1 INFO-VAX	Wed, 17 Oct 2001	Volume 2001 : Issue 577       Contents:' Re: (newbie) what can i learn from vms? & Re: -- Question from a newbie sysadmin& Re: -- Question from a newbie sysadmin& Re: -- Question from a newbie sysadmin& Re: -- Question from a newbie sysadmin Re: A free VMS implementation? Re: A free VMS implementation? Re: A free VMS implementation? Re: Alphaserver 1000a  Re: AMD's Itanion  Re: AMD's Itanion  Re: AMD's Itanion P Re: BACKUP and /NOALIAS (was: Re: [OpenVMS] V7.2 VAX satellite doesn't find V7.3* cancel <3BCCF96A.B0658B88@cableinet.co.uk> CETS-2001 Presentations  CETS-2001 Presentations  Re: DECNET ping equivalent?  Re: DECNET ping equivalent? C Re: Developers and End-Users? (was: Re: A free VMS implementation?)  Dont hesitate on this one  8267 2 ES45 Announced as "UNIX server" - VMS absent again6 Re: ES45 Announced as "UNIX server" - VMS absent again6 Re: ES45 Announced as "UNIX server" - VMS absent again6 Re: ES45 Announced as "UNIX server" - VMS absent again GBLPAGES in VMS V7.3 Re: GBLPAGES in VMS V7.3 Re: Global symbol  Re: Global symbol  Re: Global symbol  Re: Global symbol & Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors& Re: Higher prices for Alpha processors, Re: HWP/CPQ: (...internal morale is poor...)0 Re: Ingres (Was: Re: jobs? ingres vax/vms cobol)G Linux distribution incompatibility (was Re: A free VMS implementation?) " Mozilla 0.9.5 and file protections Quotas for Oracle 8i User  Re: Quotas for Oracle 8i User 7 Re: sd comman (or who has the biggest cd.com or sd.com) 2 RE: Slightly OT - how failed computers kill ships.2 Re: Slightly OT - how failed computers kill ships.2 Re: Slightly OT - how failed computers kill ships.2 Re: Slightly OT - how failed computers kill ships. Re: Stop a process Re: Stop a process Re: Stop a process Re: user authentification  Re: user authentification  Re: user authentification  Re: VMS & Unix connectivity  Re: VMS & Unix connectivity & VMS7.3,AS7.3 & external authentication* Re: VMS7.3,AS7.3 & external authentication, Re: We've burned our boats say Compaq and HP& Windows Fails To Storm the Data Centre* Re: Windows Fails To Storm the Data Centre! Re: X programming tools (UIL etc) C Re: [OpenVMS] V7.2 VAX satellite doesn't find V7.3 Alpha bootserver 0 Re: [VMS V7.3 Alpha] my first crash since months0 Re: [VMS V7.3 Alpha] my first crash since months  F ----------------------------------------------------------------------  % Date: Tue, 16 Oct 2001 14:35:50 -0400 - From: "Richard D. Piccard" <piccard@ohio.edu> 0 Subject: Re: (newbie) what can i learn from vms?( Message-ID: <3BCC7DFC.284DE49E@ohio.edu>  E Because the use of the HELP command is itself not all that obvious at  first,  I would say   
 Start with   $ HELP  HELP    +                                         RDP        Arne Vajhj wrote:   > fullforce wrote:F > > I'm interested in learning the commands and what not for vms. JustG > > need some pointers on where i should start. I have a shell account, ? > > but am a little lost with some of the commands. any help is  > > appreciated. > 
 > Start with:  >  > $ HELP >  > Arne   --B ==================================================================B Dick Piccard                           Academic Technology ManagerB piccard@ohio.edu                                 Computer ServicesB http://oak.cats.ohiou.edu/~piccard/                Ohio University   ------------------------------  % Date: Tue, 16 Oct 2001 14:40:59 -0400 - From: "Richard D. Piccard" <piccard@ohio.edu> / Subject: Re: -- Question from a newbie sysadmin ( Message-ID: <3BCC7F31.F55BCBD0@ohio.edu>  K The obvious question is, "What, exactly, does the SHOW DEVICE/FULL response  look like?"   +                                         RDP      Bert Medley wrote:  L > I inherited a DecServer 3000 and some 9GB disks.  OpenVMS "sees" the disksK > and I can do a SHOW DEVICE /FULL but when I try to INIT it I get "invalid  > media format"  HELP!!!!!   --B ==================================================================B Dick Piccard                           Academic Technology ManagerB piccard@ohio.edu                                 Computer ServicesB http://oak.cats.ohiou.edu/~piccard/                Ohio University   ------------------------------  # Date: Tue, 16 Oct 2001 20:17:41 GMT % From: medleyb@flash.net (Bert Medley) / Subject: Re: -- Question from a newbie sysadmin < Message-ID: <Xns913C9BC75383Amedleybflashnet@207.115.63.150>  0 "Richard D. Piccard" <piccard@ohio.edu> wrote in <3BCC7F31.F55BCBD0@ohio.edu>:   C >The obvious question is, "What, exactly, does the SHOW DEVICE/FULL  >response look like?"  > , >                                        RDP >  >  >Bert Medley wrote:  > G >> I inherited a DecServer 3000 and some 9GB disks.  OpenVMS "sees" the I >> disks and I can do a SHOW DEVICE /FULL but when I try to INIT it I get $ >> "invalid media format"  HELP!!!!! >  >-- C >================================================================== C >Dick Piccard                           Academic Technology Manager C >piccard@ohio.edu                                 Computer Services C >http://oak.cats.ohiou.edu/~piccard/                Ohio University  >  >  >    $ sh dev/fu dkb300  I Disk $1$DKB300: (SSD82), device type SEAGATE ST19171N CLAR09, is online,   file- G     oriented device, shareable, available to cluster, error logging is   enabled.  M     Error count                   16    Operations completed                   12G     Owner process                 ""    Owner UIC                        [SYSTEM]<     Owner process ID        00000000    Dev Prot             S:RWPL,O:RWPL,G:R,W L     Reference count                0    Default buffer size                  512 L     Total blocks            17551152    Sectors per track                    168 M     Total cylinders             5268    Tracks per cylinder                    20$     Allocation class               1   ------------------------------    Date: 16 Oct 2001 13:46:24 -0700( From: bob@instantwhip.com (Bob Ceculski)/ Subject: Re: -- Question from a newbie sysadmin < Message-ID: <d7791aa1.0110161246.ce6b2cb@posting.google.com>  h medleyb@flash.net (Bert Medley) wrote in message news:<Xns913C64773127medleybflashnet@207.115.63.150>...M > I inherited a DecServer 3000 and some 9GB disks.  OpenVMS "sees" the disks  L > and I can do a SHOW DEVICE /FULL but when I try to INIT it I get "invalid  > media format"  HELP!!!!!   are these vms compatible disks? 4 what is the exact vms init command you are entering? what version of vms?1 what firmware version are you running (pal code)?  are these scsi disks?   B sounds like the disks are not supported or your firmware or device drivers need updated!    ------------------------------  # Date: Wed, 17 Oct 2001 03:26:56 GMT 3 From: Tim Llewellyn <tim.llewellyn@cableinet.co.uk> / Subject: Re: -- Question from a newbie sysadmin / Message-ID: <3BCCF9F4.7FAE7F24@cableinet.co.uk>    Fabio Cardoso wrote: > / > What machine are you connecting these disks ?  > 6 > Alpha ? Vax ? You should check if the controller and1 > firmware support these disks (model), and maybe  > these disks can be bad.  >   6 Also check VMS version. If its a turbochannel alpha asF I suspect then the VMS might be old enough to not support those disks.  	 > Regards  >  > FC, > --- Bert Medley <medleyb@flash.net> wrote:4 > > I inherited a DecServer 3000 and some 9GB disks. > > OpenVMS "sees" the disks6 > > and I can do a SHOW DEVICE /FULL but when I try to > > INIT it I get "invalid > > media format"  HELP!!!!!    --   Tim.Llewellyn@cableinet.co.uk     C Standard disclaimer applies. My views in no way represent those of  ! my employers or service provider.    ------------------------------  # Date: Wed, 17 Oct 2001 03:06:05 GMT 3 From: Tim Llewellyn <tim.llewellyn@cableinet.co.uk> ' Subject: Re: A free VMS implementation? / Message-ID: <3BCCF508.9D4F26D6@cableinet.co.uk>   ! Steve.Spires@yellgroup.com wrote:  > N > Contact:   Tel: 3063  -  IS - Infrastructure, 1st Floor, Bridge Street Plaza > P > And of course the Unix for VMS Users book DOES compare both cp [and cpall] and > rcp to COPY along with ftp.  > Q > Let's not forget that the book isn't intended as a definitive Unix manual, it's O > there to give VMS users a handle on how to perform the same work they do on a L > VMS box on a Unix box for which they have little or no skill. Once you get, > started, you can build on that from there. >   H agreed, it's also not bad for unix people trying to understand some VMS,
 but I willD agree with Bill that I was hmmm, surprised when some of the commands didn't work D on ultrix or OSF/1. Then again, shell mayhem is something one gladly doesn't get  to deal with often on VMS.   regards    > Cheers >  > Steve Spires > = > David Mathog <mathog@caltech.edu> on 10/15/2001 06:03:22 PM  > " > To:        Info-VAX@Mvb.Saic.Com- > cc:         (bcc: Steve Spires/YellowPages) J > From:      David Mathog <mathog@caltech.edu>, 15 October 2001, 6:03 p.m. >   > Re: A free VMS implementation? >  > Bill Gunshannon wrote: > + > > . Comparing ftp to COPY?  Unix can copy I > > from machine to machine just like VMS does, You don't need to use ftp C > > if the tha mchines share a security model (just like with VMS.)  > G > I agree with some of what you say, but ftp for COPY isn't much (if at  > all) off.  > K > VMS/RMS/DECNET allow the same file syntax to be used transparently across  > nodes.  So > doing: > 0 > $ program "access string"node::device:filename > P > works pretty much without regard to what "program" is.   If proxies are set up > you can leave P > off the access string too.  Which makes COPY the usual choice for moving files > from one VMS node to another. N > However, on Unix "cp" cannot  normally move files from one node to another . > cp can move across nodesN > if NFS is used, just as COPY can move files across nodes without an explicitO > node:: if those nodes are in a cluster and the disks are cross mounted. While N > I think that rcp is probably the closest program in this regard to COPY, ftp. > will often be the tool of choice - depending, > on exactly what it is that you want to do. > 
 > Regards, >  > David Mathog > mathog@caltech.edu   --   Tim.Llewellyn@cableinet.co.uk     C Standard disclaimer applies. My views in no way represent those of  ! my employers or service provider.    ------------------------------  # Date: Wed, 17 Oct 2001 03:24:02 GMT 3 From: Tim Llewellyn <tim.llewellyn@cableinet.co.uk> ' Subject: Re: A free VMS implementation? / Message-ID: <3BCCF601.6CC5030D@cableinet.co.uk>    Peter Weaver wrote:  > 7 > "David Froble" <davef@tsoft-inc.com> wrote in message ( > news:3BCB7960.5070906@tsoft-inc.com... > >...D > > As for the issue of insuring your resume is in order, I find theF > > practice disgusting.  If you're good enough, you will do well.  IfI > > you're not, then it's all bullshit anyway.  I've seen too many people L > > who were more interested in their resume than in their employer's needs.L > >   I've also seen instances when this has hurt the employer.  I would notL > > call such an employee honest.  You may guess that my perspective is that: > > of a consultant and employer, not that of an employee. > >... > H > A person within Compaq told me once that the reason Compaq Analyze was. > written in Java was to pad someone's resume. >   D surely to bolster upgrade sales also :-). Personally I don't find I F have the time to spend upgrading my CV, no matter how many agents wantC a copy, while I am gainfully employed. Then again, when it is newly > revampled like now they don't seem to want to talk much. Life.  B ANyway, back to your original point, why do you think they started implementing new bits of VMS in C?    regards    --   Tim.Llewellyn@cableinet.co.uk     C Standard disclaimer applies. My views in no way represent those of  ! my employers or service provider.    ------------------------------  # Date: Wed, 17 Oct 2001 03:24:39 GMT 3 From: Tim Llewellyn <tim.llewellyn@cableinet.co.uk> ' Subject: Re: A free VMS implementation? / Message-ID: <3BCCF96A.B0658B88@cableinet.co.uk>    Peter Weaver wrote:  > 7 > "David Froble" <davef@tsoft-inc.com> wrote in message ( > news:3BCB7960.5070906@tsoft-inc.com... > >...D > > As for the issue of insuring your resume is in order, I find theF > > practice disgusting.  If you're good enough, you will do well.  IfI > > you're not, then it's all bullshit anyway.  I've seen too many people L > > who were more interested in their resume than in their employer's needs.L > >   I've also seen instances when this has hurt the employer.  I would notL > > call such an employee honest.  You may guess that my perspective is that: > > of a consultant and employer, not that of an employee. > >... > H > A person within Compaq told me once that the reason Compaq Analyze was. > written in Java was to pad someone's resume. >   D surely to bolster upgrade sales also :-). Personally I don't find I F have the time to spend upgrading my CV, no matter how many agents wantC a copy, while I am gainfully employed. Then again, when it is newly > revampled like now they don't seem to want to talk much. Life.  B ANyway, back to your original point, why do you think they started implementing new bits of VMS in C?    regards    --   Tim.Llewellyn@cableinet.co.uk     C Standard disclaimer applies. My views in no way represent those of  ! my employers or service provider.    ------------------------------  % Date: Tue, 16 Oct 2001 23:14:39 +0200 . From: Dennis Grevenstein <dennis@pcde.inka.de> Subject: Re: Alphaserver 1000a, Message-ID: <3BCCA33F.D24F9242@pcde.inka.de>   Fred Kleinsorge wrote: > M > The Mach64 should work.  What's the problem?  I would stick a Trio64 in it, E > it is much simpler to configure (if you have a PCI slot available).   > The Mach64 does not work without the proper .CFG files for the EISA configuration utility. A A Trio64 will be my next try. I will get an older card from Elsa. / I hope it's just the graphics chip that counts.    Dennis   ------------------------------    Date: 16 Oct 2001 12:01:31 -0700( From: bob@instantwhip.com (Bob Ceculski) Subject: Re: AMD's Itanion< Message-ID: <d7791aa1.0110161101.add9147@posting.google.com>  v Fabio Cardoso <fabiopenvms@yahoo.com.br> wrote in message news:<20011016114251.28342.qmail@web20208.mail.yahoo.com>...3 > Do you know if AMD will launch their ouwn 64 bits % > processor compatible with Itanium ?  > " > Like, Athlon, Duron, Itanion ... >  > 	 > Regards  >  > FC   >  > =====  > =========================  > =  > F?io dos Santos Cardoso  > OpenVMS System Manager > Rio de Janeiro - Brazil  > fabiopenvms@yahoo.com.br > =========================  > =  > 4 >                                                    > Do You Yahoo!?. > Make a great connection at Yahoo! Personals. > http://personals.yahoo.com  L amd had a chance to buy alpha, but they blew it ... now intel will move intoN the high end business and amd will be stuck with competing for the client end! they missed a big opportunity!   ------------------------------  # Date: Tue, 16 Oct 2001 19:32:16 GMT * From: "Bill Todd" <billtodd@metrocast.net> Subject: Re: AMD's Itanion? Message-ID: <4X%y7.513635$Lw3.31449130@news2.aus1.giganews.com>   3 Bob Ceculski <bob@instantwhip.com> wrote in message 6 news:d7791aa1.0110161101.add9147@posting.google.com...   ...   5 > amd had a chance to buy alpha, but they blew it ...   K Hardly.  Just think where they'd be now, given Compaq's lack of interest in L the Alpha OSs:  it's the synergy among Alpha, VMS, and Tru64 that made AlphaL so attractive (though had DEC capitalized on its potential early-on it mightA well have succeeded as more 'generic' hardware in its own right).     now intel will move intoaK > the high end business and amd will be stuck with competing for the clientt end!  > they missed a big opportunity!  I Intel will move into the high-end business only if IBM decides it doesn'teE want it (which seems rather unlikely):  Itanic can't hold a candle topG POWER4.  And for that matter Hammer may be better positioned to move inoK there than Itanic is:  Hammer uses *far* less power for a given performancesJ level, has far better memory-access and multi-processor performance due toK its on-chip glue features, has on-chip reliability features that higher-end L installations want (and that IA32 implementations lack, though IA64 may haveJ them), and can run existing IA32 applications at native speed (rather thanF 100 MHz Pentium speed - or perhaps 200 MHz Pentium speed when McKinleyK arrives) until such time as 64-bit applications mature.  And of course it'sS2 projected to cost only about 1/4 what Itanic does.  D Given the degree to which Compaq and HP have tied their futures to aK platform that may sink without a trace, AMD may well out-live both of them.l   - bill   ------------------------------    Date: 16 Oct 2001 14:37:27 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)c Subject: Re: AMD's Itanion3 Message-ID: <Vq86+j69wR9C@eisner.encompasserve.org>s  g In article <d7791aa1.0110161101.add9147@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes:   N > amd had a chance to buy alpha, but they blew it ... now intel will move intoP > the high end business and amd will be stuck with competing for the client end!  > they missed a big opportunity!  A By "they blew it", I take it you mean the offering price was wellPE within their budget and clearly a better investment that alternativesa available to them.  I doubt it.i   ------------------------------  # Date: Wed, 17 Oct 2001 00:41:02 GMTi2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)Y Subject: Re: BACKUP and /NOALIAS (was: Re: [OpenVMS] V7.2 VAX satellite doesn't find V7.3t1 Message-ID: <ys4z7.540$RL6.4497@news.cpqcorp.net>t  j In article <3BCCD0D0.1B4AE422@ipl.demon.co.nospam.uk>, Steve Reece <SYSTEM@ipl.demon.co.nospam.uk> writes:B :I understood that it's not just performance that suffers from notE :including the /noalias qualifier - the resultant disk structure on at  :system disk would be incorrect.  C   A failing restoration would be viewed as a serious bug in BACKUP.3  C :IIRC, there was a long discussion of this qualifier in comp.os.vmsaB :either last year or the year before which included the issue thatC :/noalias is not the default but is required when backing up systeme :disks.a  C   The discussion was around using /ALIAS it to improve performance 7B   of the BACKUP -- with /ALIAS, less goes out and less comes back B   in, but the resulting disk contents are valid regardless of the    use of /ALIAS.  D   The reason /ALIAS is not the default is because of expectations ofD   selective file restorations -- on a disk with alias entries, only E   the primary entry gets copied out, meaning only the primary can be o(   selectively restored from the archive.  /   BACKUP knows how to reconstitute the entries.   @   You will want to follow the directions for performing full and+   incremental BACKUP operations, of course.t  L   I *have* seen some bizarre BACKUP commands used over the years, including M   folks that have manually reconstitited a system disk and its alias entries lI   from a non-/IMAGE BACKUP.  (This is usually due to a lack of an /IMAGE  J   BACKUP in an archive, and also arguably due to an undetected failure in K   the testing (if any, of course) that is performed before the restoration  !   sequence is actually required.)e  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Wed, 17 Oct 2001 03:46:51 GMTi3 From: Tim Llewellyn <tim.llewellyn@cableinet.co.uk> 3 Subject: cancel <3BCCF96A.B0658B88@cableinet.co.uk> 7 Message-ID: <La7z7.3364$qq3.337396@news1.cableinet.net>r  / This message was cancelled from within Mozilla.s   ------------------------------  # Date: Wed, 17 Oct 2001 00:03:09 GMTn& From: "Jeff Killeen" <Jeff@IDM-IO.com>  Subject: CETS-2001 Presentations6 Message-ID: <1V3z7.329$lX2.26166@typhoon1.gnilink.net>  8 If you register for the portal you should have access...   Dear Colleague,   B A month has now passed since we met in Anaheim for the 2001 CompaqH Engineering Technical Symposium (CETS). We've each had time to return toK work with many new perspectives regarding both the world we live in and thej2 exciting technological future we explored at CETS.  H Since my last note, we have posted presentations made available to us byH speakers on the CETS website. As a CETS portal registrant you may accessI these presentations via www.cets2001.com <http://www.cets2001.com>. Afterk8 logging in, select "CETS 2001 Conference Presentations".  G To help us improve your CETS experience year over year, we rely on youreL comments and suggestions. Your CETS 2001 impressions are important to us andL will help us ensure that future symposiums meet your needs and expectations.C Please take a few minutes to complete the on-line Post-event Survey H <https://www.travelhq.com/register/cets2001survey/index.htm> , which mayI also be accessed via the CETS 2001 site. By completing and submitting the/F survey by October 29 , 2001, you will be eligible to win a Compaq iPAQD Pocket PC. The lucky iPAQ winner of the award for submitting on-site6 evaluations is now posted on the site www.cets2001.com http://www.cets2001.com.  E Although some CETS sessions were disrupted by the tragic events which J happened on September 11th, we believe that through the generous spirit ofJ so many in our community, the CETS experience provided value for those whoL chose to take advantage. Once again, I would like to thank everyone for yourL positive participation throughout our week together at CETS and I would likeB to extend an offer of a 20% discount on a CETS 2002 full symposiumJ registration for all Compaq partners and customers who purchased full CETSK 2001 symposium registrations. This discount will be made available to those-H who register and pay within the first 30 days of CETS 2002 registration;@ details will be sent to you as registration opens in the spring.  B Although the 2001 symposium has come to an end, we look forward toK continuing the exchange of information and building on the relationships weaH have made based on the rich technical heritages that brought us togetherL originally. Please mark your calendar and training plan to include CETS 2002> being held in St Louis, MO, October 6th - 11th, 2002 and visitG www.cets2002.com <http://www.cets2002.com> starting in January to learne more.n   Regards,  
 Bill Horzempat) Senior Director, Field Technical Programs) Compaq Computer Corporation    ------------------------------  # Date: Wed, 17 Oct 2001 02:14:39 GMTr& From: "Jeff Killeen" <Jeff@IDM-IO.com>  Subject: CETS-2001 Presentations6 Message-ID: <jQ5z7.649$uZ1.44626@typhoon2.gnilink.net>  8 If you register for the portal you should have access...   Dear Colleague,3  B A month has now passed since we met in Anaheim for the 2001 CompaqH Engineering Technical Symposium (CETS). We've each had time to return toK work with many new perspectives regarding both the world we live in and theb2 exciting technological future we explored at CETS.  H Since my last note, we have posted presentations made available to us byH speakers on the CETS website. As a CETS portal registrant you may accessI these presentations via www.cets2001.com <http://www.cets2001.com>. After 8 logging in, select "CETS 2001 Conference Presentations".  G To help us improve your CETS experience year over year, we rely on youriL comments and suggestions. Your CETS 2001 impressions are important to us andL will help us ensure that future symposiums meet your needs and expectations.C Please take a few minutes to complete the on-line Post-event SurveycH <https://www.travelhq.com/register/cets2001survey/index.htm> , which mayI also be accessed via the CETS 2001 site. By completing and submitting thecF survey by October 29 , 2001, you will be eligible to win a Compaq iPAQD Pocket PC. The lucky iPAQ winner of the award for submitting on-site6 evaluations is now posted on the site www.cets2001.com http://www.cets2001.com.  E Although some CETS sessions were disrupted by the tragic events whichnJ happened on September 11th, we believe that through the generous spirit ofJ so many in our community, the CETS experience provided value for those whoL chose to take advantage. Once again, I would like to thank everyone for yourL positive participation throughout our week together at CETS and I would likeB to extend an offer of a 20% discount on a CETS 2002 full symposiumJ registration for all Compaq partners and customers who purchased full CETSK 2001 symposium registrations. This discount will be made available to thosetH who register and pay within the first 30 days of CETS 2002 registration;@ details will be sent to you as registration opens in the spring.  B Although the 2001 symposium has come to an end, we look forward toK continuing the exchange of information and building on the relationships we H have made based on the rich technical heritages that brought us togetherL originally. Please mark your calendar and training plan to include CETS 2002> being held in St Louis, MO, October 6th - 11th, 2002 and visitG www.cets2002.com <http://www.cets2002.com> starting in January to learnf more.o   Regards,  
 Bill Horzempa ) Senior Director, Field Technical ProgramsE Compaq Computer Corporationa   ------------------------------    Date: 17 Oct 2001 02:31:43 +0800, From: Paul Repacholi <prep@prep.synonet.com>$ Subject: Re: DECNET ping equivalent?- Message-ID: <873d4j5tdc.fsf@prep.synonet.com>   4 "frank brown" <frank.brown@ci.seattle.wa.us> writes:  E > Is there an equivalent to the unix ping for DECNET phase IV to test D > connectivity to a remote node without using TCP/IP?  I'm using DIR7 > NODE::DUAn:[DIR] but wonder if there's a cheaper way.   8 MC NCP SHO NODE <blah> will tell you if it is up, modulo9 routing issues. Remember DECnet has the hello and friendse9 so even if you have no links to a node, the network knowsc if it is alive and adjacent.   -- ,< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Tue, 16 Oct 2001 21:57:19 +0200 ) From: labadie <labadie.gerard@wanadoo.fr>t$ Subject: Re: DECNET ping equivalent?* Message-ID: <3BCC911F.C9813D30@wanadoo.fr>   Hellot  K I had already posted (2000/07/12) this decnet ping with my account palm6174o  # May be it should go in the freeware@   Enjoy.   Grard Labadie   $!'f$ver(0)s
 $ SET NOON $ SET CONTROL=Y  $ ON CONTROL_Y THEN GOTO FIN* $ NETNODE_REM = F$TRNLNM("NETNODE_REMOTE") $ IF NETNODE_REM .EQS. "" THEN, NETNODE_REM  :=SYS$SYSTEM:NETNODE_REMOTE.DAT $ DEBUT:
 $ FLAG_OK = 1  $ NAME = P1e% $ IF NAME .eqs. "" THEN goto HELP_FINp' $ IF F$LOCATE(".",P1) .NE. F$LENGTH(P1)a $ THEN $       FLAG_ADDR = 1l- $       AREA = F$INTEGER(F$ELEMENT(0,".",P1))o- $       IF AREA .GT. 63 THEN GOTO ERREUR_ADDRt, $       NUM = F$INTEGER(F$ELEMENT(1,".",P1)). $       IF NUM .GT. 1023 THEN GOTO ERREUR_ADDR $       ADDR = AREA*1024+NUM $       ADDH=ADDRw, $ NAME = " Address ''AREA'.''NUM' or ''addr'! (dec)"+f$fao("!XW",addh)+" (hex)"d $ ELSE $       FLAG_ADDR = 0d# $       IF F$CVUI(0,8,P1) .GT. %X39u $       THEN' $               ADDR = F$EXTRAC(0,6,P1)  $       ELSE$ $               ADDR = F$INTEGER(P1)) $               IF ( ADDR .EQ. 0 ) .OR. (m
 ADDR .EQ. 1 )k $               THEN0 $                       ADDR = F$INTEGER(%X'P1')1 $                       IF ( ADDR .EQ. 0 ) .OR. ( 
 ADDR .EQ. 1 )  $                       THEN/ $                               ADDR = F$EXTRACn (0,6,P1) $                       ELSE. $                               IF ( ADDR .GT. 65535 ) .OR. ( ADDR .LT.1 ) $ $                               THEN, $                                       GOTO ERREUR_ADDR $ $                               ELSE. $                                       AREA =	 ADDR/1024 . $                                       NUM  = ADDR-(AREA*1024)* $                                       P1 = "''AREA'.''NUM'"2 $                                       GOTO DEBUT% $                               ENDIFg $                       ENDIFo $               ELSE( $                       AREA = ADDR/1024/ $                       NUM  = ADDR-(AREA*1024)e/ $                       P1   = "''AREA'.''NUM'" " $                       GOTO DEBUT $               ENDIFt
 $       ENDIF  $ ENDIFe $ IF FLAG_ADDR $ THEN- $       OPEN/READ/SHARE NETNODE 'NETNODE_REM'n0 $       READ/KEY="''ADDH'"/ERROR=ERR_LEC NETNODE RECORD) $       NODE_NAME = F$EXTRACT(2,6,RECORD)r/ $ NAME = NODE_NAME + " Address ''p1' or ''addr'o (dec)"+f$fao("!XW",addr)+" (hexe $       ERR_LEC: $       CLOSE NETNODEs $ ELSE- $       OPEN/READ/SHARE NETNODE 'NETNODE_REM'g1 $       READ/KEY="''ADDR'"/INDEX=1/ERROR=ERR_LEC1l NETNODE RECORD1 $       ADDH = F$CVUI(0,16,F$EXTRACT(0,2,RECORD)) ) $       NODE_NAME = F$EXTRACT(2,6,RECORD)c $       AREA = ADDH/1024 $       NUM  = ADDH-(AREA*1024) 1 $ NAME = p1 + " Address ''AREA'.''NUM' or ''addr'n (dec)"+f$fao("!XW",addh)+" (hl $       ERR_LEC1:o $       CLOSE NETNODEl $ ENDIFl $ DEFINE SYS$OUTPUT NL:e $ DEFINE SYS$ERROR  NL:p' $ A=F$FILE_ATT("''ADDR'::""1=""","ORG")- $ STATUS = $STATUS $ DEASSIGN SYS$OUTPUT0 $ DEASSIGN SYS$ERROR/ $ IF STATUS .EQ. %X0000206C !%SYSTEM-F-REMRSRC,j insufficient systemresources atg/ $ THEN WRITE SYS$OUTPUT "  ''NAME' insufficiente system resources atremote node"s( $ ELSE IF STATUS .EQ. %X00000294 !REJECT+ $      THEN  WRITE SYS$OUTPUT "  ''NAME' ist# alive,but DECNET rejectconnection "v1 $      ELSE IF STATUS .EQ. %X000020A4 ! NOSUSHOBJp0 $           THEN  WRITE SYS$OUTPUT "  ''NAME' is alive on DECNET  ", $           ELSE IF STATUS .EQ. %X0000028C !
 NOSUSHNODE  $              THEN  FLAG_OK = 0 $                WRITE. SYS$OUTPUT "''NAME'''F$ELEMENT(1,",",F$MESSAGE
 (status))"/ $              ELSE IF STATUS .EQ. %X00002094 !  UNREACHABLEA" $                 THEN FLAG_OK = 0 $                WRITE. SYS$OUTPUT "''NAME'''F$ELEMENT(1,",",F$MESSAGE
 (status))" $                     ELSE $                WRITE. SYS$OUTPUT "''NAME'''F$ELEMENT(1,",",F$MESSAGE
 (status))" $                ENDIF $           ENDIFw
 $       ENDIF  $ ENDIFe) $ IF ( FLAG_OK .AND. (F$EXTRAC(0,1,F$EDITa& (P2,"UPCASE")) .EQS. "F")) THEN MC NCP $FIN:=, $ IF F$TRNLNM("SYS$OUTPUT") .EQS. "NL:" THEN DEASSIGN SYS$OUTPUTn+ $ IF F$TRNLNM("SYS$ERROR") .EQS. "NL:" THENm DEASSIGN SYS$ERROR $ EXIT
 $ERREUR_ADDR: 
 $HELP_FIN:# $ WRITE SYS$OUTPUT "Address Error." ( $ WRITE SYS$OUTPUT "Please enter an node% address : x.xxx or decimal form xxxxx - $ WRITE SYS$OUTPUT " or enter the name of thee node." $exitb    ! Steve.Spires@yellgroup.com wrote:   N > Contact:   Tel: 3063  -  IS - Infrastructure, 1st Floor, Bridge Street Plaza > 
 > $ MC NCP > NCP> LOOP NODE nodename- >- > should work for you. >e > Cheers >  > Steve Spires > H > "frank brown" <frank.brown@ci.seattle.wa.us> on 10/16/2001 03:49:17 PM >g" > To:        Info-VAX@Mvb.Saic.Com- > cc:         (bcc: Steve Spires/YellowPages)tP > From:      "frank brown" <frank.brown@ci.seattle.wa.us>, 16 October 2001, 3:49 >            p.m.m >a > DECNET ping equivalent?t >eE > Is there an equivalent to the unix ping for DECNET phase IV to test'D > connectivity to a remote node without using TCP/IP?  I'm using DIR7 > NODE::DUAn:[DIR] but wonder if there's a cheaper way.  >  > -Frank Brown > http://www.inwa.net/~frog/   ------------------------------  # Date: Tue, 16 Oct 2001 18:19:21 GMTn2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)L Subject: Re: Developers and End-Users? (was: Re: A free VMS implementation?)1 Message-ID: <JS_y7.523$RL6.4569@news.cpqcorp.net>u  e In article <3BCC6D12.D1121E5B@cableinet.co.uk>, Tim Llewellyn <tim.llewellyn@cableinet.co.uk> writes:e :p :Hoff Hoffman wrote:  L :>   I have worked with folks that have wanted to customize specific OpenVMSM :>   components, and I have acquired the necessary sanction for and have theny- :>   provided the source code to these folks.  : 9 :do they replicate the VMS engineering build environment t  H   No.  The build environment is very large and very complex, and I know J   of only a handful of internal systems that have replicated it.  (I have I   one of these replicants, and am familiar with cloning the environment.)e  6 :or do you build "custom VMS" kits for them, I wonder?     We provide sources.i    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   Date: 16 Oct 2001 19:49:32 GMT& From: kyugcm@microservice.club24.co.uk( Subject: Dont hesitate on this one  8267= Message-ID: <3bcc8f4c$1$8512$ed9e5944@reading.news.pipex.net>a  f 80 Million Fresh email addresses Available NOW with free Mass Mailer and other essential applications.  + Best prices around. Fantastic Special Offer    For more info call   UK- 0906 664 2021y   All others-  00 44 906 664 2021f  ( Select option 2 from menu when directed.  * Lines get busy at peak times - keep trying  : We accept all major credit cards -  Delivery normaly 24hrs  F euhlistsoyiescgudjprukcklpqrxdlotqgigevchcxogowvyxjtitoirvbgbjtfytmrzj   ------------------------------  % Date: Tue, 16 Oct 2001 20:29:17 -0500p1 From: "David J. Dachtera" <djesys.nospam@fsi.net>g; Subject: ES45 Announced as "UNIX server" - VMS absent againB' Message-ID: <3BCCDEED.E200CC04@fsi.net>S   Here's the URL:i  M http://www.infoworld.com/articles/hn/xml/01/10/16/011016hncompaq.xml?1016tupm    No mention of VMS.  3 Chalk up another one for Compaq marketing "no how".c  B No misspelling there: Who's gonna market VMS? As they said in "The( Wizard of Oz", "Not nobody, not no how!"   --   David J. Dachteral dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho//   ------------------------------  % Date: Tue, 16 Oct 2001 22:41:07 -0400a- From: JF Mezei <jfmezei.spamnot@videotron.ca> ? Subject: Re: ES45 Announced as "UNIX server" - VMS absent againe, Message-ID: <3BCCEFC1.1EF3CF90@videotron.ca>   "David J. Dachtera" wrote:O > http://www.infoworld.com/articles/hn/xml/01/10/16/011016hncompaq.xml?1016tupm  >  > No mention of VMS.    U Yeah, but there are lots of news in that article. The first sentense is funny though: : 				 REFRESHING ITS COMMITMENT to the company's RISC-basedL                       AlphaServer product line, Compaq on Tuesday introduced	 three new ,                       AlphaServer offerings.   (uppercase is part of article).K     But the big clincher is: ##J Compaq will support its Alpha customers indefinitely and OS transitions toM Itanium should be seamless, but around 2003  the Alpha road map will morph to J Itanium and new versions of Alpha chips will be designed by an Intel team. ##  N Got to give it to the producers of the Curly&Carly show, they sure know how to7 stimulate the media with all sorts of funny statements.a   ------------------------------  # Date: Wed, 17 Oct 2001 03:14:02 GMT-$ From: "Ed Wilts" <ewilts@ewilts.org>? Subject: Re: ES45 Announced as "UNIX server" - VMS absent again 8 Message-ID: <20011016.221440.1272469786.4525@ewilts.org>  ; In article <3BCCDEED.E200CC04@fsi.net>, "David J. Dachtera"h <djesys.nospam@fsi.net> wrote:   > Here's the URL:o > O > http://www.infoworld.com/articles/hn/xml/01/10/16/011016hncompaq.xml?1016tupmr >  > No mention of VMS. > 5 > Chalk up another one for Compaq marketing "no how".8  G However, read the announcement on http://www.theregister.co.uk.  VMS isrI clearly mentioned.  I don't think you've got a Compaq problem, but rather- an InfoWorld problem.    	.../Ed1   -- 0 Ed Wilts, Mounds View, MN, USA mailto:ewilts@ewilts.org   ------------------------------  # Date: Wed, 17 Oct 2001 04:29:46 GMT.* From: "Bill Todd" <billtodd@metrocast.net>? Subject: Re: ES45 Announced as "UNIX server" - VMS absent againd? Message-ID: <_O7z7.516818$Lw3.31656606@news2.aus1.giganews.com>a  8 JF Mezei <jfmezei.spamnot@videotron.ca> wrote in message& news:3BCCEFC1.1EF3CF90@videotron.ca... > "David J. Dachtera" wrote: > >iL http://www.infoworld.com/articles/hn/xml/01/10/16/011016hncompaq.xml?1016tup mt > >e > > No mention of VMS. >o > I > Yeah, but there are lots of news in that article. The first sentense iss
 funny though:a7 > REFRESHING ITS COMMITMENT to the company's RISC-basedeC >                       AlphaServer product line, Compaq on Tuesdayl
 introduced > three newp. >                       AlphaServer offerings. >a! > (uppercase is part of article).e >o >a > But the big clincher is: > ##L > Compaq will support its Alpha customers indefinitely and OS transitions toL > Itanium should be seamless, but around 2003  the Alpha road map will morph toL > Itanium and new versions of Alpha chips will be designed by an Intel team. > ## >sI > Got to give it to the producers of the Curly&Carly show, they sure knowl how to9 > stimulate the media with all sorts of funny statements.r  G I suspect it was simply reporter confusion about the chips that will be K starting to appear in 'AlphaServers' (if that brand is still used) at abouteF that time (which of course will be 100% Itanic in nature, save for the 'legacy' real Alphas).  ( But I particularly enjoyed the statement  K "I think the ES45 is further indication that we are absolutely committed tocK the Alpha road map, and there is no backing off that commitment," she said.u  I Now that the Alpha road map has a big "Dead End!" sign and barricade well D this side of the horizon, it's possible that *this* commitment mightK actually be adhered to - but I can't find it in my heart to give Compaq anyh credit for that.   - bill   ------------------------------  % Date: Tue, 16 Oct 2001 18:56:52 -04002) From: John Johnstone <jj_usenet@mail.com>  Subject: GBLPAGES in VMS V7.3r( Message-ID: <3BCC82F4.3BB4FA46@mail.com>  J I've been looking at a system running VMS V7.3 and it looks like the valueJ of GBLPAGES has taken quite a jump.  In sys$system:setparams.dat, GBLPAGESK started at 150000 and was 219289 after the installation.  Most systems hereiJ have it bumped with a MIN_GBLPAGES of 300000.  On my V7.3 system, GBLPAGESK started at 6499743.  It looks like the minimum specifier won't be necessarynD here!  I looked in the V7.3 Release Notes and didn't see anything toJ explain the increase.  Does anyone know why there was such a large change?   ------------------------------  # Date: Wed, 17 Oct 2001 00:03:20 GMTe2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)! Subject: Re: GBLPAGES in VMS V7.361 Message-ID: <cV3z7.538$RL6.4519@news.cpqcorp.net>o  T In article <3BCC82F4.3BB4FA46@mail.com>, John Johnstone <jj_usenet@mail.com> writes:K :I've been looking at a system running VMS V7.3 and it looks like the valuev& :of GBLPAGES has taken quite a jump...4 :Does anyone know why there was such a large change?  F   We were asked to make OpenVMS easier to manage, and this is a (very)%   small part of that (on-going) work.o  G   The setting of GBLPAGES determines the size of the Global Page Table rH   (GPT), and it is now initially sized to support larger RMS caches and &   other (large) users of global pages.  B   Now for some details on the cost of this change, in terms of theH   new physical memory requirements: the cost of sizing the GPT is three E   memory pages (assuming an eight kilobyte page size) for every four sF   million pages (well, actually, every 4,194,304 pages) configured in F   the system.  In other words, the overhead of the new higher setting F   is viewed as very low, and the value derived from easily increasing F   the available global pages and particularly in easily increasing in @   the available RMS caching are viewed as reasonable trade-offs.  I   As for how to memory cost of the GBLPAGES (GPT) setting was determined,iJ   since somebody is going to ask: on Alpha, the page table entries (PTEs) G   are 8 bytes in size and one is needed for page mapped.  Assuming the aG   standard page size of eight kilobytes that is found on current Alpha  I   systems, the page tables will need 1/1024 of the virtual address space  K   that the tables will map.  The GPT is mapped in S2 space virtual memory,  L   and these tables are initially demand zero pages -- this means the tables H   require no physical memory until the GPT is referenced and the global K   pages are used.  (This behaviour has been the case for a long time -- as mG   the global pages are actually used, the pages comprising the GPT are  H   instantiated and locked into memory.)  Therefore, the physical memory H   consumption is solely for the system page tables that are required to F   map the global page table.  This yields another reduction in memory D   requirements of a factor of 1024.  If global page table space for G   three-quarters of the total physical memory is provided -- as is the nH   case on V7.3 -- the initial (large) setting of GBLPAGES consumes threeH   pages for every 4,194,304 (4*1024*1024) pages of the of the available    system physical memory.o  H   In other words, setting up the GPT large is cheap -- and you then pay G   for what global pages you actually use, as you always have.  But you -F   don't have to manage and you don't have to resize the GPT unless you
   want to.  I   Short answer: If the new GPT sizing bothers you, you can revert to the  E   older setting and/or you can buy more memory -- OpenVMS EngineeringtK   believes this change was a good (and a cheap) trade-off for the benefits.a  I   Another related change in V7.3: OpenVMS Engineering stuffed all of the iI   pageable and the non-pageable OpenVMS executive into the same quantity SH   of memory formerly required for just the non-pageable executive -- we H   got rid of the pageable memory requirements for the executive, and we G   improved overall system performance by eliminating all paging of the i   executive.    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Tue, 16 Oct 2001 19:30:38 GMTt  From: jlsue <jlsuexxxz@home.com> Subject: Re: Global symbol8 Message-ID: <ig2pstcqh2to2nshhri1fn99edmc40lehf@4ax.com>  E On Wed, 17 Oct 2001 00:34:25 +0800, "Kenneth" <chehon@netvigator.com>g wrote:  I >I have a global system symbol "success == 1, failure ==0" defined duringTG >system startup  and everyone can use the same symbol as a check in theEL >command procedure. However, this symbol cannot be seen, ie. the system willH >treat this is an undefine symbol,  when the command file is call by rsh> >(remote shell). How can I make the symbol can be seen in rsh? >    Um, why not just use IF? ;-)   ------------------------------  % Date: Tue, 16 Oct 2001 15:43:36 -0400R0 From: "Syltrem" <syltrem@videotron.spammenot.ca> Subject: Re: Global symbol5 Message-ID: <120z7.69573$TW.365873@tor-nn1.netcom.ca>    > Um, why not just use IF? ;-)   As in: $ IF $STATUS# $ THEN WRITE SYS$OUTPUT "ALL IS OK"). $ ELSE WRITE SYS$OUTPUT "SOMETHING WENT WRONG" $ ENDIF,  I You can also do an IF .NOT. $STATUS THEN WRITE SYS$OUTPUT "SOMETHING WENT  WRONG"   HTH  --   SyltremFI http://pages.infinit.net/syltrem (OpenVMS related web site - en franais) > To reply to myself directly, remove .spammenot from my address  : "jlsue" <jlsuexxxz@home.com> a crit dans le message news:- ig2pstcqh2to2nshhri1fn99edmc40lehf@4ax.com... G > On Wed, 17 Oct 2001 00:34:25 +0800, "Kenneth" <chehon@netvigator.com>  > wrote: > K > >I have a global system symbol "success == 1, failure ==0" defined during I > >system startup  and everyone can use the same symbol as a check in theOI > >command procedure. However, this symbol cannot be seen, ie. the system  willJ > >treat this is an undefine symbol,  when the command file is call by rsh@ > >(remote shell). How can I make the symbol can be seen in rsh? > >' >E > Um, why not just use IF? ;-)   ------------------------------  % Date: Tue, 16 Oct 2001 20:18:35 -0000 / From: Michael Zarlenga <zarlenga@conan.ids.net>  Subject: Re: Global symbol/ Message-ID: <tsp5grde1d8kd2@corp.supernews.com>-  / Syltrem <syltrem@videotron.spammenot.ca> wrote:AK : You can also do an IF .NOT. $STATUS THEN WRITE SYS$OUTPUT "SOMETHING WENT  : WRONG"   Wouldn't you want    	if (.not. ($STATUS .and. 1))    ?N   -- T -- Mike Zarlenga   ------------------------------  # Date: Tue, 16 Oct 2001 20:42:49 GMT$= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)' Subject: Re: Global symbol0 Message-ID: <00A039EA.0E1DAE3E@SendSpamHere.ORG>  a In article <tsp5grde1d8kd2@corp.supernews.com>, Michael Zarlenga <zarlenga@conan.ids.net> writes:R0 >Syltrem <syltrem@videotron.spammenot.ca> wrote:L >: You can also do an IF .NOT. $STATUS THEN WRITE SYS$OUTPUT "SOMETHING WENT	 >: WRONG"A >/ >Wouldn't you want A >- >	if (.not. ($STATUS .and. 1)) r  B Only if you like adding unnecessary processing to your procedures.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COME             J   "And of course, I'm a genius, so people are naturally drawn to my fiery I   intellect.  Their admiration overwhelms their envy!" -- Calvin & HobbesC   ------------------------------  + Date: Tue, 16 Oct 2001 18:02:38 +0000 (UTC)i From: david20@alpha1.mdx.ac.uk/ Subject: Re: Higher prices for Alpha processors + Message-ID: <9qhsnu$48u$1@aquila.mdx.ac.uk>   i In article <ZJZy7.519$RL6.4554@news.cpqcorp.net>, "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> writes:R > K >But I contend that most people, aside from the zealots, do not really care K >about the processor ISA.  They care about the price/performance - and thatWL >only a handfull of people are way out on the leading edge of wanting to payJ >for absolute performance.  What they do care about is competetive pricingG >with competetive performance in the industry.  And that their softwareF >investment is protected.  >   8 But that is the point. Itanium will not be a cheap chip.H Itanium is not IA32 !! Itanium will not take over the desktop - AMD will see to that.  + See http://www.theinquirer.net/16100101.htm$  $ AMD's Hammer to smash Intel's Plans.    I >We are porting to Itanium.  We will do our best to make it as painless aIM >transition as possible.  People who need raw performance, and need it beforeRC >say 2004/2005, can safely buy EV6x and EV7x based systems with theoK >appropriate price/performance.  When Itanium comes along, it will at firstrJ >be for early adopters - and won't be as fast as existing Alpha platforms.G >But it WILL be able to run Windows as well as VMS, something we can nonF >longer say about Alpha.  People can make apples-to-apples comparisonsJ >between vendors offerings - since pretty much everyone except SUN will beL >selling Itanium systems.  And at some point - and I am not smart enough (orL >paid enough) to know *exactly* when, Itanium family systems will be as big,/ >bigger, fast, and faster than Alpha platforms.r >@  . Only because development of Alpha was stopped.    I >I think with enough money, Alpha could have been performance competetive K >with Intel and IBM.  But it would have taken a lot of money to stay out inuK >front of both, and the delta between performance might not have been largeeG >enough for someone to pick Alpha over Itanium - especially in the UNIX-M >space.  I think not having control over the fab makes it tough to compete onl >raw performance.H >   H Not owning the FAB was supposed to have saved massive ammounts of money.M Why should having someone else FAB the chips have any affect on performance ?B  
 David Webb VMS and Unix team leader CCSS Middlesex University     PS.k  J Better hope that when they do the work to port VMS to Itanium they make itI truly portable because I foresee another port in a couple of years to AMDd Hammer or POWER 4.       >n >a >Bill Todd wrote in message ...e >>@ >>Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote in message- >>news:BrEy7.440$RL6.3822@news.cpqcorp.net...t >> >>...r >>5 >>> We continue to sell Alphas into new accounts, andtH >>> continue to get new sales every day on Alpha.  The Business Critical	 >>SystemsaK >>> division is focused on selling Alphas.  There IS NO let up in trying tof >>sellJ >>> Alpha, which would be foolish since we will not have a competetive IPF: >>> offering (with EV68 or EV7 or EV79) for quite a while. >>J >>So IPF won't be competitive until some time after EV79, eh?  Which meansI >>about the time EV8 would have shipped, in which case it seems extremelySK >>likely that IPF wouldn't have been competitive then either, for some even- >>longer period. >>I >>Can this be the same Fred K. who so eloquently supported the party line  >that:L >>Alpha couldn't have maintained a sufficient performance advantage over IPFJ >>to justify continued development?  Sure sounds as if you're saying aboveH >>that Alpha could have easily held onto at least its own markets (which >couldL >>have been expanded significantly with even minimal effort) against IPF for/ >>at least 5 - 6 years, had EV8 been completed.n >> >>- bill >> >> >> >a >e   ------------------------------  # Date: Tue, 16 Oct 2001 18:29:04 GMTr* From: "Bill Todd" <billtodd@metrocast.net>/ Subject: Re: Higher prices for Alpha processors-? Message-ID: <Q%_y7.513620$Lw3.31430924@news2.aus1.giganews.com>p  > Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote in message+ news:ZJZy7.519$RL6.4554@news.cpqcorp.net...i% > Please don't put words in my mouth.n  K I didn't:  you stated very clearly that "we will not have a competetive IPF1I offering (with EV68 or EV7 or EV79) for quite a while", which is hardly ar difficult statement to parse.   #   I expect that had we continued totK > develop the EV8 that it would have remained "competetive" with Itanium in # > performance, but I doubt it cost.   I A qualification you failed to make in your statement quoted above (thoughwB admittedly it applied only to EV7 and EV79).  Of course, given theD relatively low impact individual processor cost has on even a modestI mid-range server system (compared with, for example, the system *savings* ? that needing fewer processors to handle the same load creates),aL per-processor cost for anything beyond desktop-class systems is not all that	 critical.   &   The cost to do it apparently did not > justify the investment,h  L What is apparent is that Compaq didn't care to continue development, despiteF Alpha's healthy profits compared with other Compaq endeavors (with theK exception of Tandem products).  Compaq's own numbers prove that Alpha was arI better investment than PCs - and probably better than 'industry-standard'r servers.  0 > and it only would have gotten worse over time.  C Only if Compaq had continued to refuse to promote the architecture:eK opportunity for major sales growth persisted, despite Compaq's best effortsb' to stifle it, right up until June 25th.o   > L > EV6 was a pretty good leap in technology, and it was hard to do - busted a@ > lot of software along the way to get it right.  The OoO stuff, speculation,H > using cache state for multiprocessor locking, etc.  EV7 doesn't reallyE > change anything there, instead it focuses on things that are really  outsideDH > of the processor - like the cost of getting to memory, and how to glueK > together a NUMA system without a costly hierarchical switch.  EV8's focuseG > was mostly on how to leverage the resources on a single die which are0L > normally inefficiently used - which doesn't really help with single stream > performance.  J Would you care to estimate the percentage of Alpha revenue that depends onL single-stream use?  My own impression was that servers (where EV8 would haveI excelled, even in comparison with multi-core-per-die POWER4s, let alone aVI single-threaded SPECint pig like Itanium) brought in most of the revenue.   <   The processor speed, and number of units on the ship wouldH > have increased that somewhat - and made it "competetive" with Itanuim.  L You stated clearly that Itanium would *not* have been competitive with AlphaL through EV79, so to say that Alpha (especially EV8) would have *merely* beenF 'competetive' [sic] with Itanium seems disingenuous at the very least.   >mF > But how much faster does it have to be to "win" in the market place?  L Not all that much, if its owner cared to make any effort whatsoever.  But ifJ your contention is correct that EV79 would have retained a sufficient leadH that IPF wouldn't have been 'competetive', then EV8 would have blown IPFK completely away in any server-style application (and had at least some leadf. over POWER4, especially in power consumption).  K > Clearly there have been times when we have had well over a 2x performancet< > lead - but that didn't translate into it eating SUNs lunch  J To eat someone's lunch, you first have to be hungry.  Compaq instead actedG as if it was fat and happy (which, given that it in fact was not, seems  massively incompetent).d  
  - who have ayI > pathetic turd-on-a-chip called Sparc.  Yes, it made us the standard foroJ > people who were willing to pay *any* price for the highest performance - soF > the genome people love it, some animation shops love it, some of theI > military love it, and some of the supercomputer people love it. The VMSa% > people love it BECAUSE IT RUNS VMS.s  K The Tru64 people, who generate revenues 3/4 as high as VMS's despite a muchiK shorter history and were growing at a significantly faster rate, apparently G love it to a large degree because of the hardware.  And so do the Alpha I Linux folks, who could run Linux on just about anything.  So generalizingnH about why the VMS people love it seems a bit risky (especially given theI unequivocal sentiments so many of them have expressed in comp.os.vms overl the past 4 months).-   >-F > We have a lot of experience building relatively large servers out of Alpha.A > EV7 will provide really fast systems from 2 processors up to 64.H > (architecturally 256, and realistically 128 could be built).  EV7, and EV79F > will remain competetive with everything - including the Power4.  But ItaniumtJ > *will* catch up within a performance delta that is immaterial to all but% > zealots, and eventually surpass it._  H That remains to be seen, even *with* the influx of the Alpha engineeringD talent and *without* continued Alpha development; had Compaq insteadK continued with EV8 and kept its engineers, your statement would on the faceeJ of it be laughable.  Alpha had a credible performance road map through EV8K that suggested *increasing* absolute performance leads over Itanic (leavingnJ aside even greater efficiencies in power consumption, especially with EV8,G system performance, as a result of EV7's on-chip glue for MP and memorymL access, and system cost, as a result of requiring fewer processors to do the same work).V  &   We can all take wild guesses at whatL > that timeframe is.  And we can all speculate about when and how fast a EV8 > and EV9 would have been. >kL > But I contend that most people, aside from the zealots, do not really careL > about the processor ISA.  They care about the price/performance - and thatI > only a handfull of people are way out on the leading edge of wanting toe pay  > for absolute performance.o  I And, as explained above, price/performance is where Alpha would have mosttF eclipsed Itanic, save for in the low-end server market (where IA32 andH Hammer price/performance will trounce Itanic).  This would have remainedC true even if Intel charged $0 for each processor:  the system-leveld efficiencies are what matter.o  0   What they do care about is competetive pricingH > with competetive performance in the industry.  And that their software > investment is protected. >uJ > We are porting to Itanium.  We will do our best to make it as painless aG > transition as possible.  People who need raw performance, and need it  beforeD > say 2004/2005, can safely buy EV6x and EV7x based systems with theL > appropriate price/performance.  When Itanium comes along, it will at firstK > be for early adopters - and won't be as fast as existing Alpha platforms.cH > But it WILL be able to run Windows as well as VMS, something we can no > longer say about Alpha.e  0 Right:  Compaq fixed that situation a while ago.  .   People can make apples-to-apples comparisonsK > between vendors offerings - since pretty much everyone except SUN will bel > selling Itanium systems.  E I think you've been taking lessons from Mikey:  of *course* we shouldn4 protect customers from all that confusing diversity.  3   And at some point - and I am not smart enough (orlH > paid enough) to know *exactly* when, Itanium family systems will be as big,0 > bigger, fast, and faster than Alpha platforms.  J Naturally, since Alpha systems will no longer be being enhanced.  If AlphaL development had continued, the situation you describe would almost certainly never have occurred.  K Unless Intel manages to cow IBM as it did Compaq, Itanic will never own thenF mid-range-and-up server market, since POWER4 will offer at least equalI per-processor performance, at least double the per-die performance (givenrE its dual full processor cores per die), more than double the per-WattgD performance (a single two-core POWER4 die consumes less power than aH single-processor McKinley die is projected to, and also contains on-chipG glue for MP and memory access that AFAIK Merced, McKinley, Madison, andtK Deerfield do not and hence must power external chips to provide), and *far* G better system performance (due to that on-chip glue plus the very tightsG coupling possible between two threads running in the same die) and thusfK cost-effectiveness.  If the Alpha team can bring such features to Itanic atiK all, it won't be before 2005 (for such peripheral on-chip glue) or 2006 (at K the earliest, for things like OOO and SMT that would help Itanic be less of C a server pig - though it will still retain significant unproductiveiK overheads compared with POWER4 unless they scrap EPIC pretty much entirely,-C at which point the question will be just how effectively the ItanicQK instruction set can be mapped to an EV8-like architecture, sort of like theiI x86 instruction set is mapped to an underlying RISC-like implementation -e blech).o  L And unless Intel manages to drive AMD out of business, Itanic will never ownE the low end either, because a great deal of it can be handled by IA32aK processors and Hammer looks far more attractive than Itanic for the balance  that requires 64-bit support.n  I So if Itanic succeeds at all, it won't be for technical reasons.  And I'mcI guessing that there's at least a fair chance that Itanic will *never* pays' for its own development (unlike Alpha).m   >pJ > I think with enough money, Alpha could have been performance competetiveL > with Intel and IBM.  But it would have taken a lot of money to stay out inL > front of both, and the delta between performance might not have been largeH > enough for someone to pick Alpha over Itanium - especially in the UNIX > space.  L As noted above, Hammer and IA32 look like much tougher competition for AlphaL in the low end than Itanic does.  And POWER4 is certainly the competition inK the high end:  I think EV7 might have had difficulty competing with POWER4,LG but that EV8 would have had slight advantages over it - and even takingnE second-place to POWER4 in that lucrative a market would have returned 0 profits to Compaq far beyond any it's ever seen.  F   I think not having control over the fab makes it tough to compete on > raw performance.  I IBM supposedly fabs for others as aggressively as it does for itself, nowgF (with its 'anything for a buck' attitude) more than ever.  While I canL understand why having someone else (especially a competitor) responsible forH fabbing Alphas could make Alpha's owner nervous, IBM more than anyone isL sensitive to anti-trust regulation - and of course there's always Samsung to help keep them honest.  I Except for your unusually forthright statement about Itanic's inferiorityoJ that I quoted initially above, you seem to be reverting to your defense ofK Compaq's to-all-appearances completely indefensible position on this topic.-G If you insist on doing so, could you at least try to present some *new*r( information supporting your contentions?   - bill   ------------------------------  % Date: Tue, 16 Oct 2001 15:04:52 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>A/ Subject: Re: Higher prices for Alpha processorsE, Message-ID: <3BCC84D3.898D1391@videotron.ca>   Bob Ceculski wrote:.H > nt will never be on the level of vms ... are you forgetting why people > are still on vms?   H The number of people still on VMS is going down as competing systems areH getting closers and closer, they meet more and more of customer's needs.    H > another plus!  vms will be around thru 2015 or longer and the military > will be the primary users   M Yep, just like the military probably was the primary user of DG's dead AOS-VSnE until just recently when they switched to VMS on their JSTARs planes.W  , >along with us smart IT managers like myselfB > that aren't idiots ... nt catch vms, what have you been smoking?  F "smart" is the wrong word. "LUCKY" is the right word. Lots of smart ITH managers don't have the luxury of being able to choose the best possibleN solution , no matter what its price is. Many must follow corporate edicts thatN require only industry standard solutions, and others have budgets that preventN them from buying VMS. And many more can't buy VMS because the application they need doesn't run on VMS.   ------------------------------  % Date: Tue, 16 Oct 2001 16:05:41 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>e/ Subject: Re: Higher prices for Alpha processorsm, Message-ID: <3BCC9311.DE8EDAF1@videotron.ca>   Fred Kleinsorge wrote:F > But how much faster does it have to be to "win" in the market place?K > Clearly there have been times when we have had well over a 2x performanced= > lead - but that didn't translate into it eating SUNs lunch e   Pricing and marketing.  I Compaq could have fixed this when it bought Digital because at that time,sL Compaq had the marketing saavy and the power to take the leap to lower priceF to increase volume. But Compaq chose to ignore the potential of Alpha.  L > But I contend that most people, aside from the zealots, do not really careB > about the processor ISA.  They care about the price/performance   M We saw the potential for Alpha. And we saw how Compaq not only squandered itslN potential, but eventually lied about its commitments and killed the chip. ThatH is the important issue here, not whether Alpha could have had 1hz faster" performance than IA64 or the 8086.  5 The problem is that VMS seems to be in the same boat.f        J > We are porting to Itanium.  We will do our best to make it as painless a > transition as possible.n  L Woopty doo. I would have rathered see Compaq announce the death of Alpha, noN port of VMS, but a big marketing budget for VMS who would then have 5 years to prove its potential. p  I Compaq continues to ignore VMS and that hasn't changed with the murder oftL Alpha, even if there was a one liner "by the way, we'll also port that thing1 called VMS to IA64"  hidden in the announcements.a  5 > People who need raw performance, and need it beforeaD > say 2004/2005, can safely buy EV6x and EV7x based systems with the  > appropriate price/performance.  ? Correct for folks who already have alpha or who need short term I infrastructure. But someone isn't about to invest in a platform that willl8 require a port to another platform in a couple of years.  H > But it WILL be able to run Windows as well as VMS, something we can no > longer say about Alpha.   D Again, that is because Compaq wasn't interested in pushing Alpha, soV understandably MS and Compaq found a way to justify pulling the plug on Windows-Alpha.  M > paid enough) to know *exactly* when, Itanium family systems will be as big,V0 > bigger, fast, and faster than Alpha platforms.  N Digital had it and wasted it. Alpha could have been the de-facto standard. ForK Christ's sake, they  were *10 years*  ahead of Intel and they allowed theirrN own 8086 machines to eat their own Alpha. Digital cut its own limbs to a pointM of extermination and donated its remains to some small PC company who ignoredtF anything "Digital" and is now melting itself to fit into some HP bowl.    N The management incompetence that resulted in the wasting of Alpha and VMS  has1 not died with Alpha's death. That is the problem.w  L Why should anyone believe that VMS will be any more succesful on IA64 or any other chip ?F Why should anyone believe that VMS will be any more succesful under HPF management who have claimed that they would focus on industry standard solutions ?i   ------------------------------    Date: 16 Oct 2001 13:39:13 -0700( From: bob@instantwhip.com (Bob Ceculski)/ Subject: Re: Higher prices for Alpha processorsr< Message-ID: <d7791aa1.0110161239.dcc3a0c@posting.google.com>  r "Bill Todd" <billtodd@metrocast.net> wrote in message news:<RtLy7.16134$%B.2052129@bin1.nnrp.aus1.giganews.com>...@ > Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote in message- > news:BrEy7.440$RL6.3822@news.cpqcorp.net...  >  > ...l > 5 > > We continue to sell Alphas into new accounts, andeH > > continue to get new sales every day on Alpha.  The Business Critical
 >  SystemsK > > division is focused on selling Alphas.  There IS NO let up in trying toe >  sellcJ > > Alpha, which would be foolish since we will not have a competetive IPF: > > offering (with EV68 or EV7 or EV79) for quite a while. > J > So IPF won't be competitive until some time after EV79, eh?  Which meansI > about the time EV8 would have shipped, in which case it seems extremelycK > likely that IPF wouldn't have been competitive then either, for some eveno > longer period. > N > Can this be the same Fred K. who so eloquently supported the party line thatL > Alpha couldn't have maintained a sufficient performance advantage over IPFJ > to justify continued development?  Sure sounds as if you're saying aboveN > that Alpha could have easily held onto at least its own markets (which couldL > have been expanded significantly with even minimal effort) against IPF for/ > at least 5 - 6 years, had EV8 been completed.e >  > - bill  M alpha kills itanic now and would of later with ev8 ... why do you think inteloL bought it ... the puzzling question is why compaq was dumb enough to give it8 away?  i hope the shareholders are asking capellas that!   ------------------------------  % Date: Tue, 16 Oct 2001 16:44:33 -0400t5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>s/ Subject: Re: Higher prices for Alpha processors 1 Message-ID: <L_0z7.531$RL6.4740@news.cpqcorp.net>d  L david20@alpha1.mdx.ac.uk wrote in message <9qhsnu$48u$1@aquila.mdx.ac.uk>...D >In article <ZJZy7.519$RL6.4554@news.cpqcorp.net>, "Fred Kleinsorge"% <kleinsorge@star.zko.dec.com> writes:. >>L >>But I contend that most people, aside from the zealots, do not really careL >>about the processor ISA.  They care about the price/performance - and thatI >>only a handfull of people are way out on the leading edge of wanting toh payyK >>for absolute performance.  What they do care about is competetive pricing H >>with competetive performance in the industry.  And that their software >>investment is protected. >> >y9 >But that is the point. Itanium will not be a cheap chip.tI >Itanium is not IA32 !! Itanium will not take over the desktop - AMD willm
 >see to that.  >e, >See http://www.theinquirer.net/16100101.htm >r% >AMD's Hammer to smash Intel's Plans.t >o    D IMHO.  In a few years, Intel and Microsoft will force IA32 to becomeE obsolete so that they can churn the customer base and get new licenserI revenue, and new system sales.  Both depend on people buying new systems,eK and not holding onto old systems.  IMHO Hammer will die the death of Alpha,o6 Microsoft will give it limp support and let it fizzle.  F I would love to have continued to use Windows 3.1, but I was forced toJ Windows 95 - which also obsoleted my hardware.  I am now being battered toI go to WIndows 2K, which will force me to abandon almost all of my currentvL hardware, and some software that I will need to abandon, or re-purchase.  ItI is no great stretch to see this same thing happen with Itanium.  It is inlA both Microsofts and Intels interest to force another great churn.e   ------------------------------  % Date: Tue, 16 Oct 2001 17:06:08 -0400 5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> / Subject: Re: Higher prices for Alpha processorsf1 Message-ID: <_i1z7.535$RL6.4702@news.cpqcorp.net>e  I What are *Alphas* "peculiar capabilities that don't exist on wintel".  Be L specific.  Although I think the question is idiotic, since you are comparingL hardware (Alpha) to hardware *and* software ("Wintel").  The question reallyF should compare Alpha hardware versus Intel hardware, built by the sameH people (what, you think that Fenwick and company have retired?), for theI same market.  So that the real difference in the end becomes nothing morew than the ISA of the chip.v  J What is *left* is absolute, and to a great degree *speculated* performanceI of chip generations at any particular point in time.  Frankly, for 97% ofoJ people using VMS - I don't think they care if the hardware is the absoluteI fastest in the world at the moment they buy it.  But they would really bee" interested in seeing it *cheaper*.      = JF Mezei wrote in message <3BCBCAF3.D9323503@videotron.ca>...  >"Main, Kerry" wrote: F >> As stated in previous thread, not selling Alpha's until competitive5 >> flavours of IPF are available would be ridiculous.r >>K >> Does not HP still continue to market and sell PA very agressively today?h >> Of course they do.w >cI >The difference is that Compaq never wanted to sell Alpha, they wanted to  sellJ >Wintel servers. The June 25 announcement removed any wind left in Alpha'sE >sales, except for those who have no choice because they need Alpha's  peculiar) >capabilities that don't exist on wintel.l   ------------------------------  % Date: Tue, 16 Oct 2001 17:11:49 -0400 5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>l/ Subject: Re: Higher prices for Alpha processors 1 Message-ID: <jo1z7.536$RL6.4702@news.cpqcorp.net>   = JF Mezei wrote in message <3BCC84D3.898D1391@videotron.ca>...e >Bob Ceculski wrote:I >> nt will never be on the level of vms ... are you forgetting why peoplee >> are still on vms? >fI >The number of people still on VMS is going down as competing systems areeI >getting closers and closer, they meet more and more of customer's needs.t >n >oI >> another plus!  vms will be around thru 2015 or longer and the militaryd >> will be the primary users >*G >Yep, just like the military probably was the primary user of DG's deadg AOS-VSF >until just recently when they switched to VMS on their JSTARs planes. >     J Please explain this to me.  The original JSTARS was VMS on VAX and Alphas,E the prototype JSTARS flew in Desert Storm.  The latest is all Alphas.e4 Exactly where does this "DG" crap keep comming from?  - >>along with us smart IT managers like myselfaC >> that aren't idiots ... nt catch vms, what have you been smoking?e >eG >"smart" is the wrong word. "LUCKY" is the right word. Lots of smart IT I >managers don't have the luxury of being able to choose the best possiblelJ >solution , no matter what its price is. Many must follow corporate edicts thatG >require only industry standard solutions, and others have budgets thate preventcJ >them from buying VMS. And many more can't buy VMS because the application they >need doesn't run on VMS.t  L Yup.  We're a more of a niche player today than any time in the past.  LargeJ niches that are profitable to all, and where VMS's strengths are worth theH cost.  Odds are if a third party application doesn't run on VMS, and youL aren't in the right vertical segment to cause us to pay to get it supported,: then you won't be able to use VMS as a viable alternative.   ------------------------------  % Date: Tue, 16 Oct 2001 16:39:03 -0400 5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>a/ Subject: Re: Higher prices for Alpha processorsa1 Message-ID: <zV0z7.530$RL6.4732@news.cpqcorp.net>   D Please. please. please.  Do not respond to me.  I made a mistake andF responded to your assertions.  You want to parse things into nice neatJ little bundles.  I'm a liar or an idiot, or both.  Not a  __ing thing willE alter that in your universe.  This isn't a court of law, or a ColumboeJ episode - and the truth (and opinion) is seldom self consistant.  You wantG to postulate performance characteristics, time to market and cost - alloH fabricated out of pretty much hearsay - good *and* bad.  Spin on.  Enjoy	 yourself.m  I I'll stick to these loose assertions:  Alpha was great.  Loved it.  Had aeH great time.  A decision I had no part in making was that it was not costL effective to invest in the Alpha ISA long term (at this point I really don'tL care if the reason is valid or not, or if *you* agre with it), and to switchH to the Itanium architecture - for good or ill.  People making a lot moreJ money than I do believe that Itanium will be competetive, and we have EV6xC and EV7x until it is.  I myself will hedge on the exact date.   The,L performance of the mythical EV8 or EV9 are not interesting, as they will notF be built.  We will be on an industry standard architecture, offered byB everyone except (perhaps) SUN.  "Hammer" will never be embraced byK Microsoft, and will quietly die.  IA32 will be phased out by the end of the H first decade of the 2000's, as both Intel and Microsoft "churn" the userL base to force PCs to be replaced.  Linux will follow (or lead).  Most of ourK customers will be happy with the VMS to Itanium transition, which will giveeL them compatetively priced hardware.  A minority will make a decision to moveH onto another vendors OS.  We will do everything in our power to make the transition long, and smooth.  K It was foolish of me to A) say anything about Itanuim, and worse B) respondb to anything you write.  K I promise.  promise.  promise to not respond to you, and to ignore you fromt now on.  It was my mistake.I  I Really.  I have figured out how to work the newsgroup filters, I just hadi been avoiding it in this group.    _Fredp   Bill Todd wrote in message ... > ? >Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote in message-, >news:ZJZy7.519$RL6.4554@news.cpqcorp.net...& >> Please don't put words in my mouth. >,L >I didn't:  you stated very clearly that "we will not have a competetive IPFJ >offering (with EV68 or EV7 or EV79) for quite a while", which is hardly a >difficult statement to parse. >4$ >  I expect that had we continued toL >> develop the EV8 that it would have remained "competetive" with Itanium in$ >> performance, but I doubt it cost. >nJ >A qualification you failed to make in your statement quoted above (thoughC >admittedly it applied only to EV7 and EV79).  Of course, given thetE >relatively low impact individual processor cost has on even a modestrJ >mid-range server system (compared with, for example, the system *savings*@ >that needing fewer processors to handle the same load creates),H >per-processor cost for anything beyond desktop-class systems is not all that
 >critical. > ' >  The cost to do it apparently did notE >> justify the investment, > E >What is apparent is that Compaq didn't care to continue development,  despite G >Alpha's healthy profits compared with other Compaq endeavors (with the L >exception of Tandem products).  Compaq's own numbers prove that Alpha was aJ >better investment than PCs - and probably better than 'industry-standard'	 >servers.  >e1 >> and it only would have gotten worse over time.  >pD >Only if Compaq had continued to refuse to promote the architecture:L >opportunity for major sales growth persisted, despite Compaq's best efforts( >to stifle it, right up until June 25th. >e >>K >> EV6 was a pretty good leap in technology, and it was hard to do - busted  anA >> lot of software along the way to get it right.  The OoO stuff,e
 >speculation,nI >> using cache state for multiprocessor locking, etc.  EV7 doesn't reallyeF >> change anything there, instead it focuses on things that are really >outsideI >> of the processor - like the cost of getting to memory, and how to gluelL >> together a NUMA system without a costly hierarchical switch.  EV8's focusH >> was mostly on how to leverage the resources on a single die which areF >> normally inefficiently used - which doesn't really help with single stream >> performance.e >.K >Would you care to estimate the percentage of Alpha revenue that depends onsH >single-stream use?  My own impression was that servers (where EV8 would haveJ >excelled, even in comparison with multi-core-per-die POWER4s, let alone aJ >single-threaded SPECint pig like Itanium) brought in most of the revenue. >e= >  The processor speed, and number of units on the ship wouldeI >> have increased that somewhat - and made it "competetive" with Itanuim.  > G >You stated clearly that Itanium would *not* have been competitive withb AlphasH >through EV79, so to say that Alpha (especially EV8) would have *merely* beenG >'competetive' [sic] with Itanium seems disingenuous at the very least.y >f >>G >> But how much faster does it have to be to "win" in the market place?t >sJ >Not all that much, if its owner cared to make any effort whatsoever.  But ifK >your contention is correct that EV79 would have retained a sufficient leadwI >that IPF wouldn't have been 'competetive', then EV8 would have blown IPFsL >completely away in any server-style application (and had at least some lead/ >over POWER4, especially in power consumption).l >tL >> Clearly there have been times when we have had well over a 2x performance= >> lead - but that didn't translate into it eating SUNs luncho >iK >To eat someone's lunch, you first have to be hungry.  Compaq instead acted3H >as if it was fat and happy (which, given that it in fact was not, seems >massively incompetent). >w > - who have aJ >> pathetic turd-on-a-chip called Sparc.  Yes, it made us the standard forK >> people who were willing to pay *any* price for the highest performance -  >sonG >> the genome people love it, some animation shops love it, some of theeJ >> military love it, and some of the supercomputer people love it. The VMS& >> people love it BECAUSE IT RUNS VMS. > L >The Tru64 people, who generate revenues 3/4 as high as VMS's despite a muchL >shorter history and were growing at a significantly faster rate, apparentlyH >love it to a large degree because of the hardware.  And so do the AlphaJ >Linux folks, who could run Linux on just about anything.  So generalizingI >about why the VMS people love it seems a bit risky (especially given thesJ >unequivocal sentiments so many of them have expressed in comp.os.vms over >the past 4 months). >e >>G >> We have a lot of experience building relatively large servers out ofn >Alpha.wB >> EV7 will provide really fast systems from 2 processors up to 64I >> (architecturally 256, and realistically 128 could be built).  EV7, anda >EV79oG >> will remain competetive with everything - including the Power4.  But  >ItaniumK >> *will* catch up within a performance delta that is immaterial to all butt& >> zealots, and eventually surpass it. >aI >That remains to be seen, even *with* the influx of the Alpha engineeringdE >talent and *without* continued Alpha development; had Compaq insteadeL >continued with EV8 and kept its engineers, your statement would on the faceK >of it be laughable.  Alpha had a credible performance road map through EV8aL >that suggested *increasing* absolute performance leads over Itanic (leavingK >aside even greater efficiencies in power consumption, especially with EV8, H >system performance, as a result of EV7's on-chip glue for MP and memoryI >access, and system cost, as a result of requiring fewer processors to dot ther >same work). >A' >  We can all take wild guesses at whatuI >> that timeframe is.  And we can all speculate about when and how fast a  EV8* >> and EV9 would have been.  >>H >> But I contend that most people, aside from the zealots, do not really careH >> about the processor ISA.  They care about the price/performance - and thatJ >> only a handfull of people are way out on the leading edge of wanting to >pay >> for absolute performance. >iJ >And, as explained above, price/performance is where Alpha would have mostG >eclipsed Itanic, save for in the low-end server market (where IA32 andtI >Hammer price/performance will trounce Itanic).  This would have remainedrD >true even if Intel charged $0 for each processor:  the system-level >efficiencies are what matter. >h1 >  What they do care about is competetive pricingsI >> with competetive performance in the industry.  And that their softwaree >> investment is protected.  >>K >> We are porting to Itanium.  We will do our best to make it as painless a H >> transition as possible.  People who need raw performance, and need it >before@E >> say 2004/2005, can safely buy EV6x and EV7x based systems with theoG >> appropriate price/performance.  When Itanium comes along, it will at  firstoL >> be for early adopters - and won't be as fast as existing Alpha platforms.I >> But it WILL be able to run Windows as well as VMS, something we can nom >> longer say about Alpha. >71 >Right:  Compaq fixed that situation a while ago.  > / >  People can make apples-to-apples comparisonseL >> between vendors offerings - since pretty much everyone except SUN will be >> selling Itanium systems.u > F >I think you've been taking lessons from Mikey:  of *course* we should5 >protect customers from all that confusing diversity.e >i4 >  And at some point - and I am not smart enough (orI >> paid enough) to know *exactly* when, Itanium family systems will be as  >big,i1 >> bigger, fast, and faster than Alpha platforms.i >hK >Naturally, since Alpha systems will no longer be being enhanced.  If AlphaeC >development had continued, the situation you describe would almost 	 certainly. >never have occurred.o >dL >Unless Intel manages to cow IBM as it did Compaq, Itanic will never own theG >mid-range-and-up server market, since POWER4 will offer at least equal J >per-processor performance, at least double the per-die performance (givenF >its dual full processor cores per die), more than double the per-WattE >performance (a single two-core POWER4 die consumes less power than acI >single-processor McKinley die is projected to, and also contains on-chipeH >glue for MP and memory access that AFAIK Merced, McKinley, Madison, andL >Deerfield do not and hence must power external chips to provide), and *far*H >better system performance (due to that on-chip glue plus the very tightH >coupling possible between two threads running in the same die) and thusL >cost-effectiveness.  If the Alpha team can bring such features to Itanic atL >all, it won't be before 2005 (for such peripheral on-chip glue) or 2006 (atL >the earliest, for things like OOO and SMT that would help Itanic be less ofD >a server pig - though it will still retain significant unproductiveL >overheads compared with POWER4 unless they scrap EPIC pretty much entirely,D >at which point the question will be just how effectively the ItanicL >instruction set can be mapped to an EV8-like architecture, sort of like theJ >x86 instruction set is mapped to an underlying RISC-like implementation - >blech). >cI >And unless Intel manages to drive AMD out of business, Itanic will nevere own F >the low end either, because a great deal of it can be handled by IA32L >processors and Hammer looks far more attractive than Itanic for the balance >that requires 64-bit support. >:J >So if Itanic succeeds at all, it won't be for technical reasons.  And I'mJ >guessing that there's at least a fair chance that Itanic will *never* pay( >for its own development (unlike Alpha). >t >>K >> I think with enough money, Alpha could have been performance competetiveoJ >> with Intel and IBM.  But it would have taken a lot of money to stay out inG >> front of both, and the delta between performance might not have beenz large I >> enough for someone to pick Alpha over Itanium - especially in the UNIXd	 >> space.i >aG >As noted above, Hammer and IA32 look like much tougher competition for  AlphaiJ >in the low end than Itanic does.  And POWER4 is certainly the competition inL >the high end:  I think EV7 might have had difficulty competing with POWER4,H >but that EV8 would have had slight advantages over it - and even takingF >second-place to POWER4 in that lucrative a market would have returned1 >profits to Compaq far beyond any it's ever seen.F > G >  I think not having control over the fab makes it tough to compete onr >> raw performance.e >gJ >IBM supposedly fabs for others as aggressively as it does for itself, nowG >(with its 'anything for a buck' attitude) more than ever.  While I cannI >understand why having someone else (especially a competitor) responsibled for*I >fabbing Alphas could make Alpha's owner nervous, IBM more than anyone isrJ >sensitive to anti-trust regulation - and of course there's always Samsung to >help keep them honest.? >fJ >Except for your unusually forthright statement about Itanic's inferiorityK >that I quoted initially above, you seem to be reverting to your defense ofeL >Compaq's to-all-appearances completely indefensible position on this topic.H >If you insist on doing so, could you at least try to present some *new*) >information supporting your contentions?  >t >- billa >h >o >t   ------------------------------  % Date: Tue, 16 Oct 2001 17:00:10 -0400.5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>d/ Subject: Re: Higher prices for Alpha processorst1 Message-ID: <pd1z7.533$RL6.4757@news.cpqcorp.net>m  K Seems to me that both you and Bill are wasted - *you* should be CEO's.  HowoG is it that your raw pearls of wisdom have gone unnoticed?  At least you 8 should be the head of Strategy for a Fortune 50 company.  E You know, I have a lot of complaints about a lot of decisions made by5I Digital, and by Compaq.  But *I* am not being paid to be the CEO, or CTO.oK And it's damn easy to look back in 20-20 and say "if only they had done X".tK I don't have all the information that the people making the decisions have.1  ; Alpha should have.  Compaq could have.  Digital might have.d  H VMS is never going to take over the world.  It had it's chance.  DigitalF blinked (a LONG time ago) and it's over.  But VMS *can* play a role inI specific areas, and specific markets.  You can buy into that or not.  ButtK aside from a few newsgroups, and pretty much the same handfull of people in G them - the rest of the world doesn't seem to see this as the end of theeK world.  I have been talking to *real* customers, spending *real* money, who J just want information on timing, and plans, but who firmly plan on stayingK with VMS.  And yes, we have people buying Alpha today, and porting to AlphalK today - knowing that down the road Itanium will be the platform.  The pointsJ is that there IS a future.  And it isn't dependent on Compaq being able to sustain an Alpha business.      = JF Mezei wrote in message <3BCC9311.DE8EDAF1@videotron.ca>...a >Fred Kleinsorge wrote:cG >> But how much faster does it have to be to "win" in the market place? L >> Clearly there have been times when we have had well over a 2x performance= >> lead - but that didn't translate into it eating SUNs lunche >g >Pricing and marketing.  >mJ >Compaq could have fixed this when it bought Digital because at that time,G >Compaq had the marketing saavy and the power to take the leap to lower  pricerG >to increase volume. But Compaq chose to ignore the potential of Alpha.r >nH >> But I contend that most people, aside from the zealots, do not really careB >> about the processor ISA.  They care about the price/performance >rJ >We saw the potential for Alpha. And we saw how Compaq not only squandered itswJ >potential, but eventually lied about its commitments and killed the chip. ThatI >is the important issue here, not whether Alpha could have had 1hz fasters# >performance than IA64 or the 8086.u >n6 >The problem is that VMS seems to be in the same boat. >- >- >- >-K >> We are porting to Itanium.  We will do our best to make it as painless ak >> transition as possible. >uJ >Woopty doo. I would have rathered see Compaq announce the death of Alpha, noL >port of VMS, but a big marketing budget for VMS who would then have 5 years to >prove its potential.y > J >Compaq continues to ignore VMS and that hasn't changed with the murder ofG >Alpha, even if there was a one liner "by the way, we'll also port thatu thing.2 >called VMS to IA64"  hidden in the announcements. >n6 >> People who need raw performance, and need it beforeE >> say 2004/2005, can safely buy EV6x and EV7x based systems with thes! >> appropriate price/performance.m > @ >Correct for folks who already have alpha or who need short termJ >infrastructure. But someone isn't about to invest in a platform that will9 >require a port to another platform in a couple of years.  >tI >> But it WILL be able to run Windows as well as VMS, something we can non >> longer say about Alpha. > E >Again, that is because Compaq wasn't interested in pushing Alpha, soeH >understandably MS and Compaq found a way to justify pulling the plug on Windows-Alpha. > I >> paid enough) to know *exactly* when, Itanium family systems will be ase big,1 >> bigger, fast, and faster than Alpha platforms.V >rK >Digital had it and wasted it. Alpha could have been the de-facto standard.t ForaL >Christ's sake, they  were *10 years*  ahead of Intel and they allowed theirI >own 8086 machines to eat their own Alpha. Digital cut its own limbs to a  pointtF >of extermination and donated its remains to some small PC company who ignoredhG >anything "Digital" and is now melting itself to fit into some HP bowl.M >  >oJ >The management incompetence that resulted in the wasting of Alpha and VMS has 2 >not died with Alpha's death. That is the problem. >kI >Why should anyone believe that VMS will be any more succesful on IA64 orr anyr
 >other chip ?rG >Why should anyone believe that VMS will be any more succesful under HPuG >management who have claimed that they would focus on industry standardr >solutions ?   ------------------------------  # Date: Tue, 16 Oct 2001 21:30:21 GMTa* From: "Bill Todd" <billtodd@metrocast.net>/ Subject: Re: Higher prices for Alpha processorsd? Message-ID: <NF1z7.513798$Lw3.31489676@news2.aus1.giganews.com>.  > Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote in message+ news:pd1z7.533$RL6.4757@news.cpqcorp.net...eH > Seems to me that both you and Bill are wasted - *you* should be CEO's. HowlI > is it that your raw pearls of wisdom have gone unnoticed?  At least you : > should be the head of Strategy for a Fortune 50 company.  I Actually, at least some of those pearls were noticed but discarded, sinceEJ they were presented to Capellas (and then redirected to Marcello) about 17 months ago.    >sG > You know, I have a lot of complaints about a lot of decisions made byeK > Digital, and by Compaq.  But *I* am not being paid to be the CEO, or CTO.tI > And it's damn easy to look back in 20-20 and say "if only they had donem X".n  K 17 months ago, we were looking forward, not back.  Ditto 2+ years ago, whennI the dual NT on Alpha/Win64 on Alpha fiascos occurred.  Ditto even earlieroG when I was asserting that Compaq had no interest in being a hardware or 5 software manufacturer, just in being a box-assembler.e  J Just because *you* may not have had a vision of a better path then doesn't mean that no one else did.  G > I don't have all the information that the people making the decisionsi have.a  D But we have more than enough to evaluate those decisions now, and inF retrospect seem to have had more than enough to have made better ones.   > = > Alpha should have.  Compaq could have.  Digital might have.l  K Yes, it's too bad, but it's water under the bridge.  The relevance *now* isfI that the same people responsible for this bungling are by and large stilliD steering the ship:  either they should be removed (and the decisionsK reevaluated where they still might be profitably reversed), or people (botheE customers and employees) should find better alternate transportation.s   - bill   ------------------------------  # Date: Tue, 16 Oct 2001 21:17:01 GMTe* From: "Bill Todd" <billtodd@metrocast.net>/ Subject: Re: Higher prices for Alpha processors C Message-ID: <gt1z7.775421$NK1.70078158@bin3.nnrp.aus1.giganews.com>h  > Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote in message+ news:zV0z7.530$RL6.4732@news.cpqcorp.net...w >h0 > Please. please. please.  Do not respond to me.  F Sorry:  when I see fuzzy logic (or worse) presented as fact by someoneD others might be inclined to pay attention to, I correct it.  Present. something sensible, and I'll applaud suitably.     I made a mistake andH > responded to your assertions.  You want to parse things into nice neat > little bundles.E  I Funny about that - perhaps you prefer nebulous implications and feel-goodsK reassurances.  Me, I like to think my opinions have some basis in fact, and 9 try to uncover such facts as will prove or disprove them.   ;   I'm a liar or an idiot, or both.  Not a  __ing thing willl > alter that in your universe.  F While my respect for your acumen has indeed plummeted of late, my realE suspicion is that you're just not very acute analytically rather thanw actively dishonest or stupid.   )   This isn't a court of law, or a ColumbotB > episode - and the truth (and opinion) is seldom self consistant.  K But when such inconsistencies are pointed out, intellectually honest people 3 attempt to reconcile them rather than just get mad.w  
   You wantI > to postulate performance characteristics, time to market and cost - allsJ > fabricated out of pretty much hearsay - good *and* bad.  Spin on.  Enjoy > yourself.n  K I try to enjoy myself, and often succeed.  But the spinning originated withrG Compaq (and its apologists), and my efforts have been to inject as mucht0 reality into the discussion as can be uncovered.   >aK > I'll stick to these loose assertions:  Alpha was great.  Loved it.  Had amJ > great time.  A decision I had no part in making was that it was not costH > effective to invest in the Alpha ISA long term (at this point I really don'thG > care if the reason is valid or not, or if *you* agre with it), and toi switchJ > to the Itanium architecture - for good or ill.  People making a lot moreL > money than I do believe that Itanium will be competetive, and we have EV6xE > and EV7x until it is.  I myself will hedge on the exact date.   TheoJ > performance of the mythical EV8 or EV9 are not interesting, as they will nott > be built.e  H Probably, though the option to build them remains open.  Should Compaq'sK leadership be replaced (as seems increasingly likely, especially should the K merger fall through) the decision could be reevaluated (in desperation, butm: that's only appropriate given the magnitude of the error).  J And even if they're not built, their potential remains interesting insofarF as it reflects upon the incompetence of Compaq's current management inE making the decision to abort them.  While EV9 may merit the adjectivedG 'mythical', EV8's design was sufficiently completed for its performancegJ potential to be fairly well understood - and by all indications from thoseI directly responsible for it that design was well worth bringing to marketaE ('hearsay', perhaps, but from the people best-qualified to venture anh opinion on the matter).n  =   We will be on an industry standard architecture, offered byn  > everyone except (perhaps) SUN.  H Or you may simply be out of business:  capitalism has a way of punishing gross incompetence.a  $   "Hammer" will never be embraced by" > Microsoft, and will quietly die.  G Care to provide substantiation for that insight into Microsoft's future,G actions?  Since Hammer is so much more attractive a platform for 64-bit I low-end and mid-range servers than Itanic is, and since Linux is activelyfH extending its influence into that space (and already has code running inH simulation on Hammer), it's not clear that Microsoft would be willing toL relinquish that advantage to Linux rather than compete on the same platform.  +   IA32 will be phased out by the end of theeJ > first decade of the 2000's, as both Intel and Microsoft "churn" the user# > base to force PCs to be replaced.   I That could happen *if* AMD goes away, but see above.  If Hammer succeeds,eH IA32 per se will fade away (just as IA16 did), but not to the benefit of Itanic.v  I And a great deal that's hard to anticipate can happen between now and thenH end of the decade, which is after all just about the same amount of time, that's elapsed since Alpha first appeared...     Linux will follow (or lead).  L Not sure how to parse the above.  If you're saying Linux will also be phased out, I suspect you're wrong.  
   Most of oursH > customers will be happy with the VMS to Itanium transition, which will give% > them compatetively priced hardware.   K By definition, most of your *remaining* customers will be.  The question isa how many of them there'll be.s  )   A minority will make a decision to movesJ > onto another vendors OS.  We will do everything in our power to make the > transition long, and smooth.  E I'm sure you will.  But the good intentions of VMS development aren't @ sufficient to erase the neglect and incompetence of VMS's owner.   - bill   ------------------------------    Date: 16 Oct 2001 21:00:51 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)a/ Subject: Re: Higher prices for Alpha processors-3 Message-ID: <VV4McVy4xFkM@eisner.encompasserve.org>l  g In article <d7791aa1.0110161239.dcc3a0c@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes:t  8 > alpha kills itanic now and would of later with ev8 ...   That is one theory.   & > why do you think intel bought it ...   Marketing perception.u  5 > the puzzling question is why compaq was dumb enoughd   Financial reality.   >  to give it away?y  C The fact that the price was not disclosed does not mean it was low.v   ------------------------------  # Date: Wed, 17 Oct 2001 03:15:08 GMTl3 From: Tim Llewellyn <tim.llewellyn@cableinet.co.uk> / Subject: Re: Higher prices for Alpha processorsn/ Message-ID: <3BCCF730.77660535@cableinet.co.uk>R   Bob Ceculski wrote:q    > H > nt will never be on the level of vms ... are you forgetting why peopleG > are still on vms?  it's security and reliability!  clustering is justpH > another plus!  vms will be around thru 2015 or longer and the militaryG > will be the primary users along with us smart IT managers like myselfgB > that aren't idiots ... nt catch vms, what have you been smoking?  G though sometimes JF's cynicism gets me down to, Bob, I think you really*H should consider how lucky you are to be in a position where you have theE budget and power to maintain such a policy. Most of the IT world doesn not,G and neither Compaq or Digital before them were ever willing to make VMSi5 competitive with the alternatives at the entry level.l  % I do find your enthusiasm refreshing.,   regards    -- O Tim.Llewellyn@cableinet.co.uk  e  C Standard disclaimer applies. My views in no way represent those of d! my employers or service provider.W   ------------------------------   Date: 17 Oct 2001 04:37:51 GMT0 From: Dave Cherkus <cherkus777@777unimaster.com>/ Subject: Re: Higher prices for Alpha processorsr1 Message-ID: <Xns913D5C56416Eidtoken@199.125.85.9>i  8 "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote in* news:ZJZy7.519$RL6.4554@news.cpqcorp.net: F > But how much faster does it have to be to "win" in the market place?? > Clearly there have been times when we have had well over a 2xmH > performance lead - but that didn't translate into it eating SUNs lunch4 > - who have a pathetic turd-on-a-chip called Sparc.  < Remember when DEC had a pathetic turd-on-a-chip called VAX, 3 and was the number two computer maker in the world?   : As you hint above, it's not about being fast.  Being fast ; (mainly at crunching numbers, not doing IO) is about one ofv8 the few things DECpaq was able to do.  Creating markets,9 satisfying customer needs, getting product into the handse9 of customers, making profits, staying focused, etc. were o; things DECpaq didn't do much of for the last decade or so, d7 and the rest is history.  I can feel the huge amount ofi7 bitterness being radiated around here, but honestly, it 7 wasn't hard to see the current situation coming for at n8 least the last five years or so.  Thankfully, some folks9 seem to be able to accept it for what it is, and still beh; happy and move on with life, with or without their favoriteR system software.   Dave   ------------------------------   Date: 16 Oct 2001 17:56:54 GMT) From: leslie@clio.rice.edu (Jerry Leslie)v5 Subject: Re: HWP/CPQ: (...internal morale is poor...)m' Message-ID: <9qhsd6$3al$1@joe.rice.edu>r  / Fabio Cardoso (fabiopenvms@yahoo.com.br) wrote:h
 : Click at :fA : http://news.cnet.com/news/0-1003-201-7542492-0.html?tag=3Dcd_mh  :i/ : To check the recommendations of a HWP and CPQi" : shareholde to stop the merger... :A From:o  4    http://news.cnet.com/news/0-1003-200-7540457.htmlA    Shareholder denounces HP-Compaq union -  Tech News -  CNET.coml  F    "We ask our contacts at HP whether they can identify any successful3    (large) tech merger, and they can't," Katz said.a   --Jerry Leslie   ------------------------------  % Date: Tue, 16 Oct 2001 20:18:07 -0400n. From: Chuck McCrobie <mccrobie@cablespeed.com>9 Subject: Re: Ingres (Was: Re: jobs? ingres vax/vms cobol)*. Message-ID: <3BCCCE3F.C5E0E72B@cablespeed.com>  H I've used PostgreSQL on FreeBSD - it seemed stable and rich enough for a; Java JDBC development project I was working on at the time.t  D I've heard rumors that it might be ported to OpenVMS, but who knows?   Kevin Handy wrote: >  > John Eisenschmidt wrote: > >m3 > > Ahh, another product who's sad story ends with:  > >.< > > "And then Computer Associates bought them <DOT DOT DOT>" > >A@ > > I have heard of Ingres but never had the pleasure to use it. > >  > A > Ingres started out as an academic exercise.  The same group who5@ > did the original Ingres (source is available in various places@ > on the net for that version) later went on to create Postgres.; > Postgres was later modified to have an SQL interface, and A > renamed to Postgres95 (or some such number).  It was then laterq? > renamed to PostgreSQL, which is currently being developed ando > used.o > A > > Ingres was a major database company back in the 80s and earlyoG > > 90s.  If I remember correctly Ingres was originally developed in an @ > > academic environment and then taken commerical by RelationalC > > Technology in Alameda (Oakland) California.  In its early yearst> > > VMS was it's main development platform.  After significant@ > > market success the company's name was changed to Ingres.  InA > > the early 90s it fell into financial trouble and was acquirednF > > by ASK Computer Systems.  A few years later ASK got into financialA > > problems and was acquired by Computer Associates.  I think CAt> > > just milked the product and didn't do much marketing.  The  > > last version I used was 6.4.   -- t --n   ------------------------------  % Date: Tue, 16 Oct 2001 19:44:33 +0100r3 From: "Malcolm" <malcolm@neverness.freeserve.co.uk> P Subject: Linux distribution incompatibility (was Re: A free VMS implementation?)/ Message-ID: <9qhvqb$4lr$1@newsg3.svr.pol.co.uk>n  7 "Tony Scandora" <scandora@cmt.anl.gov> wrote in message % news:9qfg95$tte$1@milo.mcs.anl.gov...m >C [...]oI > Even with the Linux kernel under the tight control of Mr. Torvalds, oneaC > cannot expect system management tools or application installationr
 procedures4 > developed for one distribution to work on another.   That's to be expected.  @ The kernel is the only part of Linux which is central to all theC distributions and is maintained by Linus. It only makes a few basicrG assumptions about the file system, such as there is a root directory, arK /etc, and a /etc/init (or is it /sbin/init.d?). Then it runs the 'init' ande/ the rest depends on what you set up init to do.hD 'init' and all programs onwards can vary wildly from distribution to distribution...p  D The major problem is with file paths, which can vary wildly from one distribution to another.  I No real standards in the Linux world. It's as if sites just built VMS the 8 way they wanted, from SYS$SYSTEM:STARTUP.COM upwards ;-(  I [Not that I wouldn't mind *if* there were some standards for this sort ofu thing in the Linux world]p  8 Sorry to go off topic, but I thought I should clarify...  	 -Malcolm. A [part-time Linux admin (at home), full-time VMS systems admin :-]b   ------------------------------  % Date: Tue, 16 Oct 2001 22:35:25 -0500hC From: "Craig A. Berry" <craig.berry@nospam.SignalTreeSolutions.com> + Subject: Mozilla 0.9.5 and file protections I Message-ID: <craig.berry-F8CD41.22352416102001@newsrump.sjc.telocity.net>C  C The newest version of Mozilla still has problems if you install it t> under the SYSTEM account and then try to run it from another, G non-privileged account.  It gives the "Starting..." message, chugs for tF a bit, and then just quits with no error messages.  By using security 6 auditing, I determined that the following files under D SYS$COMMON:[MOZILLA] are installed with protections other than what 
 they need:  (                            has     needs( component.reg              W:RE    W:RWE& [.components]xpti.dat      W:      W:R( [.components]xptitemp.dat  W:RE    W:RWE  B Once I change the protections, Mozilla runs fine.  I suspect most H people don't encounter this because they run from the same account they E install from and thus own the files, but if this doesn't get fixed I  H bet it will generate a lot of support calls when Mozilla 1.0 is finally D released.  In any case, it would be nice to get a NOPRIV error when : Mozilla can't even start due to a file protection problem.  H Quite possibly write access is not really needed for one or more of the G files that currently requires it, so it may be that the right thing to .B do here would involve a code change as well as (or instead of) an  installer change.    ------------------------------  % Date: Tue, 16 Oct 2001 15:16:43 -0400f- From: "John Eisenschmidt" <jeisensc@aaas.org> " Subject: Quotas for Oracle 8i User+ Message-ID: <sbcc4f6f.093@AAASMTA.aaas.org>-  I So far, the performance difference between Oracle 7.3.4 on an AS800 and =rK Oracle 8i on a DS10 has been pleasing, but I think the DS10 can do better =1J (particularly in the disk I/O department). The max disk i/o I've seen on =G the monitor is ~80 on the DS10, but I've seen the AS800 ~400 before.=20o  K My suspicion is that the quotas for the 8i user are too low. Am I way off =E base? Any suggestions?  = Username: ORACLE8                          Owner:  ORACLE DBA H Account:                                   UIC:    [200,100] ([ORACLE8])  9 Maxjobs:         0  Fillm:       100  Bytlm:       256000 9 Maxacctjobs:     0  Shrfillm:      0  Pbytlm:           0c9 Maxdetach:       0  BIOlm:       150  JTquota:       8192t9 Prclm:           8  DIOlm:       150  WSdef:         4000,9 Prio:            4  ASTlm:       250  WSquo:         8000r9 Queprio:         4  TQElm:        10  WSextent:     65536"9 CPU:        (none)  Enqlm:      8192  Pgflquo:     250000i Authorized Privileges:C CMKRNL       IMPERSONATE  LOG_IO       NETMBX       PFNMAP      =20 A PRMGBL     PRMMBX       SYSGBL       SYSLCK       SYSNAM      =20d SYSPRV       TMPMBX     WORLDe    6 Username: ORACLE7                Owner:  ORACLE 7 USERH Account:  ORACLE7                  UIC:    [200,202] ([AMGUI71T,ORACLE7]9 Maxjobs:         0  Fillm:       100  Bytlm:       256000i9 Maxacctjobs:     0  Shrfillm:      0  Pbytlm:           0 9 Maxdetach:       0  BIOlm:       150  JTquota:       4096w9 Prclm:           8  DIOlm:       150  WSdef:         4000a9 Prio:            4  ASTlm:       250  WSquo:         8000u9 Queprio:         4  TQElm:        10  WSextent:     65536'9 CPU:        (none)  Enqlm:      8192  Pgflquo:     256000e Authorized Privileges:C CMKRNL       GROUP        GRPNAM       GRPPRV       IMPERSONATE =20rK IMPORT    LOG_IO       MOUNT        NETMBX       OPER         PFNMAP      =t =20r? PRMGBL   PRMMBX       READALL      SETPRV       SYSGBL      =20oL SYSLCK       SYSNAM      SYSPRV       TMPMBX       WORLD                   =
        =20   Thanks,1 John                   =20   ------------------------------   Date: 16 Oct 2001 21:17:39 GMT' From: "Jim Strehlow" <jims@data911.com>7& Subject: Re: Quotas for Oracle 8i User0 Message-ID: <9qi85j$lea@dispatch.concentric.net>  ) We encountered OpenVMS user authorization61 size problems when going from Oracle 8.0.5 to 8i.l  " We use the following for Oracle8i:  + ADD ORACLE8 /UIC=[101,1] /DEVICE= DISKORA -s-  /DIRECTORY= [ORACLE8i] /OWNER="Oracle DBA" -a  /MAXJOBS= 0 /MAXACCTJOBS= 0 -(  /MAXDETACH= 0 /PRCLM= 10 /PRIORITY= 4 -3  /FILLM= 500 /SHRFILLM= 0 /BIOLM= 250 /DIOLM= 250 -o)  /ASTLM= 3000 /TQELM= 3000 /ENQLM= 3000 - +  /BYTLM= 151000 /PBYTLM= 0 /JTQUOTA= 8192 -b"  /WSDEFAULT= 2048 /WSQUOTA= 4096 -&  /WSEXTENT= 16384 /PGFLQUOTA= 250000 -I  /PRIVILEGES=(    CMKRNL, DETACH, LOG_IO, NETMBX, OPER, PFNMAP, PRMGBL, -oJ                   PRMMBX, SYSGBL, SYSLCK, SYSNAM, SYSPRV, TMPMBX, WORLD) -I  /DEFPRIVILEGES=( CMKRNL, DETACH, LOG_IO, NETMBX, OPER, PFNMAP, PRMGBL, - J                   PRMMBX, SYSGBL, SYSLCK, SYSNAM, SYSPRV, TMPMBX, WORLD) -,  /FLAGS=( DisPwdDic, DisPwdHis, NODisUser) -6  /PWDMINIMUM= 8 /PWDLIFETIME= 0 /NOPWDEXP /PASSWORD= X    ADD   /IDENTIFIER  ORA_DBA &  GRANT /IDENTIFIER  ORA_DBA    ORACLE8  3 Jim Strehlow, Systems Manager, Assistant Oracle DBA  Data911, Alameda, CA  / "Let them do their worst. We will do our best."o    8 "John Eisenschmidt" <jeisensc@aaas.org> wrote in message% news:sbcc4f6f.093@AAASMTA.aaas.org... G So far, the performance difference between Oracle 7.3.4 on an AS800 and I Oracle 8i on a DS10 has been pleasing, but I think the DS10 can do betterdL (particularly in the disk I/O department). The max disk i/o I've seen on the@ monitor is ~80 on the DS10, but I've seen the AS800 ~400 before.  I My suspicion is that the quotas for the 8i user are too low. Am I way off  base? Any suggestions?  = Username: ORACLE8                          Owner:  ORACLE DBAaH Account:                                   UIC:    [200,100] ([ORACLE8])  9 Maxjobs:         0  Fillm:       100  Bytlm:       256000 9 Maxacctjobs:     0  Shrfillm:      0  Pbytlm:           0e9 Maxdetach:       0  BIOlm:       150  JTquota:       8192o9 Prclm:           8  DIOlm:       150  WSdef:         4000e9 Prio:            4  ASTlm:       250  WSquo:         8000 9 Queprio:         4  TQElm:        10  WSextent:     65536m9 CPU:        (none)  Enqlm:      8192  Pgflquo:     250000  Authorized Privileges:: CMKRNL       IMPERSONATE  LOG_IO       NETMBX       PFNMAP8 PRMGBL     PRMMBX       SYSGBL       SYSLCK       SYSNAM SYSPRV       TMPMBX     WORLDr    6 Username: ORACLE7                Owner:  ORACLE 7 USERH Account:  ORACLE7                  UIC:    [200,202] ([AMGUI71T,ORACLE7]9 Maxjobs:         0  Fillm:       100  Bytlm:       256000r9 Maxacctjobs:     0  Shrfillm:      0  Pbytlm:           0p9 Maxdetach:       0  BIOlm:       150  JTquota:       4096e9 Prclm:           8  DIOlm:       150  WSdef:         4000e9 Prio:            4  ASTlm:       250  WSquo:         8000t9 Queprio:         4  TQElm:        10  WSextent:     65536r9 CPU:        (none)  Enqlm:      8192  Pgflquo:     256000w Authorized Privileges:? CMKRNL       GROUP        GRPNAM       GRPPRV       IMPERSONATEuD IMPORT    LOG_IO       MOUNT        NETMBX       OPER         PFNMAP6 PRMGBL   PRMMBX       READALL      SETPRV       SYSGBL8 SYSLCK       SYSNAM      SYSPRV       TMPMBX       WORLD   Thanks,  John   ------------------------------  % Date: Tue, 16 Oct 2001 13:16:15 +1000r/ From: "Phil Howell" <phowell@snowyhydro.com.au>u@ Subject: Re: sd comman (or who has the biggest cd.com or sd.com)0 Message-ID: <LR7z7.228$Qj7.15976@ozemail.com.au>  4 <SWARS@mailer.MPI-STUTTGART.MPG.DE> wrote in message6 news:01K9J0870W4Y8WXSPT@MAILER.MPI-STUTTGART.MPG.DE... > hallo all,< > we have a customer how needs the sd command and I have not< > such a dcl script for changing the directory is there some
 > around ?	 > h.swarsr  ( this one claims to be the "ultimate" one (I don't use it myself though) Phil   _TNA62:typ cd.come $! CD.COM v6.09s) $! The Ultimate Change Directory Command.  $!D $  hdir     = f$trnlnm("SYS$LOGIN")                 ! Home DirectoryD $  ndir     = f$edit(p1,"UPCASE")                   ! New  DirectoryD $  odir     = f$environment("DEFAULT")              ! Old  DirectoryB $  prompton = (f$edit(f$trnlnm("SYS$PROMPT"),"UPCASE") .eqs. "ON") $!< $  if (ndir .eqs. "")           then goto DISPLAY   ! No DirE $  if (ndir .eqs. "*")          then goto DIRSEARCH ! Search for Dirs B $  if (ndir .eqs. "?")          then goto HELP      ! Instructions $!	 $  PARSE: A $  length   = f$length(ndir)                        ! Fix up ndir * $  if (f$location("@",ndir) .eq. 0) .or. -H       (f$location("$",ndir) .eq. 0) then ndir = f$extract(1, length - 1, ndir)m& $  right    = f$location("]",ndir) + 1< $  if (right .gt. length) then right = f$location(">", ndir)@ $  if (right .le. length) then ndir  = f$extract(0, right, ndir) $!F $  if (f$trnlnm(ndir) .eqs. "") then goto CASESYM   ! Not Logical NameB $     ndir   = f$trnlnm(ndir)                       ! Logical Name $     goto PARSE $! $  CASESYM:e@ $  if ("''&ndir'" .eqs. "")     then goto CASE0     ! Not Symbol< $     ndir = 'ndir'                                 ! Symbol $     goto PARSE $!	 $  CASE0:lA $  len_ndir = f$length(ndir)                        ! Regular Dir 2 $  if (f$location("[", ndir) .lt. len_ndir) .or. -<       (f$location("<", ndir) .lt. len_ndir) then goto SETDIR $!> $  CASE1:                                           ! Home DirB $  if ((ndir .nes. "HOME") .and. (ndir .nes. "\")) then goto CASE2 $     ndir = hdirt $     goto SETDIRe $!? $  CASE2:                                           ! . .. .dir 5 $  if (f$location(".", ndir) .nes. 0) then goto CASE3t* $     if (ndir .eqs. "..") then ndir = "-"- $     if (f$extract(0, 2, ndir) .eqs. "..") --;          then ndir = "-" + f$extract(1, len_ndir - 1, ndir)y $     ndir = "[" + ndir + "]"k, $     if (ndir .eqs. "[.]") then ndir = odir $     goto SETDIR  $!7 $  CASE3:                                           ! : ; $  if (f$location(":", ndir) .ge. len_ndir) then goto CASE4r) $     left    = f$location(":", ndir) + 1'( $     symbol  = f$extract(left, 1, ndir)= $     if (symbol .eqs. ":")  then goto CASE3B       ! :: Node F $     if ((symbol .eqs. "[") .or. (symbol .eqs. "<")) then goto SETDIR0 $        ndir = f$extract(0, left, ndir) + "[" ->               + f$extract(left, len_ndir - left+1, ndir) + "]" $     goto SETDIR  $!C $  CASE3B:                                          ! NODE::nothinga3 $  if (f$length(ndir)-1 .gt. left) then goto CASE3Cn $     ndir = ndir + "[000000]" $     goto SETDIR, $!E $  CASE3C:                                          ! NODE::directory @ $  if ((f$location("[", ndir) - f$location("<", ndir)) .ne. 0) -       then goto SETDIR $gH $     ndir = f$parse(ndir,,,"NODE") + "[" + f$parse(ndir,,,"NAME") + "]" $     goto SETDIRi $!9 $  CASE4:                                           ! dirv $  ndir = "[" + ndir + "]" $!
 $  SETDIR: $  set default 'ndir':/ $  if (f$parse("") .eqs. "") then goto DIRERRORe $! $  DISPLAY:R: $  if ((ndir .nes. "") .and. prompton) then goto NODISPLAY" $     hnode = f$getsyi("NODENAME"); $     cnode = f$parse(f$trnlnm("SYS$DISK"),,,"NODE") - "::"Z, $     if (cnode .eqs. "") then cnode = hnode& $     cdir  = f$environment("DEFAULT") $     write sys$output " "5 $     write sys$output "          Home Node: ", hnodea4 $     write sys$output "     Home Directory: ", hdirG $     if (cdir .eqs. hdir) .and. (cnode .eqs. hnode) then goto DISPSKIPp5 $     write sys$output "       Current Node: ", cnode 4 $     write sys$output "  Current Directory: ", cdir $  DISPSKIP: $     write sys$output " " $!
 $  NODISPLAY: " $  ndir = f$environment("DEFAULT")" $  if .not. prompton then goto END $!0 $  if (f$length(ndir) .ge. 32) then goto TOOLONG $!
 $  SETPROMPT:c $  set prompt = 'ndir'" "u $! $  END:a $  exitc $! $  DIRERROR: $  write sys$output " "eD $  write sys$output "          ", ndir, " Directory does not exist!" $  write sys$output " "r $  set default 'odir'  $  ndir = odir $  goto NODISPLAYe $!	 $! PromptbD Problems------------------------------------------------------------ $! $  TOOLONG:aK $! Prompt is too long. Get rid of everything to the left of [ or <. If thatrH $! doesn't work, get rid of a subdirectory at a time.  As a last resort, $! set the prompt back to $. $!# $  left     = f$location("[", ndir)- $  len_ndir = f$length(ndir); $  if (left .ge. len_ndir) then left = f$location("<",ndir)R0 $  if (left .gt. 0) .and. (left .lt. len_ndir) -8       then ndir = f$extract(left, len_ndir - left, ndir) $! $  STILLTOOLONG:4 $    if (f$length(ndir) .lt. 32) then goto SETPROMPT) $    left     = f$location(".", ndir) + 1  $    len_ndir = f$length(ndir)+ $    if left .ge. len_ndir then ndir = "$ "r $    if left .ne. len_ndir -A         then ndir = "[*" + f$extract(left, len_ndir - left, ndir)c $    goto STILLTOOLONG $! $! WildcardoB Directory--------------------------------------------------------- $!
 $  DIRSEARCH:)+ $  error_message = f$environment("MESSAGE")   $  on control_y then goto DIREND  $  on control_c then goto DIREND& $  set message/nosev/nofac/noid/notext $  write sys$output " ".
 $  dispct = 1R
 $  dirct  = 0  $  pauseflag = 1 $! $  DIRLOOP:e! $    userfile = f$search("*.dir")pB $    if (userfile .eqs. "") .and. (dirct .ne. 0) then goto DIRMENU- $    if (userfile .eqs. "") then goto DIRNONE  $    dispct = dispct + 1 $    dirct  = dirct  + 1* $    on severe then $ userprot = "No Priv"1 $    userprot = f$file_attributes(userfile,"PRO")d4 $    if userprot .nes. "No Priv" then userprot = " "> $    userfile'dirct' = "[." + f$parse(userfile,,,"NAME") + "]" $    userprot'dirct' = userprota5 $    lengthflag = (f$length(userfile'dirct') .gt. 18).* $    if lengthflag then write sys$output -J         f$fao("  !3SL   !34AS  ", dirct, userfile'dirct'), userprot'dirct'2 $    if (.not. lengthflag) then write sys$output -J         f$fao("  !3SL   !20AS  ", dirct, userfile'dirct'), userprot'dirct') $    if (dispct .lt. 8) then goto DIRLOOPg $    dirct  = dirct  + 1 $    userfile'dirct' = ""e $    dirct  = dirct  + 1 $    userfile'dirct' = ""e# $    if pauseflag then goto DIRMENUa $    dispct = 0o $    goto DIRLOOP0 $! $  DIRMENU:  $  write sys$output " "c, $  if (userfile .eqs. "") then goto DIRMENU24 $     write sys$output "    M   More subdirectories" $  if pauseflag then -= $     write sys$output "    N   More subdirectories/No pause"  $! $  DIRMENU2:: $     write sys$output "    R   Re-Display subdirectories"/ $     write sys$output "    Q   Quit (default)"h $a $  DIRINQUIRE: $  write sys$output " "d# $  inquire dirchoice "  Select One"y $  write sys$output " "  $!# $  if (dirchoice .gt. 0)    .and. -w3       (dirchoice .le. dirct) then goto DIRCASEDIGIT ) $  dirchoice = f$edit(dirchoice,"UPCASE")m" $  if (dirchoice .eqs. "")  .or. -3       (dirchoice .eqs. "Q")  then goto DIRCASEBLANK-" $  if (dirchoice .eqs. "M") .or. -2       (dirchoice .eqs. "N")  then goto DIRCASEMORE1 $  if (dirchoice .eqs. "R")  then goto DIRCASEREDt $! $  DIRCASERROR:e. $  if (dirct .eq. 1)   then write sys$output -B       "  Select 1 to change to the ", userfile1, " subdirectory. " $  revdirct = dircts2 $  if (dispct .eq. 8) then revdirct = revdirct - 2. $  if (dirct .gt. 1)   then write sys$output -C       "  Valid subdirectory selections are 1 through ", revdirct, "p	 (Octal)."s $  goto DIRINQUIRE $! $  DIRCASEDIGIT:: $  if (userfile'dirchoice' .eqs. "") then goto DIRCASERROR $  ndir = userfile'dirchoice'  $  goto DIREND $! $  DIRCASEBLANK:1 $  write sys$output "  Subdirectory not changed."a $  write sys$output " "i $  goto DIREND $! $  DIRCASEMORE:e
 $  dispct = 0r. $  if (dirchoice .eqs. "N") then pauseflag = 0- $  if (userfile .nes. "")   then goto DIRLOOPe: $  write sys$output "  No more subdirectories to display." $  goto DIRINQUIRE $! $  DIRCASERED:
 $  dispct = 1i $  DISPLOOP:6 $     if (userfile'dispct' .eqs "") then goto DISPDONT7 $     lengthflag = (f$length(userfile'dispct') .gt. 18)n+ $     if lengthflag then write sys$output -s=          f$fao("  !3SL   !34AS  ", dispct, userfile'dispct'),u userprot'dispct'3 $     if (.not. lengthflag) then write sys$output -f=          f$fao("  !3SL   !20AS  ", dispct, userfile'dispct'),  userprot'dispct' $     DISPDONT:f $     dispct = dispct + 1i/ $     if (dispct .le. dirct) then goto DISPLOOPd $  goto DIRMENU- $! $  DIRNONE:-A $  write sys$output "No subdirectories to choose, or no directorye privileges." $  write sys$output " "t $  goto DIREND $!
 $  DIREND: $  set message 'error_message' $  on control_y then exit2 $  on control_c then exite( $  if (ndir .eqs. "*") then goto DISPLAY
 $  goto PARSEr $!L $!-Help--------------------------------------------------------------------- -- $! $  HELP: $  type sys$input   >                CD.COM  Version 6  VMS Change Directory Command  5                          Usage:  CD command/directoryi  L CD         Display home directory,       CD ..       Change directory to theK            current directory, node.      CD [-]      dir above current dir.   J CD \       Change directory to your      CD ..sub    Change directory to a? CD HOME    SYS$LOGIN directory.          CD [-.sub]  "sideways" 
 subdirectory.h  G CD dir     Change directory to the       CD *        Display/select thet> CD [dir]   [dir] directory.                          available subdirectories.i  B CD .sub    Change directory to the       CD .        Reset current
 directory.? CD [.sub]  [.sub] subdirectory.          CD ?        Display CDb
 instructions.   D      CD :== @SYS$LOGIN:CD.COM                 DEFINE SYS$PROMPT "ON"F      To make CD available from                To have the VMS $ promptL      any directory you change to.             display the current directory.  +                               By The Mentorw $  goto ENDs   ------------------------------  % Date: Tue, 16 Oct 2001 14:12:25 -0400 * From: WILLIAM WEBB <WWEBB1@email.usps.gov>; Subject: RE: Slightly OT - how failed computers kill ships.c- Message-ID: <0033000038507894000002L042*@MHS>a  ; =0AGives a whole new meaning to "kernel panic", doesn't it..   WWWebb   > -----Original Message-----1 > From: Info-VAX-Request@Mvb.Saic.Com at INTERNETt* > Sent: Tuesday, October 16, 2001 10:51 AMD > To: Webb, William W Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNET9 > Subject: Slightly OT - how failed computers kill ships.i >h >t" > To:        Info-VAX@Mvb.Saic.Com > cc: : > Contact:   Tel: 3063  -  IS - Infrastructure, 1st Floor, > Bridge Street Plazan > 0 > Slightly OT - how failed computers kill ships. >  >U> > For those in the UK, or who have Channel 4 available via SKY > there is a program= > tonight at 9:30pm called 'Going Critical' which details thec > series of errors and? > failures [including the computer systems shutting down] which  > led to HMS: > Coventry [of which I am an ex-stoker] being sunk by old, > relatively slow planes" > with bombs during the Falklands. > 8 > It should be compulsory viewing for all those who make > decisions on using > computers to run ships.  >PE > If anyone's interested, you can get some information from this URL;  > C http://www.channel4.com/plus/going_critical/hms_coventry/index.html   H And if anyone is STILL interested, let me know and I'll compare what th=	 ey decide 7 caused the problems with what we were told at the time.j   Cheers   Steve S:     [Information] -- PostMaster:D This transmission is intended solely for the addressee(s) and may beH confidential. If you are not the named addressee, or if the message has=  been H addressed to you in error, you must not read, disclose, reproduce, dist=	 ribute oru use this transmission.  H Delivery of this message to any person other than the named addressee i= s not:H intended in any way to waive confidentiality.  If you have received thi= sGF transmission in error please contact the sender or delete the message.  
 Thank you.  H Yell Limited, Queens Walk, Oxford Road, Reading, Berkshire, RG1 7PT. Re= gistered0 in England and Wales, registered number 4205228.  H Yellow Pages Sales Limited, Queens Walk, Oxford Road, Reading, Berkshir= e, RG1A 7PT. Registered in England and Wales, registered number 1403041.=G   ------------------------------    Date: 16 Oct 2001 13:41:28 -0700( From: bob@instantwhip.com (Bob Ceculski); Subject: Re: Slightly OT - how failed computers kill ships.z= Message-ID: <d7791aa1.0110161241.78d7f12f@posting.google.com>e  W Steve.Spires@yellgroup.com wrote in message news:<00256AE7.0055AFE9.00@quegw01.btyp>...R" > To:        Info-VAX@Mvb.Saic.Com > cc:JN > Contact:   Tel: 3063  -  IS - Infrastructure, 1st Floor, Bridge Street Plaza > 0 > Slightly OT - how failed computers kill ships. >  > Q > For those in the UK, or who have Channel 4 available via SKY there is a program2R > tonight at 9:30pm called 'Going Critical' which details the series of errors andJ > failures [including the computer systems shutting down] which led to HMSQ > Coventry [of which I am an ex-stoker] being sunk by old, relatively slow planesA" > with bombs during the Falklands. > K > It should be compulsory viewing for all those who make decisions on using) > computers to run ships.s > E > If anyone's interested, you can get some information from this URL;X > E > http://www.channel4.com/plus/going_critical/hms_coventry/index.html  > R > And if anyone is STILL interested, let me know and I'll compare what they decide9 > caused the problems with what we were told at the time.i >  > Cheers > 	 > Steve S  >  >  > [Information] -- PostMaster:F > This transmission is intended solely for the addressee(s) and may beN > confidential. If you are not the named addressee, or if the message has beenR > addressed to you in error, you must not read, disclose, reproduce, distribute or > use this transmission. > N > Delivery of this message to any person other than the named addressee is notJ > intended in any way to waive confidentiality.  If you have received thisH > transmission in error please contact the sender or delete the message. >  > Thank you. > Q > Yell Limited, Queens Walk, Oxford Road, Reading, Berkshire, RG1 7PT. Registered22 > in England and Wales, registered number 4205228. > O > Yellow Pages Sales Limited, Queens Walk, Oxford Road, Reading, Berkshire, RG1 B > 7PT. Registered in England and Wales, registered number 1403041.  K hopefully someone from the u.s. navy dept. watches it and begins to realize < how dumb it was to pick nt over vms for our ship operations!   ------------------------------  # Date: Tue, 16 Oct 2001 21:17:42 GMT_+ From: "Andy Proctor" <aproctor@hotmail.com>B; Subject: Re: Slightly OT - how failed computers kill ships. 5 Message-ID: <1003266441.531675@ananke.eclipse.net.uk>   - <Steve.Spires@yellgroup.com> wrote in message ) news:00256AE7.0055AFE9.00@quegw01.btyp...j" > To:        Info-VAX@Mvb.Saic.Com > cc:  which led to HMSJ > Coventry [of which I am an ex-stoker] being sunk by old, relatively slow planes" > with bombs during the Falklands.  
 Hey Steve,H There are many more horror stories like this, maybe not so fatal, that I6 know of. Ex CWEA (actually a mech - not a tiffy) here!; (Invincible/'Collingrad'/Northwood and Beaver to name some)  Mail me if you like for a chat!r  8 aproctor@hotmail.com or work andy.proctor@spirentcom.com   Cheers   Andy   ------------------------------  # Date: Wed, 17 Oct 2001 03:34:00 GMTP3 From: Tim Llewellyn <tim.llewellyn@cableinet.co.uk>G; Subject: Re: Slightly OT - how failed computers kill ships. / Message-ID: <3BCCFB91.1E4856EB@cableinet.co.uk>-   WILLIAM WEBB wrote:0 > : > Gives a whole new meaning to "kernel panic", doesn't it. >  > WWWebb >   A you wouldn't want to run out of swapfile space in such situations7 either.7    6$ > > To:        Info-VAX@Mvb.Saic.Com > > cc:.< > > Contact:   Tel: 3063  -  IS - Infrastructure, 1st Floor, > > Bridge Street Plazal > >, > R > And if anyone is STILL interested, let me know and I'll compare what they decide9 > caused the problems with what we were told at the time.r >  go on, thenm    b   -- u Tim.Llewellyn@cableinet.co.uk  e  C Standard disclaimer applies. My views in no way represent those of t! my employers or service provider.    ------------------------------  # Date: Tue, 16 Oct 2001 18:17:03 GMTi2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: Stop a processn1 Message-ID: <zQ_y7.521$RL6.4355@news.cpqcorp.net>   Z In article <9qhp37$jq72@imsp212.netvigator.com>, "Kenneth" <chehon@netvigator.com> writes:L :If I cannot stop the process by STOP PROCE/ID=XXX, is there any other way IL :can stop the process, just like UNIX to use kill -9 to force the process to :die?   L   As STOP/ID=pid does what you want, I will guess that you have encountered J   a process stuck in a mutex (mutual exclusion) or miscellaneous resource L   wait (mwait) state.  Is this the case here?  Does the target process show J   up as MWAIT or RWAST or other scheduling state on a SHOW SYSTEM command?   	-->  L   Please do not specify a command from another operating system in isolationJ   as your desired goal, please let us know what you want to do and why youK   want to do it -- in this case, I've used kill -9, so I know what it does.RJ   But in the general case, folks might not be familiar with any particularI   command -- the more details and the more background you provide on your E   goal(s), the more likely you will get a solution to your problem.     L   When posting a question, please remember to include details -- the shorterJ   your question, the less likely that an the answer will address the topicH   that you intended.  Please remember to include the OpenVMS version andK   platform, the exact command used, error message(s) (if any) seen, and anyoM   other relevent information -- such as the details of the target process in  I   this particular case.  For additional details on the background details.-   often required, please see the OpenVMS FAQ.2  K   Put another way, if you ask a terse question, you should expect to get a dJ   terse answer.  You might not get an answer to the question you intended    to ask, too.    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Tue, 16 Oct 2001 20:19:12 -0000)/ From: Michael Zarlenga <zarlenga@conan.ids.net>r Subject: Re: Stop a process / Message-ID: <tsp5i0ht8rk0e2@corp.supernews.com>   & Kenneth <chehon@netvigator.com> wrote:M : If I cannot stop the process by STOP PROCE/ID=XXX, is there any other way I"M : can stop the process, just like UNIX to use kill -9 to force the process to  : die?  ( What's the message you get when you try?   -- n -- Mike Zarlenga   ------------------------------  # Date: Tue, 16 Oct 2001 23:29:15 GMT[) From: rob.buxton@wcc.govt.nz (Rob Buxton)o Subject: Re: Stop a process 2 Message-ID: <3bccc20c.1394962870@news.wcc.govt.nz>  E On Wed, 17 Oct 2001 01:05:26 +0800, "Kenneth" <chehon@netvigator.com>e wrote:  L >If I cannot stop the process by STOP PROCE/ID=XXX, is there any other way IL >can stop the process, just like UNIX to use kill -9 to force the process to >die?$ >l >"B The product AMDS or the newer Availability Manager (both availableD free from the Compaq Web site) have Process deletion mechanisms that* seem to work where the stop proc/id fails.   They're worth a look.r  F But, as others have noted there are times when nothing will get rid of the process short of a reboot.   Rob.   ------------------------------  % Date: Tue, 16 Oct 2001 22:21:37 +0200y2 From: martin@radiogaga.harz.de (Martin Vorlaender)" Subject: Re: user authentification; Message-ID: <3bcc96d1.524144494f47414741@radiogaga.harz.de>   3 Hoff Hoffman (hoffman@xdelta.zko.dec.nospam) wrote:36 > Mike Rechtman <michael.rechtman@digital.com> writes:- > :Depending on the language you plan to use:RF > :1  call SYS$HASH_PASSWORD with the users parameters and the entered > :password ? > :2  get the quadword password field from the SYSUAF ($GETUAI) ' > :3  compare the two quadword results.  >P) >   The crackers will love this approach.-  E The only problem they'll have is that you need privileges for step 2..   cu,d   Martin -- tD                     | Martin Vorlaender    |    VMS & WNT programmer-   Smiert Spamionem  | work: mv@pdv-systeme.depD                     |       http://www.pdv-systeme.de/users/martinv/4                     | home: martin@radiogaga.harz.de   ------------------------------  # Date: Tue, 16 Oct 2001 20:47:24 GMT)2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)" Subject: Re: user authentification1 Message-ID: <w11z7.532$RL6.4443@news.cpqcorp.net>   p In article <3bcc96d1.524144494f47414741@radiogaga.harz.de>, martin@radiogaga.harz.de (Martin Vorlaender) writes:4 :Hoff Hoffman (hoffman@xdelta.zko.dec.nospam) wrote:7 :> Mike Rechtman <michael.rechtman@digital.com> writes: . :> :Depending on the language you plan to use:G :> :1  call SYS$HASH_PASSWORD with the users parameters and the entered  :> :password@ :> :2  get the quadword password field from the SYSUAF ($GETUAI)( :> :3  compare the two quadword results. :>* :>   The crackers will love this approach. : F :The only problem they'll have is that you need privileges for step 2.  G   Actually not, depending on exactly what you are up to.  You *really* t@   need to be careful here, lest you expose more than you expect.    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Wed, 17 Oct 2001 03:01:53 GMT 3 From: Tim Llewellyn <tim.llewellyn@cableinet.co.uk>," Subject: Re: user authentification/ Message-ID: <3BCCF415.DF094E3F@cableinet.co.uk>-   Kevin Andreoli wrote:  >  > Mike Rechtman wrote: > >,. > > Depending on the language you plan to use:G > > 1  call SYS$HASH_PASSWORD with the users parameters and the enteredd > > password@ > > 2  get the quadword password field from the SYSUAF ($GETUAI)( > > 3  compare the two quadword results. > >  > Alternative view:- >  > 1. read word from dictionary$ > 2. call SYS$HASH_PASSWORD as above > 3. get quadword from SYSUAFi > 4. compare two quadwords# > 5. If quadwords different goto 1. 5 > 6. Success! - a VMS system account has been hacked!  > H > What other reason is there for such a program?  VMS has perfectly good+ > authentication if it is set up correctly.: >    sure, my thoughts too.  J > Isn't that what the guy in 'The Cuckoo's Egg' was doing with U**x boxen? >   D yes, but it does require read access to the sysuaf on VMS, something which arH secure VMS system does not allow without priviledged access. C2 security onB unix accomplishes the same thing but without it /etc/passwd can be vulnerable..   regardsd > -- > Kev. > NNNN   -- o Tim.Llewellyn@cableinet.co.uk  q  C Standard disclaimer applies. My views in no way represent those of  ! my employers or service provider.    ------------------------------  % Date: Tue, 16 Oct 2001 19:24:33 +0100 3 From: "Malcolm" <malcolm@neverness.freeserve.co.uk>t$ Subject: Re: VMS & Unix connectivity/ Message-ID: <9qhujl$3pk$1@newsg3.svr.pol.co.uk>t  : "Jack Patteeuw" <jjpatteeuw@peoplepc.com> wrote in message& news:3BCB9584.BDCDCDE8@peoplepc.com... >' >c > Matt Muggeridge wrote: > <snip>J > > You will need to decide which system will be your NFS server and which will > > be your NFS client.r > <snip> >.L > Well, actually, no you don't.  It is possible to configure your VMS and/orI > Unix system to be both a NFS server and a client of each other.  It amy  seem. > strange but it can be done and it does work. >s >iF I must say that I am quite impressed with the NFS support in UCX 5. ItL allows RMS file attributes to be stored in a separate file on the NFS sever;I when you go to retrieve your file from the NFS disk, it uses this file toaK supply VMS with the correct RMS attributes. [Note: reading the file on Unixl is a different matter :-]c  L Presently, I am running (on a test system) a (non-production) DBMS database,H which is stored on an NFS disk. Despite dire warnings of doom, the wholeJ thing seems to work beautifully[*]. I haven't tried running a large numberB of connections against the database; that might cause a problem ;)  K Just out of interest, what file operations AREN'T supported on NFS? The NFStJ disk reports itself as ODS-2, so presumably, everything works (or pretendsE to work :-), but what about things like SET FILE /NOMOVE and SET FILEe /ENTER?S  C The NFS server in question is a Network Appliances NetApp box. It's-L primarily used by our Unix and NT systems, but I was short of disk space, so I gave it a go...r  F One thing I am thinking of is setting up interconnectivity between theL NetApp and VMS. I am toying with the idea of setting up NFS access to the NTI home directories from the VMS system. Looks like this would be complex (I L want to set up write access, so I'd need to set up a lot of NFS proxies...).H Once I have it done, I could set it up with NT: or NT_HOME: as a logicalG name, and users could dump database reports directly into their NT home J drives from the database application... The intention though is to "merge"H the Unix and NT home directories on the NetApp box, so I'll have to wait till that's sorted out first.r  K Does anyone have any programs to read /etc/passwd, compare the real name inCH a SYSUAF.LIS file with that in the Unix password file, and create an NFSI proxy if the two match using the UID and GID from the Unix password file? G Just wondering... (that approach would take care of about 70-75% of theu work)i  	 -Malcolm.   J [*] Although RMS is probably irrelevant here, since DBMS bypasses RMS fileK organisation (Record format: Undefined) - as long as you could consistentlyW$ read what you wrote it would work... > -- >- > Jack Patteeuwt   ------------------------------  # Date: Tue, 16 Oct 2001 19:53:45 GMTv4 From: "Matt Muggeridge" <Matt.Muggeridge@compaq.com>$ Subject: Re: VMS & Unix connectivity@ Message-ID: <df0z7.173861$bY5.814874@news-server.bigpond.net.au>  K Yep, I realise you can run both, though I was attempting to make a point on@K management.  There are different things you need to do if you manage an NFSs server versus an NFS client.  ) Thanks for making it clear to the thread.l   Cheers,I Matt.r  : "Jack Patteeuw" <jjpatteeuw@peoplepc.com> wrote in message& news:3BCB9584.BDCDCDE8@peoplepc.com... >  >  > Matt Muggeridge wrote: > <snip>J > > You will need to decide which system will be your NFS server and which will > > be your NFS client.  > <snip> >sL > Well, actually, no you don't.  It is possible to configure your VMS and/orI > Unix system to be both a NFS server and a client of each other.  It amy, seem. > strange but it can be done and it does work. >a >. > -- >t > Jack Patteeuws   ------------------------------  % Date: Tue, 16 Oct 2001 20:30:52 +0100n+ From: "Tim Jackson" <tim.jackson@amsjv.com>a/ Subject: VMS7.3,AS7.3 & external authentication/$ Message-ID: <3bcc8857@pull.gecm.com>  E System is DS10 & DS20E VMScluster running VMS 7.3 and AS 7.3 as a BDCw, for a domain where the PDC is an NT4 server.  - "NT" usernames are hostmaped to VMS accounts.j  G External authentication is set for the users but those users whose "NT" D username is > 15 characters cannot logon to their VMS accounts via aG terminal emulator, while all others work fine.  I thought the max. "NT"   username size was 20 characters.  
 Any ideas?   TIA,D ------------------ Purely Personal Opinion -------------------------D Tim Jackson                                    tim.jackson@amsjv.com Air Systems Groupn Alenia Marconi Systems Ltd.c   ------------------------------  # Date: Tue, 16 Oct 2001 23:31:52 GMT ) From: rob.buxton@wcc.govt.nz (Rob Buxton) 3 Subject: Re: VMS7.3,AS7.3 & external authentication.2 Message-ID: <3bccc2d4.1395162147@news.wcc.govt.nz>  1 On Tue, 16 Oct 2001 20:30:52 +0100, "Tim Jackson"l <tim.jackson@amsjv.com> wrote:  F >System is DS10 & DS20E VMScluster running VMS 7.3 and AS 7.3 as a BDC- >for a domain where the PDC is an NT4 server.  >O. >"NT" usernames are hostmaped to VMS accounts. >BH >External authentication is set for the users but those users whose "NT"E >username is > 15 characters cannot logon to their VMS accounts via avH >terminal emulator, while all others work fine.  I thought the max. "NT"! >username size was 20 characters.v >a >Any ideas?J   Not really but...0  A Why do you use Host Mapping? are you're NT Usernames different tog their VMS Counterparts?u  ? Here we have External Authentication, but because the Usernamesl3 between VMS & NT are the same we have no Host Maps.t  D Does anything show up under Opcom, I would have thought there'd be aD Login Failure recorded somewhere, this might indicate what was going wrong.   Rob. >f >TIAE >------------------ Purely Personal Opinion -------------------------bE >Tim Jackson                                    tim.jackson@amsjv.coml >Air Systems Group >Alenia Marconi Systems Ltd. >r >s >s   ------------------------------   Date: 16 Oct 2001 19:36:51 GMT3 From: bobd@araminta.uts.ohio-state.edu (Bob DeBula)c5 Subject: Re: We've burned our boats say Compaq and HPs: Message-ID: <9qi28j$142$1@charm.magnus.acs.ohio-state.edu>  C D. Aaron Sawyer <aaron@110.net> carefully crafted electrons to say:  > > L > > And how much money does that person get for destroying the company?  HowN > > often can CEO's make more money destroying a company rather than making it > > successful?e >  > 1. US$Millions.  > 0 > 2. Every time they are hired for that purpose.  A And remember the progression we've been seeing, with the standard F "safety net" seeming to be a cushy job within the protective sphere ofK the  Wintel Hegemony for these Caesar salads after the victim compan(y|ies) I is/are essentially defunct.  Makes you wonder sometimes....Poor DEC, poorgB SGI, poor wHolit Packherd, poor DECPAQ, poor cHomPac...Rocket RickB appears to have softened up at least two and appears to be leadingG the pack for the honor of the most billyuns pounded in (to the ground).O  F ======================================================================D (Note: Author not responsible for content effects caused by neuronicE distortions due to the close proximity to several largish EMR sourcese at the time of composition).   ------------------------------   Date: 16 Oct 2001 22:18:56 GMT) From: leslie@clio.rice.edu (Jerry Leslie)a/ Subject: Windows Fails To Storm the Data Centreo' Message-ID: <9qibog$h45$1@joe.rice.edu>6% Keywords: vms,unix,windows,datacenterS   From:e  9    http://news.zdnet.co.uk/story/0,,t269-s2096347,00.html@E    ZDNet |UK| - News - Story - Windows fails to storm the data centre       "11:38 Monday 1st October 2001)    Computerwire.com  Source: ComputerwirelG    So far, Microsoft's high-end Windows Datacenter Server hasn't been aaE    success. It will be a while before IT managers trust their largestE    systems to Microsoft   I    Microsoft's high-end answer to Unix - Windows Datacenter Server - willeF    fail to see widescale adoption until late in 2002, nearly two years@    after its launch, because of concern among customers over the2    operating system's reliability and suitability.    .    .    .A    Observers believe that Windows 2000 Datacenter Server will seefE    increased adoption by customers as they come to accept Microsoft'scH    claims of reliability - the company is aiming the operating system at7    mission-critical environments, such as data centres.n  C    One reason for growing acceptance could come as more third-party C    applications and hardware configurations are certified under the F    Windows Datacenter Program - a year old this month. Ten ISVs and 12G    OEMs are already certified, with more expected in coming months, andeC    applications and hardware configurations are undergoing rigorous     testing..."    I Given Microsoft's track record to-date on "reliability and suitability", .> this is wishful thinking without further technology donations.  J The pro-Microsoft PHBs are fond of saying "Well, NT is where VMS 3.x was".H A fairer comparision is to compute the age of WNT/W2K/WXP, and then lookF to see where VMS was after that many years. VMS 3.x is still ahead of  today's NT in reliability.  F If Microsoft takes over computing, it's because the companies who sell7 both Microsoft systems and competing products allow it.h  4 --Jerry Leslie     (my opinions are strictly my own)   ------------------------------  # Date: Wed, 17 Oct 2001 02:32:18 GMTw! From: "ClaudeVMS" <unix@dev.null>o3 Subject: Re: Windows Fails To Storm the Data Centrei@ Message-ID: <S46z7.64974$w62.39588900@news1.denver1.co.home.com>  I It is very depressing to see the worlds' best operating system sitting onc theiG sidelines while the Spartans (NT) and the Legless Wonders (UNIX) battleoJ to provide what was available in 1978 under VMS - many more times reliable
 at that!!!        6 "Jerry Leslie" <leslie@clio.rice.edu> wrote in message! news:9qibog$h45$1@joe.rice.edu...  > From:o >R; >    http://news.zdnet.co.uk/story/0,,t269-s2096347,00.html G >    ZDNet |UK| - News - Story - Windows fails to storm the data centre  >v" >   "11:38 Monday 1st October 2001+ >    Computerwire.com  Source: Computerwire I >    So far, Microsoft's high-end Windows Datacenter Server hasn't been asG >    success. It will be a while before IT managers trust their largestt >    systems to Microsofte > K >    Microsoft's high-end answer to Unix - Windows Datacenter Server - willmH >    fail to see widescale adoption until late in 2002, nearly two yearsB >    after its launch, because of concern among customers over the4 >    operating system's reliability and suitability. >    . >    . >    .C >    Observers believe that Windows 2000 Datacenter Server will see G >    increased adoption by customers as they come to accept Microsoft'snJ >    claims of reliability - the company is aiming the operating system at9 >    mission-critical environments, such as data centres.- >-E >    One reason for growing acceptance could come as more third-party.E >    applications and hardware configurations are certified under the-H >    Windows Datacenter Program - a year old this month. Ten ISVs and 12I >    OEMs are already certified, with more expected in coming months, and1E >    applications and hardware configurations are undergoing rigorousn >    testing..." >o > J > Given Microsoft's track record to-date on "reliability and suitability",@ > this is wishful thinking without further technology donations. >EL > The pro-Microsoft PHBs are fond of saying "Well, NT is where VMS 3.x was".J > A fairer comparision is to compute the age of WNT/W2K/WXP, and then lookG > to see where VMS was after that many years. VMS 3.x is still ahead of  > today's NT in reliability. > H > If Microsoft takes over computing, it's because the companies who sell9 > both Microsoft systems and competing products allow it.  >W6 > --Jerry Leslie     (my opinions are strictly my own)   ------------------------------  % Date: Tue, 16 Oct 2001 17:19:34 -0400s5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>e* Subject: Re: X programming tools (UIL etc)1 Message-ID: <Av1z7.537$RL6.4704@news.cpqcorp.net>i  ' ICS Builder Xcessory (BX).  www.ics.com   I I've met with these guys, and seen their tools, and it seems to be what Ie& would start a Motif application using.      " Martin Vorlaender wrote in message+ <9qh0dq$nidh1$1@ID-56200.news.dfncis.de>...hJ >>>> I forget the details, and am in no hurry to re-learn.  But if you canI >>>> generate a UID, is that portable?  I would suspect so; I assume it'snF >>>> only a binary input file.  There are many .UIDs sitting on Alpha. >>>4A >>> No need to port UID files. There is an UIL compiler on Alpha.o >>H >> On my VAX, we have a licence for VUIT, was a licence required for the >> UIL compiler on the Alpha?u >a; >Nope. It's part of DECwindows nee OSF/Motif ($ UIL/MOTIF).e >sF >Look what I've found about generating UIL files (in the online help): >t >  ... First, you G >  describe a user interface by creating a UIL module with one  of  theeG >  editors  available  on  VAX  VMS (such as the VAX Language-Sensitivee >  Editor), ...  >uJ >So there's an LSE module for UIL. Makes sense (in a non-grpahical way...) >r >cu,	 >  Martint >--iK >One OS to rule them all       | Martin Vorlaender  |  VMS & WNT programmers8 >One OS to find them           | work: mv@pdv-systeme.deK >One OS to bring them all      |   http://www.pdv-systeme.de/users/martinv/0? >And in the Darkness bind them.| home: martin@radiogaga.harz.dea >  >j   ------------------------------  % Date: Wed, 17 Oct 2001 00:29:04 +0100p1 From: Steve Reece <SYSTEM@ipl.demon.co.nospam.uk>aL Subject: Re: [OpenVMS] V7.2 VAX satellite doesn't find V7.3 Alpha bootserver6 Message-ID: <3BCCD0D0.1B4AE422@ipl.demon.co.nospam.uk>  A I understood that it's not just performance that suffers from noteD including the /noalias qualifier - the resultant disk structure on a system disk would be incorrect.>  B IIRC, there was a long discussion of this qualifier in comp.os.vmsA either last year or the year before which included the issue thatkB /noalias is not the default but is required when backing up system disks.   Steve.   Peter LANGSTOEGER wrote: > l > In article <3BC63E6E.6B490D52@ipl.demon.co.nospam.uk>, Steve Reece <SYSTEM@ipl.demon.co.nospam.uk> writes:G > >It may not be the cause of the problem, but the image backup command-K > >should have a /NOALIAS in it too since around 6.2 days.  Otherwise thereQK > >can be problems with the aliased directory entries on the restored disk.i > D > AFAIK, that would only be performance problems during the restore.; > The disk structure is correct in both ways. That's /IMAGE-K > And ANAL/DISK (and also a SHOW DEVICE/FILES) and a running VMS proves it.e >  > --> > Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651= > Network and OpenVMS system manager  Fax.    +43 1 81111-888 > > <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netJ > A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"   -- nG "A shadow fell over her face; clear, as if the composure were rent likeUE a veil.  And her lips parted, but only with a short intake of breath.tA Then she said, 'Well, then you are right.  Indeed, we are even.'"P% 		Louis, "Interview with the Vampire"m   ------------------------------  % Date: Wed, 17 Oct 2001 00:03:43 +0100 1 From: Steve Reece <SYSTEM@ipl.demon.co.nospam.uk>d9 Subject: Re: [VMS V7.3 Alpha] my first crash since monthsM6 Message-ID: <3BCCCADF.7900DDED@ipl.demon.co.nospam.uk>   Kor Rinkens wrote: > c > axica@yahoo.com (Safir) wrote in message news:<e85d7983.0110050050.8cb49c6@posting.google.com>...e` > > eplan@kapsch.net (Peter LANGSTOEGER) wrote in message news:<3bbcdf72$1@news.kapsch.co.at>... <trimed> > P > On vms 7.3 there are problems with snmp, you have to install tcp/ip v5.1 eco 2 > O > If you did enable the service snmp, your system may hang or crash on the snmp 	 > servicey > 
 > Regards Korq  H Last I heard Kor, TCP/IP Services for OpenVMS v5.1 Eco 2 had been put on= hold pending some changes to smtp.  It may be worth checking.-   Steve. -- ,G "A shadow fell over her face; clear, as if the composure were rent likeeE a veil.  And her lips parted, but only with a short intake of breath.$A Then she said, 'Well, then you are right.  Indeed, we are even.'"m% 		Louis, "Interview with the Vampire"D   ------------------------------  # Date: Wed, 17 Oct 2001 00:07:32 GMTi2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)9 Subject: Re: [VMS V7.3 Alpha] my first crash since months 1 Message-ID: <8Z3z7.539$RL6.4803@news.cpqcorp.net>c  j In article <3BCCCADF.7900DDED@ipl.demon.co.nospam.uk>, Steve Reece <SYSTEM@ipl.demon.co.nospam.uk> writes:  : :...TCP/IP Services for OpenVMS v5.1 Eco 2 had been put on> :hold pending some changes to smtp.  It may be worth checking.  B    ECO3 is out, and the SMTP problem introduced in ECO2 was fixed.    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   End of INFO-VAX 2001.577 ************************