fred Kleinsorge wrote:N > IA32 market.  So the question really gets down to:  Can Opteron IA32 competeN > with Intels IA32.  The big problem that AMD has is that Intel can underpriceN > them, and has a significant market clout both with consumers and with system > builders.l  K From an overall company point of view, which do you think should be able to " offer the most competitive price ?  H The company that has a single architecture that spanes the low end (high- volume) to some of the high end (low volume)     or  N The company that has two architectures. The first one goes from the low end toN the mid level, and the second one goes from mid level to some of the high end.    N Intel will EVENTUALLY have to pay for IA64. The sums of money sunk into it areG huge. They cannot continue to spend inordinate amounts of money on IA64eM without any return because that money will also be needed to work on the 8086e' line to make it compete against Hammer.i   ------------------------------  % Date: Mon, 28 Oct 2002 15:46:29 -08000$ From: Shane Smith <ssmith@icius.com>% Subject: RE: Immutable laws of the PC 0 Message-ID: <01C27E99.55D89AB0@sulfer.icius.com>  E I can't post it's, it's binary and enormous, but for those of you whoeC are interested go looking for an AVI file of Doom 3 (probably spelteB "Doom ]|["). There's one floating around with interviews with John@ Carmack and several other developers, and footage of the in-gameF rendering. It's practically movie quality, especially the bit with theE creature from hell eating the fat zombie.* This stuff has been demo'dRF live a few months ago on a P4 with a Radeon 9700 pro graphics card, so it's for real.  G Will I upgrade to whatever's needed to play that game smoothly? "Is the B Pope Catholic/ is Luxembourg small/and do bears sh[alalala] in the	 woods?"**9  G But Fred's right, Mr. Carmack is cutting edge, but even he designed the2G new engine to work with a GeForce 3 graphics card, which is generations0G out of date already. Most companies are only just starting to require a C graphics card with onboard T&L, and that dates back to the original1E GeForce 1. It'll be some time before any game company releases 64-bit0H only software. It will happen eventually. Practically everything expectsG a 3D accelerator of some sort now, and those only really waded into the9E mainstream about 4 years ago. It'll take longer for a CPU change, but 
 it'll happen.    Shane1  " *= Yes, viewer discretion advised.  E **= Since we don't need another OT thread about who sang that, it was:A Rowan Atkinson and Kate Bush at one of the Comic Relief concerts.1   -----Original Message-----4 From: JF Mezei [mailto:jfmezei.spamnot@videotron.ca]& Sent: Monday, October 28, 2002 1:53 PM To: Info-VAX@Mvb.Saic.Com % Subject: Re: Immutable laws of the PC0     Fred Kleinsorge wrote:I > You really think that shrink wrap games will show up "in weeks" without1 > 32-bit versions?  F Do you have teenagers ? If the newest version of Quake comes in 64 bit only,aE you can bet that you'll have hundreds of thousands of young teenagers H blackmailing their parents into buying a new 643 bit PC for christmas or their , birthday (or any other day for that matter).  F You only one guy in a class room to get 64 bit Quake and the rest will follow.k  H Do not underestimate the power of gaming in the home PC market. Have you seenC those games lately ? They have live 3d rendering of objects that isu bloodyC impressive.  given 64 bit, and they will get evben more impressive.i  F What used to take days on a VAX to generate is now done live/real time on aG home PC. And what used to take months to generatye (such as some scenes  onH titanic) will soon be incorporated into some video game. And it won't be long> before you start to get 3D games that you use with 3D glasses.   ------------------------------  % Date: Mon, 28 Oct 2002 23:27:17 +0000.' From: Elliott Roper <elliott@yrl.co.uk> % Subject: Re: Immutable laws of the PCh2 Message-ID: <281020022327175380%elliott@yrl.co.uk>  B In article <zWgv9.40$b9.1806442@news.cpqcorp.net>, Fred Kleinsorge$ <kleinsorge@star.zko.dec.com> wrote:  G > Atlant Schmidt wrote in message <3DBD8598.245A9CA5@mindspring.com>...W <snip>  I > Really?  I can easily see running into a 32-bit *physical* limit, but I M > wanna know what single application virtual address space requires that 33rdRK > bit on a typical home computer today, or even in 3 years.  We can come up@M > with a list of things that potentially *could* use it... but that isn't theeJ > same as requiring it.  Nor do I believe that any ISV in their right mindJ > would target the 64-bit space at the exclusion of the 32-bit space.  TooL > many IA32 systems in the world.   So far, I haven't heard what the "killer > app" for desktop 64 bits is. >  > 5 > >You *BET* I'll stretch to buy those retread 64-bit'/ > >Power4 chips when IBM starts selling them toa! > >Apple to fit into Macintoshes!  > >m > : > Yeah, but Atlant I *know* you are the lunatic fringe ;-)   I'll be lining up too! > + > >I wouldn't expect the PC world to be anym
 > >different.r > > 6 > >And there's really no need to wait for applications8 > >to port; it'd be an entirely obvious thing to provide0 > >a 32-bit environment within the larger 64-bit- > >environment just like MacOS/X provides theh3 > >"Classic" (MacOS/9) operating environment withina0 > >a much-richer MacOS/X environment. And within5 > >Classic is a full MC68K emulation so programs lastc4 > >compiled for the pre-RISC architecture still run! > >e > G > You don't need a 64-bit machine, to add more than 32 bits of physicaljH > addressing, as long as the virtual addressing remains limited within a > single VA to 32-bits.p  D Sounds like the PDP-11 all over again. More physical memory than you? can address. What will they call UMRs and APRs this time round?s  D Are you old enough to remember TKB and ODL Fred? We don't want to go through *that* again do we?n   Elliott    ------------------------------  # Date: Mon, 28 Oct 2002 23:35:14 GMTe# From: "John Smith" <a@nonymous.com>a% Subject: Re: Immutable laws of the PCeJ Message-ID: <SQjv9.118684$%h2.103732@news02.bloor.is.net.cable.rogers.com>  @ "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote in message, news:NPgv9.39$F8.1783936@news.cpqcorp.net...L > Yeah, sure.  Cool Quake-64 for Windows.  How much "extra" will you pay forK > it?  How many people buy PC's just as gaming stations (as opposed to that0A > just being one additional thing you do) rather than, say a PS2?R >FI > You really think that shrink wrap games will show up "in weeks" withouti > 32-bit versions?  L It's not necessarily that 64-bits makes a technical difference. It does make& a marketing and perceptual difference.  6 Sometimes it's the sizzle that sells before the steak.   ------------------------------   Date: 29 Oct 2002 02:19:45 GMT, From: bill@gw5.cs.uofs.edu (Bill Gunshannon)% Subject: Re: Immutable laws of the PC 5 Message-ID: <apkr80$2hf4v$2@ID-135708.news.dfncis.de>R  2 In article <281020022327175380%elliott@yrl.co.uk>,* 	Elliott Roper <elliott@yrl.co.uk> writes: > F > Sounds like the PDP-11 all over again. More physical memory than youA > can address. What will they call UMRs and APRs this time round?h > F > Are you old enough to remember TKB and ODL Fred? We don't want to go > through *that* again do we?  >     C Hey.....  Let's not go dising the PDP-11.  There are still a lot ofu3 us very pleased with the ones we have running.  :-)`   bill   -- iJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  % Date: Mon, 28 Oct 2002 21:03:04 -0500 * From: "Bill Todd" <billtodd@metrocast.net>% Subject: Re: Immutable laws of the PC 2 Message-ID: <2nGdnVEDz7LHcSCgXTWcqQ@metrocast.net>  @ "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote in message- news:5cfv9.27$lT7.1196154@news.cpqcorp.net..., >d > David Mathog wrote in messaget1 > <20021028091705.3fff90df.mathog@caltech.edu>...  > >sD > >This is exactly why Opteron is going to drop depth charges on theL > >already submerged Itanic.  Joe Gamer or Mary Officeuser can upgrade theirG > >machine from a Pentium or Athlon to an Opteron machine and still run F > >all the same software.  Sure, it won't take advantage of the 64 bit stuff,H > >but it will run faster than what they had.  And when Windows XP64 (orD > >whatever they call it) ships they will still be able to run theirC > >existing applications and then add any new 64bit ones they need.O > >0K > >In other words, for at least 95% of the computer market it's a darn nearrI > >painless migration to 64bit land with an increase in 32bit performancet > thrownJ > >in to sweeten the pot.  Or turning that argument over, Opteron is goingE > >to have a lot of sales that have nothing whatsoever to do with itsv > >64bit capabilities. > >e >eL > This is where it gets interesting.  The assumption is that Opteron will beJ > able to compete on price/performance with whatever generation of IA32 isB > then current when it finally appears.  As you imply, there is no
 compelling2 > reason at this point for 64-bits on the desktop.  L First, both of you seem to be using 'Opteron' to refer to a desktop-suitable? Hammer, whereas in fact the Opteron designation is reserved forfK multi-processor server (or possibly high-end workstation) use.  The desktoprH Hammers will be called something like 'Athlon-64', will be smaller, lessF expensive chips (smaller cache, only one HyperTransport link, only oneL on-chip memory channel), and while they'll nominally appear at the end of Q1F along with Opteron AMD seems to be planning to emphasize higher-marginL Opteron sales until production capacity has ramped up fully some time in H2.  D Now that that's out of the way, while it may be true that there's noJ compelling reason *today* for 64 bits on the desktop you ignored what DaveJ said above:  "when Windows XP64 (or whatever they call it) ships they willK still be able to run their existing applications and then add any new 64bit@ ones they need."  J In other words, for Hammer to have an advantage over IA32 buyers need onlyG suspect that at some time before their new systems are obsolete (a timee@ which they may reasonably feel they can push out by purchasing aI 64-bit-capable platform) they *may* wish to run *some* 64-bit applicationaG (or system) on them.  Now, I wouldn't necessarily expect the average PC.J buyer to be thinking in such terms on his/her own, but I do tend to expect9 that AMD ads will steer their thoughts in this direction.      We can construct some.G > examples where 64-bits might potentially be used, but the mass-market  ISV'ssL > are not going to retarget their applications for the IA64x and exclude the > IA32 market.  K So what?  It takes only the potential that *one* popular 64-bit application J will materialize to make Hammer more attractive in this respect than IA32.  C What's a good candidate for such an application?  For starters, anynJ application that costs upwards of, say, $4K/seat:  at that price, the costE of the platform on which to run it (at least if it's a desktop Hammer J platform) starts to become negligible, so if there's any real advantage toG be gained from coding it in 64 bits there's little reason to spend timeh shoe-horning it into 32.  G I can't guarantee that such an application will materialize in the nexttL couple of years, but there's certainly a good chance that one will.  Why notF buy a platform that will be able to handle it, most other things being equal?  @   So the question really gets down to:  Can Opteron IA32 compete > with Intels IA32.u  F (Not to keep you in undue suspense:  the answer is emphatically "Yes." Details below.)u  ;   The big problem that AMD has is that Intel can underprice  > them,t  J Not at all clear.  After all, Intel certainly doesn't underprice AMD todayD for comparable processors, rather the reverse.  For a given level ofH performance, not only are AMD's processors priced considerably lower butH they cost considerably less to fab, at least based on the fact that theyJ occupy signficantly less chip area:  in the same process (and AMD plans toK start migrating to 90 nm. only shortly after Intel does), Athlons are aboutcI half the size of P4s (and AMD will continue to produce K7 Athlons for the D low end of the market for quite a while yet - they've even got a newJ version, Barton, slated to ship in Q1), desktop Hammers are about the sameI size as Athlons, and even full-fledged Opterons are significantly smallert. than P4s despite having larger on-chip caches.  L It may well be that the only way Intel could underprice AMD would be to sellG below cost.  And while Intel has the financial resources to do so for a2I while, even the current Administration's FTC might feel compelled to stepe5 into such blatant leveraging of its market dominance.   8 > and has a significant market clout both with consumers  I That's debatable.  My guess would be that at least close to a majority ofnJ purchasers haven't a clue what processor is actually inside their computerH (even if they've heard of Intel but not AMD), and when they just look at@ prices the lowest-priced boxes tend to be those with AMD inside.  J And of the rest, at least some percentage (gamers and other 'enthusiasts')H has an active preference for AMD processors, since they offer far better cost/performance.     and with system > builders.o  B Depends on who you call a 'system builder'.  For the white box (orK 'third-tier') crowd, AMD wins hands down, and their second-tier penetration8K is at least good.  And those market segments have been gaining market shareiK at the expense of the first-tier vendors (except perhaps Dell) for the paste couple of years now.  >   So it's unlikely that Opteron will be able to compete with a > better price,   J Wrong - see above.  The interesting question will be just how far Intel isD willing to drop Xeon prices (and profits) in order to deprive AMD ofK comparable Opteron prices and profits:  AMD appears to anticipate a ramp-uptD period for Opteron before sufficient units become available to startJ diverting them to the lower-profit desktop Hammer products, and can likelyH sell as many as it can make during that time at very tidy profits unless  Intel drops Xeon prices a bunch.  <  and we'll get to find out just how much that "Intel Inside"$ > sticker really means to consumers.   Exactly - as I suggested above.    > L > That leaves performance for Opteron to compete on, and if it's performanceG > will be significant enough for people to pay a premium for it...  howu largeoI > is the potential base of super-high-end desktop PC users?  If the priceeH > point today for an entry PC is about $600, and a high-end is say $1500G > (ignoring options) - will the user pay more for "64-bit ready Opterona
 > Inside"?  I Again, you seem to be confusing Opteron with the desktop Hammers.  Last InL knew, Intel price Xeons in the $1K - $4K range (i.e., similarly to Itanics),H and AMD will likely price bare Opteron processors well over $1K at leastI initially unless (as noted above) Intel slashes Xeon prices (and possiblyaF even if Intel *does* slash Xeon prices, if the reception of Opteron is0 anything close to what it appears likely to be).  I When enough production capacity exists to start moving desktop Hammers asuJ well, their prices will presumably be similar to high-end P4 prices (whichK currently range up to over $600/processor, IIRC).  K7 Athlons will continue 5 to hold down the mid-to-low-price segments just fine.i  D   How much more?  Or is it more likely that the system builder needsL > to hit a price point - and needs to eat the incremental cost for the chip? >:J > So, I believe that the only logical target for Opteron is the higher-endK > IA32 server space (say up to 8-way SMP boxes) - where people will pay forgF > the performance - NOT the desktop which is far more price sensitive. WhichrE > if I am right, reduces that "95%..." you speak of to 95% of a smalll segmenty > of the IA32 space.  K You are only right by virtue of having restricted your comments to Opteron, K which is by definition not a desktop product.  Desktop Hammers will competeuK well down into the desktop market, though not to its bottom for a while yett2 (where K7 Athlons will continue to hold the fort).   >pI > The 64-bit capabilities may take years for anyone to seriously exploit.h" > Probably database vendors first.  J No, they'll take years for a *lot* of people to exploit.  But a *few* will start immediately.   >tJ > And note - the only real business for "shrink wrap software" (outside ofL > perhaps Apple) is Windows.  So that is really the only place the IA32-ness > is an advantage.  H Since that's what the overwhelming majority of desktops are sold to run, what's your point?  H It is, however, worth noting that Hammer offers the attractive option ofJ having a fast 32-bit Windows box that can also be booted into 64-bit LinuxD run a special 64-bit app if necessary - even if you never use 64-bit Windows.  <   Linux (despite Linus's dislike if IPF) will happily run on, > IPF, or IA32, or one would imagine x86-64. >fJ > >With Itanic, assuming any actually ever appear for sale, you're looking at$ > >all new software from the get go. > 
 > Not really.u  I More like 'yes and no'.  If you're talking a desktop box (which we mostly I seem to have been up until now), then 'yes':  most of what you're runningpI will likely be the same 32-bit software you run today (with an occasionaltK 64-bit application mixed in, since otherwise the choice of Itanic over IA32rF on your desktop would have been really dumb, a problem Hammer does notK have), and unless you convert to 64-bit versions of all that software it'll-G run like a slug.  OTOH, if you bought the Itanic box primarily to run amG specific 64-bit server app, there's a good chance that you can tolerateD; running ancillary 32-bit apps on it at 300 MHz PPro speeds.p  <   You need new software for the applications you really careJ > about performance for, and others can use the IA32 capabilities.  So, if you H > are buying a 8-way Windows Server to run Oracle - you really only care thatI > Oracle is native.  Not that Solitare doesn't run faster.  So Intels jobd hasnK > to be to convince the major server ISV's to produce IPF versions of their: > software for Windows64.i >o( > >  And the only reason to go to Itanic@ > >is to get a 64 bit machine. It's a jarring transition and oneE > >that there's no compelling reason to make outside the data center.  > 1 > There is no jarring transition for most people.o  K Indeed:  they'll just get Hammers instead, at least for all uses other thansI dedicated servers where the existence of native 64-bit Itanic applicationnF versions will minimize the pain (though Itanic will still have troubleC competing with Hammer on cost/performance - and likely even on pure-
 performance).s   ...k  H > IPF *does* look very appealing to customers.  It competes head-to-head withJ > the fastest CPU's out there, including Alpha.  Perhaps it's not bleeding- > edge performance, but it outperforms SPARC.a  L Now, which is it?  Head-to-head with the fastest CPUs, or not bleeding edge?B I do give you credit for allowing both possibilities, and I agree.  I Just how fast is McKinley?  Running HP-UX, using the HP-UX compiler, with,K feedback optimization, on HP's zx1 chipset, its SPECint2K score (810) looks-J very good.  Take away the HP-UX compiler, and the scores drop toward SPARCK territory (680 for the MS compiler, 619 for the Intel compiler - don't know6I if feedback optimization was used to obtain those scores).  Take away thegK zx1 chipset, and the platform evaporates (if/when Intel and IBM offer theirdJ own chipsets then we'll see whether zx1 is as superior as HP has claimed -K but if it is, that will just be another strike against volume acceptance ofiJ Itanic, since non-HP Itanics will have much more difficulty competing with the rest of the world).   K For that matter, how much would the competition's SPECint scores improve ifa6 *they* started to use feedback-directed optimizations?  K About the most that can be said is that Itanic is not going to win based oniJ pure performance, but it probably isn't going to lose based on it either -$ at least until Hammer comes along...     The Opteron numbers are vaport( > until there is a real chip and system.  E You might want to talk with Fred Weber about that.  He claims to have,H obtained the numbers on a 2 GHz Opteron system in the lab.  Partners areL running systems at at least 1.6 GHz.  MP systems are running at at least 1.4> GHz.  User systems are out in the field with 'early adopters'.  L That sounds pretty real to me, at least as far as getting credible benchmark
 numbers goes.i      And by that time there will be > faster versions of IPF.m  J Not according to both current schedules.  Opteron is scheduled for releaseI at the end of Q1.  Madison is scheduled for release in Q2.  And, for thattI matter, according to Intel's projection that Madison will exceed McKinleyeF performance by 30% - 50% Madison will *at best* match Opteron's 32-bitJ SPECint performance (and thus lag behind Opteron's expected 64-bit SPECint
 performance).   2   What it doesn't look appealing to is the typical4 > non-server IA32 user **AND NEITHER DOES OPTERON**.   But the desktop Hammers do.t   - bill   ------------------------------  # Date: Mon, 28 Oct 2002 20:19:00 GMTt# From: "John Smith" <a@nonymous.com>i@ Subject: Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeekI Message-ID: <UYgv9.86381$mxk1.78654@news04.bloor.is.net.cable.rogers.com>@  6 "Barry Treahy, Jr." <Treahy@MMaz.com> wrote in message! news:3DBD8644.5070705@MMaz.com...@ >sJ > If Intel were to release an 64-bit X86, they would undermine their IA-64I > efforts!  This could create serious problems for us VMS folks expectingeI > to run on IA-64 because what was the Compaq rational for killing Alpha?r9 >  Not enough volume to support the design and fab costs.  >eD > So, if Intel does introduce a 64-bit X64 for the desktop, you knowJ > manufacturers will introduce servers based on that chip set, essentially+ > killing high volume demand for the IA-64.y >lH > Am I the only one that sees the irony of this?  A major porting effortF > of VMS to another 'low volume and essentially proprietary' chip set?H >  Worse yet, a chip set that HP/Compaq doesn't directly control and may6 > not even be as fast as the Alpha that was abandoned?    = No Barry, you aren't the only one who sees the irony of this.   K However all that do see the irony don't occupy executive level positions at  HP.t   ------------------------------    Date: 28 Oct 2002 18:27:58 -0500- From: Rich Alderson <alderson+news@panix.com>p@ Subject: Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek. Message-ID: <mddbs5ep2ip.fsf@panix5.panix.com>  & Shane Smith <ssmith@icius.com> writes:  5 > Shouldn't that be 1982 performance at a 2020 price?s   > -----Original Message-----0 > From: David Mathog [mailto:mathog@caltech.edu]) > Sent: Monday, October 28, 2002 10:07 AM6  ; > All the performance you'd expect in 2002 at a 1982 price!oM > HP is pleased to announce the availability of the Whizbang Itanium II entryaJ > Level VMS server.  This server comes with a single user VMS license, oneD > Itanium II chip, 2Gb memory, 200Gb serial ATA disk, floppy, cdrom. > Prices start at 110K$  e    A 2020 cost lots more than that!   -- nN Rich Alderson                                          alderson+news@panix.comL   "You get what anybody gets.  You get a lifetime."  --Death, of the Endless   ------------------------------    Date: 29 Oct 2002 02:32:19 +0800, From: Paul Repacholi <prep@prep.synonet.com>9 Subject: Re: Now where's the OpenVMS OpenOffice software?P- Message-ID: <87pttuifd8.fsf@prep.synonet.com>i  ? "Colin Butcher" <colinDOT.butcherAT@xdeltaDOT.coDOT.uk> writes:c  * > <http://www.newmonday.co.uk/Job/3705704>  2 Isn't that the one of the `big 5' that IBM bought?   -- n< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.e@                                              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: 28 Oct 02 14:41:01 EST From: grant@rigel.cc.wmich.edu% Subject: RE: OpenVMS 7.3-1 and CSWINGp/ Message-ID: <P2kozYs7RzFc@mcduck.acs.wmich.edu>o  C Just for the heck of it, I updated the copy of the cSWING sources IhF found so it compiles without errors.  I think I got past the few casesD of routines with inconsistent numbers of arguments, several cases ofH referencing non-existent structure members, and a library function whichF doesn't exist in the C headers (lib$erase_page). I also cleaned up allH the IOSBs, corrected some longs which should have been shorts (channels,E and lengths) and a few improperly initialized descriptors, but, thereAD were a few  coding errors I don't have time to pursue, including twoF global routines named "rest_cursor" and an attempt to assign a pointer to an integer.  D However, I don't have the time or ambition to do the kind of rewrite1 which is indicated by Craig Berry's observation. s  ? If anyone would like to either take on that job starting with a D compilable copy of the source, or just try the compilable version toE see whether it works any better than the old executable, you can maili> be and I'll ship you the zip with the original and my changes.  	 Good luckc  ` In article <JFEPKAPBPMDFDBOIANGDGEKPCBAA.dallen@nist.gov>, "Dan Allen" <dallen@nist.gov> writes: >  >  >> -----Original Message-----N7 >> From: Carlos Costa [mailto:carlos.costa@datawest.ca],) >> Sent: Friday, October 25, 2002 6:21 PM: >> To: Info-VAX@Mvb.Saic.Com( >> Subject: Re: OpenVMS 7.3-1 and CSWING >> w >> eh >> "Dan Allen" <dallen@nist.gov> wrote in message news:<JFEPKAPBPMDFDBOIANGDCEJFCBAA.dallen@nist.gov>...! >> > > -----Original Message-----t >> > eP >> >      	Are you serious! Do you really think that ANY operating system vendorI >> > 	can or will commit to maintaining compatability with latent bugs in @ >> > 	all of the application software ever developed for the OS! >> gD >> Excuse me, but yes, I am serious. If you read the above statementG >> carefully you will see that that is exactly what HP/Compaq has done.lI >> They have committed to maintaining 100% binary compatability. I didn'tiH >> say it, THEY did. If they didn't mean it they shouldn't have said it.I >> Binary compatibility is a pretty tight compatibility. It means that noaG >> program needs to be re-compiled or re-linked. It will work as it is.M% >> They even say "no testing needed".  >> eG >> Secondly, one of the reasons I pointed that statement out is because-D >> that statement means, to me, that there is a bug in the VMS code. > L > 	See Craig Berry's post on the misuse of the context value and name stringL > 	in the FIB in the CSWING source code. Most likely the problem.  Note alsoJ > 	that this is what I would call a latent application bug, i.e. it is notJ > 	providing the proper input to the QIO call.  The fact that it worked byI > 	chance in 7.3 but fails in 7.3-1 is not what I would call a bug in the>G > 	OS and I doubt VMS engineering would be motivated to restore the oldl > 	behavior. > I >> >If not I'd suggest you quit whining about what MIGHT be an OS bug and 
 >> give usI >> >some hard info on what's actually happening. If I had the environmentt
 >> to testD >> >the damn thing myself I would have already done it for you! This >> isn't rocketrB >> >science, it's not even computer science, it's programming 101. >> OH >> I don't know about that, the particular code is recursive and full ofB >> VMS specifics. Not for the faint of heart. And please don't getF >> personal about this. You're hinting that I'm not as smart as you toD >> not have fixed this myself, and you could very well be right. ButE >> "Don't judge someone until you've walked a mile in his shoes." But H >> you've only seen a small code fragment. There are many more thousands >> of lines. > K > 	I said absolutely nothing about your "smarts".  The comment was directedeG > 	at your focus on your believe that you had uncovered an OS bug basedlH > 	solely on YOUR interpretation what binary compatability meant and theF > 	lack of additional data with regard to problems identified by folksE > 	more familiar with VMS coding practices.  I encourage you to learnrG > 	more about the VMS operating system. I've walked many a mile in your I > 	shoes and I learned fairly early on that these problems are invariably * > 	a latent problem in MY code not the OS. > L > 	I've lurked on this list for quite some time now and I believe the peopleJ > 	here are more than willing to give of their time and experience helpingN > 	anyone from the pure neophyte to the seasoned internals programmer diagnoseN > 	and correct problems with their codes. But the way to accomplish that is toK > 	post a complete problem description and source code when available. TheniP > 	follow up and test problems identified by this process and report the results@ > 	to the list if you are unable to make sense of them yourself. > P > 	And as Craig Berry has pointed out (see below) the problem would appear to beP > 	an application error in CSWING.  Had you posted a more extensive code snippetJ > 	including the error processing of the SS$_NOSUCHFILE condition code you7 > 	likely could have shortcut this effort considerably.o > I > 	*********** The following analysis thanks to Craig Berry *************r > 6 > 	From get_dirs in swing.c after an IO$_ACCESS qio:   >    >         /*  . >         **  file in dir but not indexf.sys   >         */  * >         if (iosb[0] == SS$_NOSUCHFILE)  
 >         {     >             fib.fib$l_wcc++;   >             continue;  s
 >         }  l >   E > 	Isn't the "wild card context" of the file information block one oflG > 	those opaque longword values that one should not depend on coming in,H > 	sequence, much less increment oneself and hope that doing so will notH > 	get you in trouble?  I certainly don't see anything in the I/O User'sI > 	Guide that suggests I could predict or should attempt to influence therI > 	value of the wcc.  I have a strong suspicion that the code I've quotedtD > 	is the source of the looping problem.  Section 1.3.1.2 of the I/O > 	User's Guide says:t >   G > 	"On the first lookup, FIB$L_WCC should contain zero entries. On eachhF > 	lookup, the ACP returns a nonzero value in FIB$L_WCC, which must beF > 	passed back on the next lookup call. In addition, you must pass theE > 	resultant name string returned by the previous lookup using the P4 E > 	result string buffer, and its length in the P3 result length word.tG > 	This string is used together with FIB$L_WCC to continue the wildcarde4 > 	search at the correct position in the directory." >    ------------------------------   Date: 28 Oct 2002 23:38:25 GMT) From: davis@space.mit.edu (John E. Davis) % Subject: Re: OpenVMS 7.3-1 and CSWING 1 Message-ID: <slrnarrijg.av5.davis@aluche.mit.edu>d  H On Wed, 23 Oct 2002 13:05:28 +0200, Rainer Giese <waste.not@welcome.net> wrote:M >It apparently was John E.Davis (found in main.c), but it is no longer on his I >homepage (http://space.mit.edu/~davis/index.html). Modifications made byoF >Foteos Macrides (http://www.ku.edu/cwis/people/Foteos.Macrides.html).  E I had nothing to do with CSWING other than writing MOST, which CSWINGaF used for paging files.  You will find that I still maintain MOST-- see+ http://www.jedsoft.org/most/ for more info.p   --John   ------------------------------   Date: 29 Oct 2002 00:59:43 GMT# From: hoff@hp.nospam (Hoff Hoffman)h% Subject: Re: OpenVMS 7.3-1 and CSWING,) Message-ID: <apkmhv$7l$1@web1.cup.hp.com>   L In article <B9DCEA92.18D0%calo@shaw.ca>, Carlos Costa <calo@shaw.ca> writes:J :Yes, I see what they are saying now. I should add, though, that a "panic"M :has not occurred. I've gotten "panics" in the past and they output a message L :and exit the application. This has not happened. So I guess that means thatD :SS$_NORMAL is always being returned, which is a good thing I guess.  K   You cannot generalize quite this far.  You just haven't gotten caught yet K   by this bug.  You have clearly gotten caught by at least one other latent2L   bug, and the specified code path will mishandle all successful completions   other than SS$_NORMAL.  M :SS$_NOMOREFILES is being returned (sometimes anyways) as I've seen it duringa7 :my own feeble attempts to figure out what is going on.s  K   Please start by skimming the programming concepts manual and some of the tK   related documentation.  If you don't know the rules and some of the basictI   assumptions of OpenVMS programming as yet -- and no offense is intendedvI   here, as every experienced OpenVMS programmer once had no clue how thissH   stuff works and how it fits together -- you won't know a latent coding/   bug even when you are looking directly at it.A  I   For a list of some of the more common programming bugs found lurking inoK   various code modules, please see topic (1661) in the Ask The Wizard area.b    0 On Wed, 23 Oct 2002 09:30:27 UTC, Paddy O'Brien $ <paddy.o'brien@tg.nsw.gov.au> wrote:  G :> Regardless, they are names that I know I have encountered elsewhere 2F :> in the VMS world and the person who said that it was poorly writtenI :> and should never have worked (paraphrasing) is denigrating some people ? :> who, IIRC, have helped to make the VMS world a better place.r    D   The cited chunk of C error-handling code is poorly written, and isE   (IMNSHO) insufficiently defense.  There is no shortage of bad code.nF   If we take this or other discussions of mistakes personally, we willD   never learn how to avoid the problems.  If we learn what defensiveF   error handling and proper synchronization and other such code shouldH   look like, we will have fewer (repeat) problems with application code,:   and hopefully fewer folks having troubles with upgrades.  J   (Thought question: how do you think that I know that the code construct H   that was shown was questionable?  Do you think I might have once codedJ   this particular construct into my own application code, and subsequently/   encountered problems with the construct?  :-)n  N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.come   ------------------------------  % Date: Mon, 28 Oct 2002 22:09:59 -0600 7 From: "Craig A. Berry" <craigberry@nospam.telocity.com>w% Subject: Re: OpenVMS 7.3-1 and CSWINGdG Message-ID: <craigberry-6F60D1.22095928102002@news.directvinternet.com>a  : In article <JFEPKAPBPMDFDBOIANGDGEKPCBAA.dallen@nist.gov>,%  "Dan Allen" <dallen@nist.gov> wrote:l  P > 	And as Craig Berry has pointed out (see below) the problem would appear to be$ > 	an application error in CSWING.    C I want to make clear that my intention was to point aficionados of  E CSWING toward a possible solution.  The ratio of blame to insight in  & this thread has been a bit depressing.   ------------------------------  % Date: Mon, 28 Oct 2002 16:25:47 -0500r From: CINDY@BILBO.UINDY.EDUn Subject: Question on XFC0 Message-ID: <021028162547.4c933@BILBO.UINDY.EDU>   Hi All,   L I just applied the newest XFC ECO.  Set the VCC_FLAGS back to 2 in SYSGEN.  J Ran Autogen and rebooted the system.  Our system is an Alpha ES40 running  VMS 7.3.  4 When I execute a SHOW MEMORY/CACHE command it shows:    Vols in Full XFC Mode 0G    Vols in VIOC Compatible Mode 3 (we have 3 disk drives on our system)t  8 When I execute a SHOW MEMORY/CACHE=VOL=DKA100: it shows:#    Caching mode is VIOC Compatible.i  G What else do I need to do to the system to cause it to begin using XFC?h   Thanks.e  M *****************************************************************************eJ Cindy Steinmetz                               E-Mail: steinmetz@.uindy.eduL Director Administrative Computing Services            System@Bilbo.Uindy.EduA University of Indianapolis                    Fax: (317) 788-3300tG 1400 E Hanna Avenue                           Telephone: (317) 788-3361< Indianapolis, IN  46227-3630M *****************************************************************************iB          In Music there is harmony .... In harmony there is peace.M *****************************************************************************    ------------------------------    Date: 28 Oct 2002 14:49:50 -0600+ From: young_r@encompasserve.org (Rob Young)y Subject: Re: Question on XFC3 Message-ID: <L3o1c7GpxmgB@eisner.encompasserve.org>o  N In article <021028162547.4c933@BILBO.UINDY.EDU>, CINDY@BILBO.UINDY.EDU writes:	 > Hi All,  > N > I just applied the newest XFC ECO.  Set the VCC_FLAGS back to 2 in SYSGEN.  L > Ran Autogen and rebooted the system.  Our system is an Alpha ES40 running 
 > VMS 7.3. > 6 > When I execute a SHOW MEMORY/CACHE command it shows: >    Vols in Full XFC Mode 0I >    Vols in VIOC Compatible Mode 3 (we have 3 disk drives on our system)M > : > When I execute a SHOW MEMORY/CACHE=VOL=DKA100: it shows:% >    Caching mode is VIOC Compatible.  > I > What else do I need to do to the system to cause it to begin using XFC?  >   B 	That "Full XFC Mode" throws folks for a loop.  You are using XFC. 	Several tidbits to show that:  F <node>$ show mem/cache/full  ! Small box , limited memory, single disk  @               System Memory Resources on 28-OCT-2002 16:30:20.84  B Extended File Cache  (Time of last reset: 25-OCT-2002 10:59:32.11)L   Allocated (MBytes)            7.06    Maximum size (MBytes)          64.00L   Free (MBytes)                 0.71    Minimum size (MBytes)           3.12M   In use (MBytes)               6.35    Percentage Read I/Os              86%nM   Read hit rate                   95%   Write hit rate                     0%aL   Read I/O count              133463    Write I/O count                21584L   Read hit count              127661    Write hit count                    0L   Reads bypassing cache           37    Writes bypassing cache             0L   Files cached open              454    Files cached closed               43L   Vols in Full XFC mode            0    Vols in VIOC Compatible mode       1L   Vols in No Caching mode          0    Vols in Perm. No Caching mode      0  1 I/O Statistics - Distributions (MAX_IO_SIZE: 127) 1 -------------------------------------------------h. Transfer Size:      Reads  Read Hits    Writes.    1 Block IO:      54038      53264      4160.    2 Block IO:       2445       2049      1562.    3 Block IO:      38591      38100       459.    4 Block IO:        801        656      1176.    5 Block IO:      19038      18950       189.    6 Block IO:        767        670        25.    7 Block IO:       1133       1052        12.    8 Block IO:       1771       1643        73   [snip]  * <node>$ mcr sysgen help sys_param vcc_flag   Sys_Parameters     VCC_FLAGS-   	[snip]a  G        2      Enables file system data caching and selects the Extended @               File Cache. This is the default for Alpha systems.     <node>$ analyze/system  " OpenVMS (TM) Alpha system analyzer   SDA> xfc XFC> show summaryr
 XFC> <return>o
 XFC> <return>    XFC Summary  -----------l, System Wide Read Statistics since last reset, --------------------------------------------, Virtual Reads:                        134288,   Sum of Read Around Count:               37,   and Read Through Count:             134251, Reads Completed:                      134287, Read Hits:                            128314. Read Cache Hit Percentage:             95.55 %, Total Synch Completion Count:         107837, Read Around due to Het. Cluster:           0, Read Around due to Modifiers:              0, Read Around due to Size:                  36, Total reads past EOF:                   1145, Total I/Os with read-ahead:              473, Read Hits due to read-ahead:             277, Paging I/Os:                           20236   [snip]   	Here is what VIOC looks like:  ( <node2>$ show mem/cache/full   ! VMS 7.2@               System Memory Resources on 28-OCT-2002 16:35:37.48   Virtual I/O Cache5N     Total Size (Kbytes)         ___000    Read IO Count                8860660N     Free Kbytes                 ___992    Read Hit Count               8134133N     Kbytes in Use               ___008    Read Hit Rate                    91%N     Write IO Bypassing Cache    187784    Write IO Count               5319619N     Files Retained                  91    Read IO Bypassing Cache       286716  C 	99 files max retained in cache, several hundred above on 7.3-1 XFCd' 	node.  Much smaller VIOC display, etc.c   				Robp   ------------------------------  % Date: Mon, 28 Oct 2002 13:02:23 +0000d( From: Nic Clews <sendspamhere@127.0.0.1>1 Subject: Re: Quorum disk appears to not be votinga) Message-ID: <3DBD355F.95D6414B@127.0.0.1>o   Stephen McElduff wrote:r > F > I am trying to configure a 2 Node DecNet based cluster with a quorumG > disk. As best I can tell, I have set all of the sysgen parameters (in F > the modprarams.dat) files on each system such that DKA1 should be my > quorum disk. > F > I have configured it to give each node a single vote and to give theA > quorum disk a single vote, so that if any single node dies, thecG > cluster will transition but remain available (after a transition) for 
 > my apps....p > G > I believe that my quorum disk is not contributing a vote (see detailsr > below)  * Can you post the results of the following?  
 $ ANAL/SYS SDA> SHOW CLUSTERs (few pages)u  C Here I'm expecting to see from each systems' point of view what theo state of the quorum disk is.   $ SHOW DEV DKA1:/fullw (on each system)  @ Here we'll have host name (and alternate host name if there's anA alternate path), just so we know both systems can see the disk. I > presume the ALLOCLASS figure is for some local disks, and your% DISK_QUORUM isn't really "$200$DKA1:"   E For the quorum disk's vote to be counted, it has to be seen, and alsotG trusted, all members of the cluster with a direct connection need to ber/ able to make their mark in the QUORUM.DAT file.r   --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciencess nclews at csc dot come   ------------------------------   Date: 29 Oct 2002 01:28:05 GMT# From: hoff@hp.nospam (Hoff Hoffman) 1 Subject: Re: Quorum disk appears to not be votinge) Message-ID: <apko75$7l$3@web1.cup.hp.com>e  w In article <1a941731.0210251401.2e518336@posting.google.com>, stephen.mcelduff@transcore.com (Stephen McElduff) writes:f8 :I am trying to configure a 2 Node DecNet based cluster   7   There is no such thing as a "DecNet based cluster".  a  ?   DECnet and IP and LAT are not involved in cluster operations.I  J   The MOP protocol does get involved in the initial download and bootstrapK   of cluster satellite nodes and this does provide some confusion.  The MOPsK   support itself is provided by DECnet Phase IV, by DECnet-Plus, and by theoK   OpenVMS LANCP tool on V6.2 and later.  Once the MOP download of a clusterhK   satellite is complete, MOP is no longer involved (barring an upline dump,HB   but I digress) in cluster operations.  MOP is not involved with =   non-satellite OpenVMS bootstraps within an OpenVMS Cluster.o  *   Clustering does not operate over DECnet.  &   Clustering does not operate over IP.  '   Clustering does not operate over LAT.m  M   Clustering operates over the System Communications Services (SCS) protocol.   F :with a quorum disk. As best I can tell, I have set all of the sysgen H :parameters (in the modprarams.dat) files on each system such that DKA1  :should be my quorum disk. :pE :I have configured it to give each node a single vote and to give thet@ :quorum disk a single vote, so that if any single node dies, theF :cluster will transition but remain available (after a transition) for :my apps....  B   Please see the OpenVMS FAQ for setting VOTES and EXPECTED_VOTES.E   For your configuration with two nodes and a quorum disk, the quorumhH   disk MUST be configured on a multi-host shared interconnect to operate   the way you expect.'  F :I believe that my quorum disk is not contributing a vote (see details :below)   I   Please see the existing discussions of two-node clusters and of settingrG   setting VOTES and EXPECTED_VOTES in the OpenVMS manuals, here in the  I   newsgroup archives, in the Ask The Wizard area, and particularly in thenI   OpenVMS FAQ.  Unless you have a multi-host disk involved here, there is H   no reason and no benefit of using a quorum disk in this configuration.  < :A cut/paste from my modparams.dat file follows: (the votes,B :expected_votes, disk_quorum values are the same on both machines)  B   OpenVMS version and platform, and the specific disk and the disk&   controller type(s) involved, please?  A   The DK device name implies the use of SCSI, and multi-host SCSIw@   disk storage supported on specific SCSI controllers on OpenVMS   Alpha systems.  N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comu   ------------------------------   Date: 29 Oct 2002 02:41:48 GMT, From: bill@gw5.cs.uofs.edu (Bill Gunshannon)! Subject: Re: Rejuvenating OpenVMSr5 Message-ID: <apkshc$2e30n$1@ID-135708.news.dfncis.de>i  A Probably need an education program for companies trying to market  VMS Software too.   ? I received SPAM from Mom Software Services.  I sent them a copyr? of The Boulder Pledge.  They responded by SPAMing me yet again. @ VMS doesn't have enough available software that it can afford to have any of them blacklisted.o   billB PS.  I would expect everyone on c.o.v is going to get at least one copy of this SPAM if not more.   -- iJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   j   ------------------------------  % Date: Mon, 28 Oct 2002 14:20:16 -0600 * From: "Dave Gudewicz" <dgudewicz@core.com>H Subject: Re: Rejuvenating OpenVMS [was Re: So I went to the HP IT forum]8 Message-ID: <3dbd9c01$0$1449$1dc6e903@news.corecomm.net>  I If VMS is considered legacy, then what does one label UNIX?  Its about 10a years older than VMS.a   -- Dave...t  G It is noble to teach oneself, but still nobler to teach others-and lessi trouble. -----Mark TwainI  5 "Bob Ceculski" <bob@instantwhip.com> wrote in messagec7 news:d7791aa1.0210271029.54740f6c@posting.google.com...a: > JF Mezei <jfmezei.spamnot@videotron.ca> wrote in message' news:<3DBB2985.B45A651@videotron.ca>...y > > Brian Chase wrote:K > > > I think getting OpenVMS back into universities as a general computinghD > > > platform would be very difficult.  VMS has long since lost its momentum > > > in academia. > >-I > > Unix is a clustering wannabe.  VMS has what Unix will have years fromu now. > > J > > Teachers and graduate students currently wouldn't want a legacy system calledI > > VMS. But what if Digital/Compaq/HP/whetever were to volunteer its VMSaK > > ambassadors to give lectures on clustering and disaster tlleran issue ?e InF > > those lectures, obviously VMS' strengths would be used as example. > >s > > This will have 2 effects:s > >tK > > -students, graduate students and teachers will find that those lecturest aresI > > giving students information/experience on state of the art clusteringb thatK > > their current unix environmment cannot provide. So those guest lecturesI willD > > provide the univeristy with a more complete curriculum. And give students an K > > edge when they get jobs or start volunteering to ikrpove Linux etc etc.h > >pJ > > Having the best clustering system doesn give you any payback unless to makeK > > sure people know you have the best clustering system. If students don't  knowI > > VMS is 10 years ahead of Unix in clustering, then students won't tell  theirhJ > > future employers that the proposed unix solution is far behind VMS for1 > > clustering and mission critical applications.e > >yI > > And by prosenting VMS as a stand of the art OS to universities, those H > > ambassadors will get a foot in the door and work to change attitudes within2 > > academia to be more open/positive towards VMS. > >,E > > Giving away VMS boxes is pointless if the teachers don't want VMSi because theyJ > > see VMS as a legacy system. Present VMS as something the teachers need to haveoK > > in order to teach the state of the art, then all of a sudden, VMS gains*< > > stature and the door starts to open for VMS in academia. > > L > > Note that I had made such a suggestion/proposal to Compaq some time ago, butp( > > of course, no response was received. >nA > vms is a legacy system ... its been around 25 years now with no @ > competition, and it will be around another 25 years ... that's > a heck of a legacy ...   ------------------------------  # Date: Mon, 28 Oct 2002 23:37:22 GMTb# From: "John Smith" <a@nonymous.com>dH Subject: Re: Rejuvenating OpenVMS [was Re: So I went to the HP IT forum]H Message-ID: <SSjv9.99238$Q3S.24882@news01.bloor.is.net.cable.rogers.com>  5 "Dave Gudewicz" <dgudewicz@core.com> wrote in message62 news:3dbd9c01$0$1449$1dc6e903@news.corecomm.net...K > If VMS is considered legacy, then what does one label UNIX?  Its about 10  > years older than VMS.k    
 Desirable.   Because it is marketed.p   ------------------------------  + Date: Mon, 28 Oct 2002 12:36:49 -0800 (PST)t. From: Fabio Cardoso <fabiopenvms@yahoo.com.br>0 Subject: Re: SAP? still a market for consultants@ Message-ID: <20021028203649.15410.qmail@web20202.mail.yahoo.com>   Didier  3 I will move from OVMS job soon ! But I am searchingmE for some BEA stuff. I think these middlewares like Weblogic, MQSerieshF and etc ... is the better choice for us from the VMS community change  jobs ! y   Regards  FC f9 --- Didier Morandi <Didier.Morandi.nospam@Free.fr> wrote:lD > I have been suggested to think "go SAP" in case I find no more VMS > missions in Europe.r >  > I have a few questions:l >  > 1. what do you think?o@ > 2. how long it will take to have some (real) knowledge to sellH > 3. Should I try to join a Big 6, a small (total_nr - 6), a third-party > vendor, or stay independent? >  > Thanks for your advice.I >  > D. >  > ---------------------------o > Posted with MacSoup 2.4.6  > Remove .nospam to email me.f     =====4 ========================== Fbio dos Santos Cardoso OpenVMS System Manager Rio de Janeiro - Brazili fabiopenvms@yahoo.com.br ==========================  2 __________________________________________________ Do you Yahoo!?2 Y! Web Hosting - Let the expert host your web site http://webhosting.yahoo.com/   ------------------------------  % Date: Mon, 28 Oct 2002 22:46:00 +0100 6 From: Arne =?ISO-8859-1?Q?Vajh=F8j?= <arne@vajhoej.dk>0 Subject: Re: SAP? still a market for consultants) Message-ID: <3DBDB018.2010002@vajhoej.dk>p   Bill Hobbs wrote:    > Didier.Morandi.nospam@Free.fr (Didier Morandi) wrote in message news:<1fkr7q9.14qsdfd611rnkN%Didier.Morandi.nospam@Free.fr>...D >>I have been suggested to think "go SAP" in case I find no more VMS >>missions in Europe.h    F > For a product that is supposed to already do everything, my previous6 > employer is in its second year of SAP customization.    E > SAP appears to be the Microsoft of the ERP sector: great marketing,wH > crap product which one can render as they are great at marketing crap.G >  SAP sales people apparently have some of the best reality-distortiono > generators on the planet.r  8 Why spend money fixeing bugs, when you can get customers to pay you fixing them ?  
 :-) or :-(   Arne   ------------------------------  % Date: Mon, 28 Oct 2002 22:59:02 +0100e4 From: Didier.Morandi.nospam@Free.fr (Didier Morandi)0 Subject: Re: SAP? still a market for consultantsA Message-ID: <1fkscpt.s3rqt7l8s2ioN%Didier.Morandi.nospam@Free.fr>e  6 Bill Hobbs <bdhobbs18@acm.org> and others wrote a lot: ../..i    E So, as a 20+ VMS specialist, after having read the 250 first pages of H Jos Antonio Fernandez's SAP Handbook (980 pages, McGraw-Hill 1997), whoH is, btw, a former DECcie, and all your answers in here, I may understandC that a set of VMS-like tools could be useful to the poor SAP systema mgrs.'   That could be a start.  E About training, it seems that there are two routes to follow: project F manager, with knowledge of the Client's business (FI, CO, HR, etc) andD implementation specialist. As I do not know/like UNIX, the choice is done.o  D As far as my soul is concerned, if we are a few to find a good idea,C maybe we could produce and sell an SAPxorcist package :-) such likesE mySAP or so, but with the VMS touch, which may be summarized (as Rossy> Perot would say) in three words: quality, quality and quality?   D. ---------------------------e Posted with MacSoup 2.4.6n Remove .nospam to email me.o   ------------------------------  # Date: Mon, 28 Oct 2002 23:32:34 GMT # From: "John Smith" <a@nonymous.com> 0 Subject: Re: SAP? still a market for consultantsI Message-ID: <mOjv9.118660$%h2.29676@news02.bloor.is.net.cable.rogers.com>M  ? "John Eisenschmidt" <jweisen@eisenschmidt.org> wrote in messagea. news:20021028184529.C10239@eisenschmidt.org...D > Unless the Voices are Mistaken, John Smith (a@nonymous.com) Wrote: > > C > > "John Eisenschmidt" <jweisen@eisenschmidt.org> wrote in messageo2 > > news:20021028172109.I31845@eisenschmidt.org... > > > F > > > I'm a 20-something IT professional who is both jaded and foolishK > > > enough to think I can change the world. Someone loan me a few milliongJ > > > and I'll go develop an account system that doesn't suck. We all need1 > > > dreams, but mine are pretty much fantasies.  > > >BI > > > Though I can see the ads: Britney Spears likes my financial system,h > > > and so should you! > > & > > http://www.sunsystemsamericas.com/ >eD > I smell vapor. That much Javascript can't be the sign of a healthy@ > company. I have a rule that if the site doesn't render well inC > Mozilla, then it's not a company worth considering. That, and the4D > sheer lack of technical details scares me. I'd rather you lied andF > tried to impress me with buzz words than said nothing -- a strategicE > partnership with Microsoft and Oracle does not a good product make.  >oH > I was bitching about these applications not being abstracted properly.F > Case in point: websites should be abstracted to run on any platform. >nD > Say I build a site and the boss tells me it needs to run on IIS. IE > should be able to go to the NT server and check the code out of theoD > source control system (because we all use source control, RIGHT?).H > Three months later boss says: it needs to move to Webstar on MacOS. GoB > to the Mac, check out the latest source, boom it works. A coupleF > months later we start sleeping with IBM, check the latest source out$ > on my AIX server running Webspere. >tC > "But lunatic, " you ask, "what about server side programming? Fora. > where shall my scripts and the like reside?" > H > Well, if you designed it in something portable, like Perl or whatever,C > you just load the tool you need on the platform of choice. If you9H > designed it on something that isn't portable, you should be shot. ThisC > is what the CGI paradigm is for: running non-markup applications.e >eE > You want speed? Then look at something like mod_perl for Apache. AtkH > least Apache will run on VMS/Unix/Windows/MacOS/ToasterOS/etc. If not,G > it's fine with me, just pick something that will allow you to turn ond	 > a dime.p >xD > Now, I know some of you hate Perl, so YMMV. Pick your poison, withH > CGIs you could compile ADA and have it executed with your form action.E > There is no law that says you can't write a program in assembler towE > dynamically generate the HTML that your web server spews forth, but@F > that would kind of shoot my "abstraction" idea, so don't do that ok? >kH > But don't take my word for it, I think I saw a team of ASP programmersF > hanging out in front of 7-11 the other day with "will code for food" > signs.    I Their accounting engine has been around for nearly 20 years and is a gooda@ one. It used to run on VMS - I don't know about that any longer.  J You could purchase the product from them as a package or as an engine that" you could bolt into your own apps.  I  I looked at it off and on over the years and was very impressed with therG quality of the basic engine. I was considering it for possible use in asD derivatives accounting system and other related applications, but myK customers never decided to bite, preferring instead to run their businesseseI on Lotus and Excel spreadsheets with manual, no audit trail, entries intob their accounting systems.a  I As to their web site, I cannot comment. Many web sites of companies you'di< think ought to know better, are often built by 16-year olds.   ------------------------------  % Date: Mon, 28 Oct 2002 20:32:25 -0500 + From: John Johnstone <jj_usenet2@yahoo.com> / Subject: SMTP_CFGETERRORs with TCPIP V5.3 ECO 1 ) Message-ID: <3DBD9ED9.5CED23E4@yahoo.com>i  E This is an update on an ongoing problem that I've had with TCPIP SMTPn mail.t  D I'm seeing a problem with mail delivery of some messages on an AlphaF with TCPIP V5.3 ECO 1.  I was first seeing the problem with TCPIP V5.1@ but the problem still exists with V5.3.  The messages fail to be@ delivered with the following errors showing in the SMTP logfile:  @ A permanent error makes this mail undeliverable. Must bounce it.C %TCPIP-E-SMTP_CFGETERROR, error getting address/domain from controle!  file, type of tridentusa.com<CR>P5 -RMS-W-RTB, 0 byte record too large for user's buffer<  lE The domain name that's shown in the error message is our domain.  TheCD <CR> that I've shown on the end is actually a single carriage-return5 character in the log file when viewed with an editor.S  = I've confirmed that the problem is definitely due to multiple<@ occurrences of successive carriage-return characters in the mail< message.  I've captured packets via TCPTRACE that show this.  H Most, if not all, of the messages are coming from Microsoft mail serversF for whatever that means.  Most, possibly all, of the messages are sentF via automated processes, not mail that's composed by human hands.  The messages will show lines like:  =    A line of text<CR><CR><LF>Another line of text<CR><CR><LF>y   rather thanh  5    A line of text<CR><LF>Another line of text<CR><LF>e  B It's seems that the two successive carriage-returns cause the SMTPH receiver to create a bogus control file entry for the message.  When the9 SMTP symbiont then goes to deliver the message, it bombs.   G I've been able to re-process the mail messages via the SMTP SFF utilityiC without errors.  If I take the original message from the _TEXT fileeA which includes these <CR> strings and use it to create a new mailtG message with the SMTP commands for SFF added, SFF sends it and the SMTPa% symbiont delivers it with no problem.n  H Although the 5.3 release notes state that the SMTP control files are now8 different, that hasn't made any difference for this bug.   ------------------------------  % Date: Mon, 28 Oct 2002 21:04:36 -0500u- From: JF Mezei <jfmezei.spamnot@videotron.ca> 3 Subject: Re: SMTP_CFGETERRORs with TCPIP V5.3 ECO 1l, Message-ID: <3DBDECB3.5C8955DB@videotron.ca>  L I recently had problems with SMTP. the local MX information is stored in theK ROUTES database for some strange reason. And I had no local MX records, nor>= could I add one due to some RMS problem with the routes file.i  J I deleted the routes file, crecreated a new one, repolulated the permanentF routes and added a single local MS record. And since then no problems.   ------------------------------  % Date: Mon, 28 Oct 2002 21:42:35 -0500r+ From: John Johnstone <jj_usenet2@yahoo.com>i3 Subject: Re: SMTP_CFGETERRORs with TCPIP V5.3 ECO 1 ) Message-ID: <3DBDAF4B.783FF04D@yahoo.com>    JF Mezei wrote:  > N > I recently had problems with SMTP. the local MX information is stored in theM > ROUTES database for some strange reason. And I had no local MX records, nory? > could I add one due to some RMS problem with the routes file.  > L > I deleted the routes file, crecreated a new one, repolulated the permanentH > routes and added a single local MS record. And since then no problems.  > That's probably not related to my problem here.  The MX recordB definition for our domain is provided by a DNS server that is on aA different system.  I agree that it's not obvious why the local MXe" records reside in the routes file.   ------------------------------  % Date: Mon, 28 Oct 2002 22:27:42 -0500I- From: JF Mezei <jfmezei.spamnot@videotron.ca>e3 Subject: Re: SMTP_CFGETERRORs with TCPIP V5.3 ECO 1O, Message-ID: <3DBE0026.C31D97F0@videotron.ca>   John Johnstone wrote:v@ > That's probably not related to my problem here.  The MX recordD > definition for our domain is provided by a DNS server that is on a > different system.   K That was my case too. But the SMTO server first tries the route database to L resolve a domain name and then goes to the defined DNS server. If the routes( database was empty, it would fail there.  R have you tried TCPIP SHOW MX ?  If it fails, then you have the same problem I had.   ------------------------------  # Date: Mon, 28 Oct 2002 14:48:08 GMT # From: "John Smith" <a@nonymous.com>t) Subject: Re: So I went to the HP IT forumaI Message-ID: <I6cv9.85679$mxk1.56063@news04.bloor.is.net.cable.rogers.com>n  5 "Bill Gunshannon" <bill@cs.uofs.edu> wrote in messagel/ news:apjde3$224id$1@ID-135708.news.dfncis.de...  > >*D > > Mark Gorham appears to read comp.os.vms/info-vax on a regular orH > > semi-regular basis. I have had email responses from Mark to comments > > made here. > I > So maybe we should just post something asking him if he is reading thisSC > and soliciting his comments??  Oh wait, I think I just did!!  :-)     K He couldn't make any comments here without an army of lawyers first readingoL what he proposed to say, then having the lawyers modify his intended remarksF until they said nothing committal one way or the other, and then thoseI edited remarks would have to be voted on by the BOD. After all, this is a'& public forum and not subject to a NDA.  J Expect any replies to questions posted today to Mark to appear sometime in the 2011 timeframe.a   ------------------------------  % Date: Mon, 28 Oct 2002 09:58:17 -0500n- From: JF Mezei <jfmezei.spamnot@videotron.ca>') Subject: Re: So I went to the HP IT forumu, Message-ID: <3DBD5075.E7D7B54E@videotron.ca>   Bill Gunshannon wrote:D > > Mark Gorham appears to read comp.os.vms/info-vax on a regular orH > > semi-regular basis. I have had email responses from Mark to comments > > made here. > I > So maybe we should just post something asking him if he is reading thistC > and soliciting his comments??  Oh wait, I think I just did!!  :-)L   No, you post something like: "sJ Some lady with a blue dress in a new england restaurant told me on a cold,O raisy automn day that Mark Gorham is the one who caused Sue's shoulder ailment.D "s  J And then you wait for a message from Sue saying that Mr Goham asked her to deny this allegation   :-) :-) :-) :-) :-):   ------------------------------  % Date: Mon, 28 Oct 2002 15:17:25 +0000J% From: Alan Greig <a.greig@virgin.net>s) Subject: Re: So I went to the HP IT forum.8 Message-ID: <1lkqru85e57mm5es09sf4oco0v63ode5bo@4ax.com>  F On 28 Oct 2002 13:17:57 GMT, bill@cs.uofs.edu (Bill Gunshannon) wrote:   >re. >dH >So maybe we should just post something asking him if he is reading thisB >and soliciting his comments??  Oh wait, I think I just did!!  :-)  F I think he suspects that if he starts posting publicly he'll get drawnC in too far and never get any real work done. And I don't know if hehE just skims or reads regularly. But he has responded directly by email. to a post of mine.   -- Alan -- Alan   ------------------------------  % Date: Mon, 28 Oct 2002 14:58:01 +0000r' From: Andrew Harrison SUNUK Consultancya- Subject: Re: TCPIP 5.3 and SMTP authorizations. Message-ID: <3DBD5079.7010801@nospamn.sun.com>   David Webb wrote:T > In article <3DB97816.8040108@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> writes: >  >> >>David Webb wrote:  >>[ >>>In article <apbcts$v5s@newton.cc.rl.ac.uk>, "Richard Brodie" <R.Brodie@rl.ac.uk> writes:o >>>h >>>te >>>>"JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message news:3DB91ED3.2A473E2D@videotron.ca...R >>>> >>>> >>>>: >>>>>Now, going to my ISP with a "real" SMTP server: (sun)S >>>>>220 VL-MS-MR004.sc1.videotron.ca -- Server ESMTP (iPlanet Messaging Server 5.2b >>>>>iP >>>>>So the SMTP server that comes with TCPIP Services isn't keeping up with the& >>>>>Joneses in terms of the commands. >>>>Z >>>>That's a bit apples and oranges. Having a quick look at Sun's web site, the list price' >>>>for their server starts at $20 000.h >>>> >>>> >>>i >>> P >>>And you can get pretty much the same MTA as in the SUNone/Iplanet product for8 >>>VMS, Tru64 (and even WindowsNT) from Process as PMDF. >>>o >>C >>Users of both would find that they are similar but they have bothlC >>diverged and have different capabilites, direct LDAP for example.  >> >  > 	 > Andrew,C > N > Are you saying that SunOne doesn't have LDAP support - since PMDF has had it > for quite sometime ? >     Nope I said direct LDAP support.   >  > = >>PMDF is the SunOne/iPlanet messaging server MTA, a previouse> >>version  was also the MTA integrated into SIMS Sun's earlier >>Messaging Server.o >>> >>PMDF was integrated into SIMS 3.x and 4.0 and was called IMS" >>if my memory serves me correctly >>E >>The PMDF MTA uses SMTP/ESMTP to transfer mail to other SMTP serverse2 >>and has an API to allow people to program to it. >>? >>The message store uses HTTP/IMAP4/POP to talk to its clients.w >>C >>I hope this may go some way to clear up a long standing confusiont@ >>on your part as to what function PMDF actually performs in theB >>SIMS/iPlanet Messaging server. It is not the IMAP server its the >>MTA. >> > : > I've always referred in our discussions to the PMDF MTA.N > You are the one who has in the past referred to SIMS as just an IMAP server.A > I'm glad we both agree that the common element is the PMDF MTA.o >   < Actually this isn't quite true, you refered to SIMS as being< based on PMDF, it isn't. The MTA from 3.x onwards is but the rest isn't.-  ( This is an entire posting that you made.  E "Although I agree with much of what you say. This is now a done deal.-H Compaq will definitely lose sales, upgrades and revenue from maintenanceH as a result of this if they do nothing. Hopefully this will spur them to do something about it.A Note. Sun survived quite happily for a number of years with theira: flagship mail product SIMS using an OEM'd version of PMDF.D Hence at least in the short term a similar deal might be struck with iPlanet.B Indeed if Sun is serious about wanting iPlanet to support multipleH platforms and not appear to be a poodle of Sun then it might be possibleA for Compaq to do a deal to make sure all the iPlanet products areNE supported on VMS and Tru64. (Eg providing Compaq staff and systems to ( help in porting/ maintaining software)."  / The previous thread makes no reference to MTA's   7 In fact it is clear that you have undergone a change of 0 view because it we go back a bit further we get.   Question from Jerry Alan BragaL  >What I am really looking for is an IMAP type package that will allow email  >anywhere over the WEB.  >L  >I know innosoft is out there but now SUN owns it and most likely thats the  >end of the VMS version.a  >   Your response.  I "SUN have passed support and development of PMDF over to process softwareF* see http://www.process.com/tcpip/pmdf.html  L (Seems like SUN only wanted PMDF so that they had ownership of the SIMS codeG - which could then be given to Iplanet - and the Innosoft staff (?) )."I  > Strangely when I refer to SIMS as Sun's IMAP/POP3 server which< is entirely true as POP/IMAP are protocols supported by SIMS) you have seen fit to try to correct this.   < I say strangely because in a discussion you were having with
 Arne he said:v  D "1)  As far as I know, then SUN bougth Innosoft to get the directory?      server, which is a key component in an enterprise solutionA      today.   > 2)  Again as far sas I know, then SUN did not create their ownA      mail product, they made an OEM version of PMDF. And that was*,      available before they bougth Innosoft."   You didn't correct him !!e   Odd.   Regards* Andrew Harrison*   ------------------------------  + Date: Mon, 28 Oct 2002 17:46:22 +0000 (UTC)-+ From: david20@alpha1.mdx.ac.uk (David Webb)r- Subject: Re: TCPIP 5.3 and SMTP authorizationj+ Message-ID: <apjt5e$qah$1@aquila.mdx.ac.uk>G   In article <3DBD5079.7010801@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> writes: >  >t >David Webb wrote: >> In article <3DB97816.8040108@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> writes:M >> . >>>  >>>David Webb wrote: >>>O\ >>>>In article <apbcts$v5s@newton.cc.rl.ac.uk>, "Richard Brodie" <R.Brodie@rl.ac.uk> writes: >>>> >>>>f >>>>>"JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message news:3DB91ED3.2A473E2D@videotron.ca... >>>>>s >>>>>o >>>>>e; >>>>>>Now, going to my ISP with a "real" SMTP server: (sun) T >>>>>>220 VL-MS-MR004.sc1.videotron.ca -- Server ESMTP (iPlanet Messaging Server 5.2 >>>>>>Q >>>>>>So the SMTP server that comes with TCPIP Services isn't keeping up with thet' >>>>>>Joneses in terms of the commands.y >>>>>m[ >>>>>That's a bit apples and oranges. Having a quick look at Sun's web site, the list price2( >>>>>for their server starts at $20 000. >>>>>  >>>>>  >>>> >>>>Q >>>>And you can get pretty much the same MTA as in the SUNone/Iplanet product fort9 >>>>VMS, Tru64 (and even WindowsNT) from Process as PMDF.  >>>> >>>RD >>>Users of both would find that they are similar but they have bothD >>>diverged and have different capabilites, direct LDAP for example. >>>  >> t >>  
 >> Andrew, >>  O >> Are you saying that SunOne doesn't have LDAP support - since PMDF has had it  >> for quite sometime ?a >>   > ! >Nope I said direct LDAP support.  >   > Please explain exactly what you mean by "Direct" LDAP support.       >>   >> 3> >>>PMDF is the SunOne/iPlanet messaging server MTA, a previous? >>>version  was also the MTA integrated into SIMS Sun's earliere >>>Messaging Server. >>> ? >>>PMDF was integrated into SIMS 3.x and 4.0 and was called IMS-# >>>if my memory serves me correctlyh >>>nF >>>The PMDF MTA uses SMTP/ESMTP to transfer mail to other SMTP servers3 >>>and has an API to allow people to program to it.0 >>> @ >>>The message store uses HTTP/IMAP4/POP to talk to its clients. >>> D >>>I hope this may go some way to clear up a long standing confusionA >>>on your part as to what function PMDF actually performs in thekC >>>SIMS/iPlanet Messaging server. It is not the IMAP server its the  >>>MTA.  >>>  >> i; >> I've always referred in our discussions to the PMDF MTA.eO >> You are the one who has in the past referred to SIMS as just an IMAP server.gB >> I'm glad we both agree that the common element is the PMDF MTA. >>   > = >Actually this isn't quite true, you refered to SIMS as beinga= >based on PMDF, it isn't. The MTA from 3.x onwards is but theo >rest isn't.  > Ahhhhh Do we have to keep going through this again and again ?       >s) >This is an entire posting that you made.- >-F >"Although I agree with much of what you say. This is now a done deal.I >Compaq will definitely lose sales, upgrades and revenue from maintenanceoI >as a result of this if they do nothing. Hopefully this will spur them to  >do something about it.rB >Note. Sun survived quite happily for a number of years with their; >flagship mail product SIMS using an OEM'd version of PMDF.oE >Hence at least in the short term a similar deal might be struck with 	 >iPlanet. C >Indeed if Sun is serious about wanting iPlanet to support multipleeI >platforms and not appear to be a poodle of Sun then it might be possibleIB >for Compaq to do a deal to make sure all the iPlanet products areF >supported on VMS and Tru64. (Eg providing Compaq staff and systems to) >help in porting/ maintaining software)."5 >.0 >The previous thread makes no reference to MTA's >   F This is about SUN's purchase of Innosoft. And nowhere does it say that
 PMDF == SIMS.2L All it says is that SIMS used OEM'd PMDF Technology. It doesn't specify what7 technology since that was irrelevent to the discussion.B    8 >In fact it is clear that you have undergone a change of1 >view because it we go back a bit further we get.I >y >Question from Jerry Alan Braga M > >What I am really looking for is an IMAP type package that will allow emails > >anywhere over the WEB.  > >	M > >I know innosoft is out there but now SUN owns it and most likely thats thed > >end of the VMS version. > >2 >. >Your response.: >oJ >"SUN have passed support and development of PMDF over to process software+ >see http://www.process.com/tcpip/pmdf.htmly > M >(Seems like SUN only wanted PMDF so that they had ownership of the SIMS code H >- which could then be given to Iplanet - and the Innosoft staff (?) )." >e    J I was responding directly to the question about the future of the InnosoftN products after SUN had purchased Innosoft. PMDF includes it's own IMAP server.4 However that does not make PMDF just an IMAP server.    ? >Strangely when I refer to SIMS as Sun's IMAP/POP3 server whichy= >is entirely true as POP/IMAP are protocols supported by SIMSt* >you have seen fit to try to correct this. >o  ( No I have corrected you when you stated    "SIMS was SUN's IMAP server"   and   K "Solaris Internet Mail Server was replaced by Netscape's IMAP server as thet' standard IMAP server that we supplied."l  K As evidenced by a later posting from Arne Vajhoej this was taken by many toh; mean that iPlanet no longer included any code from Innosofte  ! "Process is selling PMDF for VMS.   I And SUN has dumped all the mail stuff they got from Innosoft, so probably:; Process can do whatever they belive they can make money on.s   "a  M After I pointed him at Docs showing that iPlanet still included the PMDF-MTA r his response was r   "OK.  /  I got my impression from our local SUN FUDist:a "a  - And then precedes to quote your IMAP posting.a      = >I say strangely because in a discussion you were having with  >Arne he said: > E >"1)  As far as I know, then SUN bougth Innosoft to get the directory @ >     server, which is a key component in an enterprise solution >     today. >2? >2)  Again as far sas I know, then SUN did not create their ownmB >     mail product, they made an OEM version of PMDF. And that was- >     available before they bougth Innosoft."a >  >You didn't correct him !! >   N The discussion was about whether any mail products could survive or whether asK Arne seemed to think the future was all in the hands of Microsoft Exchange.lF If you look further down the post you will have seen he's inclusion of* the following from my previous response :-  L "Since it is obvious there is no way I can convince you otherwise I will not9 be responding to any more of your posts on this subject."e  > Hence it is hardly suprising that I did not make any response.      
 David Webb VMS and Unix team leader CCSS Middlesex University       >Odd.O >V >Regards >Andrew Harrison >e   ------------------------------   Date: 28 Oct 2002 22:43:21 GMT# From: hoff@hp.nospam (Hoff Hoffman)rB Subject: Re: The VMSNET group of forums (was: Re: Acrobat Reader?)* Message-ID: <apkei9$om8$1@web1.cup.hp.com>  d In article <ap8qh6$sf1g3$1@ID-135708.news.dfncis.de>, bill@gw5.cs.uofs.edu (Bill Gunshannon) writes:D :In article <1fkjrte.1xzgh9g1weeh98N%Didier.Morandi.nospam@free.fr>,8 :	Didier.Morandi.nospam@Free.fr (Didier Morandi) writes:M :> Thank you Hoff, I stand accused. But I'm happy to hear about the VMSNET...r  B :Becuase of the ease with which code sources can be made availableA :through the web the use of USENET as a source distribution media. :has all but ceased...    H   The FAQ maintainer (eg: me) was notified of this thread (thanks to theF   person that provided the email pointer, as I probably would not haveH   found this FAQ correction otherwise), and will reword the existing andG   rather old references to these .source groups within the text of the o"   next edition of the OpenVMS FAQ.    N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comt   ------------------------------  # Date: Mon, 28 Oct 2002 22:47:43 GMTa. From: peter@langstoeger.at (Peter LANGSTOEGER)) Subject: Re: Timezone-change observations 1 Message-ID: <j8jv9.4460$aa2.93267@news.chello.at>d  V In article <3DBD81BC.C10356B1@pacbell.net>, Don Sykes <annonymous@pacbell.net> writes:" >My Alpha didn't change correctly.4 >I'm running VMS 7.2 w/ TCPIP 5.1. My logicals show:' >  "SYS$TIMEZONE_DAYLIGHT_SAVING" = "1"-) >  "SYS$TIMEZONE_DIFFERENTIAL" = "-25200"V >  "SYS$TIMEZONE_NAME" = "PDT"6 >  "SYS$TIMEZONE_RULE" = "PST8PDT7,M4.1.0/2,M10.5.0/2" >-* >Still PDT. I wonder wht this didn't work?  / $ WRITE SYS$OUTPUT F$GETSYI ("AUTO_DLIGHT_SAV")    Do you run DECdts ?-   --   Peter "EPLAN" LANGSTOEGER8% Network and OpenVMS system specialistG E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------   Date: 29 Oct 2002 01:14:01 GMT$ From: hoff@xhp.nospam (Hoff Hoffman)% Subject: Re: TSM on the Freeware disk ) Message-ID: <apkncp$7l$2@web1.cup.hp.com>r  l In article <168270-2200210525201948789@M2W041.mail2web.com>, "kenrbnsn1@rcn.com" <kenrbnsn1@rcn.com> writes:  J :...I thought I would try to use the TSM from the Freeware (CD & Web area)G :Downloading from the Web or using the files off the CD, gives the sameo	 :results:|  4   From the OpenVMS Frequently Asked Questions (FAQ):   		...nI                     o  http://www.openvms.compaq.com/freeware/freeware50/n  G                      o  Beware: The TSM saveset shipped on the Freeware I                         V5.0 disk media is busted. Download a new copy ofr>                         the saveset from the Freeware website. 		...n    G   If you are getting problems with the FTP website download, then checkeG   that you are using the correct FTP file transfer mode.  Browser-based H   FTP downloads can get this binary-vs-ASCII transfer choice wrong, and I   FTP file transfers through other platforms are notorious for clobberingl   the file attributes.  H   The http://www.openvms.compaq.com/freeware/freeware50/tsm/ zip archiveI   can be another way to transfer the files around, and without clobbering.=   the file attributes when transfering through other systems.   7   ...(or hints on getting into the decserver manually?)i  ,   Connect to the first serial port directly.  C   Alternatively, use the LANCP console connection or the DECnet NCPm   CONNECT command.  D   There have been a number of existing discussions of remote consoleG   connections via LAT out to DECserver devices, please check the Googled@   newsgroup archives for specifics.   If you have configured theC   DECserver devices per the DECserver software kit's configuration eD   tools and are successfully downloading the DECserver software, you@   should be able to use the DECnet NCP CONNECT command directly.  0   I am assuming DECnet Phase IV here, of course.  N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comn   ------------------------------  # Date: Tue, 29 Oct 2002 03:59:02 GMTx2 From: Stephen Eickhoff <seickhoff@mind-spring.com>3 Subject: Unique problem involving virtual terminalsu. Message-ID: <3DBE0773.1070309@mind-spring.com>  G I run a site which hosts, among other things, the classic game GalacticnB Trader. Galactic Trader was intended for reliable LANs and not theG internet, so handling players who disconnect has been a real problem. IcG thought the best solution for this was to enable virtual terminals, but F after some testing I realized that these are controlled by user ID andG not originating IP. This is a problem because, so that I didn't have tosH create a new account for every new player, I created one captive accountD with no password (and no privs) for the players to logon with. So ifH someone disconnects, the next player to telnet in could resume the other player's session if they wish.  D I'm wondering if there's some way around this. If I could only allowG reconnects to a virtual terminal from the same IP, I could still handle G most of the accidental (and not so accidental) disconnects. Only people E who are on dial-up and lose their PPP connection  would be stuck with H the timed lockout I have in place. That would be a great improvement. Is- there any way I can enforce this restriction?-  D This is really the only way I can think of to eliminate this form ofA cheating without totally rewriting the game to use TCPIP sockets.h   ------------------------------  # Date: Tue, 29 Oct 2002 00:03:59 GMTb. From: peter@langstoeger.at (Peter LANGSTOEGER)< Subject: [TCPware V5.6-2] Default Route gets lost (on Alpha)2 Message-ID: <Pfkv9.5706$aa2.104637@news.chello.at>  3 You may remember my posting from about 2 weeks ago. C 	[TCPware V5.6-2] Problems with SET GATEWAY (on OpenVMS Alpha V7.3)-  B Today I verified again, that my PWS433au (with OpenVMS Alpha V7.3)4 has a default gateway set during startup of TCPware.   ...P1 %STARTNET-I-DEBUG, NETCU START/IP LPB-0 127.0.0.11[ %STARTNET-I-DEBUG, NETCU START/IP EWA-0 192.168.1.3 /MASK=255.255.255.0 /FLAGS=(NOTRAILERS)B" %STARTNET-I-DEBUG, NETCU START/TCP" %STARTNET-I-DEBUG, NETCU START/UDP# %STARTNET-I-DEBUG, NETCU START/INETe" %STARTNET-I-DEBUG, NETCU START/UCX* %STARTNET-I-DEBUG, NETCU SET TIMEZONE "UT"e %STARTNET-I-DEBUG, NETCU DEFINE TIMEZONE/FILE=TCPWARE:TIMEZONES.DAT/SELECT=/SELECT=("EUROPE") MET-DST 7 %STARTNET-I-DEBUG, NETCU SET DOMAIN luna.langstoeger.at 0 %STARTNET-I-DEBUG, NETCU SET GATEWAY 192.168.1.1 ...   . TCPware(R) for OpenVMS Internet Routing Table:  J Destination          Gateway              Flags    RefCnt     UseCnt  LineJ -----------          -------              -----    ------     ------  ----K 192.168.1.0          192.168.1.3          UNIL          0          1  EWA-0 K 127.0.0.0            127.0.0.1            UNIL          0          0  LPB-0oK all others (default) 192.168.1.1          UNG           0          0  EWA-0m  I then after some more TCPware startup modules an IP Multicast Address gets_ added_  . TCPware(R) for OpenVMS Internet Routing Table:  J Destination          Gateway              Flags    RefCnt     UseCnt  LineJ -----------          -------              -----    ------     ------  ----K 224.0.0.9            127.0.0.1            UHA           0          0  LPB-0nK 192.168.1.0          192.168.1.3          UNIL          0          5  EWA-0 K 127.0.0.0            127.0.0.1            UNIL          0          0  LPB-0qK all others (default) 192.168.1.1          UNG           0          1  EWA-0o  8 and this table keeps the same until startup is finished.  G But then seconds/minutes after startup is completed the default gatewayr7 seems to get lost (and some problems starts to pop up).t  . TCPware(R) for OpenVMS Internet Routing Table:  J Destination          Gateway              Flags    RefCnt     UseCnt  LineJ -----------          -------              -----    ------     ------  ----K 224.0.0.9            127.0.0.1            UHA           0          0  LPB-0eK 192.168.1.0          192.168.1.3          UNIL          0          5  EWA-0 K 127.0.0.0            127.0.0.1            UNIL          0          0  LPB-00    ' Do you have any idea why this happens ?d@ Does or did anyone with a support contract (I'm only a hobbyist)* see the same behaviour on his/her system ?H Currently I'm stuck (and live with doing a NETCU SET GATEWAY 192.168.1.1, then, which OTOH doesn't get lost later on).   TIAh   -Peter   -- 0 Peter "EPLAN" LANGSTOEGERn% Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  % Date: Mon, 28 Oct 2002 21:47:40 -0500   From: John Santos <JOHN@egh.com>@ Subject: Re: [TCPware V5.6-2] Default Route gets lost (on Alpha)4 Message-ID: <1021028214107.400A-100000@Ives.egh.com>  - On Tue, 29 Oct 2002, Peter LANGSTOEGER wrote:o  5 > You may remember my posting from about 2 weeks ago.yE > 	[TCPware V5.6-2] Problems with SET GATEWAY (on OpenVMS Alpha V7.3)d > D > Today I verified again, that my PWS433au (with OpenVMS Alpha V7.3)6 > has a default gateway set during startup of TCPware. >  > ...a3 > %STARTNET-I-DEBUG, NETCU START/IP LPB-0 127.0.0.1.] > %STARTNET-I-DEBUG, NETCU START/IP EWA-0 192.168.1.3 /MASK=255.255.255.0 /FLAGS=(NOTRAILERS)-$ > %STARTNET-I-DEBUG, NETCU START/TCP$ > %STARTNET-I-DEBUG, NETCU START/UDP% > %STARTNET-I-DEBUG, NETCU START/INETR$ > %STARTNET-I-DEBUG, NETCU START/UCX, > %STARTNET-I-DEBUG, NETCU SET TIMEZONE "UT"g > %STARTNET-I-DEBUG, NETCU DEFINE TIMEZONE/FILE=TCPWARE:TIMEZONES.DAT/SELECT=/SELECT=("EUROPE") MET-DST29 > %STARTNET-I-DEBUG, NETCU SET DOMAIN luna.langstoeger.ate2 > %STARTNET-I-DEBUG, NETCU SET GATEWAY 192.168.1.1 > ...  > 0 > TCPware(R) for OpenVMS Internet Routing Table: > L > Destination          Gateway              Flags    RefCnt     UseCnt  LineL > -----------          -------              -----    ------     ------  ----M > 192.168.1.0          192.168.1.3          UNIL          0          1  EWA-0gM > 127.0.0.0            127.0.0.1            UNIL          0          0  LPB-0iM > all others (default) 192.168.1.1          UNG           0          0  EWA-0  > K > then after some more TCPware startup modules an IP Multicast Address getsw > addedm > 0 > TCPware(R) for OpenVMS Internet Routing Table: > L > Destination          Gateway              Flags    RefCnt     UseCnt  LineL > -----------          -------              -----    ------     ------  ----M > 224.0.0.9            127.0.0.1            UHA           0          0  LPB-0dM > 192.168.1.0          192.168.1.3          UNIL          0          5  EWA-0tM > 127.0.0.0            127.0.0.1            UNIL          0          0  LPB-0iM > all others (default) 192.168.1.1          UNG           0          1  EWA-0: > : > and this table keeps the same until startup is finished. > I > But then seconds/minutes after startup is completed the default gatewayb9 > seems to get lost (and some problems starts to pop up).  > 0 > TCPware(R) for OpenVMS Internet Routing Table: > L > Destination          Gateway              Flags    RefCnt     UseCnt  LineL > -----------          -------              -----    ------     ------  ----M > 224.0.0.9            127.0.0.1            UHA           0          0  LPB-0 M > 192.168.1.0          192.168.1.3          UNIL          0          5  EWA-0eM > 127.0.0.0            127.0.0.1            UNIL          0          0  LPB-0  >  > ) > Do you have any idea why this happens ? B > Does or did anyone with a support contract (I'm only a hobbyist), > see the same behaviour on his/her system ?J > Currently I'm stuck (and live with doing a NETCU SET GATEWAY 192.168.1.1. > then, which OTOH doesn't get lost later on). >  > TIA  >  > -Peter  E Doesn't happen on mine.  I have 2 Alpha's and 2 Vaxes running TCPWareiL V5.6-2.  One Alphas is VMS V7.2-1.  The other Alpha and both Vaxes are V7.3.  D I do have a real router (with a real routable IP address) set as theE gateway, rather than a local, non-routable address, but I don't thinkiE that should matter.  I also don't have the 224... route (is that fromh	 netware?)   G Are you sure nothing is deleting the route?  Check TCPWARE:ROUTING.COM.t HTH.   -- v John Santosh Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------   End of INFO-VAX 2002.597 ************************