0 INFO-VAX	Fri, 17 Jan 2003	Volume 2003 : Issue 33      Contents:* Re: *** DANGER, Will Robinson! DANGER! **** Re: *** DANGER, Will Robinson! DANGER! ***4 Re: Access linux and/or windows directories with vms4 Re: Access linux and/or windows directories with vms Alpha 3000 ... Re: Alpha 3000 ... Re: Alpha 3000 ... Re: Alpha 3000 ... RE: Alpha 3000 ... Re: Alpha 3000 ... Re: Alpha 3000 .... Alpha, Itanic and Opteron benchmark comparison2 Re: Alpha, Itanic and Opteron benchmark comparison2 Re: Alpha, Itanic and Opteron benchmark comparison2 RE: Alpha, Itanic and Opteron benchmark comparison2 Re: Alpha, Itanic and Opteron benchmark comparison2 Re: Alpha, Itanic and Opteron benchmark comparison2 Re: Alpha, Itanic and Opteron benchmark comparison2 Re: Alpha, Itanic and Opteron benchmark comparison2 Re: Alpha, Itanic and Opteron benchmark comparison2 Re: Alpha, Itanic and Opteron benchmark comparison" Re: DCPS "green-bar" look printing Re: DECnet-Plus vs DECnet IV Re: DECnet-Plus vs DECnet IV Re: DECnet-Plus vs DECnet IV	 ES40 Hang 
 Re: ES40 Hang 
 Re: ES40 Hang 
 RE: ES40 Hang  FA: Systems & Options Catalog  Foreign mail question (code 19) % GENERIC_POLL messages in Mozilla/CSWB J Re: How much disk I/O can I do with a sys$QIOW IO$_WRITEVBLK/IO$_READVBLK?- Re: HP Sets the Stage for Alpha's Last Hurrah - Re: HP Sets the Stage for Alpha's Last Hurrah - Re: HP Sets the Stage for Alpha's Last Hurrah + Re: Independent VMS crash analysis service.  Re: Is Galaxy an EXOKERNEL ?- Re: Itanium with alpha goodies on the way ...  Montecito slips to 2005  Re: Montecito slips to 2005  Re: Montecito slips to 2005  Re: Montecito slips to 2005 . Re: Most common OpenVMS versions in use today?. RE: Most common OpenVMS versions in use today?. RE: Most common OpenVMS versions in use today? NT on Alpha  Re: NT on Alpha  RE: NT on Alpha  Re: NT on Alpha  Re: NT on Alpha  RE: NT on Alpha  RE: NT on Alpha  Re: NT on Alpha  Re: NT on Alpha  Re: NT on Alpha  RE: NT on Alpha  Re: Spiralog anyone? Re: TZ15 problems  Re: vax basic questions  RE: vax basic questions  Re: vax basic questions  Re: vax basic questions  Re: vax6k.openecs.org rebirth   F ----------------------------------------------------------------------  + Date: Thu, 16 Jan 2003 13:18:50 -0800 (PST) . From: Fabio Cardoso <fabiopenvms@yahoo.com.br>3 Subject: Re: *** DANGER, Will Robinson! DANGER! *** ? Message-ID: <20030116211850.8830.qmail@web20203.mail.yahoo.com>    Well  - Looks like the Lost in Space robot crashed !     Regards    FC  ! --- norm.raphael@metso.com wrote:  > % > Has anyone an explanation for this:  > / >    Press RETURN to continue, q RETURN to quit , > ***     DANGER, Will Robinson! DANGER! ***; >         operator new failure: insufficient virtual memory M > %SYSTEM-F-OPCCUS, opcode reserved to customer fault at PC=FFFFFFFF80A3E434, 
 > PS=0000001B  > 4 >   Improperly handled condition, image exit forced. >  > [OpenVMS V7.2-2] >  >      =====  ========================== Fbio dos Santos Cardoso OpenVMS System Manager Rio de Janeiro - Brazil  fabiopenvms@yahoo.com.br ==========================  2 __________________________________________________ Do you Yahoo!?5 Yahoo! Mail Plus - Powerful. Affordable. Sign up now.  http://mailplus.yahoo.com    ------------------------------  % Date: Thu, 16 Jan 2003 19:34:11 -0600 @ From: "David J. Dachtera" <djesys.nospam@earthlink.spamfree.net>3 Subject: Re: *** DANGER, Will Robinson! DANGER! *** 6 Message-ID: <3E275D93.5BE2C322@earthlink.spamfree.net>   norm.raphael@metso.com wrote:  > % > Has anyone an explanation for this:  > / >    Press RETURN to continue, q RETURN to quit , > ***     DANGER, Will Robinson! DANGER! ***; >         operator new failure: insufficient virtual memory M > %SYSTEM-F-OPCCUS, opcode reserved to customer fault at PC=FFFFFFFF80A3E434, 
 > PS=0000001B  > 4 >   Improperly handled condition, image exit forced. >  > [OpenVMS V7.2-2]   ...and the image is ... ?    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------    Date: 16 Jan 2003 14:45:27 -0500& From: fdc@columbia.edu (Frank da Cruz)= Subject: Re: Access linux and/or windows directories with vms 1 Message-ID: <b0724n$hme$1@watsol.cc.columbia.edu>   ' In article <3e26f3e5@uni-wuerzburg.de>, < Franz-Jrgen Tollmann <tollmann@vim.uni-wuerzburg.de> wrote:; : "Frank da Cruz" <fdc@columbia.edu> schrieb im Newsbeitrag - : news:b06iqp$bku$1@watsol.cc.columbia.edu... + : > In article <3e26a5ae@uni-wuerzburg.de>, @ : > Franz-Jrgen Tollmann <tollmann@vim.uni-wuerzburg.de> wrote:J : > If they do not have a terminal session, then a more "modern" approach,K : > such as NFS, Samba, PATHWORKS, etc, might be appropriate, but these are M : > also more difficult to install and manage, might expose your hosts to new D : > risks, and probably don't handle cross-platform record-format or? : > character-set conversion of text files very well or at all.  : B : Are you joking ? The big enterprise solutions for cross-platform1 : networksharing bear graeater risks than kermit?  : L No, I'm not joking.  Kermit is a user-mode application program that requiresG no special privileges to run.  It's not a system service or file-system L extension.  If you follow newsgroups like comp.os.linux.networking or almostL any other Unix-or-Windows sysadmin or network oriented group, you'll find itG full of questions about and problems with Samba, NFS, and all the other M fashionable solutions.  One might conclude that they are not easy to install,  administer, or keep safe.   0 : How secure do you suggest a telnet connection? : K Remember we are talking about PCs with Windows and a VMS server.  Kermit 95 B for Windows supports the following secure TCP/IP connection types:    . SSL/TLS terminal sessions0  . Telnet, FTP, and HTTP secured with SSL or TLSA  . Telnet, Rlogin, and FTP secured with Kerberos IV or Kerberos V "  . Telnet and FTP secured with SRP  . SSH v1 terminal sessions   . SSH v2 terminal sessions   K Thus you can use practically any kind of security that is installed on VMS. H Again, if you have a Kermit 95 terminal session to VMS, nothing could beL easier than downloading a file through the terminal session you already haveJ -- regardless of the transport, protocol, and security -- it's the same inH all cases: dialup, SSH, clear-text Telnet, secure Telnet.  You also get:  *  . Excellent VT320 emulation (ask anybody)#  . ISO 2022 character-set switching   E Kermit 95 is one of the few emulators that handles ISO 2022 switching D among ASCII, ISO 8859-1 (or DEC MCS), DEC Special, and DEC TechnicalG character sets, letting applications like EDIT/TPU as well any properly G coded forms-filling application work correctly in German (or Russian or  Greek for that matter).   H Plus an integrated cross-platform script language that lets you automateK any task you could do by hand, and lots more besides that.  You should look J through the Kermit website and base your complaints and criticisms on whatE actually exists, rather than an impression you might have formed many 
 years ago.  !   http://www.columbia.edu/kermit/    - Frank    ------------------------------   Date: 16 Jan 2003 23:33 CDT ' From: carl@gerg.tamu.edu (Carl Perkins) = Subject: Re: Access linux and/or windows directories with vms - Message-ID: <16JAN200323331429@gerg.tamu.edu>    bill@cs.uofs.edu writes...( >In article <3E26926E.F2A4EF95@aaa.com>,* >	Jan-Erik Sderholm <aaa@aaa.com> writes:E >> Hm, re-reading the subject of this thread, it look like you'd like 5 >> to access linux and/or windows *from* VMS, right ?  >>   >> VMS -> Linux   : FTP J >> VMS -> Windows : FTP to one of your "PC-Servers" where the users access5 >>                  the files through a common share.  >>   > < >Contrary to popular belief, most users do not consider that >"accessing" files.   E I don't see why this wouldn't be "accessing"the files. They log in to C the VMS system and run the thingy that they need to run. The thingy G produces the output file which it then automatically FTPs to the proper D destination (asking for the password if you don't want to store it).D The user looks in the destination folder on the server via the usualF commands/point-n-click and sees the data file which they can access as needed.   ! Where's the "not accessing" part?   & >We won't even go into how quickly one9 >would end out with multiple copies of the same files all   >containing something different. >  >bill   C Sure, lets go into it. The answer is "not very quickly, if at all". F Why? Because Windows/Linux will only let you have 1 version of a file.H Since the FTP would be automated so that at the end of the job it pushesD the file over to the other system it will obviously be putting it inA the proper target directory - it should also be giving the file a A relevant name, or a user supplied name. You might overwrite files 1 occasionally, but you won't have multiple copies.   F Well, if the users move them around willy-nilly and/or always have theF proceedure produce files that have the same name and it writes to moreF than one destination (like perhaps one per project or some such thing)I then it will happen (although they would still be identifiable by project F and filename, just not by filename alone). But that is their own faultK for being stupid. You can't protect stupid people from their own stupidity.    --- Carl   ------------------------------    Date: 16 Jan 2003 11:54:07 -0800' From: timasmith@hotmail.com (Tim Smith)  Subject: Alpha 3000 ... = Message-ID: <a7234bb1.0301161154.2a968559@posting.google.com>    Hi, F    I got my Alpha 3000 on EBay for a few hundered bucks - very excited# about that. It is on its way to me.   B    Then I thought, hmm what about the power requirements - so doesF anyone know do I need to setup some special amp/voltage in my basementF to get this up and running - anything else I need apart from dry, cool and ideally static free?   thanks!    Tim    ------------------------------  % Date: Thu, 16 Jan 2003 21:11:37 +0100 6 From: =?ISO-8859-1?Q?Arne_Vajh=F8j?= <arne@vajhoej.dk> Subject: Re: Alpha 3000 ... ) Message-ID: <3E2711F9.3010700@vajhoej.dk>    Tim Smith wrote:H >    I got my Alpha 3000 on EBay for a few hundered bucks - very excited% > about that. It is on its way to me.  > D >    Then I thought, hmm what about the power requirements - so doesH > anyone know do I need to setup some special amp/voltage in my basementH > to get this up and running - anything else I need apart from dry, cool > and ideally static free?  ; There are no special requirements for a DEC 3000 model x00.   < A model 400 takes 300 W which is the same as an ordinary PC.   Arne   ------------------------------  % Date: Thu, 16 Jan 2003 21:36:29 +0100 " From: "Hans Vlems" <hvlems@iae.nl> Subject: Re: Alpha 3000 ... 5 Message-ID: <b0754k$modbd$1@ID-143435.news.dfncis.de>   6 "Tim Smith" <timasmith@hotmail.com> schreef in bericht7 news:a7234bb1.0301161154.2a968559@posting.google.com...  > Hi, H >    I got my Alpha 3000 on EBay for a few hundered bucks - very excited% > about that. It is on its way to me.  > D >    Then I thought, hmm what about the power requirements - so doesH > anyone know do I need to setup some special amp/voltage in my basementH > to get this up and running - anything else I need apart from dry, cool > and ideally static free? > G The 3000 is about as problematic as an average PC with respect to power  requirements, humidity% and temperature, so no worries there. K The point is, where are you located and where did the system come from? Not  all Alphas have anA external switch to change between US and European voltages AFAIK.    Hans   ------------------------------  % Date: Thu, 16 Jan 2003 21:52:01 +0100 6 From: =?ISO-8859-1?Q?Arne_Vajh=F8j?= <arne@vajhoej.dk> Subject: Re: Alpha 3000 ... ) Message-ID: <3E271B71.2000205@vajhoej.dk>    Arne Vajhj wrote: > Tim Smith wrote:I >>    I got my Alpha 3000 on EBay for a few hundered bucks - very excited & >> about that. It is on its way to me. >>E >>    Then I thought, hmm what about the power requirements - so does I >> anyone know do I need to setup some special amp/voltage in my basement I >> to get this up and running - anything else I need apart from dry, cool  >> and ideally static free?  >  > = > There are no special requirements for a DEC 3000 model x00.  > > > A model 400 takes 300 W which is the same as an ordinary PC.  $ Working temperatur 15-32 C / 59-90 F  " Store temperatur 5-50 C / 41-122 F   No problem I assume.   Arne   ------------------------------  % Date: Thu, 16 Jan 2003 12:49:04 -0800 $ From: Shane Smith <ssmith@icius.com> Subject: RE: Alpha 3000 ... 0 Message-ID: <01C2BD5D.CA696C30@sulfer.icius.com>  ) My 3000/300 ran off ordinary wall supply.    Shane    -----Original Message-----: From: timasmith@hotmail.com [mailto:timasmith@hotmail.com]) Sent: Thursday, January 16, 2003 11:54 AM  To: Info-VAX@Mvb.Saic.Com  Subject: Alpha 3000 ...      Hi, F    I got my Alpha 3000 on EBay for a few hundered bucks - very excited# about that. It is on its way to me.   B    Then I thought, hmm what about the power requirements - so doesF anyone know do I need to setup some special amp/voltage in my basementF to get this up and running - anything else I need apart from dry, cool and ideally static free?   thanks!    Tim    ------------------------------    Date: 16 Jan 2003 15:29:33 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)  Subject: Re: Alpha 3000 ... 3 Message-ID: <ZFXJvaKjBQIr@eisner.encompasserve.org>   g In article <a7234bb1.0301161154.2a968559@posting.google.com>, timasmith@hotmail.com (Tim Smith) writes:  > Hi, H >    I got my Alpha 3000 on EBay for a few hundered bucks - very excited% > about that. It is on its way to me.   H  Which model?  My 300 never needed anything more than an ordinary 15 amp  115 Volt plug.    ------------------------------    Date: 16 Jan 2003 19:13:41 -0800' From: timasmith@hotmail.com (Tim Smith)  Subject: Re: Alpha 3000 ... = Message-ID: <a7234bb1.0301161913.337d8d75@posting.google.com>   _ "Hans Vlems" <hvlems@iae.nl> wrote in message news:<b0754k$modbd$1@ID-143435.news.dfncis.de>... 8 > "Tim Smith" <timasmith@hotmail.com> schreef in bericht9 > news:a7234bb1.0301161154.2a968559@posting.google.com...  > > Hi, J > >    I got my Alpha 3000 on EBay for a few hundered bucks - very excited' > > about that. It is on its way to me.  > > F > >    Then I thought, hmm what about the power requirements - so doesJ > > anyone know do I need to setup some special amp/voltage in my basementJ > > to get this up and running - anything else I need apart from dry, cool > > and ideally static free? > > I > The 3000 is about as problematic as an average PC with respect to power  > requirements, humidity' > and temperature, so no worries there. M > The point is, where are you located and where did the system come from? Not  > all Alphas have anC > external switch to change between US and European voltages AFAIK.  >  > Hans  D Cool, thanks for the replies to all.  I'm located in Canada so apart. from being a little cold up I should be ok :-)   ------------------------------  % Date: Thu, 16 Jan 2003 11:31:22 -0800 $ From: Shane Smith <ssmith@icius.com>7 Subject: Alpha, Itanic and Opteron benchmark comparison 0 Message-ID: <01C2BD52.CCA02350@sulfer.icius.com>  F Unfortunately (at least for me) in German, but the picture says enough to be interesting.  4 http://www.heise.de/newsticker/data/as-16.01.03-002/  " It's labeled SAP SD @ tier, 4 way:   HPServer rx5670 Itanium 2   470  Alpha Server ES45           426  Proliant DL590 Itanium      206  IBM xSeries 440             330  AMD Opteron 1.6ghz        ca600   A Don't know what that "ca" means. Maybe one of our German speaking G colleagues could read the article and tell us? Babelfish just sat there  when I tried it.   Shane    ------------------------------  % Date: Thu, 16 Jan 2003 21:07:09 +0100 6 From: =?ISO-8859-1?Q?Arne_Vajh=F8j?= <arne@vajhoej.dk>; Subject: Re: Alpha, Itanic and Opteron benchmark comparison ) Message-ID: <3E2710ED.8050605@vajhoej.dk>    Shane Smith wrote:H > Unfortunately (at least for me) in German, but the picture says enough > to be interesting. > 6 > http://www.heise.de/newsticker/data/as-16.01.03-002/ > $ > It's labeled SAP SD @ tier, 4 way: > ! > HPServer rx5670 Itanium 2   470 ! > Alpha Server ES45           426 ! > Proliant DL590 Itanium      206 ! > IBM xSeries 440             330 ! > AMD Opteron 1.6ghz        ca600  > C > Don't know what that "ca" means. Maybe one of our German speaking I > colleagues could read the article and tell us? Babelfish just sat there  > when I tried it.  > In danish ca. is an abbreviation equal to approx. in english !   ARne   ------------------------------  % Date: Thu, 16 Jan 2003 21:13:10 +0100 6 From: =?ISO-8859-1?Q?Arne_Vajh=F8j?= <arne@vajhoej.dk>; Subject: Re: Alpha, Itanic and Opteron benchmark comparison ) Message-ID: <3E271256.1040803@vajhoej.dk>    Shane Smith wrote:H > Unfortunately (at least for me) in German, but the picture says enough > to be interesting. > 6 > http://www.heise.de/newsticker/data/as-16.01.03-002/ > $ > It's labeled SAP SD @ tier, 4 way: > ! > HPServer rx5670 Itanium 2   470 ! > Alpha Server ES45           426 ! > Proliant DL590 Itanium      206 ! > IBM xSeries 440             330 ! > AMD Opteron 1.6ghz        ca600    So: 
 1. Itanium 2. Alpha
 3. PowerPC 4. x86  : (Opteron is just a chip not a system I can go out and buy)   Arne   ------------------------------  % Date: Thu, 16 Jan 2003 13:05:35 -0800 $ From: Shane Smith <ssmith@icius.com>; Subject: RE: Alpha, Itanic and Opteron benchmark comparison 0 Message-ID: <01C2BD5F.FA8BE490@sulfer.icius.com>  H There is no x86 on that list, there's an early Itanic and an Itanic 2. I= made the same mistake initially: saw "Proliant" thought Xeon.    Shane    -----Original Message-----, From: Arne Vajh=F8j [mailto:arne@vajhoej.dk]) Sent: Thursday, January 16, 2003 12:13 PM  To: Info-VAX@Mvb.Saic.Com ; Subject: Re: Alpha, Itanic and Opteron benchmark comparison      Shane Smith wrote:H > Unfortunately (at least for me) in German, but the picture says enough > to be interesting. >=206 > http://www.heise.de/newsticker/data/as-16.01.03-002/ >=20$ > It's labeled SAP SD @ tier, 4 way: >=20! > HPServer rx5670 Itanium 2   470 ! > Alpha Server ES45           426 ! > Proliant DL590 Itanium      206 ! > IBM xSeries 440             330 ! > AMD Opteron 1.6ghz        ca600    So: 
 1. Itanium 2. Alpha
 3. PowerPC 4. x86  : (Opteron is just a chip not a system I can go out and buy)   Arne   ------------------------------  + Date: Thu, 16 Jan 2003 13:20:56 -0800 (PST) . From: Fabio Cardoso <fabiopenvms@yahoo.com.br>; Subject: Re: Alpha, Itanic and Opteron benchmark comparison @ Message-ID: <20030116212056.40498.qmail@web20210.mail.yahoo.com>  % I had sent this link to my manager !  / The company is planning to buy two IBM x440 !!!    Regards    FC  ) --- Shane Smith <ssmith@icius.com> wrote: H > Unfortunately (at least for me) in German, but the picture says enough > to be interesting. > 6 > http://www.heise.de/newsticker/data/as-16.01.03-002/ > $ > It's labeled SAP SD @ tier, 4 way: > ! > HPServer rx5670 Itanium 2   470 ! > Alpha Server ES45           426 ! > Proliant DL590 Itanium      206 ! > IBM xSeries 440             330 ! > AMD Opteron 1.6ghz        ca600  > C > Don't know what that "ca" means. Maybe one of our German speaking I > colleagues could read the article and tell us? Babelfish just sat there  > when I tried it. >  > Shane  >  >      =====  ========================== Fbio dos Santos Cardoso OpenVMS System Manager Rio de Janeiro - Brazil  fabiopenvms@yahoo.com.br ==========================  2 __________________________________________________ Do you Yahoo!?5 Yahoo! Mail Plus - Powerful. Affordable. Sign up now.  http://mailplus.yahoo.com    ------------------------------  % Date: Thu, 16 Jan 2003 22:25:30 +0100 6 From: =?ISO-8859-1?Q?Arne_Vajh=F8j?= <arne@vajhoej.dk>; Subject: Re: Alpha, Itanic and Opteron benchmark comparison ) Message-ID: <3E27234A.3040403@vajhoej.dk>    Shane Smith wrote:J > There is no x86 on that list, there's an early Itanic and an Itanic 2. I? > made the same mistake initially: saw "Proliant" thought Xeon.    Ah.    I missed that one.  9 [I never thougth the first Itanic's ever hit the streets]    Arne   ------------------------------  % Date: Thu, 16 Jan 2003 21:04:10 +0100 9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> ; Subject: Re: Alpha, Itanic and Opteron benchmark comparison ' Message-ID: <3E27103A.812CCF7F@aaa.com>    ca. = Aprox.   ------------------------------  % Date: Thu, 16 Jan 2003 15:29:00 -0500 > From: "Ray Fusci" <rxfxuxsxcxix(at)xcxhxaxrxtxexrx(dot)xnxext>; Subject: Re: Alpha, Itanic and Opteron benchmark comparison / Message-ID: <v2e57lk2r3bt07@corp.supernews.com>   G "ca" isn't German, it's Latin. It's the abbreviation for "circa", which > means "around". Also"approximately", or "roughly", or "about".  1 "Shane Smith" <ssmith@icius.com> wrote in message * news:01C2BD52.CCA02350@sulfer.icius.com...H > Unfortunately (at least for me) in German, but the picture says enough > to be interesting. > 6 > http://www.heise.de/newsticker/data/as-16.01.03-002/ > $ > It's labeled SAP SD @ tier, 4 way: > ! > HPServer rx5670 Itanium 2   470 ! > Alpha Server ES45           426 ! > Proliant DL590 Itanium      206 ! > IBM xSeries 440             330 ! > AMD Opteron 1.6ghz        ca600  > C > Don't know what that "ca" means. Maybe one of our German speaking I > colleagues could read the article and tell us? Babelfish just sat there  > when I tried it. >  > Shane  >  >    ------------------------------  % Date: Thu, 16 Jan 2003 21:32:05 +0100 1 From: SAP Trainee <Didier.Morandi.nospam@Free.fr> ; Subject: Re: Alpha, Itanic and Opteron benchmark comparison & Message-ID: <3E2716C5.1070109@Free.fr>   ca = circa = more or less  This is latin.   D.   What did I win?    Shane Smith a crit:H > Unfortunately (at least for me) in German, but the picture says enough > to be interesting. > 6 > http://www.heise.de/newsticker/data/as-16.01.03-002/ > $ > It's labeled SAP SD @ tier, 4 way: > ! > HPServer rx5670 Itanium 2   470 ! > Alpha Server ES45           426 ! > Proliant DL590 Itanium      206 ! > IBM xSeries 440             330 ! > AMD Opteron 1.6ghz        ca600  > C > Don't know what that "ca" means. Maybe one of our German speaking I > colleagues could read the article and tell us? Babelfish just sat there  > when I tried it. >  > Shane  >  >    ------------------------------  % Date: Thu, 16 Jan 2003 21:36:14 +0100 1 From: SAP Trainee <Didier.Morandi.nospam@Free.fr> ; Subject: Re: Alpha, Itanic and Opteron benchmark comparison & Message-ID: <3E2717BE.8050700@Free.fr>   Shane Smith a crit:H > Unfortunately (at least for me) in German, but the picture says enough > to be interesting. > 6 > http://www.heise.de/newsticker/data/as-16.01.03-002/ > $ > It's labeled SAP SD @ tier, 4 way: > ! > HPServer rx5670 Itanium 2   470 ! > Alpha Server ES45           426 ! > Proliant DL590 Itanium      206 ! > IBM xSeries 440             330 ! > AMD Opteron 1.6ghz        ca600  > C > Don't know what that "ca" means. Maybe one of our German speaking I > colleagues could read the article and tell us? Babelfish just sat there  > when I tried it.  H Already early prototypes of the AMD Opteron processor knew in four-fold G systems the met competition -- all the same whether 32 or 64-bittig --  E with the SAP SD bench mark in the shade place, and with only 1.6 GHz  F clock (2 GHZ are planned) with the Launch. With the SAP R/3 Sales and C distribution bench mark (SD) becomes typical Business transactions  G (orders to give up, etc. corrects.) accomplished in a large data base.  I As measure serves the number of the SD users, who can be served within a  G given time interval. Bench mark for different hardware configurations,  I approximately for "2 animal 4way servers" are led individually. For this  F for example HP for the Itanium indicates 2 in the rx5670 470 SD user. B First, still few optimized attempts a few weeks ago were with the G quadruple Opteron system with R/3 and DB2-Datenbank under 64-Bit-Linux  G already with 550 SD Usern, which were served of middle period of reply  H with 1.9 seconds, about 600 SD Usern lay the time still scarcely over 2 I seconds. In the meantime the 600-SD-User-Marke was probably exceeded. No  ? bad prospects thus for AMDs entrance with servers with Opteron  C processors. Above diagram is however no longer whole on the newest  H conditions. In the meantime HP "drove" the Itanium-2-System under Hp-ux G likewise on over 600-SD-User. Intel announces meanwhile for the coming  H Madison processor for Sap apo on a four-fold prototype system (with 1,3 K GHz clock) an increase of 28 per cent opposite the current 1-GHz-Itanium 2.    ------------------------------  % Date: Thu, 16 Jan 2003 14:30:38 -0500 * From: Paul Anderson <paul.anderson@hp.com>+ Subject: Re: DCPS "green-bar" look printing 5 Message-ID: <160120031430387634%paul.anderson@hp.com>   7 In article <3E2327DB.638E65B@yahoo.com>, John Johnstone  <jj_usenet2@yahoo.com> wrote:   E > Thanks for the help.  I didn't think to try the HELP.  If the PRINT I > Parameters are in the documentation, I must have missed it.  If they're & > not there, they certainly should be.  C The documentation available online via HELP PRINT_PARAMETERS is not E included in the DCPS documentation.  I think I know why this was done C in the past, but I agree it should be in both places (with a single D source of course).  I'll see if we can do this for a future release.   Paul   --    Paul Anderson   OpenVMS Engineering    Hewlett-Packard Company    ------------------------------  # Date: Thu, 16 Jan 2003 18:33:08 GMT A From: "Colin Butcher" <colin_DOT.butcher_AT@xdelta_DOT.co_DOT.uk> % Subject: Re: DECnet-Plus vs DECnet IV = Message-ID: <EVCV9.6044$qe3.48453853@news-text.cableinet.net>   I Don't know why Lookout Fast (Outlook Express) sent that blank message. Oh 6 well. I probably hit some key inadvertently. Anyway...    K There are a couple of things I wrote a while back which might help you with E some of the background at www.xdelta.co.uk - they're based on a DECUS H session I've given a few times. I should redo it sometime and update theH notes. I might offer it as a one day seminar at the European DECUS eventJ with live hands-on demos. Please let me know (privately, see munged e-mail! address) if there are any takers.   I The problem with multiple NICs is that Phase IV will set all NICs running I DECnet to have the same MAC address - which is based on the Phase IV node L address, so you can't have those NICs connected to the same network segment,L or those network segments interconnected by anything bridging layer 2 stuff.K Interconnecting by routers is OK, but not bridges or layer 2 switches. It's K OK to have Phase IV with multiple NICs if you're running true dual-rail (or I multi-rail) ethernet where the rails are independent or interconnected by > routers - a typical design for very high availability systems.  K This 'Phase IV changes the MAC address' behaviour is behind what's meant by L Phase IV compatible addressing in Phase V - and it's why NET$CONFIGURE won'tK let you (by default) enable Phase IV Compatible addressing on more than one  NIC.  L Node names / addresses are set up using the DECNET_REGISTER executable. NoteI that you can export to a text file, edit the text file and then re-import F the address list. I can't remember the format, but it's pretty easy toL figure out once you see it. I generally advise people to mangle the "addressJ towers" (a set of addresses by which a node can be reached) to only have aJ single address entry (typically either a Phase IV or a Phase V or a DECnetK over IP entry) because otherwise a failure to reach a node with one address I entry can time out, then it tries the next, and so on. This can lead to a K lot of confusion about timers, retransmit factors, timeouts etc. because it F doesn't do what you're expecting it to do - i.e. it takes 2 to 3 timesJ longer to time out than you expected. Also beware that the default timeoutK interval is something like 600 seconds (see NET$CONFIGURE). I usually slash F that down to more like 30, but it's really only a problem in a network* without routers providing routing updates.  J You can run other protocols (DECnet V address format - won't talk to PhaseJ IV machines though because of the MAC address issue, TCP/IP, LAT, SCS, MOPC ...) over the other adapters to get greater overall network traffic I throughput. I typically use a couple of adapters for each protocol if I'm H after serious throughput. Don't forget that Gigabit ethernet is probably
 best for SCS.   G Another common DECnet-Plus (aka Phase V) issue is the naming cache - it I keeps cropping up in questions. The naming cache is 'sticky' over reboots J and will eventually clean itself up. The recommended 'best practice' is toI use the NCL FLUSH SESSION CONTROL NAMING CACHE ENTRY "*" command when you F change any of the naming information. I assume that you're using localL naming option, so whenever you mess with the local namespace content throughK DECNET_REGISTER, even for nodes other than the one acting at the 'executor' J (to borrow some Phase IV terminology), you are advised to flush the namingJ cache so that subsequent namespace lookups aren't retrieved from the cacheH but are retrieved from the local namespace data (and thus populating the  naming cache with the new data).  I I now much prefer Phase V to Phase IV because of all the good information K you can get out of it for tracing problems, but it took a while to get used  to it all. Good luck!    -- Hope this helps. Cheers, Colin. ' (colinDOT.butcherAT@xdeltaDOT.coDOT.uk)     ; "Lyndon Bartels" <lbartels@pressenter.com> wrote in message ( news:3E25ADDF.56C069BC@pressenter.com... > Hello, > J > I've been using DECnet Phase IV on all my machines. But it seems that to@ > accomplish some of my goals, I need to move to DECnet Plus.... > E > I just finished an install of 7.3-1 and 'Plus' on my workstation to H > start testing things out. So I haven't dug into any documentation yet. >  > 	 > But....  > D > Is there any documentation talking about the comparisons? Just theH > rudimentery stuff... If I wanted to define a node on IV. What would be > the analaous stuff on Plus.  >  > G > Any pointers would be great. I hate calling up CSC for questions like > > that. I'll be reading docs for a long time before I do that. >  >  >  > Thanks in advance, >  > Lyndon >  >  > --I > My opinions are mine and mine alone. They seldom align with those of my  > employer.  >  > J > The only good thing about putting the cart before the horse is you don't > have > to look at the horse's butt.   ------------------------------  # Date: Thu, 16 Jan 2003 23:41:23 GMT ! From: rob.buxton@wcc.spam.govt.nz % Subject: Re: DECnet-Plus vs DECnet IV % Message-ID: <3e2741b7.967268208@news>   2 On Wed, 15 Jan 2003 18:52:16 -0600, Lyndon Bartels  <lbartels@pressenter.com> wrote:   >Hello,  > I >I've been using DECnet Phase IV on all my machines. But it seems that to ? >accomplish some of my goals, I need to move to DECnet Plus....  > D >I just finished an install of 7.3-1 and 'Plus' on my workstation toG >start testing things out. So I haven't dug into any documentation yet.  >  > 	 >But....   > C >Is there any documentation talking about the comparisons? Just the G >rudimentery stuff... If I wanted to define a node on IV. What would be  >the analaous stuff on Plus. >  > F >Any pointers would be great. I hate calling up CSC for questions like= >that. I'll be reading docs for a long time before I do that.  >   F A couple of quick points as there's a ton of useful stuff in the other	 postings. 9 Adding nodename translations is done by using the program ( sys$system:decnet_register. Just run it.  8 Another command I've found useful, define a symbol e.g.   7   DNM == "$SYS$UPDATE:DECNET_MIGRATE CONVERT COMMAND ""    And then you can do    NodeA  dnm show know obj   2 ! *** Converting the following NCP command to NCL: !     SHOW KNOW OBJ 4 SHOW NODE 0 SESSION CONTROL APPLICATION * ALL STATUS  E There's also an X front end to ncl, which I played with briefly but I2C no longer use and no longer have it set up under excursion so can'tn
 give details.*   HTH*   Rob.   ------------------------------  % Date: Thu, 16 Jan 2003 19:24:31 -0600i@ From: "David J. Dachtera" <djesys.nospam@earthlink.spamfree.net>% Subject: Re: DECnet-Plus vs DECnet IVh6 Message-ID: <3E275B4F.DB4F51B8@earthlink.spamfree.net>   Lyndon Bartels wrote:  >  > David J. Dachtera wrote: > >e > > Lyndon Bartels wrote:h > > >i > > > Hello, > > >rN > > > I've been using DECnet Phase IV on all my machines. But it seems that toD > > > accomplish some of my goals, I need to move to DECnet Plus.... > >r > > Such as ... ?l > >I@ > > That is, what do you need to do that DECnet-IV can't handle? > >e > I > I want to add a second pair if NICS to my servers. Retain TCP/IP on thehI > "public" (read "current") network paths. Create a "private" network foreE > DECnet, and eventually SCS traffic. This will make the network guys  > happy. > . > BUT.... I have to have everything redundant. > E > I was told by CSC that in order to accomplish this, I'd need DECnet  > Plus.   F If you want to have both DECnet circuits active at the same time, thatG may be true. Otherwise, in DECnet-IV you can set up both lines/circuitsED and DEFINE one STATE ON and the other STATE OFF, and make the switch' manually when the active circuit fails.a   --   David J. Dachterao dba DJE Systemsr http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/p   ------------------------------  % Date: Thu, 16 Jan 2003 20:16:12 -0600o. From: Lyndon Bartels <lbartels@pressenter.com> Subject: ES40 Hang. Message-ID: <3E27130B.1EFFE0F8@pressenter.com>   I had an interesting happening.   H I had one of my ES40s hang. I tried to force a crash at the console, butH it was so hung, that wouldn't work. It's cluster-mate lost connection to
 it as well...E    # I'm running v7.2-1. 4 cpus. 6G Mem.o  H Each ES40 has redundant NICs, Redundant CIPCAs, and redundant KGPSA-CAs.    $ I had to press the reset button....   3 Anyway, because of that, I didn't get a crash-dump.      So I have two questions..n   1.  Any ideas what happened?   2.  What things can I look at? -     Thanks in advance,   Lyndon     --  G My opinions are mine and mine alone. They seldom align with those of my>	 employer.R    H The only good thing about putting the cart before the horse is you don't have to look at the horse's butt.   ------------------------------  + Date: Fri, 17 Jan 2003 03:07:09 +0000 (UTC)m5 From: "John Wallace" <johnwallace4@yahoo.dotco.dotuk>t Subject: Re: ES40 Hang0 Message-ID: <b07s0t$6km$1@sparta.btinternet.com>  ; "Lyndon Bartels" <lbartels@pressenter.com> wrote in messagew( news:3E27130B.1EFFE0F8@pressenter.com...! > I had an interesting happening.o >sJ > I had one of my ES40s hang. I tried to force a crash at the console, butJ > it was so hung, that wouldn't work. It's cluster-mate lost connection to > it as well...= >= >=% > I'm running v7.2-1. 4 cpus. 6G Mem.  >tJ > Each ES40 has redundant NICs, Redundant CIPCAs, and redundant KGPSA-CAs. >= >=% > I had to press the reset button...._ >_5 > Anyway, because of that, I didn't get a crash-dump.w >u >A > So I have two questions..p >m > 1.  Any ideas what happened? >-  > 2.  What things can I look at? >  >3 > Thanks in advance, >D > Lyndon >e >  > --I > My opinions are mine and mine alone. They seldom align with those of myA > employer.e >g >DJ > The only good thing about putting the cart before the horse is you don't > have > to look at the horse's butt.   ------------------------------  + Date: Fri, 17 Jan 2003 03:15:18 +0000 (UTC)i5 From: "John Wallace" <johnwallace4@yahoo.dotco.dotuk>e Subject: Re: ES40 Hang0 Message-ID: <b07sg6$715$1@sparta.btinternet.com>    ooops, trackpad trouble again...   What I meant to say was...  I The underlying hardware cause for a PCI Alpha hang like that is generallydI that some PCI-related device is misbehaving; a PCI transaction is neitherwK completing nor being timed out and consequently the PCI bus(es) is/are hung.E . All I/O goes via PCI, so when that happens there is no input and no4$ output, no control P, no crash dump.  G The operator log on the other node may tell you a little about what wasm& going on. It probably won't be enough.  * All other suggestions gratefully received.   regards  john  @ "John Wallace" <johnwallace4@yahoo.dotco.dotuk> wrote in message* news:b07s0t$6km$1@sparta.btinternet.com... >a= > "Lyndon Bartels" <lbartels@pressenter.com> wrote in messagec* > news:3E27130B.1EFFE0F8@pressenter.com...# > > I had an interesting happening.r > >.L > > I had one of my ES40s hang. I tried to force a crash at the console, butL > > it was so hung, that wouldn't work. It's cluster-mate lost connection to > > it as well...m > >i > >l' > > I'm running v7.2-1. 4 cpus. 6G Mem.r > >sL > > Each ES40 has redundant NICs, Redundant CIPCAs, and redundant KGPSA-CAs. > >  > >c' > > I had to press the reset button....l > >t7 > > Anyway, because of that, I didn't get a crash-dump.n > >  > >- > > So I have two questions..  > >t  > > 1.  Any ideas what happened? > > " > > 2.  What things can I look at? > >n > >r > > Thanks in advance, > >p
 > > Lyndon > >  > >  > > --K > > My opinions are mine and mine alone. They seldom align with those of myo
 > > employer.i > >  > >tL > > The only good thing about putting the cart before the horse is you don't > > have  > > to look at the horse's butt. >b >N   ------------------------------  % Date: Thu, 16 Jan 2003 22:01:15 -0600y0 From: arturo saavedra <arturo.saavedra@wcom.com> Subject: RE: ES40 HangB Message-ID: <NDBBJFNBJJHKADILJHIJKELNLAAA.arturo.saavedra@mci.com>  I We had this issue on all our es40s over a period of 1 year.  Turned out =oH we needed to replace every single CPU due to rev problems with shipped = CPUs  F Here is our current info.. can't remember what we used to have.. I'd =C have to search 2 gig worth of mail :) Which i'll try to do shortly.                System Configuration:v ---------------------  System Information:tH System Type   Compaq AlphaServer ES40                   Primary CPU ID = 00H Cycle Time    1.5 nsec (666 MHz)                        Pagesize       =	 8192 Byter   Memory Configuration: D Cluster    PFN Start    PFN Count         Range (MByte)        UsageF  #00             0          256         0.0 MB -     2.0 MB    ConsoleE  #01           256       130731         2.0 MB -  1023.3 MB    System F  #02        130987           85      1023.3 MB -  1024.0 MB    ConsoleE  #03        131072       262140      1024.0 MB -  3071.9 MB    System F  #04        393212            4      3071.9 MB -  3072.0 MB    Console  # Per-CPU Slot Processor Information: 7 CPU ID         00                        CPU State    =( rc,pa,pp,cv,pv,pmv,plo& CPU Type       EV67  Pass 2.6 (21264A)G PAL Code       1.96-103                  Halt PC      00000000.20000000dG CPU Revision   ....                      Halt PS      00000000.00001F00m= Serial Number  				     Halt Code    "Bootstrap or Powerfail"uD Console Vers   V6.2-1                    Halt Request "Default, No = Action"     7 CPU ID         01                        CPU State    =- rc,pa,pp,cv,pv,pmv,pl & CPU Type       EV67  Pass 2.6 (21264A)G PAL Code       1.96-103                  Halt PC      00000000.00000000dG CPU Revision   ....                      Halt PS      00000000.00001F00iE Serial Number                            Halt Code    "Bootstrap or =a
 Powerfail"D Console Vers   V6.2-1                    Halt Request "Default, No = Action"o  7 CPU ID         02                        CPU State    =* rc,pa,pp,cv,pv,pmv,pl>& CPU Type       EV67  Pass 2.6 (21264A)G PAL Code       1.96-103                  Halt PC      00000000.00000000 G CPU Revision   ....                      Halt PS      00000000.00001F00bE Serial Number                            Halt Code    "Bootstrap or = 
 Powerfail"D Console Vers   V6.2-1                    Halt Request "Default, No = Action"t  7 CPU ID         03                        CPU State    =u rc,pa,pp,cv,pv,pmv,plt& CPU Type       EV67  Pass 2.6 (21264A)G PAL Code       1.96-103                  Halt PC      00000000.00000000oG CPU Revision   ....                      Halt PS      00000000.00001F00n           -----Original Message-----: From: John Wallace [mailto:johnwallace4@yahoo.dotco.dotuk]( Sent: Thursday, January 16, 2003 9:15 PM To: Info-VAX@Mvb.Saic.Comu Subject: Re: ES40 Hang      ooops, trackpad trouble again...   What I meant to say was...  A The underlying hardware cause for a PCI Alpha hang like that is =B	 generallyoC that some PCI-related device is misbehaving; a PCI transaction is =o neithertH completing nor being timed out and consequently the PCI bus(es) is/are = hungD  All I/O goes via PCI, so when that happens there is no input and no$ output, no control P, no crash dump.  G The operator log on the other node may tell you a little about what wasl& going on. It probably won't be enough.  * All other suggestions gratefully received.   regardsr john  @ "John Wallace" <johnwallace4@yahoo.dotco.dotuk> wrote in message* news:b07s0t$6km$1@sparta.btinternet.com... >a= > "Lyndon Bartels" <lbartels@pressenter.com> wrote in messaget* > news:3E27130B.1EFFE0F8@pressenter.com...# > > I had an interesting happening.e > >aJ > > I had one of my ES40s hang. I tried to force a crash at the console, = buto@ > > it was so hung, that wouldn't work. It's cluster-mate lost =
 connection toa > > it as well...5 > >- > > ' > > I'm running v7.2-1. 4 cpus. 6G Mem.  > >cC > > Each ES40 has redundant NICs, Redundant CIPCAs, and redundant =.
 KGPSA-CAs. > >i > > ' > > I had to press the reset button....n > > 7 > > Anyway, because of that, I didn't get a crash-dump.  > >  > >  > > So I have two questions..o > >e  > > 1.  Any ideas what happened? > >n" > > 2.  What things can I look at? > >m > >e > > Thanks in advance, > >i
 > > Lyndon > >e > >e > > --J > > My opinions are mine and mine alone. They seldom align with those of = my
 > > employer.  > >e > >,H > > The only good thing about putting the cart before the horse is you = don'td > > have  > > to look at the horse's butt. >3 >1   ------------------------------  % Date: Thu, 16 Jan 2003 14:42:37 -0500B3 From: "Homer J. Simpson" <hsimpson@burnsenergy.com>h& Subject: FA: Systems & Options Catalog6 Message-ID: <9QDV9.33350$Jm2.15209@news.bellsouth.net>  / part 2 (final) of the collection is now online:a   Jan-Mar '91nJ http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=616&item=2905775148   Apr-Jun '91nJ http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=616&item=2906082959   Oct-Dec '91 J http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=616&item=2906344449   Supplement '92J http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=616&item=2906355144   Nov '92uJ http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=616&item=2906526223   Apr '93TJ http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=616&item=2905720220   Nov '93@K http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=3519&item=2301271438n   Jan. '95J http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=616&item=2905652541   Dec '95 SupplementJ http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=616&item=2906856285   Nov '96gK http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=3519&item=2301644528r   May '97 SupplementK http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=3519&item=2301646213h  + Intel-based Server Config Guide Mar-Apr '96 K http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=3519&item=2300432032h  + Intel-based Server Config Guide Sep-Dec '96h= http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=2301718691g   ------------------------------  % Date: Thu, 16 Jan 2003 14:31:17 -040000 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>( Subject: Foreign mail question (code 19)/ Message-ID: <3E26FA74.126CB55B@vl.videotron.ca>e  0 A few questions about the foreign mail protocol:  > I get an function code value of 19  which I do not know about.  N It is sent after the message has been written, between the item code of 4 (the/ subject) and item code of 5 (the message text).n  I Anyone have any information on this ? The short/incomplete document I had B found describing the foreign mail protocol doesn't talk about it.   K Also, I find myself bit at a loss with regards to the passing back of errorEN messages if one recipient failed. (during item codes 2 (validate) and 6 (send)N ). Something about passing back the address of a routine which will supply theI error text. Any examples ? (why not just return a pointer to a descriptora containing the text?).  G Also, while I am at it.  The protocol works this way: you have a single K subroutine that is called repeatedly with different item codes. Using the C K language, is the only way to keep context between calls to use the "static"t8 keyword to keep those structures "alive" between calls ?  I Or is there a way to store a pointer to dynamically allocated memory at a03 specific offset to the "context" provided by MAIL ?a   ------------------------------  # Date: Thu, 16 Jan 2003 22:16:31 GMT ' From: Colin Blake <colin@theblakes.com>o. Subject: GENERIC_POLL messages in Mozilla/CSWB, Message-ID: <3E272F3E.4060605@theblakes.com>  I If you are using Mozilla or CSWB (any version) and you frequently see an  I abundance of GENERIC_POLL messages accompanied by frequent hangs, please - read on.  H If your browser frequently fails to load a page, or hangs, and messages H similar to the following appear on the screen, then you are most likely D being affected by a networking problem which only seems to manifest % itself on relatively fast processors:o  8 GENERIC_POLL: unknown condition, errno=16 vaxc$errno=7088 GENERIC_POLL: unknown condition, errno=4 vaxc$errno=20961 GENERIC_POLL: warning, unable to find the bad FDs 4 GENERIC_POLL: warning, expected 1 events but found 2  H The following appears to prevent the problem, but disables IPv6 support ? (therefore you can not use this workaround if you require IPv6 A3 functionality). Edit MOZILLA.COM and find the line:8  4 $ define /user GETIPNODEBYNAME TCPIP$GETIPNODEBYNAME   Change the definition to:k  + $ define /user GETIPNODEBYNAME NO_SUCH_NAMES   Then restart Mozilla.c  I Please provide feedback. I'd like to know if this improves your browsing m experience.    Colin Blakem OpenVMS Engineeringf   ------------------------------   Date: 16 Jan 2003 22:57 CDTi' From: carl@gerg.tamu.edu (Carl Perkins)hS Subject: Re: How much disk I/O can I do with a sys$QIOW IO$_WRITEVBLK/IO$_READVBLK?p- Message-ID: <16JAN200322570738@gerg.tamu.edu>o  7 "Richard Maher" <maher_rj@hotspamnotmail.com> writes...d5 >; (c) Copyright Tier3 Software. All rights reserved.0 >;A >;     Ownership of this software and all associated intellectualuA >;     property rights remain vested in Tier3 Software Ltd.  This A >;     software  or any other copies thereof  may not be provided 7 >;     or otherwise made available to any other person.e  H I hate to tell you this (well, not really), but you need to sue yourselfJ for copyright violation. It seems that you forgot the "without permission"D part that such notices normally include. As it is, you yourself have( violated the terms you put on the thing.   --- Carl   ------------------------------  # Date: Thu, 16 Jan 2003 23:04:21 GMTg# From: "John Smith" <a@nonymous.com>.6 Subject: Re: HP Sets the Stage for Alpha's Last HurrahA Message-ID: <VTGV9.18$ej1.0@news02.bloor.is.net.cable.rogers.com>e  D "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com> wrote in0 message news:3e26edfe$1_2@hpb10302.boi.hp.com... > 5 > > and the recent Inquirer article purports to quoterD > > an internal cHumPaq source as saying that indeed EV7 would *not* haveD > > shipped save for some government commitments with teeth in them. Now, I+ > > don't presume to vouch for that source,r > E > Inquirer reports should always be taken with a large grain of salt.h I haveB > seen everything from remarkably accurate statements that imply a theft of@ > internal slides, to statements that can only have been made by people; > without a clue, probably based on some third hand hallwayo conversation, and & > also by people with an axe to grind.    F Isn't there that old New England saying, "Where there's smoke, there's fire"?   ------------------------------  % Date: Thu, 16 Jan 2003 19:05:16 -0400u0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>6 Subject: Re: HP Sets the Stage for Alpha's Last Hurrah/ Message-ID: <3E273A9A.BE5C5B82@vl.videotron.ca>    John Smith wrote:   tH > Isn't there that old New England saying, "Where there's smoke, there's > fire"?    N Not necessarily. It could be smoledering, producing lost of smokes and mirrors) well before the fire is actually started.    ------------------------------  % Date: Thu, 16 Jan 2003 19:05:44 -05000* From: "Bill Todd" <billtodd@metrocast.net>6 Subject: Re: HP Sets the Stage for Alpha's Last Hurrah2 Message-ID: <vKSdneAHPJd71bqjXTWcpQ@metrocast.net>  . "John Smith" <a@nonymous.com> wrote in message; news:VTGV9.18$ej1.0@news02.bloor.is.net.cable.rogers.com...  >oF > "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com> wrote in2 > message news:3e26edfe$1_2@hpb10302.boi.hp.com... > >e7 > > > and the recent Inquirer article purports to quote F > > > an internal cHumPaq source as saying that indeed EV7 would *not* > haveF > > > shipped save for some government commitments with teeth in them. > Now, I- > > > don't presume to vouch for that source,n > >uG > > Inquirer reports should always be taken with a large grain of salt.  > I haveD > > seen everything from remarkably accurate statements that imply a
 > theft ofB > > internal slides, to statements that can only have been made by > people= > > without a clue, probably based on some third hand hallwayn > conversation, ande( > > also by people with an axe to grind. >h >eH > Isn't there that old New England saying, "Where there's smoke, there's > fire"?  L Fred's assessment of L'Inq strikes me as directly on target:  sometimes MikeL sees real smoke and reports it, others he just blows it.  In this particularK case the content is believable and alleged to come from an internal source,u; but that's a considerable distance away from hard evidence.k   - bill   ------------------------------   Date: 17 Jan 2003 00:05:43 GMT' From: prosullivan@aol.com (PROSULLIVAN) 4 Subject: Re: Independent VMS crash analysis service.9 Message-ID: <20030116190543.17493.00000051@mb-bd.aol.com>   N Two problems with this a) the only people who would use these services are notN those under CPQ VMS support. Such people would go to third party suppliers andH get them to do the software support. You should try and consult to theseM providers as well as customers, buy offering, hey an on-line web-based experth! system they can use (for a fee). 0  K b) is obvious. The VMS market is not growing. Faced between porting off VMSnH onto Itanium away from Alpha, most corporation think 1) how long has the; application got to live? 2) can it be ported to other OS's?t  N How many 'New VMS customer announcements' do you get from HP/CPQ per year? 10?B 5? (on the installed (some would be say "marooned' 400k user base)   good luck, nonetheless.    ------------------------------  + Date: Thu, 16 Jan 2003 13:22:46 -0800 (PST) . From: Fabio Cardoso <fabiopenvms@yahoo.com.br>% Subject: Re: Is Galaxy an EXOKERNEL ? @ Message-ID: <20030116212246.40820.qmail@web20210.mail.yahoo.com>  = I think the only one commercial product which is a EXOKERNEL  = is VMWare ESX ! At least, the documentation is similar to the  EXOKERNEL philosophy.=   Regards=   FCD --- Fred Kleinsorge <kleinsorge@star-dot-zko-dot-dec-dot-com> wrote:I > I don't think so, although I'm not entirely sure that I have a complete 5 > grasp of what the MIT folks mean by the expression.o > F > My interpretation is that the exokernel provides the direct hardwareN > abstraction, which would imply to me an interface to all HW devices.  On topG > of this, you create a "user mode" UNIX abstraction - one which can be J > bypassed by the application, and can directly talk to this new exokernelM > abstraction.  Hence, it looks like an attempt to be able to create a "fast"i9 > single application system, with UNIX-like abstractions.s >  > That isn't Galaxy in my mind.  >  > = > "Fabio Cardoso" <fabiopenvms@yahoo.com.br> wrote in messager< > news:20030116161100.76139.qmail@web20210.mail.yahoo.com...
 > > Dear Sirs  > >e* > > Is the Galaxy software an EXOKERNEL  ? > >k > >o > > Regards6 > >l > > FC > >c	 > > =====c > > ========================== > > Fbio dos Santos Cardoso > > OpenVMS System Manager > > Rio de Janeiro - Brazil- > > fabiopenvms@yahoo.com.br > > ========================== > > 6 > > __________________________________________________ > > Do you Yahoo!?9 > > Yahoo! Mail Plus - Powerful. Affordable. Sign up now.- > > http://mailplus.yahoo.come >  >      =====n ========================== Fbio dos Santos Cardoso OpenVMS System Manager Rio de Janeiro - Brazilo fabiopenvms@yahoo.com.br ==========================  2 __________________________________________________ Do you Yahoo!?5 Yahoo! Mail Plus - Powerful. Affordable. Sign up now.c http://mailplus.yahoo.comn   ------------------------------  % Date: Thu, 16 Jan 2003 18:58:49 -0500 * From: "Bill Todd" <billtodd@metrocast.net>6 Subject: Re: Itanium with alpha goodies on the way ...2 Message-ID: <x4edncBp06Pa2rqjXTWcpQ@metrocast.net>  5 "Bob Ceculski" <bob@instantwhip.com> wrote in message 6 news:d7791aa1.0301160720.bb0b8d5@posting.google.com...( > Intel confirms dual core Itanium ahead >u > Battle of the "brains" >u0 > By Mike Magee: Thursday 16 January 2003, 09:37 >  >h > from the inquirer ...- >-E > CHIP GIANT INTEL will release a version of the Itanium in 2005 thatrG > includes two processors in the same package - confirming reports fromt > last year.H > And, apparently, according to Mike Kanellos at news.com it will up theC > level three cache on Itanium Madison processors to 9MB next year.c >:H > That's not shown on current Intel roadmaps seen by PC manufacturers soE > far. Madison and Deerfield are launched in the middle of this year,.H > and the cache for the former is as much as 6MB and will run at a speed > of 1.5GHz. > H > According to the Kanellos piece, the Intel Montecito Itanium will haveE > the dual processor cores. Recently we reported that AMD could scale 7 > its future Hammer technology to a two core model too.e >nH > Montecito will be made on a 90 nanometer process. It is this chip thatH > is the first to be likely to include elements from the Alpha processor= > that Intel magically swiped off Compaq before HP bought it.  > 
 > See Also* > Itanic will use Alpha, EV8, SMT features  D I'm afraid you're still just very confused, Bob (and still intent onK parading your confusion for the world to snicker at).  But at least in thisGK case you've got a bit of company, because Mike clearly hasn't had much of aiG clue about when Alpha technology could possibly first show up in Itanic  either.   $ So I'll set you straight once again:  K The processor core for Madison, Deerfield, and Montecito is essentially therL same as McKinley's - absolutely zero Alpha influence there.  The differencesK between these 4 chips are primarily in the amount of on-chip cache, so zerorJ Alpha influence there as well.  The 2005 version of Montecito has now beenL announced to have dual cores, but with essentially the same system interfaceK (chip pinout) as all its siblings, so zero Alpha influence there too (I hadtK thought 18 or so months ago that Intel might have been able at least to getZG some of the EV7 on-chip glue into its 2005 product, but that would havebI dramatically changed the chip pinout and it now seems you'll have to wait L yet another generation to see that, let alone any Alpha contributions to the actual processor core).   F Since you've proved completely ineducable for years now I really don'tL expect any of the above to sink in, but it seemed appropriate to clarify theC situation for others who might not realize the degree to which your  statements are complete drivel.s   - bill   ------------------------------  % Date: Thu, 16 Jan 2003 21:21:45 -0500w* From: "Bill Todd" <billtodd@metrocast.net>  Subject: Montecito slips to 20052 Message-ID: <waKdnQ_9Y69V9bqjXTWcog@metrocast.net>  ) http://news.com.com/2100-1001-980898.htmll  I This article is as good an example of spin as I've ever seen.  Instead oflL questioning what caused Intel to slip Montecito (the Itanic shrink to 90 nm,L with 12 MB of on-chip cache) from 2004 to 2005, it paints the new roadmap as2 a significant march stolen on Intel's competition.  L But the facts, if one examines them, tell a very different story.  DeferringA the shrink for a year has a major negative impact on yield, powereE consumption, cache size, and performance for that year.  Adding a newcJ Madison variant (still in 130 nm) with an additional 3 MB of on-chip cacheL to fill the gap increases the negative impact on yield and power consumptionL (especially if it's clocked faster, as claimed), and the 50% added cache mayL or may not be all that helpful (they claim it will clock faster than MadisonK I, but that only drives up power consumption even more).  And claiming that J they've moved up the dual-core Itanic by two years may well be an outrightK lie, since around a year ago when Montecito in 2004 was added to the publiceK roadmap Intel mentioned that a 'bump' was scheduled for 2005 - and while atCI the time I thought it might be EV7-style on-chip glue, the claim that therA dual-core Montecito will retain chip-interface compatibility with L McKinley/Madison/Deerfield removes that option, so what's left but to assume: that the dual-core variant was planned for 2005 all along?  J Bottom line:  Itanic performance won't be improving at the rate previouslyI anticipated (and, incidentally, this also implies that there won't be anyCD EV7-style on-chip glue until at least 2006).  Funny - several ItanicK boosters were quick to suggest that EV8 might have slipped (despite lack oflF any actual problem to point to), but none of them seemed to be able toK conceive of the possibility that the rosy Itanic roadmap they'd bought intow might do the same.   - bill   ------------------------------  % Date: Thu, 16 Jan 2003 20:48:49 -0600m@ From: "David J. Dachtera" <djesys.nospam@earthlink.spamfree.net>$ Subject: Re: Montecito slips to 20056 Message-ID: <3E276F11.82B87859@earthlink.spamfree.net>   Bill Todd wrote: > + > http://news.com.com/2100-1001-980898.html  > K > This article is as good an example of spin as I've ever seen.  Instead ofEN > questioning what caused Intel to slip Montecito (the Itanic shrink to 90 nm,N > with 12 MB of on-chip cache) from 2004 to 2005, it paints the new roadmap as4 > a significant march stolen on Intel's competition. > N > But the facts, if one examines them, tell a very different story.  DeferringC > the shrink for a year has a major negative impact on yield, poweruA > consumption, cache size, and performance for that year.  [snip]e  H Seems congruent with reports today that Intel is cutting its chip-making budget for the next year.   B Yet more evidence that The AlphaCide was a "bungle in the jungle"?   -- m David J. Dachterai dba DJE Systemst http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/n   ------------------------------  % Date: Thu, 16 Jan 2003 23:16:24 -0400c0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>$ Subject: Re: Montecito slips to 2005/ Message-ID: <3E277586.B43C15C4@vl.videotron.ca>b   Bill Todd wrote: > + > http://news.com.com/2100-1001-980898.htmle  J Interesting. The heading says Intel is moving Montecito ahead from 2007 toN 2005. But later in the article, it is written Montecito was originally planned: for 2004. So in the end, they have delayed it by one year.  J While, in the end, performance is what counts, I am puzzled as to why IA64J needs so much cache to achieve its performance. Is IA64 so flawed that theI only way to achieve palatable performance to load the chip up with a hugegK about of fast cache memory ? Or do other processors have the same amount of  cache memory ?   ------------------------------  % Date: Fri, 17 Jan 2003 00:33:36 -0500G* From: "Bill Todd" <billtodd@metrocast.net>$ Subject: Re: Montecito slips to 20052 Message-ID: <dFydnagSI9UjCLqjXTWcog@metrocast.net>  = "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in messagen) news:3E277586.B43C15C4@vl.videotron.ca...e > Bill Todd wrote: > >,- > > http://news.com.com/2100-1001-980898.htmlp >sL > Interesting. The heading says Intel is moving Montecito ahead from 2007 toH > 2005. But later in the article, it is written Montecito was originally planned < > for 2004. So in the end, they have delayed it by one year. >CL > While, in the end, performance is what counts, I am puzzled as to why IA64L > needs so much cache to achieve its performance. Is IA64 so flawed that theK > only way to achieve palatable performance to load the chip up with a huge J > about of fast cache memory ? Or do other processors have the same amount of > cache memory ?  K No, they have from considerably less to a hell of a lot less:  EV68 had 128sJ KB on chip (though up to 16 MB off chip, which is supposedly over half theH speed of the new EV7 on-chip L2), EV7 has 1.75 MB on chip, POWER4 and 4+I have about 1.5 MB on chip, Athlons have 384 KB on chip, Pentiums have 256 L KB - 512 KB on chip, Xeons have 512 KB - 2 MB on chip, Hammers will have 3842 KB - 1 MB on chip, PA-RISC has 1.5 MB on chip, ...  H I've heard it said that one reason Itanic requires so much fast cache isJ because it's an in-order processor, which means that any cache miss stalls@ the processor entirely while it waits for main memory (unlike anB out-of-order processor which can often keep crunching ahead in theH instruction stream executing independent instructions that don't requireG uncached data while the request to main memory completes).  That soundsaJ reasonable, at any rate.  SPARC is an in-order processor as well, and only2 recently got a decent-sized on-chip cache (1 MB?).   - bill   ------------------------------    Date: 16 Jan 2003 13:17:06 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) 7 Subject: Re: Most common OpenVMS versions in use today?t3 Message-ID: <rRT1UgH4Iob4@eisner.encompasserve.org>   g In article <c113b52c.0301160947.755a9d0d@posting.google.com>, jbecker@ui.urban.org (Jim Becker) writes:b  D > The current Encompass instapoll, running today and running for two& > weeks, addresses this question. See: > http://www.encompassus.org/a  F Netscape Communicator's "search" capability, as well as my own survey,8 tells me the string "poll" does not appear on that page.   ------------------------------  % Date: Thu, 16 Jan 2003 11:22:34 -0800 $ From: Shane Smith <ssmith@icius.com>7 Subject: RE: Most common OpenVMS versions in use today?00 Message-ID: <01C2BD51.BA71B550@sulfer.icius.com>  C They haven't used the word "poll", but there is a box full of radio8B buttons under a heading "Which version of OpenVMS are you running,H predominantly?". They only have major versions listed though, 7.x Alpha," 6.x Alpha, 1.x Alpha, 7.x VAX etc.  C Clicking the view results link, 85.3% of the answers are 7.x at theu( moment. Next highest is 8.8% on 6.x VAX.   Shaned   -----Original Message-----: From: Kilgallen@SpamCop.net [mailto:Kilgallen@SpamCop.net]) Sent: Thursday, January 16, 2003 11:17 AMh To: Info-VAX@Mvb.Saic.Comn7 Subject: Re: Most common OpenVMS versions in use today?     = In article <c113b52c.0301160947.755a9d0d@posting.google.com>,t) jbecker@ui.urban.org (Jim Becker) writes:e  D > The current Encompass instapoll, running today and running for two& > weeks, addresses this question. See: > http://www.encompassus.org/s  F Netscape Communicator's "search" capability, as well as my own survey,8 tells me the string "poll" does not appear on that page.   ------------------------------    Date: 16 Jan 2003 15:10:32 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)N7 Subject: RE: Most common OpenVMS versions in use today? 3 Message-ID: <xDVKukbtnoF3@eisner.encompasserve.org>l  W In article <01C2BD51.BA71B550@sulfer.icius.com>, Shane Smith <ssmith@icius.com> writes: E > They haven't used the word "poll", but there is a box full of radiorD > buttons under a heading "Which version of OpenVMS are you running, > predominantly?".  / Ok, I see -- off the right edge of the page :-)-   ------------------------------  % Date: Thu, 16 Jan 2003 20:57:02 +0100  From: "stm" <stm@pt.lu>s Subject: NT on Alpha% Message-ID: <3e270e8d$1_1@news.vo.lu>-  8 I have an Alpha Server 1000 with VMS 7.2  (as Hobbyist).( I want to try to put NT on this server .( Can both systems run on the same server?$ What's the best way to implement NT.A I tried some official Web sites for NT or VMS but without result.   
 Need help....s   Rgds,  STM,   ------------------------------  % Date: Thu, 16 Jan 2003 21:31:55 +0100t" From: "Hans Vlems" <hvlems@iae.nl> Subject: Re: NT on Alpha5 Message-ID: <b07523$lpmn1$1@ID-143435.news.dfncis.de>r  H Even though it is beyond me why you'd want this, I'll try and answer the	 question. H The AS1000 uses different console firmware than TRU64 and VMS. IIRC it's called ARC. D ARC is entirely menu driven. The AS1000 supports both console modes.J One can change between the two, the procedure is listed in the AS1000 user guide. The AS1200 user guide states:i   >>> set os_type NT
 >>> alphabios   I This will put the system in ARC mode on subsequent reboots. If you do notI' want that omit the set os_type command.aK Once in ARC mode (or AlphaBIOS mode) you can return to SRM (the >>> prompt). in the window calledJ Advanced CMOS Setup. Select either OpenVMS Console or DIGITAL UNIX Console switch off; wait 30 secs3 and power on the unit. Next set os_type to openVMS.,   Hans  D "stm" <stm@pt.lu> schreef in bericht news:3e270e8d$1_1@news.vo.lu...: > I have an Alpha Server 1000 with VMS 7.2  (as Hobbyist).( > I want to try to put NT on this server* > Can both systems run on the same server?& > What's the best way to implement NT.C > I tried some official Web sites for NT or VMS but without result.a >e > Need help....u >n > Rgds,  > STM  >M >o   ------------------------------  % Date: Thu, 16 Jan 2003 13:00:17 -0800.$ From: Shane Smith <ssmith@icius.com> Subject: RE: NT on Alpha0 Message-ID: <01C2BD5F.47D0E350@sulfer.icius.com>  @ To take a quote from a recent post out of context: "Danger, WillB Robinson!!!". Windows NT reaches out to all available disks duringF installation, and "marks" them. It claims this "mark" is harmless, andB it is to other MS operating systems, but I've seen reports that itF totally borks VMS format disks. Before starting, back up the VMS disksH to external media. You'll probably have to reinitialize and restore them after NT installation.  F You'll need a separate disk for the NT system, and you'll need to haveH the NT compatible version of the firmware installed. I don't know if theE 1000 has a big enough eprom to hold both, if it doesn't you'd have tosG flash the eprom every time you swapped OS's. This would be a Bad Thing.a  G I did install NT on an AS200 once, but my memory is dim on the subject.m> I'll shut up now before I start setting you on the wrong path.   Shane    -----Original Message----- From: stm [mailto:stm@pt.lu]) Sent: Thursday, January 16, 2003 11:57 AM  To: Info-VAX@Mvb.Saic.Comb Subject: NT on Alpha    8 I have an Alpha Server 1000 with VMS 7.2  (as Hobbyist).( I want to try to put NT on this server .( Can both systems run on the same server?$ What's the best way to implement NT.A I tried some official Web sites for NT or VMS but without result.   
 Need help....m   Rgds,t STMs   ------------------------------  + Date: Thu, 16 Jan 2003 21:32:57 +0000 (UTC)r5 From: "John Wallace" <johnwallace4@yahoo.dotco.dotuk>e Subject: Re: NT on Alpha0 Message-ID: <b078e8$9l6$1@sparta.btinternet.com>  L A handful of additional points, plus a general comment: dual-boot between NTL and VMS shouldn't be too difficult, but what follows is based on memory fromJ a long time ago. Note that this exercise will get more difficult if you're not already confident with NT.   1) Supported Hardwareh  L It's probably obvious, but... to dual boot, your chosen hardware needs to beL supported by both OSes. E.g. Adaptec SCSI adapters are fine for NT but don'tL do VMS. You're going the other way so should probably be OK if you just haveL run of the mill hardware. You'll find out soon enough... I'm assuming you've2 got a graphical console already which NT supports?   2) EISA Config Utility (ECU)  K In theory you need to run the appropriate ECU (as well as change OStype) asoL part of OS switching on any EISA-based Alpha box (even if you don't have anyH EISA cards, I believe). Sometimes the IRQ etc setup is different betweenG OSes, sometimes you're lucky and end up with identical IO/IRQ settings.   
 3) Partitionsn  I Take some care with partitioning the boot drive. You need a small (10MB?)sL FAT16 partition to boot from; the rest can be partitioned as you like duringF the install. An NT/Alpha install CD should have a program on it calledJ ARCINST which will do the boot partition for you. Install can mysteriously= hang if the partitioning  isn't right. See AlphaNT FAQ below.n   4) NT Disk Signaturesd  F Another common gotcha in a dual-boot setup: NT Disk Administrator willG happily overwrite the first few blocks of non-Windows disks: "writing aaG signature is a safe operation", says Bill. Not on a VMS or Unix disk it  isn't.   5) AlphaNT websitesd  D Have a look at Aaron Sakovich's NT/Alpha website at www.alphant.com,I particularly the FAQ (e.g. File and Disk Systems section). This site alsouG suggests means of using Windows 2000 beta on your Alpha, which might or-L might not be relevant. Some of the web sites he references are long gone but+ the main site still contains valuable info.:  	 Good lucke regards  john    - "Hans Vlems" <hvlems@iae.nl> wrote in message / news:b07523$lpmn1$1@ID-143435.news.dfncis.de...oJ > Even though it is beyond me why you'd want this, I'll try and answer the > question.sJ > The AS1000 uses different console firmware than TRU64 and VMS. IIRC it's
 > called ARC.aF > ARC is entirely menu driven. The AS1000 supports both console modes.L > One can change between the two, the procedure is listed in the AS1000 user > guide. > The AS1200 user guide states:  >  > >>> set os_type NT > >>> alphabiosi > K > This will put the system in ARC mode on subsequent reboots. If you do note) > want that omit the set os_type command. E > Once in ARC mode (or AlphaBIOS mode) you can return to SRM (the >>>  prompt)  > in the window calledL > Advanced CMOS Setup. Select either OpenVMS Console or DIGITAL UNIX Console > switch off; wait 30 secs5 > and power on the unit. Next set os_type to openVMS.3 >  > Hans > F > "stm" <stm@pt.lu> schreef in bericht news:3e270e8d$1_1@news.vo.lu...< > > I have an Alpha Server 1000 with VMS 7.2  (as Hobbyist).* > > I want to try to put NT on this server, > > Can both systems run on the same server?( > > What's the best way to implement NT.E > > I tried some official Web sites for NT or VMS but without result.  > >0 > > Need help....  > >i	 > > Rgds,  > > STM  > >  > >l >S   ------------------------------  % Date: Thu, 16 Jan 2003 15:00:23 -0800r% From: Dean Woodward <deanw@rdrop.com>  Subject: Re: NT on Alpha( Message-ID: <3E273987.1040603@rdrop.com>   Shane Smith wrote:B > To take a quote from a recent post out of context: "Danger, WillD > Robinson!!!". Windows NT reaches out to all available disks duringH > installation, and "marks" them. It claims this "mark" is harmless, andD > it is to other MS operating systems, but I've seen reports that it! > totally borks VMS format disks.i  E I don't believe this happens on installation- rather, the first time  E you run Disk Manager.  Regardless, the obvious solution is to unplug c# the VMS disks before installing NT.E  G And yes, letting NT write a signature to a VMS disk will make you wish   you hadn't.0   ------------------------------  % Date: Thu, 16 Jan 2003 15:01:11 -08001$ From: Shane Smith <ssmith@icius.com> Subject: RE: NT on Alpha0 Message-ID: <01C2BD70.2F4A4AE0@sulfer.icius.com>  F I distinctly remember seeing the message about doing the mark, and theD mark being harmless during installation - I never did run DM on thatF machine. It may do the mark when you run disk manager, in fact I thinkH it's quite likely. Billy would expect you to run that whenever you put aH new disk in, so it'd have to add the mark then. I guess that means you'd0 have to avoid running DM on a dual-boot machine.   Shane]   -----Original Message-----, From: Dean Woodward [mailto:deanw@rdrop.com]( Sent: Thursday, January 16, 2003 3:00 PM To: Info-VAX@Mvb.Saic.Coma Subject: Re: NT on Alpha     Shane Smith wrote:B > To take a quote from a recent post out of context: "Danger, WillD > Robinson!!!". Windows NT reaches out to all available disks duringH > installation, and "marks" them. It claims this "mark" is harmless, andD > it is to other MS operating systems, but I've seen reports that it! > totally borks VMS format disks.   E I don't believe this happens on installation- rather, the first time lE you run Disk Manager.  Regardless, the obvious solution is to unplug f# the VMS disks before installing NT.J  G And yes, letting NT write a signature to a VMS disk will make you wish $ you hadn't.b   ------------------------------  % Date: Thu, 16 Jan 2003 17:23:28 -0600 0 From: arturo saavedra <arturo.saavedra@wcom.com> Subject: RE: NT on AlphaB Message-ID: <NDBBJFNBJJHKADILJHIJIELBLAAA.arturo.saavedra@mci.com>  A Does somebody have the copy of that program that is supposed to = I "enhance" 32bit apps on the 64bit Alpha platorm?  As I recall, I had to = E let that run and it would "tweek" the 32bit app and make it perform =e better.I   Thanks!n artC     -----Original Message-----, From: Dean Woodward [mailto:deanw@rdrop.com]( Sent: Thursday, January 16, 2003 5:00 PM To: Info-VAX@Mvb.Saic.Comp Subject: Re: NT on Alpha     Shane Smith wrote:B > To take a quote from a recent post out of context: "Danger, WillD > Robinson!!!". Windows NT reaches out to all available disks duringH > installation, and "marks" them. It claims this "mark" is harmless, andD > it is to other MS operating systems, but I've seen reports that it! > totally borks VMS format disks.h  G I don't believe this happens on installation- rather, the first time=201G you run Disk Manager.  Regardless, the obvious solution is to unplug=200# the VMS disks before installing NT.n  I And yes, letting NT write a signature to a VMS disk will make you wish=209 you hadn't.2   ------------------------------  % Date: Thu, 16 Jan 2003 16:16:24 -0800 % From: Dean Woodward <deanw@rdrop.com>e Subject: Re: NT on Alpha& Message-ID: <3E274B58.20905@rdrop.com>   arturo saavedra wrote:A > Does somebody have the copy of that program that is supposed to B > "enhance" 32bit apps on the 64bit Alpha platorm?  As I recall, ID > had to let that run and it would "tweek" the 32bit app and make it > perform better.=   FX!32.  ' http://www.support.compaq.com/amt/fx32/A   ------------------------------  % Date: Thu, 16 Jan 2003 16:43:39 -0800o& From: Tom Crabtree <tccrab@sunset.net> Subject: Re: NT on Alpha* Message-ID: <3E2751BB.8288E651@sunset.net>   If what you are looking for is software which will cripple your 64bit Alpha system so that it will run 32bit Windoze Applications, then FX32! is what you seek:.7 http://www.support.compaq.com/amt/fx32/fx-download.html5  l If you should ever find software which 'tweaks' 32bit apps to optimize them for 64bit hardware please share!   TomC     arturo saavedra wrote:   > Does somebody have the copy of that program that is supposed to "enhance" 32bit apps on the 64bit Alpha platorm?  As I recall, I had to let that run and it would "tweek" the 32bit app and make it perform better.  >p	 > Thanks!. > arte >I > -----Original Message-----. > From: Dean Woodward [mailto:deanw@rdrop.com]* > Sent: Thursday, January 16, 2003 5:00 PM > To: Info-VAX@Mvb.Saic.Como > Subject: Re: NT on Alpha >  > Shane Smith wrote:D > > To take a quote from a recent post out of context: "Danger, WillF > > Robinson!!!". Windows NT reaches out to all available disks duringJ > > installation, and "marks" them. It claims this "mark" is harmless, andF > > it is to other MS operating systems, but I've seen reports that it# > > totally borks VMS format disks.  >tF > I don't believe this happens on installation- rather, the first timeF > you run Disk Manager.  Regardless, the obvious solution is to unplug% > the VMS disks before installing NT.  >oH > And yes, letting NT write a signature to a VMS disk will make you wish
 > you hadn't.m   ------------------------------  + Date: Fri, 17 Jan 2003 03:22:29 +0000 (UTC)e5 From: "John Wallace" <johnwallace4@yahoo.dotco.dotuk>d Subject: Re: NT on Alpha/ Message-ID: <b07stk$d5b$1@venus.btinternet.com>   L According to  www.alphant.com, FX!32 technology was effectively incorporatedJ direct into the OS in Windows 2000/Alpha. I can't remember whether MS tookJ it as was or "improved" it.  So depending on the purpose of this exercise,K Win2K Alpha might be of interest even though it never got to final release.a   htho john  3 "Tom Crabtree" <tccrab@sunset.net> wrote in messagea$ news:3E2751BB.8288E651@sunset.net...G > If what you are looking for is software which will cripple your 64bitcJ Alpha system so that it will run 32bit Windoze Applications, then FX32! is what you seek:9 > http://www.support.compaq.com/amt/fx32/fx-download.htmld >uH > If you should ever find software which 'tweaks' 32bit apps to optimize% them for 64bit hardware please share!y >  > TomC >q >t > arturo saavedra wrote: >aC > > Does somebody have the copy of that program that is supposed toeK "enhance" 32bit apps on the 64bit Alpha platorm?  As I recall, I had to letmG that run and it would "tweek" the 32bit app and make it perform better.n > >  > > Thanks!r > > artc > >o > > -----Original Message-----0 > > From: Dean Woodward [mailto:deanw@rdrop.com], > > Sent: Thursday, January 16, 2003 5:00 PM > > To: Info-VAX@Mvb.Saic.Comt > > Subject: Re: NT on Alpha > >o > > Shane Smith wrote:F > > > To take a quote from a recent post out of context: "Danger, WillH > > > Robinson!!!". Windows NT reaches out to all available disks duringL > > > installation, and "marks" them. It claims this "mark" is harmless, andH > > > it is to other MS operating systems, but I've seen reports that it% > > > totally borks VMS format disks.l > >oH > > I don't believe this happens on installation- rather, the first timeH > > you run Disk Manager.  Regardless, the obvious solution is to unplug' > > the VMS disks before installing NT.P > >sJ > > And yes, letting NT write a signature to a VMS disk will make you wish > > you hadn't.s >-   ------------------------------  % Date: Thu, 16 Jan 2003 21:51:10 -0600n0 From: arturo saavedra <arturo.saavedra@wcom.com> Subject: RE: NT on AlphaB Message-ID: <NDBBJFNBJJHKADILJHIJEELNLAAA.arturo.saavedra@mci.com>  I That's news to me, there is an actual win2k version for Alpha? or can I =; uses the standard win2k image ?   -----Original Message-----: From: John Wallace [mailto:johnwallace4@yahoo.dotco.dotuk]( Sent: Thursday, January 16, 2003 9:22 PM To: Info-VAX@Mvb.Saic.Comr Subject: Re: NT on Alpha    A According to  www.alphant.com, FX!32 technology was effectively =n incorporatedG direct into the OS in Windows 2000/Alpha. I can't remember whether MS =  tookB it as was or "improved" it.  So depending on the purpose of this =	 exercise,2D Win2K Alpha might be of interest even though it never got to final = release.   hths john  3 "Tom Crabtree" <tccrab@sunset.net> wrote in message.$ news:3E2751BB.8288E651@sunset.net...G > If what you are looking for is software which will cripple your 64bitfI Alpha system so that it will run 32bit Windoze Applications, then FX32! =r is what you seek:9 > http://www.support.compaq.com/amt/fx32/fx-download.htmlo > H > If you should ever find software which 'tweaks' 32bit apps to optimize% them for 64bit hardware please share!r >a > TomC >  >  > arturo saavedra wrote: >aC > > Does somebody have the copy of that program that is supposed torI "enhance" 32bit apps on the 64bit Alpha platorm?  As I recall, I had to =, lettG that run and it would "tweek" the 32bit app and make it perform better.b > >s > > Thanks!i > > arta > >o > > -----Original Message-----0 > > From: Dean Woodward [mailto:deanw@rdrop.com], > > Sent: Thursday, January 16, 2003 5:00 PM > > To: Info-VAX@Mvb.Saic.Com- > > Subject: Re: NT on Alpha > >  > > Shane Smith wrote:F > > > To take a quote from a recent post out of context: "Danger, WillH > > > Robinson!!!". Windows NT reaches out to all available disks duringJ > > > installation, and "marks" them. It claims this "mark" is harmless, = and"H > > > it is to other MS operating systems, but I've seen reports that it% > > > totally borks VMS format disks.d > >oH > > I don't believe this happens on installation- rather, the first timeH > > you run Disk Manager.  Regardless, the obvious solution is to unplug' > > the VMS disks before installing NT.t > >aG > > And yes, letting NT write a signature to a VMS disk will make you =. wish > > you hadn't.@ >y   ------------------------------  % Date: Thu, 16 Jan 2003 19:50:01 -0500o* From: "Bill Todd" <billtodd@metrocast.net> Subject: Re: Spiralog anyone?c2 Message-ID: <UYydnY-wOL7VzrqjXTWcog@metrocast.net>  > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message7 news:cf15391e.0301161018.31ceaed6@posting.google.com...a7 > "Bill Todd" <billtodd@metrocast.net> wrote in messages. news:<uEidnbtUbZ0Wwr2jXTWcoQ@metrocast.net>...G > > Once you remove the assumption that most reads hit in cache, use ofrE > > log-structuring for general-purpose storage becomes less tenable.e However,I > > for *some* kinds of data it's still nearly optimal, even without that  muchC > > caching - so what you want is a hybrid system that handles data  according toJ > > how it's used (which must often be inferred from how it is structured, butbE > > once such a system exists a feedback loop exists that helps causea
 developers) > > to structure the data appropriately).r >s- > Interesting.  Care to expand on this a bit?t   OK.    >iA > The strength of log-structured storage is its speed of streamed D > writes.  The weakness is that data tends to not land in an optimalE > spot for later retrieval.  If one were to break the assumption thatmH > once data is written, it tends to stay there until replaced, one couldE > take advantage of the flexibility that is inherent in introducing a:D > mapping layer between LBNs as seen by the host file system and the > physical storage locations.m  F While that's how Spiralog did it, that mapping layer can be expensive:K you've got to maintain the map persistently, and while a lot of this can besI subsumed by just adding a bit of map-update data to each segment write itlK still introduces an additional level of indirection on subsequent reads (atbH least if the overall map is too large to cache, which in many systems it	 will be).n   >oD > One might observe typical read access patterns and re-arrange data? > into physical storage order matching the anticipated order of? > retrieval.  K People have simulated this with good results, but it's the kind of approachfD I tend to view with some skepticism (just a bit too clever).  I haveK considerably more faith in trying to maintain data in physical order (which-L the cleaners of some log-structured file systems attempt to do):  sequentialG access is pretty common, and application developers have at least often-C designed their software to try to keep accesses sequential with theh9 assumption that the underlying system will optimize this.W  <   This could speed operations such as operating system boot,G > image activation, and so forth, because the next data to be requesteddE > would be in the track read-ahead cache found in most disks (and thet4 > read-ahead cache found in some controllers) today.  K Actually, IIRC Windows rearranges its start-up files to this end.  But thatrH doesn't require any help from the file system, just something to monitorK their access order and rearrange them on disk in that order.  Of course, itNJ does nothing to optimize internal access order within each individual file either.a  D My original reference was more to data that's intrinsically accessedJ randomly, hence has nothing to gain from being updated in place.  The bestJ example I can think of would be an RMS indexed file:  if its management isF inhaled into the kernel and managed transactionally there, I believe IF understand how to handle that in a log-structured manner such that allL instantaneous update activity to all RMS indexed files on a given disk couldI be merged into a single disk write followed by a single small transactionoL log commit record, with no adverse impact on read speed (but it does requireG significant additional mechanism to track data in the system cache, and_E because it uses LBN rather than VBN space management requires similarsE log-structured management of the volume bitmap and makes the internal F on-disk structure of an indexed file markedly different from that of aL normal file - e.g., there is no file map, just the map implicit in the indexL tree, which means that file deletion has to walk the tree and block I/O copyG operations aren't possible).  You could also do the same thing within acH normal ODS-2 file envelope, but that would significantly expand the fileD size (due to internal fragmentation:  you could use a sparse file, IJ suppose, but then you'd be allocating/deallocating on both file and systemG levels at once, which starts to offset the gains to to log-structuring,oH though if system allocation were log-structured - possibly a win just byF itself, since the volume bitmaps are themselves intrinsically accessed? randomly - it still might be a win over the current mechanism).    - bill   ------------------------------  % Date: Thu, 16 Jan 2003 14:20:56 -0500m3 From: "Homer J. Simpson" <hsimpson@burnsenergy.com>  Subject: Re: TZ15 problems6 Message-ID: <0wDV9.33337$Jm2.28033@news.bellsouth.net>  C Did you try reseating all SCSI cables & terminators (to "clean" theMI contacts)?  I recall seeing that type of problem from termination issues.e  F "Brian Tillman" <tillman_brian@notnoone.notnohow.com> wrote in message news:3e26fb3f$1@news.si.com...K > I have a cluster of VAXes with four TZ15 (8mm) tape drives.  These drivesdH > have been working fine for years.  Last week, three of them started to fail,mL > reporting parity errors, device offline, and such.  All four of the drivesJ > are in StorageWorks cannisters with two drives in a BA shelf attached to theiI > SCSI bus of VAXA at positions 1 and 4 of the shelf and the other two intL > another BA shelf attached to the SCSI bus of VAXB and at positions 2 and 5, > of the shelf.  Thus, the drives appear as: >  > $ sho dev mk >wG > Device                  Device           Error    Volume         Freef Transo > MntsH >  Name                   Status           Count     Label        Blocks Counts > Cnti. > $10$MKA100:   (VAXA)  Online               0/ > $10$MKA200:    (VAXB)  Online               0 . > $10$MKA400:   (VAXA)  Online               0/ > $10$MKA500:    (VAXB)  Online               0c >gL > Our field service persion replaced the failed drives, but all of a sudden, Ih > see: >  > $ sho dev mk >tG > Device                  Device           Error    Volume         Freer Transr > MnttH >  Name                   Status           Count     Label        Blocks Countr > Cnt . > $10$MKA100:   (VAXA)  Online               0/ > $10$MKA200:    (VAXB)  Online               0a/ > $10$MKA400:    (VAXB)  Online               0p. > $10$MKA400:   (VAXA)  Online               0/ > $10$MKA500:    (VAXB)  Online               0a >tJ > Notice the extra MKA400.  It appeared when we inserted a drive in slot 5 ofF > the shelf attached to VAXB.  It responds to commands as though it is MKA400. I > We've checked the jumpers and they seem properly set to allow the shelfcI > position to determine the SCSI ID.  If I reference the real MKA400 fromrJ > VAXA, both MKA400s show up as allocated and in use.  If I move the drive toI > another slot (i.e., if I remove MKA200 and put MKA500 in its place), itoK > acquires the SCSI ID of the slot, so I know it's not hard-wired to a SCSI-F > ID.  Lastly, if I try to reference MKA500, I get device timeouts, as thoughL > it's not there at all (which it appears to be the case, since the drive is2 > getting SCSI ID 4, in spite of being in slot 5). >1J > Now, I've made these drive switches while the VAXes have been running byI > removing the power cords to the shelves and then removing/inserting theaF > drives and finally replacing the cords.  Since they're not on an HSxF > controller, there's no quiesce the bus button.  Besides, since these devices L > are the only ones on the bus besides the VAXes and no I/O is occuring when I J > change them, I'm not worried about data loss.  Hoever, something appears toI > be screwed up.  Rebooting is hard to do, since the VAXes are productioni and  > used 24 hours a day. >3J > Any suggestions?  If I need to reboot, I can arrange it, but it's a last	 > resort.c > --K > Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot come7 > Smiths Aerospace  Addresses modified to prevent SPAM.cB > 3290 Patterson Ave. SE, MS Replace "at" with "@", "dot" with "." > Grand Rapids, MI 49512-1991a: >        This opinion doesn't represent that of my company >    ------------------------------  % Date: Thu, 16 Jan 2003 17:29:13 -0500e& From: "Ed Vogel" <ed.vogel@compaq.com>  Subject: Re: vax basic questions. Message-ID: <3e27323f$1_2@hpb10302.boi.hp.com>  ?     Actually (thinking about this some more...) you may be abletC     to just declare the file as a virtual array of bytes.  Then youe,     don't have to do any gets/puts yourself.       Ed    1 "Ed Vogel" <ed.vogel@compaq.com> wrote in messagel( news:3e26f5f5$1_2@hpb10302.boi.hp.com... |t& | "Dan" <dan@vrx.net> wrote in message4 | news:46rd2vsatahnnl82ijfk26t7i2u6tnr9sm@4ax.com...G | | Sorry I'm rather rusty at vax basic, but is there a way to read and 1 | | write a binary file one byte/char at a time ?  | |i | L |     I'm a little rusty myself Dan (I keep wanting to add ";" at the end of
 | every BASICe |     statement).o |hE |     Create a 512 byte MAP, and then open the file with ORGANIZATIONo VIRTUAL, | MAP <map name>.- |-J |     (I think) the ORGANIZATION VIRTUAL will open the file in binary mode | (doing RMSF |     READS/WRITES, not GET/PUTS).  Doing this, you must get things in blocks | of 512 bytes,1J |     but then if the MAP is an array of bytes that contains 512 elements, you/ | can read one2 |     byte at a time by looking through the array. |nK |     There may be other ways of doing this, but I think what I've proposedt | will work. |  |     Ed VogelI |     Compaq C Engineering (but worked on BASIC compilers for many years)a |s |  |i |f   ------------------------------  % Date: Thu, 16 Jan 2003 14:46:56 -0800m$ From: Shane Smith <ssmith@icius.com>  Subject: RE: vax basic questions0 Message-ID: <01C2BD6E.28B3A7F0@sulfer.icius.com>  G Virtual arrays are something of an afterthought in DEC BASIC, they were E provided to be compatible with the PDP11's Basic ++. Last time I usedg( them, their performance sucked noodles.    Shanee   -----Original Message-----+ From: Ed Vogel [mailto:ed.vogel@compaq.com]J( Sent: Thursday, January 16, 2003 2:29 PM To: Info-VAX@Mvb.Saic.Como  Subject: Re: vax basic questions    ?     Actually (thinking about this some more...) you may be able2C     to just declare the file as a virtual array of bytes.  Then youj,     don't have to do any gets/puts yourself.       Ed    1 "Ed Vogel" <ed.vogel@compaq.com> wrote in messaget( news:3e26f5f5$1_2@hpb10302.boi.hp.com... |n& | "Dan" <dan@vrx.net> wrote in message4 | news:46rd2vsatahnnl82ijfk26t7i2u6tnr9sm@4ax.com...G | | Sorry I'm rather rusty at vax basic, but is there a way to read andi1 | | write a binary file one byte/char at a time ?l | |  | E |     I'm a little rusty myself Dan (I keep wanting to add ";" at them end of
 | every BASICs |     statement).6 |oE |     Create a 512 byte MAP, and then open the file with ORGANIZATION  VIRTUAL, | MAP <map name>.f |sE |     (I think) the ORGANIZATION VIRTUAL will open the file in binary  mode | (doing RMSF |     READS/WRITES, not GET/PUTS).  Doing this, you must get things in blocks | of 512 bytes,r@ |     but then if the MAP is an array of bytes that contains 512	 elements,q youo | can read one2 |     byte at a time by looking through the array. |iB |     There may be other ways of doing this, but I think what I've proposed | will work. |w |     Ed VogelB |     Compaq C Engineering (but worked on BASIC compilers for many years) |  |a |r |2   ------------------------------  % Date: Thu, 16 Jan 2003 14:52:44 -0800n2 From: "Randy Park" <rjpark@mindspring.nospaam.com>  Subject: Re: vax basic questions2 Message-ID: <b07d47$fvs$1@slb0.atl.mindspring.net>  " Dan <dan@vrx.net> wrote in message2 news:46rd2vsatahnnl82ijfk26t7i2u6tnr9sm@4ax.com...E > Sorry I'm rather rusty at vax basic, but is there a way to read andu/ > write a binary file one byte/char at a time ?  >o3 > input #1,a$ would read a string in, size unfixed,oD > i forget if there is a way of defining A$ as a single char length, > not sure that would help tho >uB > and print #1,a$ has the same issues when putting output back out >w1 > its been far too long since I did any of thise.n >+? > I prefer pascal myself, but not sure the state of vms pascal,sE > if there are var t : text or such, or if I have to use file of chare
 > or what? >4? > but using pascal, I have no idea how to use the system calls,dA > like LIB$... etc (in basic you can run DCL commands from withina" > without exiting and going back). >e< > Just trying to cover my options. It's been too many years. >a > Dan.  > BASIC is like most VMS high level languages in that it prefers< to read data from files in the form of records.  To properly> process data in an existing file you should execute a DIR/FULL9 command from DCL and determine its characteristics first. = Upon determining the file organization (indexed, relative, or 8 sequential) and the record format (fixed, variable, VFC,: stream, stream_lf, or stream_cr) you can construct an OPEN9 statement in your program that matches the existing file. : Using the RECORDTYPE ANY clause with allow you to get past different record formats.i  < Once you open the file you can read the records sequentially: using the GET statement.  The number of bytes read will be8 placed in the RECOUNT variable.  Be sure to define a MAP= statement large enough to hold the largest record.  Since you ; want to process the data one byte at a time (you might wanta; to reconsider this) you should define the MAP to consist ofn< an array of BYTE integers. Make sure you use integers rather, than strings if you want decent performance.  ; If, however, your file organization is Sequential, and your 9 record format is fixed with 512 byte records, you can use5; the virtual array feature.  This feature maps your array too: the file on disk.  Note that processing can be slow as I/O9 will be done 1 block (512 bytes) at a time.  See the postl8 by Ed Vogel for an example. An alternative to this is to> take advantage of your virtual memory and read the entire file7 into memory, via successive GETs, manimpulate it as youe: like, then write the entire thing back out.  In most cases9 this yields superior performance to using virtual arrays.p  : I could give you better advice if I knew the format of the; file in question, and nature of the problem you were tryingn	 to solve.d   ------------------------------  % Date: Thu, 16 Jan 2003 19:33:11 -0600b@ From: "David J. Dachtera" <djesys.nospam@earthlink.spamfree.net>  Subject: Re: vax basic questions6 Message-ID: <3E275D57.959F6BE8@earthlink.spamfree.net>  
 Dan wrote: > E > Sorry I'm rather rusty at vax basic, but is there a way to read andy/ > write a binary file one byte/char at a time ?  > 3 > input #1,a$ would read a string in, size unfixed, D > i forget if there is a way of defining A$ as a single char length, > not sure that would help tho > B > and print #1,a$ has the same issues when putting output back out > 1 > its been far too long since I did any of thise.f > ? > I prefer pascal myself, but not sure the state of vms pascal,tE > if there are var t : text or such, or if I have to use file of chare
 > or what? > ? > but using pascal, I have no idea how to use the system calls, A > like LIB$... etc (in basic you can run DCL commands from within-" > without exiting and going back). > < > Just trying to cover my options. It's been too many years.  1 I, too, would be interested in what you're up to.l  E To me, "binary" file usually means Fixed_512. So, you read a 512 bytetD record with GET, and process the buffer contents one byte at a time.  F To output single bytes, assuming you assign a single character to, forC example, A$, then "PRINT [#x;] A$;" will output the contents of the-F string variable with not trailing <CR><LF>. It's the semi-colon at the end that's does the trick.  E See PPF.BAS in http://www.djesys.com/freeware/vms/ppf012.zip for somer	 examples.o   -- i David J. Dachtera' dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/t   ------------------------------  # Date: Fri, 17 Jan 2003 02:54:24 GMTo' From: SkyWriter <skywriter@mrnutty.com>t& Subject: Re: vax6k.openecs.org rebirth+ Message-ID: <3E2799B4.FEF4CA15@mrnutty.com>i   Alan Greig wrote:T  G > On 3 Jan 2003 07:27:24 -0600, koehler@eisner.nospam.encompasserve.orgo > (Bob Koehler) wrote: >aY > >In article <2JAN200322230368@gerg.tamu.edu>, carl@gerg.tamu.edu (Carl Perkins) writes:MK > >> Way back when, roughly 1983, I got cought calling the phsyical memory,OL > >> e.g. RAM, of an 11/780 "core" by someone who thought it odd. Why did I?H > >> I picked it up from the people I worked for and with. Why did they?I > >> Because that's what they had always called that sort of thing - they0G > >> didn't care if it was real "core" type memory, it was the physical I > >> memory and they always called it "core". At that point I didn't careeJ > >> what the difference was (and was barely aware of the difference), andI > >> neither did they. The expression was apparently pretty commonly used- > >> that way back then. > > G > >   Once upon a time we had a facility with 3 PDP-11, 2 PDP-10, and 8NE > >   11/780.  One of the engineers came by and said we were going to G > >   replace the 1MB core in the 2050, I thought he was talking out ofi > >   habit. >oG > When you replaced the core memory with IC based memory in a DEC20 themD > entire cabinet could become unstable if certain access panels wereH > swung out due to  the absence of the heavy core memory.. Field ServiceG > engineers told me they had actually heard of machines tipping over iniG > the early upgrade days before the problem was fixed. I *think* we hadaC > counterweights fitted but it might have just been a large warning  > sticker. Can't recall now. >.  D core memory never was 'swung out' oh the MF20  nmos (i think) memoryM it was on the back cabinet door, with the power supplies directly beneath it.eG on MH20 core memory is was a stand alone cabinet, and the backplane was  fixed to the cabinet.s  S an any event all dec cab's of that era has slide out stabilizer feet to accound fors* any extractable cardcage, or  peripherial.  L the DN20 type cabinets has the extra lead wieghts in the bottom of the cabs.  & 11/780 never had core (in production).  & at least that's the way i remember it.       >N > >   Nope.v >. > -- > Alan   ------------------------------   End of INFO-VAX 2003.033 ************************