1 INFO-VAX	Wed, 11 Jun 2003	Volume 2003 : Issue 321       Contents: Re: alphaServer diagnostics # Re: AVAIL (EVENTS) Default Settings # Re: AVAIL (EVENTS) Default Settings & Re: CSWB 1.2.1 fails to display images Re: Default route " Re: Dell to make Itanium 2 systems$ Re: ES40 processor upgrade questionsD Re: Hobbyist or EDU license experience with multiprocessor machines?D Re: Hobbyist or EDU license experience with multiprocessor machines?D Re: Hobbyist or EDU license experience with multiprocessor machines?D Re: Hobbyist or EDU license experience with multiprocessor machines?! Re: HSZ80 / ES40 cabling question ! Re: HSZ80 / ES40 cabling question ! Re: HSZ80 / ES40 cabling question ! Re: HSZ80 / ES40 cabling question ! Re: HSZ80 / ES40 cabling question 4 Independant VMS crashdump analysis service launched.8 Re: Independant VMS crashdump analysis service launched./ Re: Intel 64 bit Pentium seems to be on its way / Re: Intel 64 bit Pentium seems to be on its way $ Re: Java Runtime.exec(String) on VMS$ Re: Java Runtime.exec(String) on VMS$ Re: Java Runtime.exec(String) on VMS< Re: JF Mezei "flamed, insulted, denigraded(sic)"! Poor baby!& re: mail to news gateways for OpenVMS?" Re: Memory limitations for Process Re: need kbd for VAX4000/90  Re: need kbd for VAX4000/90 ' Re: OT: CNN Story Not Favorable To Bush ' Re: OT: CNN Story Not Favorable To Bush  Re: perl question  PIPE output defined to a symbol # Re: PIPE output defined to a symbol  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death ? Re: Printer recommendation, hobbyist/decwindows/mozilla and VMS ? Re: Printer recommendation, hobbyist/decwindows/mozilla and VMS ? Re: Printer recommendation, hobbyist/decwindows/mozilla and VMS 
 Re: RTR/XAI Simulatenously transferring a 1GB file from VMS to two Windows 2k servers M Re: Simulatenously transferring a 1GB file from VMS to two Windows 2k servers M Re: Simulatenously transferring a 1GB file from VMS to two Windows 2k servers M Re: Simulatenously transferring a 1GB file from VMS to two Windows 2k servers M Re: Simulatenously transferring a 1GB file from VMS to two Windows 2k servers P Re: Simulatenously transferring a 1GB file from VMS to two Windows 2k servers 2kP Re: Simulatenously transferring a 1GB file from VMS to two Windows2k servers ser Re: Supporting  Wintel laptops Re: VAX VMS 7.3 - follow up II VMS Freeware CDs Re: VMS Freeware CDs Re: VMS Freeware CDs Re: VMS Freeware CDs Re: VMS Freeware CDsC Re: Will a Radeon 9000 PCI card work in the place of a Radeon 7500? C Re: Will a Radeon 9000 PCI card work in the place of a Radeon 7500? C Re: Will a Radeon 9000 PCI card work in the place of a Radeon 7500?   F ----------------------------------------------------------------------    Date: 10 Jun 2003 17:28:40 -05004 From: kuhrt@nospammy.encompasserve.org (Marty Kuhrt)$ Subject: Re: alphaServer diagnostics3 Message-ID: <bRN5rQx8eRyY@eisner.encompasserve.org>   l In article <4b6ec350.0306060859.2d7a24b7@posting.google.com>, JimStrehlow@data911.com (Jim Strehlow) writes:> > What diagnostic tools are available for an AlphaServer DS10? >  - freeware download! >  - on the software tools cd-rom # >  - for licensed purchase from HP?  > @ > We have an OpenVMS AlphaServer DS10 on sofware support but notH > hardware support which "locks up" the operator's console (keyboard and= > mouse) running OpenVMS v7.3 patched up through a month ago.  > * > Thank you in advance for your responses. >  > Jim Strehlow > Data911, Alameda, CA, USA   6 Is the whole machine locked up, or just the X stuff?    @ I had a situation recently when using the DECWindow interface to: CMS where the system "locked up".  The mouse, keyboard and; (thus) the windows no longer responded.  I telneted in from A another machine and the system was still running.  I tracked down ? the process  running CMS and killed it.  Everything was back to 	 normal.      ------------------------------   Date: 10 Jun 03 20:54:00 +0200) From: p_sture@elias.decus.ch (Paul Sture) , Subject: Re: AVAIL (EVENTS) Default Settings) Message-ID: <kdvAa4Wxj77d@elias.decus.ch>   \ In article <03061012200158@dscis6-0.dalsemi.com>, brandon@dalsemi.com (John Brandon) writes:   <snip>   > G > So I am a dreamer - wait a tick - was not the GS series touted as the  > DreamMachine??? ;) >   A The first GS320 I came across had "Dream Machine" programmed into ? the front panel display. (Don't know whether that was a factory / setting or done by the installation engineers.)    >  > 5 > Call me what you will, just do not call me Shirley!    Resists temptation :-)     --  
 Paul Sture   ------------------------------  % Date: Tue, 10 Jun 2003 14:37:50 -0500 ( From: brandon@dalsemi.com (John Brandon), Subject: Re: AVAIL (EVENTS) Default Settings1 Message-ID: <03061014375020@dscis6-0.dalsemi.com>   I > > So I am a dreamer - wait a tick - was not the GS series touted as the  > > DreamMachine??? ;) > C > The first GS320 I came across had "Dream Machine" programmed into A > the front panel display. (Don't know whether that was a factory 1 > setting or done by the installation engineers.)    Mine said LEMON ... ;)  7 > > Call me what you will, just do not call me Shirley!  >  > Resists temptation :-)  / Yea, I forgot I put that in there... oh bother!     - And to Barry from HP - thanks for your input!   3 Call me what you will, just do not call me Shirley!    John Brandon VMS Systems Administrator $ john.brandon.spam.me.not@dalsemi.com   ------------------------------  # Date: Wed, 11 Jun 2003 01:03:52 GMT 6 From: "Craig A. Berry" <craigberry@mac.com.spamfooler>/ Subject: Re: CSWB 1.2.1 fails to display images @ Message-ID: <d2f86f4b460cd3026901de4d90a3cc57@free.teranews.com>  < In article <a98cd882.0306100539.6c1b7cb@posting.google.com>,'  Bart.Zorn@xs4all.nl (Bart Zorn) wrote:   G > I could not find anything related on bugzilla, and I could not find a ' > way of posting a problem on bugzilla.   F Try clicking the link that says "Report a Bug" in a very large font.  < It's right on the main page at http://bugzilla.mozilla.org/.   ------------------------------  # Date: Wed, 11 Jun 2003 00:46:34 GMT  From: Rob.Buxton@wcc.govt.nz Subject: Re: Default route$ Message-ID: <3ee67baa.17559390@news>  E On Tue, 10 Jun 2003 07:33:15 GMT, peter@langstoeger.at (Peter 'EPLAN'  LANGSTOEGER) wrote:   g >In article <3EE53FBC.B86BB38D@firstdbasource.com>, Michael Austin <maustin@firstdbasource.com> writes: 
 >>Kees wrote: F >>> I wonder if there is a functional difference in defining a default. >>> route in the following two ways (example): >>>  >>> First way:4 >>> UCX> set route /default /gateway=192.168.100.222: >>> UCX> set route /default /gateway=192.168.100.222 /perm >>>  >>> Second way: < >>> UCX> set route 0.0.0.0 /gateway=192.168.100.222 /networkA >>> UCX> set route 0.0.0.0 /gateway=192.168.100.222 /network/perm  >>E >>no difference. but I would upgrade from UCX to latest VMS and TCPIP 
 >>versions...  > I >I second it, but I personally still use "UCX" instead of "TCPIP" because H >of laziness typing too much characters. So, assuming an old TCPIP stackC >only on the word "UCX" in above command lines fails on my site ;-)  > F >As you might know, TCPIP has "UCX" still defined for compatibility... >  >	define verb UCX  >	   image TCPIP$UCP >	   parameter P1 " >	      value (type=$rest_of_line) >	   noqualifiers  > D I think Michael noted the prompt was UCX> which would imply an older$ version of UCX/TCPIP was being used.   >--  >Peter "EPLAN" LANGSTOEGER& >Network and OpenVMS system specialist >E-mail  peter@langstoeger.at G >A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist    ------------------------------  # Date: Tue, 10 Jun 2003 20:39:54 GMT 9 From: "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> + Subject: Re: Dell to make Itanium 2 systems 2 Message-ID: <umrFa.2294$qb1.1789@news.cpqcorp.net>  ! At least right now, none of them.    I've explained this before...   L VMS drivers writers have had to deal with a direct DMA window, and with "mapE registers".  Ultimately, except for a very few that use BAP - drivers I sometimes use the direct dma window if possible, but use map registers as / the fallback.  Some *always* use map registers.   L To get to the point where generic hardware is supportable, we need to modifyD a core set of drivers to deal with the case where the IO needs to beG buffered and copied.  It's on our list, but will take time and priority & wise - it isn't critical at this time.  K But other than that, and barring bugs - technically it would probably boot.     < "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message! news:3EE53545.1B76D514@fsi.net...  > Keith Parris wrote: 
 > > [snip]H > > A quick Google search confirms that indeed, starting about SeptemberH > > 2002, Dell started telling customers that it plans to ship Itanium 2 > > systems in 2003. > 4 > I wonder how many of them will be OpenVMS-capable? >  > -- > David J. Dachtera  > dba DJE Systems  > http://www.djesys.com/ > * > Unofficial Affordable OpenVMS Home Page:! > http://www.djesys.com/vms/soho/    ------------------------------  # Date: Tue, 10 Jun 2003 23:39:18 GMT - From: Jack Patteeuw <jjpatteeuw@peoplepc.com> - Subject: Re: ES40 processor upgrade questions > Message-ID: <G_tFa.110893$BA.39223838@twister.columbus.rr.com>   Jonathan Boswell wrote: Q > You might also tell them they could pull a 500 MHz processor out of one, put it M > in the other, and buy two identical 667 or 833 MHz processors for the donor Q > ES40.  This yields two dual-processor ES40's, though not with identical power.  J > I pulled this stunt and it works well.  (Actually I went from two singleN > processor 667s to dual processor configurations, and then to quad processorsR > with 4 667s and 4 833s.)  Your SMP licenses don't care what speed each processor > has.  D One thing to watch for.  Old 500 Mhz boxes require an FCO that IIRC I replaces the motherboard before you can upgrade to either 667 or 833 Mhz.    jp   ------------------------------  # Date: Wed, 11 Jun 2003 01:18:34 GMT L From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr")M Subject: Re: Hobbyist or EDU license experience with multiprocessor machines? 6 Message-ID: <00A21305.2AE6DCD7@SSRL.SLAC.STANFORD.EDU>  i In article <mPuFa.2314$xm1.1472@news.cpqcorp.net>, "Mark Buda" <buda@tabasco.zko.dec.no.spam.com> writes:  > F >> Does anyone have relevant experience in running Hobbyist or EDU VMS >base G >> licenses on multiprocessor machines?  If I can point at places where  >it's G >> working, I'll feel more confident in telling my colleague to order a  >CPU >> without a license.  > F >The below URL will provide you with unit information - look for DS20E > 0 >http://licensing.hp.com/swl/view.slm?page=alpha > H >Going from a single processor to a dual will go from units 25 to 75.  IG >am not familiar with the CSLG and Hobbyist programs to be able to tell G >you what they allow or not.  You would need an additional 50 units for ( >the second processor to work correctly.   Mark, thanks for your reply.  I I see that the EDU license has /UNITS=0 on it.  My understanding is that  E means the unit count is unlimited, so multiprocessors requiring only  % additional license units should work.    Is that correct?   Thanks,    -- Alan    --  O =============================================================================== 0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056 M  Paper mail to: SSRL -- SLAC BIN 99, 2575 Sand Hill Rd, Menlo Park CA   94025 O ===============================================================================    ------------------------------    Date: 10 Jun 2003 18:31:14 -07001 From: nothome@spammers.are.scum (Malcolm Dunnett) M Subject: Re: Hobbyist or EDU license experience with multiprocessor machines? - Message-ID: <L2jQtNb0n28R@malvm7.mala.bc.ca.>   2 In article <mPuFa.2314$xm1.1472@news.cpqcorp.net>,:     "Mark Buda" <buda@tabasco.zko.dec.no.spam.com> writes:  G >> working, I'll feel more confident in telling my colleague to order a  > CPU  >> without a license.  > I > Going from a single processor to a dual will go from units 25 to 75.  I H > am not familiar with the CSLG and Hobbyist programs to be able to tellH > you what they allow or not.  You would need an additional 50 units for) > the second processor to work correctly.  > I     The Hobbyist and Edu program both provide OPENVMS-ALPHA licenses with E 0 ( unlimited ) units, so the second processor should work just fine.   E (Note: this is a technical answer - not a legal one. AFAIK HP doesn't @ care how many CPUs are in your box under these programs but I'veL never seen a definitive statement from them one way or the other. PersonallyD I would take the 0 units and lack of any restrictive statement as anE indication it's ok to use multiple processors under these programs ).    ------------------------------  # Date: Wed, 11 Jun 2003 00:35:30 GMT 4 From: "Mark Buda" <buda@tabasco.zko.dec.no.spam.com>M Subject: Re: Hobbyist or EDU license experience with multiprocessor machines? 2 Message-ID: <mPuFa.2314$xm1.1472@news.cpqcorp.net>  H ""Alan Winston - SSRL Admin Cmptg Mgr"" <winston@SSRL.SLAC.STANFORD.EDU>A wrote in message news:00A21251.84CEA5D2@SSRL.SLAC.STANFORD.EDU... H > A colleague here has a uniprocessor DS20E running as a workstation for his @ > individual use.  It came with a VMS base license, and now it's enrolled in  > CSLG.  > F > He wants to add another processor.  I'm under the impression that he
 can switchE > over to the EDU license, which provides a VMS base license, and run 
 both CPUs.G > Our local CSLG guy talked to an HP CSLG guy who says you need another  license   > to 'unlock' an additional CPU.   <snip>  E > Does anyone have relevant experience in running Hobbyist or EDU VMS  baseF > licenses on multiprocessor machines?  If I can point at places where it'sF > working, I'll feel more confident in telling my colleague to order a CPU  > without a license.  E The below URL will provide you with unit information - look for DS20E   / http://licensing.hp.com/swl/view.slm?page=alpha   G Going from a single processor to a dual will go from units 25 to 75.  I F am not familiar with the CSLG and Hobbyist programs to be able to tellF you what they allow or not.  You would need an additional 50 units for' the second processor to work correctly.    --  
 Sincerely,	 Mark Buda  Hewlett-Packard Company  VMS Engineering  110 Spitbrook Road
 MS: ZK3-4/X57  Nashua, NH 03062 Voice: (603) 884-1969  FAX: (603) 884-3451   + VMS Home Page http://www.openvms.compaq.com > OpenVMS Portal http://www.openvms.compaq.com/portal/index.html   ------------------------------  % Date: Tue, 10 Jun 2003 22:36:02 -0700 % From: Dean Woodward <deanw@rdrop.com> M Subject: Re: Hobbyist or EDU license experience with multiprocessor machines? ( Message-ID: <3EE6BFC2.8030707@rdrop.com>  < I'm running hobbyist on an AS1200 (Dual 533Mhz). No worries.  F > Does anyone have relevant experience in running Hobbyist or EDU VMS E > base licenses on multiprocessor machines?  If I can point at places F > where it's working, I'll feel more confident in telling my colleague# > to order a CPU without a license.    ------------------------------  % Date: Tue, 10 Jun 2003 22:33:46 +0200  From: Dirk Munk <munk@home.nl>* Subject: Re: HSZ80 / ES40 cabling question2 Message-ID: <bc5fl0$hi1$1@news4.tilbu1.nb.home.nl>  1 This sounds like a very odd configuration indeed!   N If I understand you correctly, one disk cabinet is exclusively coupled to one O ES40. Then of course it is perfectly logical that when the ES40 goes down, the  $ disks of that cabinet are gone too !  L  From your description I understand that both HSZ80 controllers of one disk H cabinet are connected to the same SCSI bus that runs from one ES40 SCSI  controller to the other.  
 Why I wonder.   O Suppose you would have only one ES40 and one disk cabinet with a pair of HSZ80  O controllers. In that case you should connect both HSZ controllers to their own  O SCSI controller, and place the HSZ80 pair in multibus failover. This will give  L you two independent SCSI busses for redundancy, plus double the IO capacity.  P If I understand you correctly, the ES40's are only connected to each other over @ the network. That makes this whole setup a rather slow one .....  L A 'normal' setup would be to have one disk cabinet. You would have two SCSI Q busses. A SCSI bus would go from a SCSI controller in the first ES40, to one HSZ  C controller, and from there to a SCSI controller in the second ES40.   O This would give you a SCSI cluster with multibus failover on the HSZ80. Now of  N course you could double that configuration with your second disk cabinet, and P use VMS shadow over both cabinets, but I fail to see what it would bring in the  real world.        Z wrote:7 > I have 2 clustered ES40s in a cabinet running OpenVMS 9 > 7.1-2. Each ES40 has 2 SCSI cards.  In the cabinet next 3 > to the ES40s are 2 RAID 8000/ESA12000 enclosures.  > 6 > Each enclosure has 2 HSZ80 controllers and 14 disks,9 > arranged as 7 mirrors.  Each mirror in the 2 enclosures $ > is also shadowed back at the host. >  > So, the configuration is >   > 1 Shadow Set = 2 Mirrors   and > 1 Mirror     = 2 Disks > 7 > The system is supposed to be wired in such a way that : > if either ES40 is down (or crashes) no storage redudancy > will be lost.  > : > That is, all disks in both enclosures should be visible,3 > available and mountable to the remaininging ES40.  > ; > But, what we see is that when one system is down, we lose 7 > access to all  the disks in an enclosure, essentially , > losing half of all our storage redundancy. > ; > I think our problem is cabling between the HSZ80s and the < > ES40s.  Presently, I see that both SCSI cards on each ES40< > are cabled to the same enclosure (one to Controller A, the: > other to Controller B, which A and B jumpered together). >  > ; > I've been through all the hardware documenation that came  > with the RAID array: > 6 >  STORAGEWORKS HSZ80 Array Cpntroller ACS Version 8.3A >  Compaq StorageWorks UltraSCSI RAID Enclosure (DS-BA370 Series) 8 >  RAID8000 and EAS 12000 Storage Subsystem User's Guide' >  Digital StorageWorks EK-HSXSW-WC B01  > = > ...but I can't find a definitive answer for what the proper : > cabling between the HSZs and the SCSI cards in the ES40s= > should be so that I don't lose an entire array when an ES40 : > is down.  I can't even say for sure if this is possible. > ; > Can anyone here point me to another DEC/Compaq/HP public- > > ation that cover this?  I've already tried Ask The Wizard on > the HP website, to no avail.   ------------------------------  % Date: Tue, 10 Jun 2003 20:58:40 -0000 ! From: Z  <zarlenga@conan.ids.net> * Subject: Re: HSZ80 / ES40 cabling question/ Message-ID: <vechk0t0tfiqeb@corp.supernews.com>    Dirk Munk <munk@home.nl> wrote: 3 : This sounds like a very odd configuration indeed!   P : If I understand you correctly, one disk cabinet is exclusively coupled to one Q : ES40. Then of course it is perfectly logical that when the ES40 goes down, the  & : disks of that cabinet are gone too !   Yes.  And that's what we see.     N :  From your description I understand that both HSZ80 controllers of one disk J : cabinet are connected to the same SCSI bus that runs from one ES40 SCSI  : controller to the other.  < I don't know if that's accurate.  The ES40s each have 2 SCSI> cards in them.  Let's call them M1S1 and M1S2  (Machine 1 SCSI? 1 and Machine 1 SCSI 2).  M1S1 is cabled to Controller A in en- A closure 1, M1S2 is cabled top Controller B in the same enclosure.     Q : Suppose you would have only one ES40 and one disk cabinet with a pair of HSZ80  Q : controllers. In that case you should connect both HSZ controllers to their own  Q : SCSI controller, and place the HSZ80 pair in multibus failover. This will give  N : you two independent SCSI busses for redundancy, plus double the IO capacity.  B The docs I read say that MultiBus Failover is ONLY available with ? OVMS 7.2 or greater.  We're on 7.1-2.  They discuss "Trasparent > Failover," (OVSM 6.2 or greater) and I think that's what we're supposed to be setup for.       R : If I understand you correctly, the ES40's are only connected to each other over B : the network. That makes this whole setup a rather slow one .....  = I'm not sure what cluster interconnect is used.  I THINK it's 7 LAN.  And that makes sense since the second ES40 in the = cluster is intended to sit idle until the primary experiences = a HW failure, at which time the transaction processing s/w is ? started on the other ES40 and everything is back up and online.         
 : Z wrote:8 :> I have 2 clustered ES40s in a cabinet running OpenVMS: :> 7.1-2. Each ES40 has 2 SCSI cards.  In the cabinet next4 :> to the ES40s are 2 RAID 8000/ESA12000 enclosures. :>  7 :> Each enclosure has 2 HSZ80 controllers and 14 disks, : :> arranged as 7 mirrors.  Each mirror in the 2 enclosures% :> is also shadowed back at the host.  :>   :> So, the configuration is  :>  ! :> 1 Shadow Set = 2 Mirrors   and  :> 1 Mirror     = 2 Disks  :>  8 :> The system is supposed to be wired in such a way that; :> if either ES40 is down (or crashes) no storage redudancy  :> will be lost. :>  ; :> That is, all disks in both enclosures should be visible, 4 :> available and mountable to the remaininging ES40. :>  < :> But, what we see is that when one system is down, we lose8 :> access to all  the disks in an enclosure, essentially- :> losing half of all our storage redundancy.  :>  < :> I think our problem is cabling between the HSZ80s and the= :> ES40s.  Presently, I see that both SCSI cards on each ES40 = :> are cabled to the same enclosure (one to Controller A, the ; :> other to Controller B, which A and B jumpered together).  :>   :>  < :> I've been through all the hardware documenation that came :> with the RAID array:  :>  7 :>  STORAGEWORKS HSZ80 Array Cpntroller ACS Version 8.3 B :>  Compaq StorageWorks UltraSCSI RAID Enclosure (DS-BA370 Series)9 :>  RAID8000 and EAS 12000 Storage Subsystem User's Guide ( :>  Digital StorageWorks EK-HSXSW-WC B01 :>  > :> ...but I can't find a definitive answer for what the proper; :> cabling between the HSZs and the SCSI cards in the ES40s > :> should be so that I don't lose an entire array when an ES40; :> is down.  I can't even say for sure if this is possible.  :>  < :> Can anyone here point me to another DEC/Compaq/HP public-? :> ation that cover this?  I've already tried Ask The Wizard on  :> the HP website, to no avail.    ------------------------------  % Date: Tue, 10 Jun 2003 23:33:19 +0200  From: Dirk Munk <munk@home.nl>* Subject: Re: HSZ80 / ES40 cabling question2 Message-ID: <bc5j4f$m72$1@news3.tilbu1.nb.home.nl>   Z wrote:! > Dirk Munk <munk@home.nl> wrote: 5 > : This sounds like a very odd configuration indeed!  > R > : If I understand you correctly, one disk cabinet is exclusively coupled to one S > : ES40. Then of course it is perfectly logical that when the ES40 goes down, the  ( > : disks of that cabinet are gone too ! >  > Yes.  And that's what we see.  >  > P > :  From your description I understand that both HSZ80 controllers of one disk L > : cabinet are connected to the same SCSI bus that runs from one ES40 SCSI  > : controller to the other. > > > I don't know if that's accurate.  The ES40s each have 2 SCSI@ > cards in them.  Let's call them M1S1 and M1S2  (Machine 1 SCSIA > 1 and Machine 1 SCSI 2).  M1S1 is cabled to Controller A in en- C > closure 1, M1S2 is cabled top Controller B in the same enclosure.   G In your original message you wrote that both controllers are 'jumpered  O together', I took that as a SCSI T-connector between both controllers. If that  ( is the case, then you have one SCSI bus.     >  > S > : Suppose you would have only one ES40 and one disk cabinet with a pair of HSZ80  S > : controllers. In that case you should connect both HSZ controllers to their own  S > : SCSI controller, and place the HSZ80 pair in multibus failover. This will give  P > : you two independent SCSI busses for redundancy, plus double the IO capacity. > D > The docs I read say that MultiBus Failover is ONLY available with A > OVMS 7.2 or greater.  We're on 7.1-2.  They discuss "Trasparent @ > Failover," (OVSM 6.2 or greater) and I think that's what we're > supposed to be setup for.   K True, and upgrading VMS would be very sensible. Your present version is no  Q longer officially supported AFAIK. I would have to read the manuals, but I don't  4 think transparent failover is a good option for you.   >  >  > T > : If I understand you correctly, the ES40's are only connected to each other over D > : the network. That makes this whole setup a rather slow one ..... > ? > I'm not sure what cluster interconnect is used.  I THINK it's 9 > LAN.  And that makes sense since the second ES40 in the ? > cluster is intended to sit idle until the primary experiences ? > a HW failure, at which time the transaction processing s/w is A > started on the other ES40 and everything is back up and online.   O Sounds like a kind of Unix 'cluster' setup. Isn't the software cluster aware ?  N In other words, can't you run the application on two nodes as in a 'real' VMS 	 cluster ?    >  >  >  >  > : Z wrote:: > :> I have 2 clustered ES40s in a cabinet running OpenVMS< > :> 7.1-2. Each ES40 has 2 SCSI cards.  In the cabinet next6 > :> to the ES40s are 2 RAID 8000/ESA12000 enclosures. > :>  9 > :> Each enclosure has 2 HSZ80 controllers and 14 disks, < > :> arranged as 7 mirrors.  Each mirror in the 2 enclosures' > :> is also shadowed back at the host.  > :>   > :> So, the configuration is  > :>  # > :> 1 Shadow Set = 2 Mirrors   and  > :> 1 Mirror     = 2 Disks  > :>  : > :> The system is supposed to be wired in such a way that= > :> if either ES40 is down (or crashes) no storage redudancy  > :> will be lost. > :>  = > :> That is, all disks in both enclosures should be visible, 6 > :> available and mountable to the remaininging ES40. > :>  > > :> But, what we see is that when one system is down, we lose: > :> access to all  the disks in an enclosure, essentially/ > :> losing half of all our storage redundancy.  > :>  > > :> I think our problem is cabling between the HSZ80s and the? > :> ES40s.  Presently, I see that both SCSI cards on each ES40 ? > :> are cabled to the same enclosure (one to Controller A, the = > :> other to Controller B, which A and B jumpered together).  > :>   > :>  > > :> I've been through all the hardware documenation that came > :> with the RAID array:  > :>  9 > :>  STORAGEWORKS HSZ80 Array Cpntroller ACS Version 8.3 D > :>  Compaq StorageWorks UltraSCSI RAID Enclosure (DS-BA370 Series); > :>  RAID8000 and EAS 12000 Storage Subsystem User's Guide * > :>  Digital StorageWorks EK-HSXSW-WC B01 > :>  @ > :> ...but I can't find a definitive answer for what the proper= > :> cabling between the HSZs and the SCSI cards in the ES40s @ > :> should be so that I don't lose an entire array when an ES40= > :> is down.  I can't even say for sure if this is possible.  > :>  > > :> Can anyone here point me to another DEC/Compaq/HP public-A > :> ation that cover this?  I've already tried Ask The Wizard on ! > :> the HP website, to no avail.  >    ------------------------------  % Date: Tue, 10 Jun 2003 23:07:19 -0000 ! From: Z  <zarlenga@conan.ids.net> * Subject: Re: HSZ80 / ES40 cabling question/ Message-ID: <vecp577jgnt8f8@corp.supernews.com>    Dirk Munk <munk@home.nl> wrote: I : In your original message you wrote that both controllers are 'jumpered  Q : together', I took that as a SCSI T-connector between both controllers. If that  * : is the case, then you have one SCSI bus.   I see; yes that's the case.     E :> The docs I read say that MultiBus Failover is ONLY available with  B :> OVMS 7.2 or greater.  We're on 7.1-2.  They discuss "TrasparentA :> Failover," (OVSM 6.2 or greater) and I think that's what we're  :> supposed to be setup for.  M : True, and upgrading VMS would be very sensible. Your present version is no  S : longer officially supported AFAIK. I would have to read the manuals, but I don't  6 : think transparent failover is a good option for you.  : Impossible to upgrade.  The client won't allow it and even8 if they would, we'd have to also upgrade the DB and that9 vendor no longer supports or even sells upgrades for VMS.   0 So, for better or worse, we're on 7.1-2 to stay.    Q : Sounds like a kind of Unix 'cluster' setup. Isn't the software cluster aware ?  P : In other words, can't you run the application on two nodes as in a 'real' VMS  : cluster ?   7 Correct.  The s/w is not cluster aware.  This is an old 9 TP system migrated to a very simple cluster that gives it 7 CPU and storage redundancy.  The system was delivered 9  years ago, IIRC.  9 Thank you for you help and information so far.  I'll have  more questions soon, I think.    ------------------------------  # Date: Wed, 11 Jun 2003 00:40:35 GMT  From: Rob.Buxton@wcc.govt.nz* Subject: Re: HSZ80 / ES40 cabling question$ Message-ID: <3ee67879.16742046@news>  F On Tue, 10 Jun 2003 23:07:19 -0000, Z  <zarlenga@conan.ids.net> wrote:    >Dirk Munk <munk@home.nl> wrote:J >: In your original message you wrote that both controllers are 'jumpered R >: together', I took that as a SCSI T-connector between both controllers. If that + >: is the case, then you have one SCSI bus.  >  >I see; yes that's the case. >  > F >:> The docs I read say that MultiBus Failover is ONLY available with C >:> OVMS 7.2 or greater.  We're on 7.1-2.  They discuss "Trasparent B >:> Failover," (OVSM 6.2 or greater) and I think that's what we're >:> supposed to be setup for.  > N >: True, and upgrading VMS would be very sensible. Your present version is no T >: longer officially supported AFAIK. I would have to read the manuals, but I don't 7 >: think transparent failover is a good option for you.  > ; >Impossible to upgrade.  The client won't allow it and even 9 >if they would, we'd have to also upgrade the DB and that : >vendor no longer supports or even sells upgrades for VMS. > 1 >So, for better or worse, we're on 7.1-2 to stay.  >  > R >: Sounds like a kind of Unix 'cluster' setup. Isn't the software cluster aware ? Q >: In other words, can't you run the application on two nodes as in a 'real' VMS   >: cluster ? > 8 >Correct.  The s/w is not cluster aware.  This is an old: >TP system migrated to a very simple cluster that gives it8 >CPU and storage redundancy.  The system was delivered 9 >years ago, IIRC.  > : >Thank you for you help and information so far.  I'll have >more questions soon, I think.  ? I'm struggling with my memory here. But we had a cluster of 2 x , Alpha2100s connected to two HSZ Controllers.  E At the Alpha there was a Y cable, the connection then ran to the HSZ, B I think there was a connector there that allowed a second cable toF access the same port or maybe we just used the B Port. But, that cable7 then went to the second Alpha which also had a Y cable.   C This was repeated for the second controller and HBVS used to shadow  the devices.E If either Alpha went down, all disks were still visible. The Y Cables D at the ALpha ends were required in order to maintain the SCSI Bus if# an Alpha got shutdown / turned off.   @ From what I can see from the correspondence, there's no way thatC you'll NOT lose connection to one side of the storage when an Alpha  gets shut down. ! Probably doesn't help you though!    ------------------------------  % Date: Tue, 10 Jun 2003 20:33:08 +0100 * From: "John Travell" <john@travell.uk.net>= Subject: Independant VMS crashdump analysis service launched. 5 Message-ID: <bc5bpn$fkobb$1@ID-120847.news.dfncis.de>    Folks,  I We had a bit of discussion about it six months ago, and for a while I did E nothing about it. Having discovered the hard way just how bad the job J situation really is, and with an unbelievable amount of practical help andJ support from a number of people who had been total strangers up to then, I have got it all together.e  H As the title says, I can now offer an independant VMS crashdump analysis service.  K The initial goal is to fill in the gap in what HP offers, but in practice IaH will work any crash on any version that someone is willing to pay me to.K What is the gap in HP's offering ? They cover both current and PVS versionsiF if you have a suitable contract, and will offer chargeable analysis onJ current versions only. The policy for non-contract PVS crashes was always,L and still remains today, either upgrade and reproduce or buy a PVS contract.  K With me you do not need to buy any contract. I will work by the hour, or by I the day. I am happy to place a spending cap on any crash, in which case IMG report what I have found when the limit is reached, even if there is nol  solution available by that time.  J Quality of service ? During my many years at Digital, and Compaq, I made aL specialisation of doing VMS crash analysis and became one of the most expertK in the UK. By the time I was made redundant there was only one other persondH in the UK customer services with a comparable level of skill. All of theE people I learnt my trade from had moved on to other things, some intod% engineering, some to other companies. * With the service I offer you get the best.  G Today I only have an Alpha to work with, but I can get hold of a VAX at ' short notice if the demand warrants it.   I If you are interested in what I can offer you, please look up my businesst website.L http://www.jomatech.com/ and give me a call, email or landline, where we can discuss the details.     -- John Travell  VMS crashdump expertise for hire john@jomatech.coms http://www.jomatech.com/       ---t& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.484 / Virus Database: 282 - Release Date: 27/05/2003e   ------------------------------  % Date: Tue, 10 Jun 2003 16:41:41 -0400d! From: Jim Agnew <jpagnew@vcu.edu>nA Subject: Re: Independant VMS crashdump analysis service launched.e' Message-ID: <3EE64285.998BD786@vcu.edu>i  - Best to you, John...  I'll keep you on file..r  	 Jim Agnewe   John Travell wrote:s >  > Folks, > K > We had a bit of discussion about it six months ago, and for a while I dideG > nothing about it. Having discovered the hard way just how bad the jobrL > situation really is, and with an unbelievable amount of practical help andL > support from a number of people who had been total strangers up to then, I > have got it all together.a > J > As the title says, I can now offer an independant VMS crashdump analysis
 > service. > M > The initial goal is to fill in the gap in what HP offers, but in practice IeJ > will work any crash on any version that someone is willing to pay me to.M > What is the gap in HP's offering ? They cover both current and PVS versions0H > if you have a suitable contract, and will offer chargeable analysis onL > current versions only. The policy for non-contract PVS crashes was always,N > and still remains today, either upgrade and reproduce or buy a PVS contract. > M > With me you do not need to buy any contract. I will work by the hour, or bypK > the day. I am happy to place a spending cap on any crash, in which case IeI > report what I have found when the limit is reached, even if there is no0" > solution available by that time. > L > Quality of service ? During my many years at Digital, and Compaq, I made aN > specialisation of doing VMS crash analysis and became one of the most expertM > in the UK. By the time I was made redundant there was only one other personpJ > in the UK customer services with a comparable level of skill. All of theG > people I learnt my trade from had moved on to other things, some intoc' > engineering, some to other companies. , > With the service I offer you get the best. > I > Today I only have an Alpha to work with, but I can get hold of a VAX ats) > short notice if the demand warrants it.t > K > If you are interested in what I can offer you, please look up my businessm
 > website.N > http://www.jomatech.com/ and give me a call, email or landline, where we can > discuss the details. >  > -- > John Travell" > VMS crashdump expertise for hire > john@jomatech.comn > http://www.jomatech.com/ >  > ----( > Outgoing mail is certified Virus Free.< > Checked by AVG anti-virus system (http://www.grisoft.com).C > Version: 6.0.484 / Virus Database: 282 - Release Date: 27/05/2003l   -- uF "4,000 years ago I made a mistake."  Elrond Half-Elven, in "Fellowship of the Ring"   ------------------------------  # Date: Tue, 10 Jun 2003 19:10:35 GMT & From: Rick Jones <foo@bar.baz.invalid>8 Subject: Re: Intel 64 bit Pentium seems to be on its way1 Message-ID: <L2qFa.2276$0W.1064@news.cpqcorp.net>p  P Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote: > Rick Jones wrote: , >> Bill Todd <billtodd@metrocast.net> wrote: >> yC >>>It won't tomorrow: the 1.8 GHz Opteron already beats the 1.5 GHz-% >>>Itanic (Madison) in SPECweb99_SSL,3 >> 7 >> 0F >> The published four-CPU configuration with 1.8 GHz Opteron beats theE >> published four-CPU configuration of the 1.5GHz Itanium 2 6M in the A >> rx5670.  The published two-CPU configuration does not beat thet1 >> published two-CPU configuration of the rx2600.  >>    / > I would find that very alarming if I was you.   > I am perfectly content with your being alarmed :) I was simplyE clarifying Bill's statements about the present state of SPECweb99_SSL  results.  
 rick jones -- u3 speak carefully and please avoid feeding the trollseF these opinions are mine, all mine; HP might not want them anyway... :)A feel free to post, OR email to raj in cup.hp.com  but NOT BOTH...l   ------------------------------  % Date: Tue, 10 Jun 2003 17:36:35 -0400 * From: "Bill Todd" <billtodd@metrocast.net>8 Subject: Re: Intel 64 bit Pentium seems to be on its way2 Message-ID: <m6KcnX6kCZbozXujXTWcqg@metrocast.net>  3 "Rick Jones" <foo@bar.baz.invalid> wrote in messagey+ news:L2qFa.2276$0W.1064@news.cpqcorp.net...4K > Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com>n wrote: > > Rick Jones wrote:l. > >> Bill Todd <billtodd@metrocast.net> wrote: > >>E > >>>It won't tomorrow: the 1.8 GHz Opteron already beats the 1.5 GHzl' > >>>Itanic (Madison) in SPECweb99_SSL,  > >> > >>H > >> The published four-CPU configuration with 1.8 GHz Opteron beats theG > >> published four-CPU configuration of the 1.5GHz Itanium 2 6M in the1C > >> rx5670.  The published two-CPU configuration does not beat theD3 > >> published two-CPU configuration of the rx2600.  > >> >01 > > I would find that very alarming if I was you.  >L@ > I am perfectly content with your being alarmed :) I was simplyG > clarifying Bill's statements about the present state of SPECweb99_SSLe
 > results.  K And you were entirely correct to do so:  I had forgotten the dual-processortK results, and while Madison doesn't enjoy a *significant* lead there it *is*u a lead.l  I On the other hand, if Opteron actually does field its 2 GHz version about24 when Madison ships, that lead will likely evaporate.  F On the third hand, if Opteron *doesn't* field that 2 GHz version soon,F people (including myself) will need to start reducing their short-termJ expectations for the platform:  the fact that it's available, inexpensive,J and performing extremely well is great, but if AMD doesn't get the processK whipped into shape such that they can soon realize the potential clock-rateUL increases at very reasonable power levels then its continued ability to whup/ ass in the immedate future will be compromised.    - bill   ------------------------------    Date: 10 Jun 2003 12:55:08 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)o- Subject: Re: Java Runtime.exec(String) on VMSd3 Message-ID: <lnVdhLZjrJwT@eisner.encompasserve.org>t  S In article <vebv045efstoa0@corp.supernews.com>, "Tim" <tim_a_b@hotmail.com> writes:g > 4 > Unfortunately, when I try to call a com file using/ >     Runtime.getRuntime().exec("showlogiris");- > I get the following error:B >     java.io.IOException: Child creation error: exec format error  ?    I believe exec() has no idea what "showlogiris" is.  Are you A    setting DCL$PATH in the job logical name table or something sooG    that exec() can find it?  Have you tried an explicit path and addingdD    the .com extension in the same attempt?  The release notes claims>    that the simple .com file name should be enough, can you doC    System.err.println(System.getProperty("user.dir")) to make sure a>    you're working from the directory you put the .com file in?  D    Can you post what you tried with the Release Notes?  These appearC    not to have changed since at least 1.3.1, but it's worth lookingdF    for the 1.1.8 Release Notes to see if they are different.  AlthoughC    the Java Runtime.getRuntime().exec() hasn't changed, the VMS JRE G    has and an older implementation may have provided less of the exec()oI    functionality.  This is an area where the underlying VMS implemenation H    is very different from the UNIX implementation.  IIRC the VMS C RTL'sB    exec() wouldn't do all the things the JRE's exec() does, so the%    JRE has to make up the difference.y   ------------------------------  # Date: Wed, 11 Jun 2003 01:30:48 GMT ; From: "John Gemignani, Jr." <jon-nope@thiswontworkossc.net>w- Subject: Re: Java Runtime.exec(String) on VMSn< Message-ID: <cDvFa.3326$Jw6.2369335@news1.news.adelphia.net>  , "Tim" <tim_a_b@hotmail.com> wrote in message) news:vebv045efstoa0@corp.supernews.com...e	 > Hi all,c >mG > I have written a socket program in Java for our VMS box. Basically it0 allowsH > a client to open a socket from another machine and instruct the server > socket to run ".com" files.a >04 > Unfortunately, when I try to call a com file using/ >     Runtime.getRuntime().exec("showlogiris");  > I get the following error:B >     java.io.IOException: Child creation error: exec format error >uJ > The showlogiris.com file is in the same directory as the java class thatK > opens the server socket and I have verified that the showlogiris.com filei1 > works ok by running it from the command prompt.e >aG > I have followed the instructions in the following web-page, but to nox avail: >av http://h18012.www1.hp.com/java/documentation/1.4.1/ovms/docs/release_notes.html#UsingtheRuntimeexeMethodonOpenVMSAlpha >-L > I am using jdk1.1.8, which is very old, but the Runtime class has not been@ > changed since 1.0, so that should not be a problem, should it? >s> > If anybody can help me on this one I would be very grateful, >- > Kind regards,6 >7 > TimL >S >F   Tim:J     Command procedures (scripts) don't run on VMS by juse mentioning theirH name.  They are not "executable" per se.  You may have to invoke it with "@showlogiris".H   -Johnr   ------------------------------  % Date: Tue, 10 Jun 2003 22:02:57 -0400t  From: John Santos <JOHN@egh.com>- Subject: Re: Java Runtime.exec(String) on VMSg5 Message-ID: <1030610215802.2699A-100000@Ives.egh.com>b  / On Wed, 11 Jun 2003, John Gemignani, Jr. wrote:o   > . > "Tim" <tim_a_b@hotmail.com> wrote in message+ > news:vebv045efstoa0@corp.supernews.com...s > > Hi all,w > >oI > > I have written a socket program in Java for our VMS box. Basically ita > allowsJ > > a client to open a socket from another machine and instruct the server > > socket to run ".com" files.  > >r6 > > Unfortunately, when I try to call a com file using1 > >     Runtime.getRuntime().exec("showlogiris");n > > I get the following error:D > >     java.io.IOException: Child creation error: exec format error > >uL > > The showlogiris.com file is in the same directory as the java class thatM > > opens the server socket and I have verified that the showlogiris.com filea3 > > works ok by running it from the command prompt.i > >II > > I have followed the instructions in the following web-page, but to noi > avail: > >Lx > http://h18012.www1.hp.com/java/documentation/1.4.1/ovms/docs/release_notes.html#UsingtheRuntimeexeMethodonOpenVMSAlpha > >=N > > I am using jdk1.1.8, which is very old, but the Runtime class has not beenB > > changed since 1.0, so that should not be a problem, should it? > >r@ > > If anybody can help me on this one I would be very grateful, > >j > > Kind regards,l > >c > > Timn > >c > >P >  > Tim:L >     Command procedures (scripts) don't run on VMS by juse mentioning theirJ > name.  They are not "executable" per se.  You may have to invoke it with > "@showlogiris".w  F That depends on whether or not DCL$PATH is defined.  Maybe you have itI defined as a process logical name, and it isn't getting handed off to therG subprocess created by exec()?  [This assumes a lot: that exec() createsaA a subprocess; that the process logicals aren't passed to it; that D DCL$PATH is defined, so you don't need to say "@showlogiris" in DCL;F that DCL$PATH is a process, not job nor group nor system logical; ...]   HTH.   >  > -John    -- c John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Tue, 10 Jun 2003 14:23:15 -0400 ' From: "Fastload" <fastload@hotmail.com> E Subject: Re: JF Mezei "flamed, insulted, denigraded(sic)"! Poor baby!e5 Message-ID: <nmpFa.382$eR5.2254@wagner.videotron.net>   	 Coq-Ring?r  ; "Anders" <dud@nilsson.com> a crit dans le message de news:u 3EE5D85B.2911@nilsson.com... > Jarod wrote: > >c
 > > Coq-Ombrea > >>	 > > JarodR >H > Coq au vin >c	 > /Anderse >  >h >l > >c< > > "Poor Baby JF" <poor@baby.jf> a crit dans le message de7 > > news:YM7FVA7V37780.9603703704@Gilgamesh-frog.org...n1 > > > JF Mezei <jfmezei.spamnot@istop.com> wrote:2 > > >CH > > > >> > Lighten up man.  I had a smiley after that comment.  It was a poke atn > > your > > > >> > politics anyway.u > > > >IA > > > >I have been flamed, insulted, denigraded enough on the net  > > >  > > > Oh you poor baby!r > > >.G > > > Don't worry, your mommy will be along any minute now to take your, rectalF > > > temperature and give you your suppository so you'll feel better. > > >o > > >C >  > --9 > Remove the obvious part before replying by mail please!o   ------------------------------  % Date: Tue, 10 Jun 2003 11:30:34 -0700b. From: "David D Miller" <ddmiller@raytheon.com>/ Subject: re: mail to news gateways for OpenVMS?tF Message-ID: <OF0D032785.6475384A-ON07256D41.00659B04@rsc.raytheon.com>   John  8 Check out MX at ftp://ftp.madgoat.com/mx/mx042/mx042.zip   dave.f  H ----- Forwarded by David D Miller/RWS/Raytheon/US on 06/10/2003 11:29 AM -----o  J I am looking for information on bidirectional news to e-mail gateways that will run on OpenVMS.  E What I need is something that can be used to gate a newsgroup that isYC being heavily spammed in to a mailing list that would be spam free.k  6 Naturally I do not want the spam to be passed through.  - So idealy the gateway could do the following:d  : Pass through newsgroup posters that are known to be valid.  C Pass through newsgroup posters that put a special codeword in theirn message.  C Reject postings that the I.P. address of an open proxy was found in-I the headers, or in certain DNSbls, or from news servers that allow abuse.d  = Allow some customization of blocking based on header content.   F Optionally flag unknown posters for moderator aproval before relaying.  K Open source prefered - This is a hobby operation so the price must be free.    -Johnv wb8tyw@qsl.network Personal Opinion OnlyM   ------------------------------  % Date: Tue, 10 Jun 2003 20:56:22 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>e+ Subject: Re: Memory limitations for Processa' Message-ID: <3EE68C46.E6E093D9@fsi.net>-   labadie wrote: > B > "Sandeep Yelwatkar" <Sandeep_Yelwatkar@bmc.com> wrote in message+ > news:veb3v1e8434333@corp.supernews.com...b > >$ > > Hi All,i > >nL > > I have few queries regarding the parameters associated with processes on > VMSoM > > systems. I would like to know if there is any limitations on memory usager > ofN > > the process running on VMS system. If yes, then how do I know it. How do IC > > know the memory usage of the process and available free memory.  > >oN > > If there is any limitations on the memory usage of the process then how do > I0 > > increase this memory size. > >h/ > > Any help in this regard is most appreciatedS > >d
 > > Thanks > > Sandeepp > >f > >0 >  > Hellor > - > You should use Amds or Availman to see thisl@ > http://h71000.www7.hp.com/openvms/products/availman/index.html  B See also HELP Lexical F$GETJPI(). You'll want to compare WSSIZE toG WSQUOTA and WSEXTENT, and compare PAGFILCNT to PGFLQUOTA. Note that youlE need GROUP privilege to access information about processes outside of A your own UIC within your UIC group, and WORLD privilege to accessw: information about processes outside of your own UIC group.   --   David J. Dachtera  dba DJE Systemso http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/S   ------------------------------  % Date: Tue, 10 Jun 2003 13:46:09 -0400Z! From: Jim Agnew <jpagnew@vcu.edu>l$ Subject: Re: need kbd for VAX4000/90' Message-ID: <3EE61961.24ED3BE3@vcu.edu>o   Bob Koehler wrote: > a > In article <CIEJLCMNHNNDLLOOGNJIMEPMHFAA.tom@kednos.com>, "Tom Linden" <tom@kednos.com> writes:i > > Pebble Beach CA  > >-F > >>From: Bob Koehler [mailto:koehler@eisner.nospam.encompasserve.org] > >>E > >>   I know where you can buy some, cheap.  What part of the globe?e > D >    I suspect they wouldn't be cheap if you had to include airfare.  F i have some lk201's if they would do...  just give me a ups code..  or( your version of a shipping turtle..  ;-)   --  F "4,000 years ago I made a mistake."  Elrond Half-Elven, in "Fellowship of the Ring"   ------------------------------  % Date: Tue, 10 Jun 2003 20:11:07 +0200h) From: Mikkel Lauritsen <renard@nospam.dk>b$ Subject: Re: need kbd for VAX4000/90; Message-ID: <P6pFa.15398$Jp3.572458@news010.worldonline.dk>e   Tom Linden wrote:   2 > Well, sometimes you get sloppy;-(  Lost an LK2015 > anybody got a spare or anything that will work withc= > this VAX?  It has that small (handset?) modular style jack.  >  > I think 401s work too.  
 Me too :-)  C I'm trying to resurrect a VaxStation 4000/90 that did not come withr@ keyboard nor mouse, so I'm looking for an LK401 (even 201) and a4 VSXXX - preferably in Europe (Denmark) where I'm at.  0 Any pointers would be much appreciated - thanks.      Mikkel Lauritsenr   ------------------------------    Date: 10 Jun 2003 15:06:23 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)A0 Subject: Re: OT: CNN Story Not Favorable To Bush3 Message-ID: <lNw5rDL8boSq@eisner.encompasserve.org>f  q In article <zdTg9IgdspeJ@eisner.encompasserve.org>, koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:   < >   The world is free to disagree.  Freedom is a good thing.  * Not in misusing the comp.os.vms newsgroup.   ------------------------------  % Date: Tue, 10 Jun 2003 17:49:10 -0400 * From: "Bill Todd" <billtodd@metrocast.net>0 Subject: Re: OT: CNN Story Not Favorable To Bush2 Message-ID: <Yv-cnYqd0OTFz3ujXTWcpQ@metrocast.net>  H "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in message- news:zdTg9IgdspeJ@eisner.encompasserve.org...e@ > In article <veCcnbvZyp0RU3mjXTWcqQ@metrocast.net>, "Bill Todd"  <billtodd@metrocast.net> writes: > >nL > > That's technically correct, when you consider the world as a whole:  theL > > general perception was that there was no pressing need to remove him *at	 > > all*., >e< >   The world is free to disagree.  Freedom is a good thing.  J I'd say "amen", but since I'm not religious that could be misleading.  ButI one should also note that freedom carries with it responsibility:  you'resH not 'free' to decide to walk over to your neighbor's house and shoot himK (unless possibly you have caught him aiming a loaded weapon at *your* houseaL and preparing to fire it), and we should not consider ourselves 'free' to do2 the same kind of thing on the international level.   >iJ > > But in the more limited (in multiple meanings) context of the good ol' U.S.K > > of A., the vigorously-asserted presence of WMD *and the imminent threatb ofL > > their use in large quantities* was very definitely the rationale used toF > > sell a somewhat dubious public on the need for going to war *right then*,A > > rather than allowing the U.N. inspection process to continue.l >uJ >    Not entirely relavent since no one believed those claims at the time.J >    We can hardly be expected to hold our breath waiting for something we >    didn't believe in.   K I think you may be giving our fellow citizens a bit too much credit:  whileeK they might not have been willing to bet their farms that the allegation was K true had they been told to put up or shut up, a great many of them appeared'F to find it convenient to accept the assertion at face value (since theK alternative would have required that they actually consider standing up andt% doing something about the situation).    >eK > > Of course, now that the imminence of such a threat is becoming more and: moreE > > ludicrous (and more and more information is coming out of various4D > > intelligence agencies indicating that it was highly questionable *before*G > > the war as well), Dubya and friends are back-pedaling furiously ands seekinga@ > > other rationalizations for why they should not be treated as
 internationalaL > > criminals.  And a lot of 'Muricans seem to be buying it, God bless 'em - but A > > the Brits appear at least a bit less ready to be as cynically  manipulatedn* > > after the fact as they were before it. > C >    The Brits are not and never were the only ones who saw cynicallD >    manipulation.  They do happend to be the most vocal in making aG >    political fight over it right now.  I expect to here more from theeI >    US Democracts as 2004 nears.  The Brits don't have to wait that longn2 >    since they can call for an election any time.  J Our Congress can (and should) consider impeachment and removal at any time$ too.  But I'm not holding my breath.     In the meantime I expectI >    GWB's real economic plan:  cheap oil, to come on too late to fix ther >    economic mess he's made.t >)G >    You can't expect Americans to get all fussed up and righteous whenaJ >    the administraion of the oil men, for the oil men, and by the oil men( >    does exactly as poorly as expected.  F I find myself agreeing with most of your post a great deal more than IJ expected to when I started reading it.  Unfortunately, since the characterK of this Administration was already pretty clear prior to 2002 election, thebK results of that election do not inspire confidence that the problem will betK fixed in 2004 unless people *do* get somewhat more fussed up than they have  been up until now.   - bill   ------------------------------    Date: 10 Jun 2003 17:08:45 -0700: From: craig.berry@SignalTreeSolutions.com (Craig A. Berry) Subject: Re: perl question= Message-ID: <7f15589f.0306101608.33114682@posting.google.com>3  m Chris Olive <nospam@raytheon.com> wrote in message news:<cK7Fa.3506$c6.3314@bos-service2.ext.raytheon.com>...@  * >   There are things in DCL I miss in Perl  0 Just curious.  What things do you miss the most?   ------------------------------  % Date: Tue, 10 Jun 2003 16:53:02 -0500d, From: Gary L. Ross <rossgl@parknicollet.com>( Subject: PIPE output defined to a symbol8 Message-ID: <r5kcevsovpsgdd1vm1u73meo87mvtqebul@4ax.com>  F Has anyone found a way to define a symbol from within the PIPE command without outputing to a file?   Example:  ? $ pipe multinet ping <servername> /num=1 | search sys$pipe datat+ PING servername (10.10.10.1): 56 data bytes   C I get one line of data back that I would like to define to a symbol C so I may strip data out via lexicals for displaying later.  I wouldd0 like to avoid all disk I/O if possible.  Thanks.   Gary L. Ross Park Nicollet Health Servicess  5 Cold, adj:  When your dog sticks to the fire hydrant.u   ------------------------------  % Date: Tue, 10 Jun 2003 21:07:37 -0500r1 From: "David J. Dachtera" <djesys.nospam@fsi.net>b, Subject: Re: PIPE output defined to a symbol' Message-ID: <3EE68EE9.DDAAFE76@fsi.net>    "Gary L. Ross" wrote:a > H > Has anyone found a way to define a symbol from within the PIPE command > without outputing to a file? > 
 > Example: > A > $ pipe multinet ping <servername> /num=1 | search sys$pipe datam- > PING servername (10.10.10.1): 56 data bytes  > E > I get one line of data back that I would like to define to a symbol E > so I may strip data out via lexicals for displaying later.  I wouldV2 > like to avoid all disk I/O if possible.  Thanks.    Use DEFINE/JOB in your pipeline:   $ PIPE SEARCH LOGIN.COM DIR | -w7   ( READ SYS$PIPE TMP ; DEFINE/JOB SEARCH_RESULT &TMP )  $ SH LOG SEARCH_RESULT>    "SEARCH_RESULT" = "%SEARCH-I-NOMATCHES, no strings matched" (LNM$JOB_80D11880)   -- h David J. Dachteran dba DJE Systemsp http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Tue, 10 Jun 2003 13:28:16 -0400o* From: JF Mezei <jfmezei.spamnot@istop.com>& Subject: Re: Portents of Itanium death) Message-ID: <3EE6152E.F6733021@istop.com>s   Keith Parris wrote:S8 > Intel has stated its 64-bit direction, and it's IA64.   L And Compaq had stated that it woudl keep/grow Alpha for the long term. Then,/ they revealed their true intentions on June 25.i  I The fact is that IA64 has not had the "instant success" that one had beenhM primed to expect with all the "commodity"  "industry standard"  sales pitchestK from Carly and Curly. The fact is that Intels *appears* to be struggling to>N make its Ia64 palatable. The fact is that Intel did announce its IA64 wouldn'tL make it to the "commodity, high volume" category. The fact is that there areA persistant rumours that Intel will eventually have a 64 bit 8086.   ? When you combine all that, I see a very cloudy future for IA64.n   ------------------------------  % Date: Tue, 10 Jun 2003 13:19:20 -0400n* From: JF Mezei <jfmezei.spamnot@istop.com>& Subject: Re: Portents of Itanium death) Message-ID: <3EE61317.8700F1DF@istop.com>    "Mark E. Levy" wrote: 6 > You can use your "real" name and munge your address.  E I am starting to think I should have never used my real name. I fully ; undertstand John Smith. I am thinking of becoming John Doe.t  E The idiot posted to newsgroups I don't frequent, including posting mysE telephone number etc. He sent out an "I am about to kill myself" to aeN depression newsgroup, resulting in people calling me (they don't know to check2 mail headers to see that the message is a forgery.   ------------------------------  # Date: Tue, 10 Jun 2003 20:13:19 GMTa9 From: "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> & Subject: Re: Portents of Itanium death1 Message-ID: <zZqFa.2289$071.338@news.cpqcorp.net>e   What a cop out.h  J Use whatever you want for a return address to mask yourself, put your name in plain text in the body.  I My impression is that either A) you are just one of the regulars stirringBK the pot without attribution, or B) there are other reasons why if your name , were known - your opinion would be devalued.  J Heck, at least I know that Bill is a real person - who is just here out ofK some sense of spite that Alpha will eventually go away... or perhaps just atL little disgruntled.  I know Andy's a Sun marketdroid.  You can assume that IL have a bias towards HP and towards VMS considering I work for them - you can; discount *my* opinion based on that - which is your choice.e  J You have added nothing to the dialogue in terms of opinion that differs or5 adds any nuance to several other VMS/HP/IA64 bashers.7  ; "mist dragon" <mistdragon@zdnetonebox.com> wrote in messageZ7 news:7500353b.0306092202.4afd8559@posting.google.com...o> > Paddy O'Brien <paddy.o'brien@tg.nsw.gov.au> wrote in message( news:<3EE510A4.2070405@tg.nsw.gov.au>... > > Tsk, Tsk wrote:d > > > Fred Kleinsorge wrote: > > >IJ > > >>Come on.  The "analysts" here in c.o.v. seldom disclose their names, or whatS3 > > >>their financial interest is in spreading FUD.c > > >d > > >aH > > > I would say that the majority of the pundits here have real names. Those.- > > > without fake names are in the minority.O > >. > > This does not make sense.s > > L > > I would assume that everyone here has a real name whether they use it or > > not. > >rE > > I disagree with both writers in that I think the majority on thishK > > newsgroup/mailing list -- "pundits" or "analysts" or whatever -- do usedL > > their real names, even to the extent of accepting abuse.  There are onlyI > > a handful of *real* contributors here who do not use their real namesrK > > and "John Smith" has explained his own reasons and those might apply tooK > > others in this handful.  Apart from this handful, the other posters who * > > use the cloak of anonimity are trolls. > >' >oG > Now, how do you define a real name ? If you write down something elseoD > than John Smith, picking it from telephony directory and put it toG > hotmail or set up a website for it, does it mean its a real name ? Or H > if you steal an address from 'net ? How does one prove its correct one > ?f >aG > In my view the anyonymous name is to protect its owner from spam. Not-H > that I would not trust you guys/gals :) But robots scan the usenet andH > find new spam targets and less you give up your real name and address,, > the less spam you have. As simple as that. >:H > Other reasons may be organizational, but as for me, I am not employeedA > by CPU manufacturers or vendors. I am not an analyst or venture C > capitalist. I have no monetaory or other benefits of nailing downeD > Itanium, VMS, Athlon 64/Opteron etc. I hold no stakes or shares onB > AMD, Intel or other computer companies nor do I work for one. As, > usual, you have to take my word on this :) > H > My motives are to observe the market as I have in the past 25 years inG > this business and point out things and see other people's opinions onp@ > these and expand my visions. Would I reveal myself, I would be@ > targeted on all the nasty downsides of communication, spam andF > personal attacks, so I would like to remain in my insignificant mist > :) >t > Mist   ------------------------------  # Date: Tue, 10 Jun 2003 20:30:28 GMT'9 From: "Fred Kleinsorge" <my-last-name@stardotzko.dec.com>r& Subject: Re: Portents of Itanium death1 Message-ID: <EdrFa.2292$Qa1.767@news.cpqcorp.net>f  5 "David Froble" <davef@tsoft-inc.com> wrote in messageS& news:3EE5565C.8060703@tsoft-inc.com... > Fred Kleinsorge wrote: >  >rG > Oh yes they did!  I even remember the 'promise' of a $1000 system.  I  suggestVH > you back off such revisionist history because a few searches of Google	 will bury  > you! >a  J To my knowledge none of my bosses has ever said this was a primary motive.K Secondary at best.  Aspirational.  You are taking a small thing and holdingn it up as the main thing.   > >rK > > Binary compatability with IA32 is a real issue in the x86 space, not anaL > > insurmountable one - but a tough one.  But frankly, the AMD x86-64 isn'tJ > > going to do squat in the general purpose IA32 space - at least running in > > 64-bit mode. >a >w3 > I disagree.  It will do one thing.  It will sell!e >s  L Well, it's possible.  We will see won't we.  I'm waiting for a true top tier- company to announce a major comittment to it.d  ? > >  It isn't transparent either.  Win64 instead of Windows XP. F > > New binaries for anything that needs to use the 64-bit space.  And franklyqG > > not really enough justification for the bulk of the market.  Sure -d they'll  > > sell some.  But not a lot. >p >eI > Microsoft Word and Excel IS the bulk of the market.  Well, Ok, solitarer also.nG > :-)  Microsoft seems to be supporting AMD, and if they do, the 64 bit- binaries@ > will appear.  Not saying they are required.  I doubt they are. >   G Win64 isn't needed for Word, or Excel.  I don't think a faster CPU will F speed Solitaire.  You need a killer app that will make it interesting.H Databases *are* a potential - but who has signed up to build and support- systems that scales past the promised 4-ways?o  H Opteron reminds me of the DEC/Alpha mentality - if we build it they willI come.  It's nice technology - but is it fast enough and cheap enough thath? people will abandon IA32 for it?  Or abandon Sun/HP/IBM for it?r   > D > > In general AMD has a nifty piece of technology, but is in a weak
 finiancialK > > and market share position.  Alpha was a nifty piece of technology - and  it > > lost in the market place.i >t > I > No, it lost because upper management at Compaq didn't want to be in theo CPUoK > business.  I cannot agree that Alpha lost in the market place, unless you  define( > success as being a dominant commodity.  J Alpha maybe made the 3rd tier in the marketplace.  It was consistantly outD in front in performance - but not by enough.  If cool technology andA performance was enough - then Tru64/Alpha should have buried Sun..  H Don't get me wrong, I loved Alpha.  I've got time and energy invested inF understanding a lot of nitty gritty things about the architecture, theK firmware, and how VMS works on it.  But I loved VAX too.  You gotta let go.g  ( >  IA-64 isn't and won't be that either.G > Let's try to judge them both from the same perspective, not re-defineB	 'success'I; > in a manner to allow IA-64 to succeed where Alpha failed.U >u  J IPF is way early in it's lifetime to either judge it's success or failure.L CS people hate it, because they hate it's instruction set and design.  OtherJ 64-bit manufacturers obviously want it to fail.  But it is plugging ahead,= getting faster, and picking up support from a lot of players.h   ------------------------------  # Date: Tue, 10 Jun 2003 20:51:23 GMT # From: "John Smith" <a@nonymous.com>n& Subject: Re: Portents of Itanium deathH Message-ID: <fxrFa.52410$j9%.31978@news04.bloor.is.net.cable.rogers.com>  D "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> wrote in message+ news:zZqFa.2289$071.338@news.cpqcorp.net...u >iA > You have added nothing to the dialogue in terms of opinion that(
 differs or7 > adds any nuance to several other VMS/HP/IA64 bashers.l    E That's the same attitude that Bush and the Republicans use to dismissu9 legitimate protest about their policies as 'unpatriotic'.   < Just because you may not agree with the opinion, it does notE automatically make the opinion invalid. And just because you may hear'D the same opinion voiced by others, it doesn't mean that they are allB wrong. They may well be in the fullness of time, but you have no aE priori knowledge of that unless you are a true insider, and even thenj the market may prove you wrong.    ------------------------------  % Date: Tue, 10 Jun 2003 17:28:27 -0400 * From: "Bill Todd" <billtodd@metrocast.net>& Subject: Re: Portents of Itanium death2 Message-ID: <vWudnd0gVIvj0HujXTWcpQ@metrocast.net>  D "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> wrote in message+ news:EdrFa.2292$Qa1.767@news.cpqcorp.net...  > 7 > "David Froble" <davef@tsoft-inc.com> wrote in message ( > news:3EE5565C.8060703@tsoft-inc.com... > > Fred Kleinsorge wrote: > >  > > I > > Oh yes they did!  I even remember the 'promise' of a $1000 system.  I 	 > suggesteJ > > you back off such revisionist history because a few searches of Google > will buryw > > you! > >l >eL > To my knowledge none of my bosses has ever said this was a primary motive.E > Secondary at best.  Aspirational.  You are taking a small thing and  holding. > it up as the main thing.  I So what *was* the motive, Fred - if not the (supposed) economies of scalei. that would produce far less expensive systems?  I Performance?  The most that has *ever* been claimed (well, by anyone withDJ any credibility) is that Alpha's performance lead would have been reduced,K not that Itanic would have ever become significantly superior (Madison willnH perform better than EV7 in several - but hardly all - areas, but had EV8G been completed it would have jumped *well* out in front again next year J despite several years' worth of neglect and delays for both EV7 and EV8 atI Compaq's hands).  So if Itanic enjoyed no significant cost advantage overnE Alpha, it would hardly have been any motive at all for breaking one's G promises to customers and eliminating the platform they preferred - and J since there's no evidence that Intel had any idea whatsoever where to takeH Itanic after the McKinley core (that's what the transplanted EV8 team isE doing), while Alpha had a convincing roadmap far into the future, themK likelihood is that Alpha's superiority would only have increased over time.i  H Multiple OS support (Windows, Unix, Linux, high-end systems like VMS andG NSK)?  Alpha had all but the last years ago (until Curly canned Windows E development on Alpha in 1999, which if continued would have yielded awJ working Win64 platform on Alpha a full year ahead of Merced and a full twoL years ahead of the first actually usable Itanic), and would have had NSK too' at least as soon as Itanic will get it.s  F Development cost?  We've been through that issue thoroughly.  MarcelloH pegged Alpha annual development costs at $150 million at the time of theK Alphacide - presumably split between EV7 and EV8.  Winkler's number of $300wG million clearly included a lot more than just chip development - likelySI server and possibly software development as well, as supported by careful8L reading of the 1999 and 2000 Compaq annual report's comments about projectedE R&D funding (about $500 million annually) and what the biggest singlenD recipient was (storage).  And anyone who supposed that the AlphacideL wouldn't have much effect on profits was an utter imbicile:  Marcello statedL VMS's annual revenue/profit as $4 billion/$800 million in mid-2000, a March,K 2001 slide presentation repeated the $4 billion figure (plus $3 billion foruH Tru64), and the post-Alphacide (December, 2001) official response to theK negative Gartner comments about VMS indicated that revenues had been halved L (to $2 billion annually).  It's nice that they've apparently rebounded a bitF (to $2.5 - $3 billion, with $500 million profit) more recently (recentH statements allegedly from Mark Gorham reported here just today), but theC difference in VMS profits alone *clearly* would have cost-justifiediE completing EV8 and *then* seeing whether Itanic provided a reasonablenH alternative (or whether continuing to compete would be more profitable).  K I've got a suggestion or two for what the real motives behind the Alphacide  were:   H 1.  Curly's fear for his job.  The radical change in corporate directionG when Pfeiffer got ousted suggests that Ben Rosen & Co. wanted Compaq toPK return to its box-shuffling roots rather than embrace Alpha as Pfeiffer wasgI doing.  So courageous Curly bent over and said "Yes, Sir!" rather than do-G his job and educate the BoD about what was profitable and what was not.n  L 2.  Robison clearly stated that the first thing he did after coming on boardJ was discuss his own doubts about Alpha with Curly.  And since Curly to allE appearances hadn't a clue about this industry, he was probably easilyeK influenced (even though by around that time Ben Rosen was in the process ofr departing).u  I 3.  The merger.  By the time the Alphacide actually occurred, clandestineeA merger discussions were already well under way.  Even the current K Administration might have felt obligated to take a look at anti-competitiveeG aspects had Alpha still been actively competing with HP's child prodigy(L Itanic.  And Curly needed the merger to (once again) save his job, after two# years of dismal performance as CEO.a   ...e  J > Opteron reminds me of the DEC/Alpha mentality - if we build it they will > come.e  K And you know what?  They DID, despite being 'affinitied' away from its mainUK OS, despite later neglect of Alpha as a whole, despite (justified) concernspF about DEC's future, despite continuing neglect after Compaq took over:K through all of that, Alpha remained profitable and maintained the potentialnK for *real* market penetration if only its vendor could be perceived as bothr competent and committed.  D   It's nice technology - but is it fast enough and cheap enough that" > people will abandon IA32 for it?  I Earth to Fred:  one of the greatest advantages of AMD64 is that customersgL don't *have* to abandon IA32 for it, because not only is it just as fast and7 just as cheap as IA32 but it's 100% compatible with it.     >  Or abandon Sun/HP/IBM for it?  I Some people will, others won't.  Some people won't change platforms untillJ they're pried out of their cold, dead hands.  But when people *do* change,A they examine the available alternatives, and from all *technical* H perspectives in the system ranges that Opteron covers it looks damn goodH (very competitive in performance, even more so in power consumption, andH except for IA32 there's no competition whatsoever with it in cost).  TheG main (and legitimate) concerns one might have about Opteron involve thesF health of its owner, and there's indeed something of a chicken-and-eggA problem there (which IBM could and may yet do a lot to overcome)..   ...u  J > Don't get me wrong, I loved Alpha.  I've got time and energy invested inH > understanding a lot of nitty gritty things about the architecture, theI > firmware, and how VMS works on it.  But I loved VAX too.  You gotta leti go.y  L But when you're *made* to let go against your wishes and without good reasonI (let alone actively lied to in the process), that fact is likely to enterh# into subsequent decisions you make.-   - bill   ------------------------------  % Date: Tue, 10 Jun 2003 17:17:35 -0400n" From: JF Mezei <jfmezei@istop.com>& Subject: Re: Portents of Itanium death) Message-ID: <3EE64ADE.FBA3DEE7@istop.com>e   Fred Kleinsorge wrote:N > Well, it's possible.  We will see won't we.  I'm waiting for a true top tier/ > company to announce a major comittment to it.g  K Are there top tiers who have made public announcements that they are buying J chips from AMD ? Does Dell make any wintels based on AMD chips or is it an all-intel shop ?  E Will AMD make a huge splash and significantly erode Intel's dominance(J overnight ? Perhaps not. But if AMD's wintel products are competitive withN those of Intel for the wintel market, then it doesn't matter if the chip is 64" bit or 32 bits, at least initally.  E If the new breed of AMD chips have significant performance (and cost)mN advantage over Intel's wintel products, some of the top tier makers will startE to make some of their machines with it. If AMD maintains a consistenttN performance and price advantage over Intel and manages to get good press aboutM the quality/performance of its chips, then wintel makers won't feel compelledoH to build with the "intel inside" sticker on their PCs, something they doJ because of Intel advertising giving the general public the impression that Intel is better.  K AMD doesn't need to make major breakthrough in the enterprise server market(M for its chip to be succesful. Once it has proven itself in the wintel market,aG especially with regards to compatibility, perhaps it can then grow into  enterprise segment.-  K I think that AMD is more likely to displace wintel server sales compared togI IA64 server sales initially. IA64 servers are proprietary/specialised and0K won't be very popular in the "industry standard" segment where AMD can play. much of its game.T  N The danger is that if AMD slowly builds its image, by the time Intel wakes up,M it may have already given AMD a lot of ground. And that would be a good thing * because monopolies are rarely good things.  I > Win64 isn't needed for Word, or Excel.  I don't think a faster CPU willtH > speed Solitaire.  You need a killer app that will make it interesting.  F In case you haven't noticed, today's "solitaire" games are 3D renderedF sophisticated gamesthat eat up a lot of CPU and will take any internetN bandwidth they are given. They are quite likely to be the first consumer grade to adopt 64 bit.  M On the more serious side, it will be interesting to see whether movie studiosaC and special effects firms adopt the AMD chip. I know that Intel hasaE succesfully bribed one outfit to try IA64 (and thsi was done before avM palatable IA64 was available, so it was clearly a PR ploy). But it remains toaM be seen if they stick with that IA64 thing or if they switch to a cheaper AMD.% that has equal or better performance.   J > Opteron reminds me of the DEC/Alpha mentality - if we build it they willK > come.  It's nice technology - but is it fast enough and cheap enough thatuA > people will abandon IA32 for it?  Or abandon Sun/HP/IBM for it?:  L 1- AMD's chip doesn't require people abandon the 8086 because it is the sameK architecture and can be used in place of an intel 8086 in 32 bit mode.  The @ only question is whether AMD will be priced competively or not.   M 2-AMD won't cripple its chip, or charge premium for it etc to prevent it from J competing against the 8086, as was the case for Alpha. Since it is an 8086I competing head to head against Intel's 32 bit 8086, and trying to competeiN against the IA64, AMD has every reason to market the chip, price it to compete2 head to head and grow its sales as much as it can.  3 Whether it will success or not is another question.n  L > IPF is way early in it's lifetime to either judge it's success or failure.  M Considering that it is almost 10 years old, the "early in its lifetime" is anoL interesting point of view. How old will the chip have to be before you startN saying "OK, we've waited long enough, promises that the next iteration will be- very impressive are no longer credible" ?????m  H We keep being told how great IA64 *will* be. But that usually means thatL today, the chip is crap and not a success. The day they announced the murderA of Alpha, it was TODAY that counted, not some time in the future.   N Had they waited for IA64 to prove its worth before shooting Alpha in the head,H then we could have judged IA64 more objectively because we could compareM apples to apples. But you want us to compare existing technology (Alpha) withuM some future iteration of IA64. Sorry, but that doesn't work anymore, not when G one has no confidence in a vendor such as Compaq/HP who screwed its VMSi customers already.  L > 64-bit manufacturers obviously want it to fail.  But it is plugging ahead,? > getting faster, and picking up support from a lot of players.o  L Is Dell "supportinG" IA64, or was it just coerced into building at least oneN IA64 based box ? How will it compete ion a price basis with 8086 based serversI sold by Dell ? Will there be a price premium for it ? If so, then it just2 won't sell much.   ------------------------------    Date: 10 Jun 2003 13:55:41 -0700. From: mistdragon@zdnetonebox.com (mist dragon)& Subject: Re: Portents of Itanium death= Message-ID: <7500353b.0306101255.7e19e5d8@posting.google.com>a  M > As far as I'm concerned, the credibility that one loses by using a "handle"I% > is not worth the reduction in spam.g  F Considering of the route table from where your posts are coming I dontD wonder why you are so interested of having handles to peoples names.  6 > You can use your "real" name and munge your address.  F And tell just you, your route table site and million other tv watchers here ?  
 Come on :)   Mist   ------------------------------  % Date: Tue, 10 Jun 2003 23:18:02 +0200d From: Dirk Munk <munk@home.nl>& Subject: Re: Portents of Itanium death2 Message-ID: <bc5i7r$ik9$1@news3.tilbu1.nb.home.nl>   > I > Win64 isn't needed for Word, or Excel.  I don't think a faster CPU will H > speed Solitaire.  You need a killer app that will make it interesting.J > Databases *are* a potential - but who has signed up to build and support/ > systems that scales past the promised 4-ways?- > J > Opteron reminds me of the DEC/Alpha mentality - if we build it they willK > come.  It's nice technology - but is it fast enough and cheap enough thate$ > people will abandon IA32 for it?    = Fred, the point is THEY DON'T ABANDON THE IA32 TECHNOLOGY !!!bO They still have the IA32 technology, and a damned fast one too, in fact faster  & then the 'old' AMD IA32 technology !!!% Plus as a bonus 64 bit technology !!!   O Will it be cheap enough ? I don't know, but I surely hope so. If AMD is clever  G they should make the price of the Athlon 64 (= not the Opteron !) very eP competative, and this way create facts before Intel has an answer, although I'm 0 sure Intel has an answer in the making already .   > Or abandon Sun/HP/IBM for it?o  J Why shouldn't Sun/HP/IBM use the Opteron ? I heard one HP rep say that he O expects HP to produce Opteron servers as well..... And isn't IBM assisting AMD   with Opteron chip design ?     >  > L > Alpha maybe made the 3rd tier in the marketplace.  It was consistantly outF > in front in performance - but not by enough.  If cool technology andC > performance was enough - then Tru64/Alpha should have buried Sun.e  F Yes, I agree. The problem was/is that Digital/Compaq/HP marketing for P Alpha/OpenVMS/Tru64 can't even sell a bottle of Coke to someone dying of thurst.   > J > Don't get me wrong, I loved Alpha.  I've got time and energy invested inH > understanding a lot of nitty gritty things about the architecture, the< > firmware, and how VMS works on it.  But I loved VAX too.      So do/did we all, I totaly agree   > You gotta let go.Z  : Not untill the last Alpha chip leaves the production line.   ------------------------------  # Date: Wed, 11 Jun 2003 00:14:53 GMTtL From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr")& Subject: Re: Portents of Itanium death6 Message-ID: <00A212FC.4574C980@SSRL.SLAC.STANFORD.EDU>  p In article <cf15391e.0306100816.8b1fecf@posting.google.com>, keithparris_NOSPAM@yahoo.com (Keith Parris) writes:\ >JF Mezei <jfmezei.spamnot@istop.com> wrote in message news:<3EE53A46.6D06D9F3@istop.com>...L >> However, with the "threat" of Intel also producing a 64 bit 8086, the bigP >> question now becomes: if Intel produces a 64 bit version of the 8086, will itP >> be binary compatible with AMD's version ? If not, than AMD will be in troubleI >> because it will have to change its Opteron to match the Intel chip for: >> binary compatibility. >sA >Intel has stated its 64-bit direction, and it's IA64.  If AMD isdB >smart, they probably have some engineers in a back room somewhereA >figuring out how they could do a microprocessor with the ItaniumIC >intruction set, in the event Intel wins in the 64-bit marketplace.w  K AMD needs _lawyers_ in a back room somewhere figuring out how they can do a H microprocessor with a non-Pentium instruction set.  Their ability to do L Intel-compatible processors dates from a sourcing agreement for Pentium thatL emphatically doesn't carry over to Itanium.  (That's part of why Intel likes# Itanium - it cuts out competitors.)5   -- Alan0   --  O ===============================================================================e0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056nM  Paper mail to: SSRL -- SLAC BIN 99, 2575 Sand Hill Rd, Menlo Park CA   94025dO ===============================================================================    ------------------------------  % Date: Tue, 10 Jun 2003 20:37:29 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>:& Subject: Re: Portents of Itanium death' Message-ID: <3EE687D9.F95E13AF@fsi.net>r   Keith Parris wrote:l > ] > JF Mezei <jfmezei.spamnot@istop.com> wrote in message news:<3EE50221.9A81A978@istop.com>...rL > > Remember the arguments about Alpha being single-sourced being a drawbackO > > (originally) ? Well, as long as IA64 is single sources from Intel, the samecQ > > argument should apply to it. Because AMD's solution is based on the induystryeQ > > standard which is available from multiple sources, it makes it much more of arL > > commodity and hence more politically acceptable than Intel's proprietary > > IA64 chip. > E > AMD's 64-bit extensions, x86-64 or AMD64 or whatever you call them,yF > are available only in chips from a single source -- AMD.  That makesB > Opteron just as much a single-source chip as Itanium.  And usingD > Opteron makes you dependent on a single-source supplier with fewer% > financial resources than Intel has.n  D ...until (read: if) Opteron takes the place planned for Itanic, thenD AMD's financial footing will (read: should) improve proportionately.  F One should consider the possibility that in the near future both IntelH and The Redmond Empire may need to learn to find their own "new cheese",B if you can relate to that paradigm. Failure to do so leaves one inF danger of finding one's self in a "cheese station" where the supply is quickly dwindling.   --   David J. Dachteran dba DJE Systemsc http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/d   ------------------------------  % Date: Tue, 10 Jun 2003 21:30:51 -0400o* From: "Bill Todd" <billtodd@metrocast.net>& Subject: Re: Portents of Itanium death2 Message-ID: <1hmcnZD0Nu2rGnujXTWcqQ@metrocast.net>  H ""Alan Winston - SSRL Admin Cmptg Mgr"" <winston@SSRL.SLAC.STANFORD.EDU>A wrote in message news:00A212FC.4574C980@SSRL.SLAC.STANFORD.EDU...i> > In article <cf15391e.0306100816.8b1fecf@posting.google.com>,3 keithparris_NOSPAM@yahoo.com (Keith Parris) writes:a8 > >JF Mezei <jfmezei.spamnot@istop.com> wrote in message% news:<3EE53A46.6D06D9F3@istop.com>...fJ > >> However, with the "threat" of Intel also producing a 64 bit 8086, the bigwJ > >> question now becomes: if Intel produces a 64 bit version of the 8086, will itdJ > >> be binary compatible with AMD's version ? If not, than AMD will be in trouble K > >> because it will have to change its Opteron to match the Intel chip for  > >> binary compatibility. > > C > >Intel has stated its 64-bit direction, and it's IA64.  If AMD isaD > >smart, they probably have some engineers in a back room somewhereC > >figuring out how they could do a microprocessor with the ItaniumpE > >intruction set, in the event Intel wins in the 64-bit marketplace.c >fK > AMD needs _lawyers_ in a back room somewhere figuring out how they can doa a0I > microprocessor with a non-Pentium instruction set.  Their ability to dolI > Intel-compatible processors dates from a sourcing agreement for Pentiumh thatH > emphatically doesn't carry over to Itanium.  (That's part of why Intel likes % > Itanium - it cuts out competitors.):  J Er, why on Earth would AMD have either engineers *or* lawyers spending any, time figuring out how it could clone Itanic?  I HP started work on the idea in late 1988.  HP and Intel have been workingeG *hard* on productizing it for nearly a decade.  And still the result is I considerably less than overwhelming:  how many years would it take AMD toeK make an Itanic that would even work to the degree that Merced could be saidi) to have worked, let alone be competitive?r  D If AMD were pushing Itanic and Intel were pushing Opteron the entireC industry would have written off AMD as utterly crazy.  130 W?  IA32tJ performance so bad that they'll be doing an FX!32-style software emulationK mechanism *too*?  How could that *possibly* compete with a high-performance.K 64-bit implementation that runs IA32 at full native speed as well and costse only 1/5th as much?3  D While the economic climate has left AMD in a generally uncomfortableK position, the situation with respect to Itanic is about the best they could.K possibly have wished for.  It's so late in achieving acceptable performancesG (or any market penetration whatsoever) that they've had time to field a J comparable processor at far lower cost and power consumption that offers aH natural, flexible upward-migration path from IA32.  They can continue toL leverage the industry-wide IA32 infrastructure with this new processor whileH attracting customers with the promise of its extended capabilities.  AndJ Intel *can't* afford to counter it with a similar product of its own whileH still professing to be unswervably committed (as Keith noted) to Itanic.  J Even the fact that Itanic can't legally be cloned is a double-edged sword:K Intel likes the idea, but customers may not like the power that gives IntelsL over them.  Whereas AMD has made the AMD64 architecture generally available,J so if it takes off there *will* be opportunities for second sources if AMD! should start getting out of line.e  H The fact is that AMD *has to be* unswervably committed to AMD64, becauseI that's the only strategy open to it.  But regardless of its public stanceaI Intel (unlike HP) has options with respect to Itanic:  if Itanic customerhJ acceptance continues to be chilly and AMD64 seems to be catching on, IntelL *can* create (and perhaps already *has* created) an credible Hammer clone toD compete head-on.  Whether that's likely to happen could be argued adJ infinitum, but that would be kind of silly given that in a very short time we'll be able to *see*.    - bill   ------------------------------  % Date: Tue, 10 Jun 2003 22:49:23 -0400C" From: JF Mezei <jfmezei@istop.com>& Subject: Re: Portents of Itanium death) Message-ID: <3EE698AC.EF51D208@istop.com>L  * Alan Winston - SSRL Admin Cmptg Mgr wrote:N > Intel-compatible processors dates from a sourcing agreement for Pentium thatN > emphatically doesn't carry over to Itanium.  (That's part of why Intel likes% > Itanium - it cuts out competitors.)-    M If the rationale behind dumping PA-Risc  was to choose an "industry standard".J commodity chip that would keep prices down, then you'd think that HP wouldF have negotiated with Intel to ensure that IA64 would be available fromR multiple sources, which, after all, is what drove the price down for 8086 systems.  M What was wrong with PA-Risc architecture that required it be dumped in favouriJ of a totally new ,unproven one ? Couldn't HP have simply handed Pa-Risc toC Intel and told intel to go and build a new industry standard chip ?    ------------------------------  % Date: Tue, 10 Jun 2003 20:24:58 -0700>& From: Greg Cagle <gregc@gregcagle.com>& Subject: Re: Portents of Itanium death/ Message-ID: <ved8a5deuqed03@corp.supernews.com>y   JF Mezei wrote:t  , > Alan Winston - SSRL Admin Cmptg Mgr wrote: > N >>Intel-compatible processors dates from a sourcing agreement for Pentium thatN >>emphatically doesn't carry over to Itanium.  (That's part of why Intel likes% >>Itanium - it cuts out competitors.)o >  >  > O > If the rationale behind dumping PA-Risc  was to choose an "industry standard"mL > commodity chip that would keep prices down, then you'd think that HP wouldH > have negotiated with Intel to ensure that IA64 would be available fromT > multiple sources, which, after all, is what drove the price down for 8086 systems. > O > What was wrong with PA-Risc architecture that required it be dumped in favournL > of a totally new ,unproven one ? Couldn't HP have simply handed Pa-Risc toE > Intel and told intel to go and build a new industry standard chip ?t  A Nothing was wrong with PA-RISC as such, except that there was theuA perception that it was reaching its limits and the limits of RISC@> architecture, and that a new post-RISC architecture (EPIC) was: needed. That combined with the investment cost of fabs led? HP to partner with Intel. As it turns out, HP stopped investingu> in fabs anyway (Agilent got those) and now PA-RISC is producedA by someone else altogether. And the hope was Intel's economies of = scale and fab expertise would allow lower costs than HP couldo get on its own.o   -- u
 Greg Cagle gregc at gregcagle dot com   ------------------------------  % Date: Wed, 11 Jun 2003 00:03:05 -0400n* From: "Bill Todd" <billtodd@metrocast.net>& Subject: Re: Portents of Itanium death2 Message-ID: <bticnXrAeOV0N3ujXTWcqQ@metrocast.net>  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:ZaQRKmKQshH6@eisner.encompasserve.org...n4 > In article <3EE50221.9A81A978@istop.com>, JF Mezei# <jfmezei.spamnot@istop.com> writes:m   ...i  L > > What is important here is that AMD has the *potential* to have its nicheG > > become part of the industry standard commodity market. Intel's IA643 doesn't2# > > and will remain a niche market.  > >  >r? > Upside down.  Many vendors have committed to and are shipping4> > Itanium products.  Has AMD got a tier 1 vendor on board yet?  E Both products have problems at the moment.  For Opteron a significantdJ question is, "Where are the tier-1 vendors?"  But for Itanic, the questionJ remains, as it has for the past two years, "Where are the customers?"  AndJ since Madison won't be in a particularly better position compared with itsF current competition than McKinley was with its competition a year ago,C there's no obvious reason to expect a dramatic change in that area.    - bill   ------------------------------    Date: 10 Jun 2003 23:06:02 -0500+ From: young_r@encompasserve.org (Rob Young)s& Subject: Re: Portents of Itanium death3 Message-ID: <uPL9xGJDnIjR@eisner.encompasserve.org>r  N In article <3EE64ADE.FBA3DEE7@istop.com>, JF Mezei <jfmezei@istop.com> writes: > Fred Kleinsorge wrote:O >> Well, it's possible.  We will see won't we.  I'm waiting for a true top tierf0 >> company to announce a major comittment to it. > M > Are there top tiers who have made public announcements that they are buyingaL > chips from AMD ? Does Dell make any wintels based on AMD chips or is it an > all-intel shop ? >   ' 	All Intel.  Was that a trick question?t  G > Will AMD make a huge splash and significantly erode Intel's dominancem > overnight ? Perhaps not. s  6 	Not at all.  If there are secret scents that AMD will@ 	make a "huge splash and significantly erode Intel's dominance",; 	Wall Street isn't smelling it.  Follow the money.  AMD hasa 	been recently downgraded.   > M > I think that AMD is more likely to displace wintel server sales compared totK > IA64 server sales initially. IA64 servers are proprietary/specialised andUM > won't be very popular in the "industry standard" segment where AMD can playe > much of its game.r >   @ 	Intel has big speed bumps ahead with Xeon MP.  Should keep Xeon; 	competetive with Opteron, especially in those pesky tpmCs.a+ 	For example, this 4-way Dell config gives:h  C http://www.tpc.org/tpcc/results/tpcc_result_detail.asp?id=103042201k   	Exec summary:  J http://www.tpc.org/results/individual_results/Dell/dell_6600_210403_es.pdf  A 	78000 tpmC.  The configuration for that is 4 - 2.0 GHz with 2 MB : 	L2 with each CPU.  As it just so happens, Intel is set to) 	ship at month-end an upgrade of Xeon MP:w  ( http://www.theinquirer.net/?article=9823  O "As we've written earlier, the Xeon MP 2.8GHz with 2MB cache will be introduced0+ at the end of this month at around $3,700."m  > 	Now it isn't rocket science to presume that a 40% increase in3 	CPU speed will translate in quite a bit more tpmC.D   	Here is an observation.  ; 	AMD launches Opteron in April, posts tpmC to tpc.org for a.> 	4 processor box that you can take delivery of in July.  Quite@ 	good numbers - to look at, but you won't be running that config+ 	for a while.  Meanwhile, as part of a truee< 	launch campaign with savvy marketing Intel is set to launch@ 	a CPU when Opteron's become available that knock Opteron around9 	on tpmC.  The Intel boxes will be orderable.  Why tease?   @ 	AMD does good on paper, Intel delivers in reality.  Wall Street 	sees these things too.e  ? 	Just as significant, back to the speed bumps for Xeon.  In thee? 	next year Intel is set to take Xeon to 3.6 GHz with an 800 MHzeC 	front-side bus and 4 MB on-chip cache.  While others are quick to  E 	point out Opteron will be at 2.6 GHz whenever, you can see that AMD l2 	will be quickly falling behind the speed curve.    A 	Today, a 1.8 GHz 4-CPU Opteron does 80000 tpmC.  Today a 2.0 GHz @ 	4-CPU Xeon CPU box does 78000 tpmC.  Next year, Opteron will be8 	pushing 2.6 GHz, Xeon will be at 3.6 GHz.  Do the math.  J > We keep being told how great IA64 *will* be. But that usually means thatN > today, the chip is crap and not a success. The day they announced the murderC > of Alpha, it was TODAY that counted, not some time in the future.l  ? 	It will be great.  Intel is spending the money to ensure that.t> 	Don't discount Xeon either.  AMD has to fight a multi-pronged< 	battle.  Flash, IA32, IA64 and several generations of each.; 	AMD won't be able to keep up.  They haven't the resources.o  N > Is Dell "supportinG" IA64, or was it just coerced into building at least oneP > IA64 based box ? How will it compete ion a price basis with 8086 based serversK > sold by Dell ? Will there be a price premium for it ? If so, then it juste > won't sell much.  B 	Dell certainly doesn't want to spoil their debut of Madison.  You> 	can bet it will roll out with a nice price tag and impressiveB 	benchmarks.  Dell.  They know what they are doing.  It won't sellE 	much.  But it will be a VERY impressive 4-way number cruncher - that E 	will be a kick start for IA64 sales.  Especially that it will/should > 	come in at a good price.  Kick anything around at fp in that D 	4-cpu config.  Kick anything around in 4-cpus at tpmC (i.e. a greatB 	fit for Oracle as mentioned before per-CPU costs for Oracle means( 	you WANT to get a lot of bang per CPU).  D 	And as much as some like to bash HP, they are much larger than Dell@ 	and they didn't get there by accident.  They know what they are 	doing too.h   				Robu   ------------------------------  # Date: Tue, 10 Jun 2003 20:17:07 GMTy* From: Paul Anderson <paul.anderson@hp.com>H Subject: Re: Printer recommendation, hobbyist/decwindows/mozilla and VMS5 Message-ID: <100620031559053688%paul.anderson@hp.com>4  < In article <bc1ef3$lqu$1@news4.tilbu1.nb.home.nl>, Dirk Munk <munk@home.nl> wrote:i  ? > Please keep in mind that no HP inkjet Postscript printers arecF > officially supported in DCPS, and even stranger, are not on the list( > of printers to be supported in future.  G We have not gotten many requests for DCPS support for inkjet printers. bE If we added support, we would probably support the printers that come G with PostScript and a network card, like the Business InkJet 2280, 2600-	 and 3000.-  C The cost of these printers puts them in the same price range as the< low-end color laser printers.P  / > I don't even know if DCPS supports color ????j  C DCPS supports a bunch of color printers.  There's nothing DCPS does-7 that would enable or disable color files from printing.a  C > Perhaps someone at HP would be so kind to think about their loyal A > hobbiest customers, and come up with a nice VMS driver for a HPo > inkjet printer ???  C Are there specific models you would like to see supported by DCPS? aD Support for PCL printers would have to come from somewhere else than DCPS.    Paul   -- e  Paul Anderson   OpenVMS Engineering    Hewlett-Packard Companye   ------------------------------  % Date: Tue, 10 Jun 2003 23:54:48 +0200e From: Dirk Munk <munk@home.nl>H Subject: Re: Printer recommendation, hobbyist/decwindows/mozilla and VMS2 Message-ID: <bc5kcq$ada$1@news2.tilbu1.nb.home.nl>   Paul Anderson wrote:> > In article <bc1ef3$lqu$1@news4.tilbu1.nb.home.nl>, Dirk Munk > <munk@home.nl> wrote:  >  > ? >>Please keep in mind that no HP inkjet Postscript printers arejF >>officially supported in DCPS, and even stranger, are not on the list( >>of printers to be supported in future. >  > I > We have not gotten many requests for DCPS support for inkjet printers. nG > If we added support, we would probably support the printers that comeeI > with PostScript and a network card, like the Business InkJet 2280, 2600  > and 3000.t  Q I assume that HP is using the same Postscript RIP software in its entire line of aO   Postscript inkjet printers. In that case it seems there is no reason why the  K 1220 and other cheaper inkjet printers could not be used. I don't know how a7 difficult it is to add parallel or USB support to DCPS.u     > E > The cost of these printers puts them in the same price range as the  > low-end color laser printers.h  L True, but inkjet printers are far better in printing photographs etc., so I @ would say both types are for a different kind of color printing.   >  > / >>I don't even know if DCPS supports color ????l >  > E > DCPS supports a bunch of color printers.  There's nothing DCPS doest9 > that would enable or disable color files from printing.r   Good !!r   >  > C >>Perhaps someone at HP would be so kind to think about their loyal A >>hobbiest customers, and come up with a nice VMS driver for a HP  >>inkjet printer ??? >  > E > Are there specific models you would like to see supported by DCPS? oF > Support for PCL printers would have to come from somewhere else than > DCPS.m  P I was thinking of a kind of Windows driver for a cheap inkjet printer. The kind Q of driver where all RIP processing is done on the host. Not really something for n the DCPS team I think.     >  > Paul >    ------------------------------  % Date: Tue, 10 Jun 2003 21:43:16 -0400h* From: "Stanley F. Quayle" <stan@stanq.com>H Subject: Re: Printer recommendation, hobbyist/decwindows/mozilla and VMS/ Message-ID: <3EE650F4.10923.34331217@localhost>p  - On 10 Jun 2003 at 20:17, Paul Anderson wrote:nE > Are there specific models you would like to see supported by DCPS? vF > Support for PCL printers would have to come from somewhere else than > DCPS.u  ? I have a HP OfficeJet K80 multifunction printer with JetDirect   network box.  D It would be useful to print with the same sort of capabilities that . DCPS provides, even though it's a PCL printer.  6 And it would be way cool to scan and fax with it, too.  
 --Stan Quaylel Quayle Consulting Inc.  
 ----------C Stanley F. Quayle, P.E. N8SQ  +1 614-868-1363  Fax: +1 614 868-1671m1 8572 North Spring Ct. NW, Pickerington, OH  43147d= Preferred address:  stan@stanq.com       http://www.stanq.com8   ------------------------------    Date: 10 Jun 2003 11:13:54 -0700, From: davidjsullivan@hp.com (David Sullivan) Subject: Re: RTR/XAg< Message-ID: <e8f638a9.0306101013.804f3a2@posting.google.com>  = I've contacted Robert offline. If anyone has any questions on 8 current/furure plans for XA support in RTR send email to davidjsullivan@hp.comm   ------------------------------    Date: 10 Jun 2003 11:18:16 -0700+ From: JRumerman@hotmail.com (Joel Rumerman) R Subject: Simulatenously transferring a 1GB file from VMS to two Windows 2k servers= Message-ID: <a4437404.0306101018.1486ff08@posting.google.com>    Hi All,a  @ Okay. I've never logged onto a VMS box, I know nothing about theD architecture or I/O functionality, but I'm working on a team that is8 taking 1 large VMS file, about 1 GB, and transferring itC simultaneously to two different Windows 2K Advanced Servers through-C FTP. I was wondering if anybody knew of a problem this might cause.aD The results of the transfer are two files with different file sizes!E The differences are large with one server having 550mbs and the othervB server having 450mbs. It looks to me that it is splitting the file3 into two chunks, but I'm not sure (it's a graphic).e  . Does anybody know about a situation like this?   Thanks so much.t   R/Joel   ------------------------------  % Date: Tue, 10 Jun 2003 14:53:14 -0400a* From: "Stanley F. Quayle" <stan@stanq.com>V Subject: Re: Simulatenously transferring a 1GB file from VMS to two Windows 2k servers/ Message-ID: <3EE5F0DA.11810.32BBAB7D@localhost>n  G > The differences are large with one server having 550mbs and the other D > server having 450mbs. It looks to me that it is splitting the file5 > into two chunks, but I'm not sure (it's a graphic).g  F Are they transferring it specifying binary transfer?  The default FTP  setting on VMS is not binary.A  
 --Stan Quayle  Quayle Consulting Inc.  
 ----------C Stanley F. Quayle, P.E. N8SQ  +1 614-868-1363  Fax: +1 614 868-1671m1 8572 North Spring Ct. NW, Pickerington, OH  43147p= Preferred address:  stan@stanq.com       http://www.stanq.comH   ------------------------------  % Date: Tue, 10 Jun 2003 12:34:20 -0600v6 From: "Michael D. Ober" <obermd-@-alum-mit-edu-nospam>V Subject: Re: Simulatenously transferring a 1GB file from VMS to two Windows 2k servers0 Message-ID: <MwpFa.182$Dj.66158@news.uswest.net>  I Are the cluster sizes on the target systems the same?  Also, what happensa% when you FTP to one server at a time?p  
 Mike Ober.  8 "Joel Rumerman" <JRumerman@hotmail.com> wrote in message7 news:a4437404.0306101018.1486ff08@posting.google.com...i	 > Hi All,e >aB > Okay. I've never logged onto a VMS box, I know nothing about theF > architecture or I/O functionality, but I'm working on a team that is: > taking 1 large VMS file, about 1 GB, and transferring itE > simultaneously to two different Windows 2K Advanced Servers through6E > FTP. I was wondering if anybody knew of a problem this might cause.oF > The results of the transfer are two files with different file sizes!G > The differences are large with one server having 550mbs and the othereD > server having 450mbs. It looks to me that it is splitting the file5 > into two chunks, but I'm not sure (it's a graphic).g > 0 > Does anybody know about a situation like this? >  > Thanks so much.e >v > R/Joel   ------------------------------    Date: 10 Jun 2003 16:06:43 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)eV Subject: Re: Simulatenously transferring a 1GB file from VMS to two Windows 2k servers3 Message-ID: <NcBwcmJru1PX@eisner.encompasserve.org>s  k In article <a4437404.0306101018.1486ff08@posting.google.com>, JRumerman@hotmail.com (Joel Rumerman) writes:y  E > simultaneously to two different Windows 2K Advanced Servers throughh)                                   ^^^^^^^       There's your trouble!  B    OK, seriously, most likely one of them is transfering in binary)    and the other is transfering in ASCII.    ------------------------------  # Date: Tue, 10 Jun 2003 22:09:40 GMTt# From: "John Smith" <a@nonymous.com> V Subject: Re: Simulatenously transferring a 1GB file from VMS to two Windows 2k serversG Message-ID: <EGsFa.52884$j9%.5693@news04.bloor.is.net.cable.rogers.com>3  @ "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in5 message news:NcBwcmJru1PX@eisner.encompasserve.org...7? > In article <a4437404.0306101018.1486ff08@posting.google.com>,e- JRumerman@hotmail.com (Joel Rumerman) writes:  >r? > > simultaneously to two different Windows 2K Advanced Serverso throughh+ >                                   ^^^^^^^e >y >    There's your trouble! >lD >    OK, seriously, most likely one of them is transfering in binary+ >    and the other is transfering in ASCII.  >d   Bob," You got the trouble part wrong....  ? > > simultaneously to two different Windows 2K Advanced Serversl throughw   ^^^^^^^^^^^^^^^^^^^^^^^^   |tF                                                      This is where the trouble is. ;-)    ------------------------------  % Date: Tue, 10 Jun 2003 22:50:21 +0200d From: Dirk Munk <munk@home.nl>Y Subject: Re: Simulatenously transferring a 1GB file from VMS to two Windows 2k servers 2kr2 Message-ID: <bc5gjt$si1$1@news4.tilbu1.nb.home.nl>  / Is this one graphic of 1 GB? That is a big one.s  P Before we can tell you what to do, we need to know what kind of VMS file it is. M VMS has many file types, and sometimes you need to do something extra if you  " want to send it to another system.  H To know what kind of file it is, simply do a dir/full of the file, thus % something like DIR/FULL GRAPHIC.DAT .e  ' It should give you something like this:e  2 GRAPHIC.DAT;1                 File ID:  (7383,1,0)0 Size:          164/192        Owner:    [SYSTEM]# Created:     1-SEP-1999 23:46:38.72l' Revised:     1-SEP-1999 23:47:19.89 (2)p Expires:    <None specified>  Backup:     <No backup recorded> Effective:  <None specified> Recording:  <None specified> Accessed:   <None specified> Attributes: <None specified> Modified:   <None specified>
 Linkcount:  1  File organization:  Sequential Shelved state:      Online  Caching attribute:  WritethroughF File attributes:    Allocation: 192, Extend: 0, Global buffer count: 0:                      No version limit, Contiguous best try1 Record format:      Fixed length 512 byte recordsp Record attributes:  None RMS attributes:     None Journaling enabled: NoneA File protection:    System:RWED, Owner:RWED, Group:RWED, World:REo Access Cntrl List:  None Client attributes:  None    Total of 1 file, 164/192 blocks.        Q The important things to look at are the File Organization, the Record Format and iN   the Record Attributes. They will give us a good idea about what kind of VMS ' file organization you are dealing with.d                 Joel Rumerman wrote:	 > Hi All,  > B > Okay. I've never logged onto a VMS box, I know nothing about theF > architecture or I/O functionality, but I'm working on a team that is: > taking 1 large VMS file, about 1 GB, and transferring itE > simultaneously to two different Windows 2K Advanced Servers through E > FTP. I was wondering if anybody knew of a problem this might cause.nF > The results of the transfer are two files with different file sizes!G > The differences are large with one server having 550mbs and the othertD > server having 450mbs. It looks to me that it is splitting the file5 > into two chunks, but I'm not sure (it's a graphic).  > 0 > Does anybody know about a situation like this? >  > Thanks so much.m >  > R/Joel   ------------------------------  % Date: Tue, 10 Jun 2003 17:52:06 -0400." From: JF Mezei <jfmezei@istop.com>Y Subject: Re: Simulatenously transferring a 1GB file from VMS to two Windows2k servers serr) Message-ID: <3EE652F2.ED857A2C@istop.com>m   Dirk Munk wrote:R > The important things to look at are the File Organization, the Record Format andO >   the Record Attributes. They will give us a good idea about what kind of VMSa) > file organization you are dealing with.y  7 Another important thing to consider is file allocation.o  M You could have 1gb allocated, but only 450 meg used. So the resulting file on . the target system might be just 450 megs used.  L What you need to consider is that on each target system, file extentions mayF be different. So as the file grows during downloads, it gets allocatedL differently on both receiving systems, and in the last extension done to the> file, one may end up with different allocation than the other.  M I don't think that there would be any issues with concurrent downloads of theoL files. VMS has no problems with 2 processes reading a file at the same time.I But you should make sure that both Wintel boxes download in the same morer- (binary or in ftp terminology: "type Image").r   ------------------------------  % Date: Tue, 10 Jun 2003 20:49:23 -0500h1 From: "David J. Dachtera" <djesys.nospam@fsi.net>m' Subject: Re: Supporting  Wintel laptopsB' Message-ID: <3EE68AA3.B5D87B62@fsi.net>r   JF Mezei wrote:p >  > "David J. Dachtera" wrote:$ > > Another candidate for Samba, eh? > V > Nop. I don't want to install software "just in case someone shows up with a laptop". > M > I'd rather make use what is already available on my systems and just figures< > out a way to give the wintel laptop the services it needs.  H Unfortunately, "the services it needs" are likely to depend on SMB-based5 protocols, especially print services. Hence, Samba...    > The DHCP server on VMS   Which IP stack?n  2 > allows one to provide a computer with a lot moreP > information than just IP address. But I suspect most DHCP clients just discard  > that supplemental information.  
 Perhaps...   -- s David J. Dachtera  dba DJE Systemse http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/d   ------------------------------  % Date: Tue, 10 Jun 2003 20:29:56 -0500c1 From: "David J. Dachtera" <djesys.nospam@fsi.net>-' Subject: Re: VAX VMS 7.3 - follow up II3' Message-ID: <3EE68614.FEA378A3@fsi.net>o  	 JK wrote:c > [snip]0 > $ @SYS$UPDATE:VMSINSTAL VMS073 dkb200:[000000] > D > Does the command do an upgrade or can I perform a 'clean' install?  E It performs a "clean upgrade". You've got a clean install of V6.1, ifpE you've not made any changes since the install. This will upgrade yourh clean V6.1 to a clean V7.3.s  I > I made a mistake choosing the scs nodename, can I correct this with thed
 > upgrade?8 > Or should I change it using commands before upgrading?  E Change it MODPARAMS.DAT before you upgrade. If you need help with thepF editor(s), post again. You will find that there are two major editors:E EDT and TPU. Discussions of such tend to trigger "wars" as each has amC following as loyal as VMS's. There is also TECO, which has it's owny advantages and following.t  F If you want to get started quickly without taking the time to learn an editor:f  E $ SEARCH SYS$SYSTEM:MODPARAMS.DAT SCSNODE/MATCH=NOR/OUT=MODPARAMS.NEWO  $ APPEND SYS$INPUT MODPARAMS.NEW SCSNODE = "NODNAM" CTRL+Z- $ COPY MODPARAMS.NEW SYS$SYSTEM:MODPARAMS.DATl  H That is, press CTRL+Z to terminate input from the keyboard to the APPEND2 command. Then, invoke AUTOGEN to apply the change:  $ $ @sys$update:autogen getdata reboot  B > I would like to thank all of you in helping me getting started !  C Apologies for our squabbling, but glad you find this group helpful.n   -- r David J. Dachterao dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  # Date: Wed, 11 Jun 2003 04:52:59 GMTt) From: Andrew Balaam <abalaam@yahoo.co.uk>/ Subject: VMS Freeware CDsl3 Message-ID: <20030611.4525900.702163943@imagnu.geo>   G Does anyone know if the images of the VMS freeware CDs are available=20 G somewhere on the Internet for downloading? If so, where would I find=20  them?a   Andrew.    ------------------------------  # Date: Wed, 11 Jun 2003 05:17:31 GMTr) From: Andrew Balaam <abalaam@yahoo.co.uk>e Subject: Re: VMS Freeware CDsd4 Message-ID: <20030611.5173100.1433272736@imagnu.geo>   Leslie,u  I I saw your post, but the news server said that the message was no longer=S =20m available:-(  6 >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<  I On 11/06/03, 05:52:59, Andrew Balaam <abalaam@yahoo.co.uk> wrote regardi=f ng=20t VMS Freeware CDs:i    F > Does anyone know if the images of the VMS freeware CDs are availableF > somewhere on the Internet for downloading? If so, where would I find > them?i  	 > Andrew.t   ------------------------------  # Date: Wed, 11 Jun 2003 05:23:52 GMT:) From: Andrew Balaam <abalaam@yahoo.co.uk>8 Subject: Re: VMS Freeware CDsa3 Message-ID: <20030611.5235200.233919782@imagnu.geo>n  I I know that there are various places to download the files, but I really=i =20:I want the CD images, as I have a high bandwidth link at work, and just a =   I 56K modem at home. I thought it would be nice if I could download a whol=o e=20  CD, write it, then take it home.  6 >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<  E On 11/06/03, 06:08:05, LESLIE@JRLVAX.HOUSTON.RR.COM (leslie) wrote=20s regarding Re: VMS Freeware CDs:n    , > Andrew Balaam (abalaam@yahoo.co.uk) wrote:I > : Does anyone know if the images of the VMS freeware CDs are available=r  I > : somewhere on the Internet for downloading? If so, where would I find=y  	 > : them?i > :(  - > The VMS freeware programs are available at:   ' >   http://h71000.www7.hp.com/freeware/= >   Index of /freeware  4 > --Jerry Leslie   (my opinions are strictly my own); >   Note: leslie@jrlvax.houston.rr.com is invalid for emailI   ------------------------------  # Date: Wed, 11 Jun 2003 05:24:49 GMTm) From: Andrew Balaam <abalaam@yahoo.co.uk>  Subject: Re: VMS Freeware CDs=4 Message-ID: <20030611.5244900.2310948853@imagnu.geo>  6 I must be too impatient, the message body turned up:-)  6 >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<  I On 11/06/03, 06:17:31, Andrew Balaam <abalaam@yahoo.co.uk> wrote regardi=e ng=20D Re: VMS Freeware CDs:i    	 > Leslie,   I > I saw your post, but the news server said that the message was no long=e er > available:-(  8 > >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<  I > On 11/06/03, 05:52:59, Andrew Balaam <abalaam@yahoo.co.uk> wrote regar=i ding > VMS Freeware CDs:,    I > > Does anyone know if the images of the VMS freeware CDs are available=   I > > somewhere on the Internet for downloading? If so, where would I find=e  	 > > them?e   > > Andrew.m   ------------------------------  % Date: Tue, 10 Jun 2003 22:19:07 -0700A! From: SloPoke <csewell@telus.net>i Subject: Re: VMS Freeware CDsr8 Message-ID: <gtedevolk3ei3uitdqcp5mj8ioqhcojhnm@4ax.com>  * Andrew Balaam <abalaam@yahoo.co.uk> wrote:  F >Does anyone know if the images of the VMS freeware CDs are available F >somewhere on the Internet for downloading? If so, where would I find  >them?  E I'm nos sure what's all on the CDs, but a lot of goodies can be foundc at:n  - http://vms.process.com/fileserv-software.htmlr   ------------------------------  % Date: Tue, 10 Jun 2003 21:51:59 +0200r From: Dirk Munk <munk@home.nl>L Subject: Re: Will a Radeon 9000 PCI card work in the place of a Radeon 7500?2 Message-ID: <bc5d6f$jrc$1@news2.tilbu1.nb.home.nl>   Petros Dafniotis wrote:fG > As subject says. I am having a hard time finding a radeon 7500 pci inaC > Europe; I was wondering if the readily available 9000 will do thei- > trick. I have an XP1000 Compaq workstation.e > % > Thank you in advance. Kind regards,- > Petros > ---  > Petros Dafniotis, PhD  > pdafniotis@yahoo.com   No, I'm afraid not.e  N I've asked that question too for the same reason, but I was told that the VMS 8 driver at present only supports the Radeon 7500 chipset.  3 Did you not succeed getting the 7500PCI in Zrich ?0   ------------------------------  % Date: Tue, 10 Jun 2003 21:56:21 +01001* From: "John Travell" <john@travell.uk.net>L Subject: Re: Will a Radeon 9000 PCI card work in the place of a Radeon 7500?5 Message-ID: <bc5gvc$fpkr3$1@ID-120847.news.dfncis.de>>   > Petros Dafniotis wrote:"G > As subject says. I am having a hard time finding a radeon 7500 pci inoC > Europe; I was wondering if the readily available 9000 will do the - > trick. I have an XP1000 Compaq workstation.I >o Did you try club-3d ? 5 http://www.club-3d.nl/products2.php?catid=1&prodid=69o     -- John Travell  VMS crashdump expertise for hire john@travell.uk.nets http://www.travell.uk.net/       ---I& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.484 / Virus Database: 282 - Release Date: 27/05/2003t   ------------------------------  # Date: Tue, 10 Jun 2003 20:41:00 GMTe9 From: "Fred Kleinsorge" <my-last-name@stardotzko.dec.com>yL Subject: Re: Will a Radeon 9000 PCI card work in the place of a Radeon 7500?2 Message-ID: <wnrFa.2295$Pa1.2272@news.cpqcorp.net>   No.c  : "Petros Dafniotis" <pdafniotis@yahoo.com> wrote in message7 news:e54adf36.0306100839.74d23c69@posting.google.com...tG > As subject says. I am having a hard time finding a radeon 7500 pci inhC > Europe; I was wondering if the readily available 9000 will do thed- > trick. I have an XP1000 Compaq workstation.n >t% > Thank you in advance. Kind regards,e > Petros > ---c > Petros Dafniotis, PhDc > pdafniotis@yahoo.com   ------------------------------   End of INFO-VAX 2003.321 ************************