1 INFO-VAX	Fri, 21 Nov 2003	Volume 2003 : Issue 645       Contents:' Re: <CR><LF> Not in FTP Transfered File  Backwards File Dump  Re: Backwards File Dump  Re: Backwards File Dump  Re: Backwards File Dump  Re: Backwards File Dump  Re: Backwards File Dump  Re: Backwards File Dump  Re: Backwards File Dump  Re: Backwards File Dump  Re: Backwards File Dump  Re: Backwards File Dump $ Batch job submission using web page?( Re: Batch job submission using web page?( Re: Batch job submission using web page?( Re: Batch job submission using web page?( Re: Batch job submission using web page?( Re: Batch job submission using web page?( Re: Batch job submission using web page?( Re: Batch job submission using web page?( Re: Batch job submission using web page?( Re: Batch job submission using web page?( Re: Batch job submission using web page?  Bugcheck in SYSMAN.EXE on reboot$ Re: Bugcheck in SYSMAN.EXE on reboot$ Re: Bugcheck in SYSMAN.EXE on reboot$ Re: Bugcheck in SYSMAN.EXE on reboot) RE: DCL Enhancement: PUSH/POP Environment ) RE: DCL Enhancement: PUSH/POP Environment 8 Re: DCL Enhancements: Error messages for OPEN and DEFINE8 Re: DCL Enhancements: Error messages for OPEN and DEFINE8 Re: DCL Enhancements: Error messages for OPEN and DEFINE Re: DSN? DSN?= Efficiency Question: Large Arrays vs. Indexed Files on Alphas A Re: Efficiency Question: Large Arrays vs. Indexed Files on Alphas A Re: Efficiency Question: Large Arrays vs. Indexed Files on Alphas  Re: EV79 CANCELED !!!!!!!!!  Re: EV79 CANCELED !!!!!!!!!  Re: EV79 CANCELED !!!!!!!!! ) Re: Hobbyist license and layered products ) Re: Hobbyist license and layered products ) Re: Hobbyist license and layered products ) Re: Hobbyist license and layered products  HP 4th qtr earnings released  Re: HP 4th qtr earnings released  RE: HP 4th qtr earnings released  Re: HP 4th qtr earnings released  Re: HP 4th qtr earnings released Re: IA64 TPC results Re: IA64 TPC results Re: lat for linux problem  Re: ODS2 file header% Re: Peoplesoft Financials and OpenVMS  Re: Printer problem < Re: Problem with performance while running FTP in batch mode< Re: Problem with performance while running FTP in batch mode Re: SQLSRV V7.1-5 Problem  RE: SQLSRV V7.1-5 Problem 9 Re: Sun to use AMD Opteron - announcement expected Monday 9 Re: Sun to use AMD Opteron - announcement expected Monday 9 Re: Sun to use AMD Opteron - announcement expected Monday 9 Re: Sun to use AMD Opteron - announcement expected Monday 4 SYS$LIBRARY:TRACE_V74.EXE on VMS 7.3-1 causing error8 Re: SYS$LIBRARY:TRACE_V74.EXE on VMS 7.3-1 causing error8 Re: SYS$LIBRARY:TRACE_V74.EXE on VMS 7.3-1 causing error8 Re: SYS$LIBRARY:TRACE_V74.EXE on VMS 7.3-1 causing error Re: TELNET/PORT  Re: TELNET/PORT  Re: TELNET/PORT  Re: TELNET/PORT  Re: TELNET/PORT P RE: The Alpha Chip Engineering Group Revival (aka: today's HP/Intel         presP Re: The Alpha Chip Engineering Group Revival (aka: today's HP/Intel presentation The Uptimes Project  Re: The Uptimes Project  Re: VMS 2003 bootcamp summary   Re: VMS Defragmentation Software  Re: VMS Defragmentation Software  Re: VMS Defragmentation Software* Re: [OT] Mac OS X: system folder not found  F ----------------------------------------------------------------------  % Date: Thu, 20 Nov 2003 21:28:15 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> 0 Subject: Re: <CR><LF> Not in FTP Transfered File& Message-ID: <3FBD864F.210ABFC@fsi.net>   Chip Smithson wrote: > F > Thank you all for all of your input. Doesn't look like anyone reallyH > knows what may have happened. I am noticing that it is only with filesH > where the max record length is at or near 2048. We had to make it thatF > length, because the actual record can be as long as 8,000, but couldG > not get mumps system or vms to write any longer than 2048. So we have / > to write out 4 lines for every actual record.   . Ah-HAH! Now, if we'd had that info before, ...  / Yes what might have changed is indeed a puzzle.    > Why might that be? Is " > there some sort of buffer issue?   MAXBUF?   " > It has been so long since I have* > administered VMS system, I can't recall.H > This should probably be a whole other post but thought maybe the issue > is realated.- > Thanks Again though...Very frustrating.....   	 Indeed...    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  # Date: Thu, 20 Nov 2003 18:59:37 GMT 1 From: "Barry in Indy" <barrymeindy@ameritech.net>  Subject: Backwards File Dump> Message-ID: <ta8vb.2479$aw2.810496@newssrv26.news.prodigy.com>  @ I recently did a file dump, and the hexadecimal part (left side)A is the reverse of the text translation (right side). How strange! > I take it this is normal, given that the help example appears:          1.$ DUMP TEST.DAT<          Dump of file DISK0:[MOORE]TEST.DAT;1 on 14-DEC-2001 15:43:26.08 A          File ID (3134,818,2)   End of file block 1 / Allocated 3 <          Virtual block number 1 (00000001), 512 (0200) bytes>           706D6173 20612073 69207369 68540033 3.This is a samp 000000>           73752065 62206F74 20656C69 6620656C le file to be us 000010>           61786520 504D5544 2061206E 69206465 ed in a DUMP exa 000020>           00000000 00000000 0000002E 656C706D mple............ 000030>           00000000 00000000 00000000 00000000 ................ 000040>           00000000 00000000 00000000 00000000 ................ 000050>           00000000 00000000 00000000 00000000 ................ 000060     --  
 Barry in Indy    Knock me out to reply    ------------------------------  # Date: Thu, 20 Nov 2003 19:59:56 GMT 1 From: "Barry in Indy" <barrymeindy@ameritech.net>   Subject: Re: Backwards File Dump> Message-ID: <%29vb.2556$aw2.819821@newssrv26.news.prodigy.com>  < "Barry in Indy" <barrymeindy@ameritech.net> wrote in message8 news:ta8vb.2479$aw2.810496@newssrv26.news.prodigy.com...< > I recently did a file dump, and the hexadecimal part (left side) : > is the reverse of the text translation (right side). How strange!@ > I take it this is normal, given that the help example appears: > [snip]  > Replying to myself, I see that the DCL Dictionary states: "The< ASCII representation is read left to right. The hexadecimal,; decimal, and octal representations are read right to left."    Okay. Why???????   --  
 Barry in Indy    Knock me out to reply    ------------------------------  # Date: Thu, 20 Nov 2003 20:50:39 GMT # From: hoff@hp.nospam (Hoff Hoffman)   Subject: Re: Backwards File Dump1 Message-ID: <zO9vb.9441$fk.4738@news.cpqcorp.net>   r In article <%29vb.2556$aw2.819821@newssrv26.news.prodigy.com>, "Barry in Indy" <barrymeindy@ameritech.net> writes: : . :...I see that the DCL Dictionary states: "The= :ASCII representation is read left to right. The hexadecimal, < :decimal, and octal representations are read right to left." :  :Okay. Why???????   H   The DUMP display of hexidecimal data allows you to read numeric valuesF   from right (LSB) to left (MSB) as this is the way numeric values areH   stored in memory and are interpreted and are addressed on OpenVMS, andF   also allows you to read English-language text using the conventionalJ   left-to-right character sequence typical of the Romance language family.H   Hence the DUMP command hex dump is right to left for lowest to highestF   address of the hexidecimal dump, and the text dump is organized fromH   left to right for the lowest to the highest address within the string.  G   A little-endian and Arabic-language DUMP tool might well dump integer C   values and Arabic text from right to left and from right to left,    respectiviely.  K   The organization of a Kanji character set DUMP is left to the reader. :-)       N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com    ------------------------------    Date: 20 Nov 2003 15:41:43 -0800# From: dooleys@snowy.net.au (dooley)   Subject: Re: Backwards File Dump= Message-ID: <1ca82fc6.0311201541.42dc2797@posting.google.com>   w "Barry in Indy" <barrymeindy@ameritech.net> wrote in message news:<ta8vb.2479$aw2.810496@newssrv26.news.prodigy.com>... B > I recently did a file dump, and the hexadecimal part (left side)C > is the reverse of the text translation (right side). How strange! " It has always been so in vms dumpsF You have to imagine a mirror in this column ---                                               |@ >           706D6173 20612073 69207369 68540033 3.This is a samp@ >           73752065 62206F74 20656C69 6620656C le file to be us@ >           61786520 504D5544 2061206E 69206465 ed in a DUMP exa@ >           00000000 00000000 0000002E 656C706D mple............@ >           00000000 00000000 00000000 00000000 ................@ >           00000000 00000000 00000000 00000000 ................@ >           00000000 00000000 00000000 00000000 ................; If it was the "other way" then it would be tricky to "read"  the value of binary data Phil   ------------------------------  % Date: Thu, 20 Nov 2003 20:03:18 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>  Subject: Re: Backwards File Dump) Message-ID: <3FBD6441.757741E3@istop.com>   
 dooley wrote: $ > It has always been so in vms dumps1 > You have to imagine a mirror in this column --- = > If it was the "other way" then it would be tricky to "read"  > the value of binary data    L I started off with IBM mainframes with "real" dumps. A millenium later, I amN still never used to the VMS style of dumps with the hex stuff reversed. It has never been natural for me.  L I think that the VMS dumps were designed for middle east users who read from right to left ;-)    ------------------------------  + Date: Thu, 20 Nov 2003 20:36:18 -0600 (CST)  From: sms@antinode.org  Subject: Re: Backwards File Dump) Message-ID: <03112020361869@antinode.org>   > From:	SMTP%"jfmezei.spamnot@istop.com" 20-NOV-2003 20:02:40.79 To:	Info-VAX@Mvb.Saic.Com  CC:	 Subj:	Re: Backwards File Dump   * Return-Path: Info-VAX-Request@mvb.saic.com, Received: from mvb.saic.com (198.151.12.104)7 	 by alp.antinode.org (V5.3-18E, OpenVMS V7.3-1 Alpha); & 	Thu, 20 Nov 2003 20:02:29 -0600 (CST)) Message-ID: <3FBD6441.757741E3@istop.com> * From: JF Mezei <jfmezei.spamnot@istop.com>* X-Mailer: Mozilla 4.76 (Macintosh; U; PPC) X-Accept-Language: en  MIME-Version: 1.0  X-Newsgroups: comp.os.vms   Subject: Re: Backwards File Dump* Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit " X-Forwarded: by - (DeleGate/8.5.4)	 Lines: 13 % Date: Thu, 20 Nov 2003 20:03:18 -0500 # X-Complaints-To: abuse@sympatico.ca V X-Trace: news20.bellglobal.com 1069376461 64.230.46.34 (Thu, 20 Nov 2003 20:01:01 EST)0 NNTP-Posting-Date: Thu, 20 Nov 2003 20:01:01 EST Organization: Bell Sympatico. Reply-to: JF Mezei <jfmezei.spamnot@istop.com> X-Gateway-From: mvb.saic.com To: Info-VAX@Mvb.Saic.Com  X-Gateway-Source-Info: USENET   
 dooley wrote: $ > It has always been so in vms dumps1 > You have to imagine a mirror in this column --- = > If it was the "other way" then it would be tricky to "read"  > the value of binary data    L I started off with IBM mainframes with "real" dumps. A millenium later, I amN still never used to the VMS style of dumps with the hex stuff reversed. It has never been natural for me.  L I think that the VMS dumps were designed for middle east users who read from right to left ;-)    ------------------------------  + Date: Thu, 20 Nov 2003 20:36:44 -0600 (CST)  From: sms@antinode.org  Subject: Re: Backwards File Dump) Message-ID: <03112020364415@antinode.org>   ?    Oops.  Please pardon that last erroneous (unedited) message.   * From: JF Mezei <jfmezei.spamnot@istop.com>  N > I started off with IBM mainframes with "real" dumps. A millenium later, I amP > still never used to the VMS style of dumps with the hex stuff reversed. It has > never been natural for me. > N > I think that the VMS dumps were designed for middle east users who read from > right to left ;-)   C    Actually, it was designed for people who expect most significant # digits on the left.  Consider this:    alp $ type dum.c #include <unistd.h>  #include <fcntl.h>   main() {  int fd;  int datum_int = 0x87654321; ' char datum_char[] = " Make sense now?";   / fd = open( "dum.dat", (O_RDWR| O_CREAT), 0777);   + write( fd, &datum_int, sizeof( datum_int)); - write( fd, &datum_char, sizeof( datum_char));    close( fd);  }    alp $ cc dum alp $ link dum
 alp $ run dum  alp $ dump /word /rec dum.dat   ? Dump of file ALP$DKA0:[SMS]DUM.DAT;1 on 20-NOV-2003 20:38:57.65 : File ID (127553,50,0)   End of file block 1 / Allocated 35  @ Record number 1 (00000001), 21 (0015) bytes, RFA(0001,0000,0000)H                               [vvvvvvvvv------------------------- Note.]@  2065 736E 6573 2065 6B61 4D20 8765 4321 !Ce. Make sense  000000@                             00 3F77 6F6E now?............ 000010    D    In the big-endian world, it makes more sense to go the other way:  
 ung% uname -a 8 SunOS ung 5.9 Generic_112233-08 sun4u sparc SUNW,Ultra-1   ung% gcc dum.c -o dum 
 ung% ./dum ung% od -cx dum.dat C 0000000 207   e   C   !       M   a   k   e       s   e   n   s   e H            [vvvv----vvvv----------------------------------------- Note.]H             8765    4321    204d    616b    6520    7365    6e73    6520 0000020   n   o   w   ?  \0               6e6f    773f    0000 0000025   H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode,org     Saint Paul  MN  55105-2547    ------------------------------  # Date: Fri, 21 Nov 2003 02:57:45 GMT 5 From: rdeininger@mindspringdot.com (Robert Deininger)   Subject: Re: Backwards File DumpL Message-ID: <rdeininger-2011032158570001@user-uinj0g9.dialup.mindspring.com>  2 In article <3FBD6441.757741E3@istop.com>, JF Mezei" <jfmezei.spamnot@istop.com> wrote:   >dooley wrote:% >> It has always been so in vms dumps 2 >> You have to imagine a mirror in this column ---> >> If it was the "other way" then it would be tricky to "read" >> the value of binary data  >  > M >I started off with IBM mainframes with "real" dumps. A millenium later, I am O >still never used to the VMS style of dumps with the hex stuff reversed. It has  >never been natural for me.  > M >I think that the VMS dumps were designed for middle east users who read from  >right to left ;-)  G No, VMS dumps were designed for a little-endian environment.  The other  way wouldn't make sense.   ------------------------------  % Date: Thu, 20 Nov 2003 22:30:00 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>  Subject: Re: Backwards File Dump) Message-ID: <3FBD869D.6AFC36F5@istop.com>    Robert Deininger wrote: I > No, VMS dumps were designed for a little-endian environment.  The other  > way wouldn't make sense.    K It still is confusing. If you're working on a dump, you are smart enough to K read it in the proper direction. But with the VMS dump, you need to read it , backwards for text, and forwards for binary.  I I'd much rather have it listed in order of memory address, no matter what  endianness there is.   ------------------------------  % Date: Thu, 20 Nov 2003 21:54:13 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>   Subject: Re: Backwards File Dump' Message-ID: <3FBD8C65.315935D8@fsi.net>    Robert Deininger wrote:  > 4 > In article <3FBD6441.757741E3@istop.com>, JF Mezei$ > <jfmezei.spamnot@istop.com> wrote: >  > >dooley wrote:' > >> It has always been so in vms dumps 4 > >> You have to imagine a mirror in this column ---@ > >> If it was the "other way" then it would be tricky to "read" > >> the value of binary data  > >  > > O > >I started off with IBM mainframes with "real" dumps. A millenium later, I am Q > >still never used to the VMS style of dumps with the hex stuff reversed. It has  > >never been natural for me.  > > O > >I think that the VMS dumps were designed for middle east users who read from  > >right to left ;-) > I > No, VMS dumps were designed for a little-endian environment.  The other  > way wouldn't make sense.  , As we often say in this group, "it depends".  F For ASCII data and packed decimal, left-to-right makes more sense. For> binary data, right-to-left makes more sense for little-endian.  4 To bad DUMP doesn't have an appropriate qualifier...   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------    Date: 20 Nov 2003 22:57:10 -0800# From: dooleys@snowy.net.au (dooley)   Subject: Re: Backwards File Dump= Message-ID: <1ca82fc6.0311202257.1cbc8d62@posting.google.com>   [ JF Mezei <jfmezei.spamnot@istop.com> wrote in message news:<3FBD869D.6AFC36F5@istop.com>...  > Robert Deininger wrote: K > > No, VMS dumps were designed for a little-endian environment.  The other  > > way wouldn't make sense. >  > M > It still is confusing. If you're working on a dump, you are smart enough to " > read it in the proper direction.% Which for characters is left-to-right  and for binary is right-to-left ) and packed decimals have always been evil   , > But with the VMS dump, you need to read it. > backwards for text, and forwards for binary.# Don't you mean the other way round? E If you had a binary field containing the equivalent of decimal 123456   would you expect it to look like 00 01 E2 40   or  00 10 2E 04   or  40 E2 01 00   or  04 2E 10 00    Phil   ------------------------------  + Date: Thu, 20 Nov 2003 20:10:12 +0000 (UTC) . From: "Insomnee" <robert.heyes@btinternet.com>- Subject: Batch job submission using web page? 2 Message-ID: <bpj733$4a8$1@hercules.btinternet.com>  H Does anyone know if & how it is possible to submit a batch job via a webI page? We're having Oracle Forms problems and thinking of solutions. If we L did web pages running apache or something, is there a link on to the Queues?   ------------------------------  + Date: Thu, 20 Nov 2003 21:19:32 +0000 (UTC) , From: lewis@mazda.mitre.org (Keith A. Lewis)1 Subject: Re: Batch job submission using web page? . Message-ID: <bpjb54$8rt$2@newslocal.mitre.org>   "Insomnee" <robert.heyes@btinternet.com> writes in article <bpj733$4a8$1@hercules.btinternet.com> dated Thu, 20 Nov 2003 20:10:12 +0000 (UTC):I >Does anyone know if & how it is possible to submit a batch job via a web J >page? We're having Oracle Forms problems and thinking of solutions. If weM >did web pages running apache or something, is there a link on to the Queues?   H Yes, the Apache CGI can run DCL scripts or binary executables, either ofE which can submit batch jobs.  CSWS comes with some examples, I think.   0 --Keith Lewis              klewis {at} mitre.org> The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Thu, 20 Nov 2003 22:15:51 +0100 " From: Didier Morandi <no@spam.com>1 Subject: Re: Batch job submission using web page? 4 Message-ID: <3fbd2f08$0$17113$626a54ce@news.free.fr>   Insomnee wrote: J > Does anyone know if & how it is possible to submit a batch job via a webK > page? We're having Oracle Forms problems and thinking of solutions. If we N > did web pages running apache or something, is there a link on to the Queues?  E Assuming that your batch procedure is named BACKUP1.COM, if you were  H using WASD (http://wasd.vsm.com.au/) you could just write a three lines  SUBMIT.COM DCL script reading:  
 $! SUBMIT.COM  $ job := 'www_query_string' " $ submit/noprint/notify/keep 'job' $ exit  $ then you type into your web browser:  0 www.mysite.com/ht_root/script/submit.com?backup1  ( (backup1 is a parameter) and you'll get:  - $ job := backup1			!passed by the WASD server @ $ submit/noprint/notify/keep backup1	!name of the .COM procedure $ exit   (Mark?)  :-)    D.   ------------------------------  % Date: Thu, 20 Nov 2003 22:27:10 +0100 " From: Didier Morandi <no@spam.com>1 Subject: Re: Batch job submission using web page? 4 Message-ID: <3fbd31ae$0$17110$626a54ce@news.free.fr>  " Hoff Hoffman wrote at 13:15:27 PST  e > In article <bpj733$4a8$1@hercules.btinternet.com>, "Insomnee" <robert.heyes@btinternet.com> writes: K > :Does anyone know if & how it is possible to submit a batch job via a web L > :page? We're having Oracle Forms problems and thinking of solutions. If weO > :did web pages running apache or something, is there a link on to the Queues?  > E >   Trivial, using a DCL command procedure as a CGI script.  The user E >   fills in a web form, and -- when submitted -- the form passes the 7 >   data along to a user-written DCL command procedure.  > H >   If you're not familiar with DCL CGI, there is an example over at theD >   OpenVMS Ask The Wizard area: http://www.hp.com/go/openvms/wizard > I >   There are certainly other approaches available, and the best approach L >   depends on the particular details of the local application requirements.  2 You beat me on this one by 25 seconds, Hoff... :-)   D.   ------------------------------  # Date: Thu, 20 Nov 2003 21:15:27 GMT # From: hoff@hp.nospam (Hoff Hoffman) 1 Subject: Re: Batch job submission using web page? 1 Message-ID: <P9avb.9445$Fj.5938@news.cpqcorp.net>   c In article <bpj733$4a8$1@hercules.btinternet.com>, "Insomnee" <robert.heyes@btinternet.com> writes: I :Does anyone know if & how it is possible to submit a batch job via a web J :page? We're having Oracle Forms problems and thinking of solutions. If weM :did web pages running apache or something, is there a link on to the Queues?   C   Trivial, using a DCL command procedure as a CGI script.  The user C   fills in a web form, and -- when submitted -- the form passes the 5   data along to a user-written DCL command procedure.   F   If you're not familiar with DCL CGI, there is an example over at theB   OpenVMS Ask The Wizard area: http://www.hp.com/go/openvms/wizard  G   There are certainly other approaches available, and the best approach J   depends on the particular details of the local application requirements.  N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com    ------------------------------  % Date: Thu, 20 Nov 2003 17:03:24 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>1 Subject: Re: Batch job submission using web page? ) Message-ID: <3FBD3A24.3E079110@istop.com>    Didier Morandi wrote: & > then you type into your web browser: > 2 > www.mysite.com/ht_root/script/submit.com?backup1D > $ job := backup1                        !passed by the WASD serverE > $ submit/noprint/notify/keep backup1    !name of the .COM procedure  > $ exit  N There is one major issue to deal with: authentication, usernames and security.  L With OSU, I am able to have a url point to a decnet object. That object runsK under a specific VMS username. So if it were to submit a job, the job would  execute under that username.  M You really do not want to give the web server CMKRNL privilege so that it can  use SUBMIT/USER command.   ------------------------------  % Date: Thu, 20 Nov 2003 17:31:07 -0500 ! From: "Kevin F." <OpenVMS Rules!> 1 Subject: Re: Batch job submission using web page? , Message-ID: <3fbd4143$1@usenet01.boi.hp.com>  G What if I would tell you I could convert all your Oracle forms to Java?    OpenVMS_Rules!    7 "JF Mezei" <jfmezei.spamnot@istop.com> wrote in message # news:3FBD3A24.3E079110@istop.com...  > Didier Morandi wrote: ( > > then you type into your web browser: > > 4 > > www.mysite.com/ht_root/script/submit.com?backup1F > > $ job := backup1                        !passed by the WASD serverG > > $ submit/noprint/notify/keep backup1    !name of the .COM procedure 
 > > $ exit > F > There is one major issue to deal with: authentication, usernames and	 security.  > I > With OSU, I am able to have a url point to a decnet object. That object  runsG > under a specific VMS username. So if it were to submit a job, the job  would  > execute under that username. > K > You really do not want to give the web server CMKRNL privilege so that it  can  > use SUBMIT/USER command.   ------------------------------  # Date: Thu, 20 Nov 2003 23:25:24 GMT L From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr")1 Subject: Re: Batch job submission using web page? 6 Message-ID: <00A29302.FB5EE505@SSRL.SLAC.STANFORD.EDU>  V In article <3FBD3A24.3E079110@istop.com>, JF Mezei <jfmezei.spamnot@istop.com> writes: >Didier Morandi wrote:' >> then you type into your web browser:  >>  3 >> www.mysite.com/ht_root/script/submit.com?backup1 E >> $ job := backup1                        !passed by the WASD server F >> $ submit/noprint/notify/keep backup1    !name of the .COM procedure	 >> $ exit  > O >There is one major issue to deal with: authentication, usernames and security.  > M >With OSU, I am able to have a url point to a decnet object. That object runs L >under a specific VMS username. So if it were to submit a job, the job would >execute under that username.    > N >You really do not want to give the web server CMKRNL privilege so that it can >use SUBMIT/USER command.    Indeed.   K CSWS, WASD, and OSU all have mechanisms to allow scripts to run under other K usernames than the server itself; I believe WASD 8.x in fact mandates this. L (That's mostly to have the scripts run in a _reduced_-privilege sandbox thatI can't subvert the web-server config files, but could certainly be used to ; support submitting the backup job from the BACKUP account.)   I All of them allow VMS authentication and access control; you can combine  J the user having to have VMS username and password with the user having to L be a member of a particular group of usernames authorized to touch that URL.   -- Alan  --  O =============================================================================== 0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056 M  Paper mail to: SSRL -- SLAC BIN 99, 2575 Sand Hill Rd, Menlo Park CA   94025 O ===============================================================================    ------------------------------  % Date: Thu, 20 Nov 2003 19:56:48 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>1 Subject: Re: Batch job submission using web page? ) Message-ID: <3FBD62BC.B43B1D0F@istop.com>   * Alan Winston - SSRL Admin Cmptg Mgr wrote:M > CSWS, WASD, and OSU all have mechanisms to allow scripts to run under other $ > usernames than the server itself;   H out of curiosity, with OSU, what are the mechanisms to allow a script toD excute under a username other that the one running the http_server ?  Z I can only think of mapping a file specification to a URL that has a decnet specification.    L And if a web server has the ability to create a subprocess operating under aI different username, doesn't mean that the web server must have cmkrnl and L could potentially create a process under the SYSTEM account and create a lot of trouble).   ------------------------------  # Date: Fri, 21 Nov 2003 02:13:39 GMT L From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr")1 Subject: Re: Batch job submission using web page? 6 Message-ID: <00A2931A.7C7A4D30@SSRL.SLAC.STANFORD.EDU>  V In article <3FBD62BC.B43B1D0F@istop.com>, JF Mezei <jfmezei.spamnot@istop.com> writes:+ >Alan Winston - SSRL Admin Cmptg Mgr wrote: N >> CSWS, WASD, and OSU all have mechanisms to allow scripts to run under other% >> usernames than the server itself;   > I >out of curiosity, with OSU, what are the mechanisms to allow a script to E >excute under a username other that the one running the http_server ?  > [ >I can only think of mapping a file specification to a URL that has a decnet specification.  >  > M >And if a web server has the ability to create a subprocess operating under a J >different username, doesn't mean that the web server must have cmkrnl andM >could potentially create a process under the SYSTEM account and create a lot 
 >of trouble).   K You keep forgetting IMPERSONATE privilege, which is how CSWS, WASD, and OSU K all do it (on VMS 6.0 and up).  Of course, this is the moral equivalent of  J CMKRNL if you can assume the persona of an account that _has_ CMKRNL.  ButE Dave Jones managed to keep the webserver proper from having the heavy  privileges.   K Here's a brief excerpt from my webservers-on-VMS book (plug: available from K Digital Press via openvms.org or via the CSWS home page or via Amazon.com - F search on my name and disregard the sex books); this is in Chapter 19:     OSU   O User scripting is an optional and perhaps underpublicized feature in OSU.  (The M only documentation I could find on it is in the comments in the user_script.c K and user_script.com files.)  Here's what happens:  From the SYSTEM account, K create a detached process that acts as a gateway between the server and the K user script execution processes.  (This can't be created by the HTTP_SERVER M process because it requires heavy privileges: OPER, SYSNAM, and SYSPRV, which > are privileges you definitely don't want HTTP_SERVER to have.)  N When the server recognizes that it needs to run a user script, it contacts theI gateway and passes it the script and username it needs to run under.  The J gateway program uses persona services to assume the needed username, then,J through a proxy from that username to itself, creates another process withL which it communicates using DECnet task-to-task communication (and then setsO itself back to the original persona).  Because the gateway process doesn't read L and parse the server configuration files, it needs to get some configurationN information from system globals.  The gateway connects to the script executionN object, WWWEXEC,  and itself uses a different script object.  This process canN also be used to run scripts using the username/password entered as a result ofN SYSUAF-based authentication, so you can run scripts as the user you're talking* to as well as run scripts as their owners.  E [Gritty details eliminated because it would make a really long post.]    -- Alan    --  O =============================================================================== 0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056 M  Paper mail to: SSRL -- SLAC BIN 99, 2575 Sand Hill Rd, Menlo Park CA   94025 O ===============================================================================    ------------------------------  % Date: Thu, 20 Nov 2003 22:14:45 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>1 Subject: Re: Batch job submission using web page? ) Message-ID: <3FBD830B.11ED193B@istop.com>   * Alan Winston - SSRL Admin Cmptg Mgr wrote:M > and user_script.com files.)  Here's what happens:  From the SYSTEM account, M > create a detached process that acts as a gateway between the server and the " > user script execution processes.  J Thanks for explanation. I hadn't heard of this at all. Seems like a lot ofV overhead though if you have the web server talking to a middleman to a script process.  E What would be needed is a way for an unprivileged process to create a I subprocess running under another username by supplying username/password.    ------------------------------   Date: 20 Nov 2003 19:48:34 GMT% From: ejheller@aol.com.com (EJHeller) ) Subject: Bugcheck in SYSMAN.EXE on reboot : Message-ID: <20031120144834.10071.00000629@mb-m01.aol.com>  L We have a new DS10 with OVMS 7.3.1 with all the latest and greatest criticalM patches. Since we turned it on, every time we do a shutdown/reboot the system H bugchecks in SYSMAN.EXE with a code of 215 (process = STARTUP). The onlyO recovery is to manually halt the machine and issue an INIT from the maintenance N prompt, then boot the computer. We could work around by do a shutdown/noreboot3 and the issue the boot from the maintenance prompt. O In looking at the patch site, I saw there is a patch for MANAGER. So, being the N psuedo-adventurous sort, I installed this patch. Now any boot after a shutdown* requires an INIT before a successful boot.- Has anyone seen this problem and resolved it?. Thanks for your input,
 Edward Heller. TransCore ITS, Inc edward.heller@transcore-.-com..o   ------------------------------  # Date: Thu, 20 Nov 2003 20:32:18 GMTa# From: hoff@hp.nospam (Hoff Hoffman)s- Subject: Re: Bugcheck in SYSMAN.EXE on reboote1 Message-ID: <mx9vb.9439$fk.5862@news.cpqcorp.net>t  b In article <20031120144834.10071.00000629@mb-m01.aol.com>, ejheller@aol.com.com (EJHeller) writes:M :We have a new DS10 with OVMS 7.3.1 with all the latest and greatest critical2N :patches. Since we turned it on, every time we do a shutdown/reboot the system/ :bugchecks in SYSMAN.EXE with a code of 215....o  H   I'm not immediately familiar with a "code of 215" in this context, andH   it does not appear to translate as either a condition status value (in)   hex or decimal) nor as a bugcheck code.a  A :..recovery is to manually halt the machine and issue an INIT ...t  F   I suspect you will need to get into direct contact with the customerI   support center here, and I expect there will be need of detailed system J   I/O configuration and driver information, and details from the bugcheck.H   (As a start, when you re-bootstrap please get the crashdump CLUE CRASHG   and CLUE CONFIG information from the dumpfile using the SDA utility.)d  H   If you bootstrap conversationally and disable device autoconfigurationC   (system parameter NOAUTOCONFIG), how far can you get?  (I am here0D   making various guesses about the question, and on the bugcheck and   the cause of the bugcheck.)0  *   This could also potentially be hardware.  N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com    ------------------------------  + Date: Thu, 20 Nov 2003 21:11:40 +0000 (UTC) , From: lewis@mazda.mitre.org (Keith A. Lewis)- Subject: Re: Bugcheck in SYSMAN.EXE on reboott. Message-ID: <bpjamc$8rt$1@newslocal.mitre.org>   ejheller@aol.com.com (EJHeller) writes in article <20031120144834.10071.00000629@mb-m01.aol.com> dated 20 Nov 2003 19:48:34 GMT:O >psuedo-adventurous sort, I installed this patch. Now any boot after a shutdowns+ >requires an INIT before a successful boot.p  B As a software engineer, my opinion is "That's a hardware problem."   For a workaround, you could      >>>SET BOOT_RESET ON0 which causes an automatic INIT before each boot.  0 --Keith Lewis              klewis {at} mitre.org> The above may not (yet) represent the opinions of my employer.   ------------------------------  # Date: Fri, 21 Nov 2003 03:44:16 GMT % From: "Mike Naime" <mnaime@kc.rr.com>e- Subject: Re: Bugcheck in SYSMAN.EXE on reboota8 Message-ID: <kSfvb.58023$Eq1.1453@twister.rdc-kc.rr.com>  E If you are seeing  the following, this means that you have a hardwaregL problem and need to get it fixed!  Sending (FTP) the ERRLOG.SYS file that isK in the SYS$ERRLOG: directory to HP will let the hardware group diagnose ther exact problem.  A **** OpenVMS (TM) Alpha Operating System V7.3-1   - BUGCHECK ****n  K ** Bugcheck code = 00000215: MACHINECHK, Machine check while in kernel moder  L INITing the box is only masking the problem.  If it is a NEW DS10, it should still be under warranty.  
 Mike Naime  0 EJHeller <ejheller@aol.com.com> wrote in message4 news:20031120144834.10071.00000629@mb-m01.aol.com...E > We have a new DS10 with OVMS 7.3.1 with all the latest and greatest. criticalH > patches. Since we turned it on, every time we do a shutdown/reboot the systemJ > bugchecks in SYSMAN.EXE with a code of 215 (process = STARTUP). The onlyE > recovery is to manually halt the machine and issue an INIT from the  maintenance-> > prompt, then boot the computer. We could work around by do a shutdown/norebooto5 > and the issue the boot from the maintenance prompt. G > In looking at the patch site, I saw there is a patch for MANAGER. So,m	 being the G > psuedo-adventurous sort, I installed this patch. Now any boot after ao shutdown, > requires an INIT before a successful boot./ > Has anyone seen this problem and resolved it?c > Thanks for your input, > Edward Hellerc > TransCore ITS, Inc! > edward.heller@transcore-.-com..d   ------------------------------  % Date: Thu, 20 Nov 2003 12:42:51 -0800 # From: "Tom Linden" <tom@kednos.com>p2 Subject: RE: DCL Enhancement: PUSH/POP Environment9 Message-ID: <CIEJLCMNHNNDLLOOGNJIIECCIIAA.tom@kednos.com>   E have a look at ftp://freja.kednos.com/misc/pd.com then if you put it  1 in your login dir, put the following in login.com   " $ POPD == "@SYS$LOGIN:PD.COM POPD"# $ PD   == "@SYS$LOGIN:PD.COM PUSHD"d     >-----Original Message-----e5 >From: Larry Kilgallen [mailto:Kilgallen@SpamCop.net]e+ >Sent: Thursday, November 20, 2003 10:02 AMp >To: Info-VAX@Mvb.Saic.Com3 >Subject: Re: DCL Enhancement: PUSH/POP Environment. >  >-C >In article <3FBBBEC1.AE19D599@applied-synergy.com>, Chris Scheers r$ ><chris@applied-synergy.com> writes:I >> There has been some recent discussion of having the capability to savejD >> and restore the entire DCL/Terminal/etc. environment in one shot. >> mK >> A mechanism that I have seen elsewhere is a CLI PUSH/POP capability thathD >> does this.  PUSH saves the current environment, POP restores it. 4 >> Obviously, the environments are saved as a stack. >> d@ >> The environment that is saved should include at least the DCLC >> environment, RMS environment, terminal characteristics (if any),tC >> symbols, and PPF logicals (possibly all supervisor process tablem >> logicals?). >a# >We have this.  It is called SPAWN.p >e >---' >Incoming mail is certified Virus Free.,; >Checked by AVG anti-virus system (http://www.grisoft.com).eA >Version: 6.0.536 / Virus Database: 331 - Release Date: 11/3/2003e >p ---h& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.536 / Virus Database: 331 - Release Date: 11/3/2003   ------------------------------  % Date: Thu, 20 Nov 2003 17:23:53 -0800Q# From: "Tom Linden" <tom@kednos.com>-2 Subject: RE: DCL Enhancement: PUSH/POP Environment9 Message-ID: <CIEJLCMNHNNDLLOOGNJIMECMIIAA.tom@kednos.com>V  D Apologies, I noticed from the logs that I had forgotten to reset the, protections on the file, it should work now.   >-----Original Message-----o) >From: Tom Linden [mailto:tom@kednos.com] + >Sent: Thursday, November 20, 2003 12:43 PMt >To: Info-VAX@Mvb.Saic.Com3 >Subject: RE: DCL Enhancement: PUSH/POP Environment  >. >iE >have a look at ftp://freja.kednos.com/misc/pd.com then if you put it 2 >in your login dir, put the following in login.com >i# >$ POPD == "@SYS$LOGIN:PD.COM POPD"e$ >$ PD   == "@SYS$LOGIN:PD.COM PUSHD" >e >j >>-----Original Message-----6 >>From: Larry Kilgallen [mailto:Kilgallen@SpamCop.net], >>Sent: Thursday, November 20, 2003 10:02 AM >>To: Info-VAX@Mvb.Saic.ComV4 >>Subject: Re: DCL Enhancement: PUSH/POP Environment >> >>C >>In article <3FBBBEC1.AE19D599@applied-synergy.com>, Chris Scheerss% >><chris@applied-synergy.com> writes: J >>> There has been some recent discussion of having the capability to saveE >>> and restore the entire DCL/Terminal/etc. environment in one shot.R >>>tL >>> A mechanism that I have seen elsewhere is a CLI PUSH/POP capability thatD >>> does this.  PUSH saves the current environment, POP restores it.5 >>> Obviously, the environments are saved as a stack.m >>>sA >>> The environment that is saved should include at least the DCL D >>> environment, RMS environment, terminal characteristics (if any),D >>> symbols, and PPF logicals (possibly all supervisor process table >>> logicals?).  >>$ >>We have this.  It is called SPAWN. >> >>---n( >>Incoming mail is certified Virus Free.< >>Checked by AVG anti-virus system (http://www.grisoft.com).B >>Version: 6.0.536 / Virus Database: 331 - Release Date: 11/3/2003 >> >---' >Outgoing mail is certified Virus Free.t; >Checked by AVG anti-virus system (http://www.grisoft.com).iA >Version: 6.0.536 / Virus Database: 331 - Release Date: 11/3/2003n >. --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.536 / Virus Database: 331 - Release Date: 11/3/2003   ------------------------------  # Date: Thu, 20 Nov 2003 18:53:45 GMT & From: jlsue <jefflsxxxz@sbcglobal.net>A Subject: Re: DCL Enhancements: Error messages for OPEN and DEFINES8 Message-ID: <s53qrv85jora1ghq0udligigmuhs1dlin8@4ax.com>  H On Wed, 19 Nov 2003 19:24:24 +0100, Paul Sture <nospam@sture.homeip.net> wrote:     > 7 >It _has_ to be compatible with current behaviour IMHO.r >/ >Let's look at HELP OPEN/ERROR >b$ >     If the /ERROR qualifier is not: >     specified, the current ON condition action is taken. >pI >Oh dear, currently working procedures could suddenly do a silent jump in G >the code if the default were changed. There's simply too much code outa	 >there...a  I But how is that different than the change that required a "$" on each DCLs
 command line?t  J If it's an error situation, and an error is not being reported, then it isD a *latent* bug in the code and/or DCL that will eventually bite you.  H In this particular case, it HAS caused problem for myself and some otherF customers.  It is entirely non-intuitive, and inconsistent, that a DCLK procedure exiting abnormally, leaving a file opened, would *appear* to openaG the file, but really just continue from the last position.  I know thathI this should be handled by "on" conditions, but that's not always going toi happen.c  J Another example that's less so obvious (as I believe Didier showed before)H was where two pieces of code open different files but use the same name.C Again, not reporting an error is a big problem.  And the programmeruG shouldn't have to know/code something special to make that error report=
 correctly.  # Latent bugs are pretty scarey, imho= --- jls=0 The preceding message was personal opinion only.6 I do not speak in any authorized capacity for anyone,  and certainly not my employer.- (get rid of the xxxz in my address to e-mail)d   ------------------------------  % Date: Fri, 21 Nov 2003 07:24:11 +0100=* From: Paul Sture <nospam@sture.homeip.net>A Subject: Re: DCL Enhancements: Error messages for OPEN and DEFINE00 Message-ID: <3FBDBD9B.45F9B829@sture.homeip.net>   Didier Morandi wrote:i >  > Paul Sture wrote:e > : > > It _has_ to be compatible with current behaviour IMHO. > > ! > > Let's look at HELP OPEN/ERROR  > >l' > >      If the /ERROR qualifier is not = > >      specified, the current ON condition action is taken.r > >yL > > Oh dear, currently working procedures could suddenly do a silent jump inJ > > the code if the default were changed. There's simply too much code out > > there... > P > This is not the point, Paul, if you permit. There IS an error, as the OPEN didQ > not occur, for whatever reason. The point is to let the user know that the OPENtF > did not occur. See also my reply to JLSUE on 19-nov-2003 09:00 GMT+1 >   G Rhubarb, Didier. I am talking about very real potential side effects of-H such a change. It's all very well having a discussion of how it _should_H have been implemented all those years ago, but in practice, changing theG default behaviour could cost customers so much grief that it is another. nail in VMS' coffin.       -- <
 Paul Sture   ------------------------------  % Date: Fri, 21 Nov 2003 07:39:28 +0100o* From: Paul Sture <nospam@sture.homeip.net>A Subject: Re: DCL Enhancements: Error messages for OPEN and DEFINEr0 Message-ID: <3FBDC130.3A2617AE@sture.homeip.net>   jlsue wrote: > J > On Wed, 19 Nov 2003 19:24:24 +0100, Paul Sture <nospam@sture.homeip.net> > wrote: >  > > 9 > >It _has_ to be compatible with current behaviour IMHO.g > >l  > >Let's look at HELP OPEN/ERROR > >n& > >     If the /ERROR qualifier is not< > >     specified, the current ON condition action is taken. > > K > >Oh dear, currently working procedures could suddenly do a silent jump incI > >the code if the default were changed. There's simply too much code out  > >there...) > K > But how is that different than the change that required a "$" on each DCL  > command line?M   IIRC that was not silent.i   > L > If it's an error situation, and an error is not being reported, then it isF > a *latent* bug in the code and/or DCL that will eventually bite you. > J > In this particular case, it HAS caused problem for myself and some otherH > customers.  It is entirely non-intuitive, and inconsistent, that a DCLM > procedure exiting abnormally, leaving a file opened, would *appear* to openeI > the file, but really just continue from the last position.  I know that K > this should be handled by "on" conditions, but that's not always going tod	 > happen.  > L > Another example that's less so obvious (as I believe Didier showed before)J > was where two pieces of code open different files but use the same name.E > Again, not reporting an error is a big problem.  And the programmer.I > shouldn't have to know/code something special to make that error reporte > correctly. > % > Latent bugs are pretty scarey, imhov  E I'll agree about the inconsistencies, and it should perhaps have been 6 fixed in the V4.0 timeframe, when plenty of DCL broke.  H It's simply been there so long that changing the default behaviour would be equally scary.g  A I've known about it for at least 20 years, and even used it to myg3 advantage on occasions (usually for DCL debugging).e  E I'm arguing that it's too late now to change the _default_ behaviour.t  D Far easier to document it as a "feature", and as suggested elsewhereG have a logical or appropriate SET command to enable enhanced OPEN errort checking (off by default).   -- n
 Paul Sture   ------------------------------  # Date: Thu, 20 Nov 2003 19:22:31 GMT / From: "Jeff Goodwin" <jgoodwin@maine.rrr-r.com>c Subject: Re: DSN?n9 Message-ID: <Xv8vb.130678$ZC4.81117@twister.nyroc.rr.com>n  > I've just used DSNlink ITS and MAIL today.  No current issues.   -Jeffs    . "John N." <JNixon@cfl.rr.com> wrote in message5 news:xs8vb.13927$86.285443@twister.tampabay.rr.com...iH > Is DSN dead yet?  I cannot get our modem to make a connection.  I have heard=A > rumors that it would go away soon, but I refused to believe it.o >- >I   ------------------------------  # Date: Thu, 20 Nov 2003 19:18:53 GMTL# From: "John N." <JNixon@cfl.rr.com>P
 Subject: DSN?S; Message-ID: <xs8vb.13927$86.285443@twister.tampabay.rr.com>   L Is DSN dead yet?  I cannot get our modem to make a connection.  I have heard? rumors that it would go away soon, but I refused to believe it.4   ------------------------------    Date: 20 Nov 2003 14:38:14 -08002 From: hoefelmeyer@hotmail.com (Cheryl Hoefelmeyer)F Subject: Efficiency Question: Large Arrays vs. Indexed Files on Alphas= Message-ID: <72a72e76.0311201438.3bdafc33@posting.google.com>h  F We have a GS80 and an ES40, not clustered, each running OpenVMS 7.3. IE am writing a program in Basic that will run on each box, and I have afE question regarding which of the following would be most efficient, inr general.  F The program will operate on three very large files, and at one point I
 can either@ (1) decide to read some fields of some records into an array and9 search for information sequentially through the array, or/F (2) read the information into another file and access it each time via single read.F The number of elements or records in this circumstance is not expected to exceed 10,000.2  ' Elsewhere in the program, I can either >F (1) maintain records that will eventually be written to an output fileD in an array because they may need to be updated some number of times$ before a final record is written, orB (2) write the first record to the output file and just update each time it is necessary.cC The first option would call for reading sequentially through a veryoF large array to find the proper record to update each time a new recordB is added. The second calls for 1-3 file operations per each record? added. The number of records maintained here is on the order ofu
 1,000,000.  F So, for each of these, which is the best option in general? Years ago,B I/O on DEC boxes was such a bottleneck that the array option wouldC generally be the faster route, but I&#8217;m no longer sure that is  such a big issue now.d   ------------------------------  # Date: Thu, 20 Nov 2003 22:59:40 GMTk# From: hoff@hp.nospam (Hoff Hoffman)DJ Subject: Re: Efficiency Question: Large Arrays vs. Indexed Files on Alphas1 Message-ID: <wHbvb.9453$Qt.1128@news.cpqcorp.net>a  r In article <72a72e76.0311201438.3bdafc33@posting.google.com>, hoefelmeyer@hotmail.com (Cheryl Hoefelmeyer) writes:G :We have a GS80 and an ES40, not clustered, each running OpenVMS 7.3...g  ?   With the OpenVMS V7.3 XFC V2.0 ECO or later, I would suggest.   G :The program will operate on three very large files, and at one point Ia :can eithers  B   Please elaborate on your particular view of "very large files".   ,   10K records is rather small, for instance.  A :(1) decide to read some fields of some records into an array andi: :search for information sequentially through the array, or  -   Sequential searches are slow.  (Obviously.)   C   Depending on what you are up to, you might be able to lay out oneaB   or more RMS indexed files, with one or more likely with multiple   keys present in each file.  G :(2) read the information into another file and access it each time vian
 :single read.fG :The number of elements or records in this circumstance is not expectedn :to exceed 10,000.  F   How big is a record here?  (Depending on what you are up to, systemsE   using files and RMS access will keep most everything in cache -- ifeH   you have enough memory, XFC will keep most everything cached for you.)  ( :Elsewhere in the program, I can either G :(1) maintain records that will eventually be written to an output filePE :in an array because they may need to be updated some number of timesh% :before a final record is written, oreC :(2) write the first record to the output file and just update each  :time it is necessary.  D   Or you can use RMS global buffers and/or XFC caching to manage and   maintain this for you.  D :The first option would call for reading sequentially through a veryG :large array to find the proper record to update each time a new recordVC :is added. The second calls for 1-3 file operations per each recordn :added.   A   You might have one to three file calls, but these might not mape@   to an equivalent number of disk I/O operations.  Additionally,A   RMS indexed file searches are binary in nature, and -- when then?   global buffers are sized appropriately -- the index trees are3B   kept in memory.  XFC provides additional capabilities here, too.  A :        The number of records maintained here is on the order of  :1,000,000.   H   A million records is slightly more serious, but still not particularly   large.  < :So, for each of these, which is the best option in general?  C   I'd tend to see if I could keep the whole mess in one or more RMS D   indexed files, and use RMS global buffers and XFC caching.  RMS isA   good at searches and caching, and its easier and quicker to usee   existing code.  E   The real question here is one of application design and applicationnE   requirements, and your proposed design is rather close to that of a-G   database application.  MySQL might be an option here, or a commercialhE   database package could potentially be pressed into service as well.e  F   Performance requirements are another obvious consideration -- if theD   cost of writing and tuning outweighs the performance requirements,F   a "dumber" and slower design can be a better choice.  If performanceE   is a more central issue, then there are other design considerationseI   that come into play and there is a corresponding incentive spend (more)cH   on the design and on the coding and tuning efforts -- and you can tuneG   RMS file access, as well -- there are some comparatively easy ways to E   tune performance, such as increasing the allocation size and extend &   size, and upping the buffer sizes...    N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faqlN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com    ------------------------------  % Date: Thu, 20 Nov 2003 15:05:26 -0800P* From: "Jack Peacock" <peacock@simconv.com>J Subject: Re: Efficiency Question: Large Arrays vs. Indexed Files on Alphas2 Message-ID: <VPadnd3HNqQq1SCi4p2dnA@mpowercom.net>  ? "Cheryl Hoefelmeyer" <hoefelmeyer@hotmail.com> wrote in message67 news:72a72e76.0311201438.3bdafc33@posting.google.com... H > We have a GS80 and an ES40, not clustered, each running OpenVMS 7.3. IG > am writing a program in Basic that will run on each box, and I have amG > question regarding which of the following would be most efficient, inO
 > general. >nH > The program will operate on three very large files, and at one point I > can eitherB > (1) decide to read some fields of some records into an array and; > search for information sequentially through the array, oraH > (2) read the information into another file and access it each time via > single read.H > The number of elements or records in this circumstance is not expected > to exceed 10,000.0 >0H If you opt for (1) why not sort the array and use a binary search ratherJ than a sequential pass?  Memory is fuzzy but aren't there some VMS or LIB$ services for this?  I I wouldn't categorize 10K records as a very large file unless the records0K are extremely large themselves.  I've run RMS indexed files in the millions-F of records and find the response time to search is barely noticable toF users, perhaps a slight hesitation of 2-3 seconds to retrieve a set ofL records distributed across the file.  Is there a reason you can't index your- primary files, or build a separate tag index?e    Jack Peacock.   ------------------------------  % Date: Thu, 20 Nov 2003 15:10:00 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>$ Subject: Re: EV79 CANCELED !!!!!!!!!) Message-ID: <3FBD1F97.45A03914@istop.com>m   jlsue wrote:K > Well, it really doesn't make total sense to expect some huge drop anyway.DK > From the business' perspective, unless they were already on the very edgeeK > of performance needs with EV68 and the current EV7 - so much so that they F > could't wait for EV7x - they don't really need to change their plansM > because the systems are still shipping and are supported according to plan.o    M That decision (to kill EV79) alone won't kill plans a customer might have had4K to stay with VMS. But it may be the straw that breaks the camel's back. AndoK there have been plenty of such straws in past years, at a time when VMS was M already struggling to retain customers sdue to lack of marketing, disapearingh applications etc.   K > matters much for most businesses.  If the business is already invested in?K > Alpha technology, this doesn't really mean much in their plans to migratea  K Yes it does. Combined with all other such decisions, this may mean that the L business has decided to stay away from anything HP because they have learned% that they cannot trust HP's promises.x  N Yes it does. Combined with all other such decisions, customers now see a clearL pattern of HP telling customers not to worry about VMS and Alpha when it hasH already decided in the back doors to act differently than promised. This5 happened with the Alpha murder and with EV7 and EV79.h  M If HP had taken a liking to VMS, mentioned it profusely in its presentations, L did some real marketing of VMS and instead of "promising" not to kill it forC at least 5 years, prmised to grow its installed base and bring backhN applications, then the whole Alpha murder wouldn't have been such a big issue.F The problem is that we are seing the OS being neglected by HP, and itsI platform being neglected by HP. Combine this with the Palmer and Capellas-E reigns, and it is bad news when a customer is forced to rething theiro2 platforms because of the forced migration to IA64.  N Like it or not, the now "accelerated" Alpha phasing out plans by HP is forcingM customers to develop a long term strategy, at first for all new applications,D7 and later on, for the remaining VMS based applications.B  K > HP is still delivering on the commitments for support and availability of.L > AlphaServer systems that should provide plenty of performance to not upset > most business plans.  M Well, your own use of the word "still" says it all. The way I read the above,s6 you're saying that HP hasn't yet changed those plans.   H > There's nothing in the EV7z announcement that changes the timeline for< > delivering OpenVMS for the Intel 64-bit processor systems.  I Yes there is. By accelerating Alpha's retirement (if you produce a slower L chip, it means it will stay marketable for a lesser time), you are bringiungN much closer the time at which customers are forced to consider the switch to a
 new platform..  M If customers are forced to do that decision making at a time when IA64 hasn't J yet proven itself to the marketplace, and at a time when its own future isN still clouded by the 64 bit 8086 everyone knows Intel has up its sleeve,  then9 customers won't be so keen to invest in an IA64 platform.2  J > long time.  The ability to take the newer Alpha CPUs and put them into aM > 64-CPU system, and even put these into VMScluster systems, provides quite a @ > lot of power and growth capability until well past 2007, imho.  E The problem is that the minute a company has decided VMS is no longer I strategic to them, new applicatiosn will go on another platform and thereeB won't be much need for growing CPU capacity for their VMS systems.   ------------------------------  % Date: Thu, 20 Nov 2003 17:37:00 -0500l! From: "Kevin F." <OpenVMS Rules!>m$ Subject: Re: EV79 CANCELED !!!!!!!!!* Message-ID: <3fbd429f@usenet01.boi.hp.com>  2 Interesting comments... Wrong.. but interesting...    7 "JF Mezei" <jfmezei.spamnot@istop.com> wrote in messagew# news:3FBBC6B4.120738E6@istop.com...3 > jlsue wrote:I > > I know of absolutely no - that's ZERO - customers who's business will	K > > suffer due to the switch from EV79's planned architecture, and the EV7zI > > architecture.. >cI > Because of Alpha's last iterations are slower than had been anticipatedV and8K > because the EV79 turns out to be a dud that is essentially the same as anD EV7,K > customers will abandon Alpha sooner rather than later. Problem is that inl/ > doing so, they also are likely to abandon HP.n >gI > One has to consider long term trends and long term strategic decisions.r YoulK > don't see sales impact the day after Carly makes an announcement for such: things.- >w >eI > > Hogwash.  HP is still delivering on the commitments for CPU upgrades.  >-J > Technically, the lawyers probably will agree with you. But in spirit, HP has2I > NOT delivered on promises. It has not made EV7 as fast as it could have0 been+ > and will not produce the shrink for EV79.- > , > What this does is affect HP's credibility. >rL > And considering that IA64 is still just a promise of something nice in theH > future, credibility is important. That is a good reason why many don't believerJ > IA64 will pan out, simply because we don't believe anything HP says with > regards to VMS.a   ------------------------------  # Date: Thu, 20 Nov 2003 23:38:22 GMTp# From: "John Smith" <a@nonymous.com>e$ Subject: Re: EV79 CANCELED !!!!!!!!!I Message-ID: <Ofcvb.128862$HoK.46503@news01.bloor.is.net.cable.rogers.com>t   JF Mezei wrote:e > jlsue wrote:D >> Well, it really doesn't make total sense to expect some huge dropF >> anyway. From the business' perspective, unless they were already onE >> the very edge of performance needs with EV68 and the current EV7 -1F >> so much so that they could't wait for EV7x - they don't really needG >> to change their plans because the systems are still shipping and are  >> supported according to plan.l >u >.F > That decision (to kill EV79) alone won't kill plans a customer mightD > have had to stay with VMS. But it may be the straw that breaks theA > camel's back. And there have been plenty of such straws in pastoF > years, at a time when VMS was already struggling to retain customers: > sdue to lack of marketing, disapearing applications etc. > @ >> matters much for most businesses.  If the business is alreadyG >> invested in Alpha technology, this doesn't really mean much in their[ >> plans to migrate" > D > Yes it does. Combined with all other such decisions, this may meanE > that the business has decided to stay away from anything HP becauser9 > they have learned that they cannot trust HP's promises.l > D > Yes it does. Combined with all other such decisions, customers nowD > see a clear pattern of HP telling customers not to worry about VMS@ > and Alpha when it has already decided in the back doors to actD > differently than promised. This happened with the Alpha murder and > with EV7 and EV79. >-@ > If HP had taken a liking to VMS, mentioned it profusely in its> > presentations, did some real marketing of VMS and instead ofF > "promising" not to kill it for at least 5 years, prmised to grow itsB > installed base and bring back applications, then the whole AlphaD > murder wouldn't have been such a big issue. The problem is that we@ > are seing the OS being neglected by HP, and its platform beingD > neglected by HP. Combine this with the Palmer and Capellas reigns,? > and it is bad news when a customer is forced to rething their54 > platforms because of the forced migration to IA64. > E > Like it or not, the now "accelerated" Alpha phasing out plans by HP-D > is forcing customers to develop a long term strategy, at first forA > all new applications, and later on, for the remaining VMS based  > applications.  > < >> HP is still delivering on the commitments for support andD >> availability of AlphaServer systems that should provide plenty of0 >> performance to not upset most business plans. >lD > Well, your own use of the word "still" says it all. The way I readB > the above, you're saying that HP hasn't yet changed those plans. >jE >> There's nothing in the EV7z announcement that changes the timelineeA >> for delivering OpenVMS for the Intel 64-bit processor systems.- > D > Yes there is. By accelerating Alpha's retirement (if you produce aG > slower chip, it means it will stay marketable for a lesser time), yourE > are bringiung much closer the time at which customers are forced toe( > consider the switch to a new platform. > C > If customers are forced to do that decision making at a time whendF > IA64 hasn't yet proven itself to the marketplace, and at a time whenC > its own future is still clouded by the 64 bit 8086 everyone knows E > Intel has up its sleeve,  then customers won't be so keen to invest, > in an IA64 platform. > D >> long time.  The ability to take the newer Alpha CPUs and put themD >> into a 64-CPU system, and even put these into VMScluster systems,F >> provides quite a lot of power and growth capability until well past >> 2007, imho. >aG > The problem is that the minute a company has decided VMS is no longer E > strategic to them, new applicatiosn will go on another platform andtA > there won't be much need for growing CPU capacity for their VMSr
 > systems.    H Also see the thread started on 2003-11-20 @ 10:42am entitled 'PeoplesoftJ Financials and OpenVMS' for a taste of the continuing decline of VMS awareE applications, and this one is one of the ones that helps sell monstern" machines to Fortune 500 companies.   HP just doesn't f*cking get it.f   ------------------------------  + Date: Thu, 20 Nov 2003 15:05:18 -0600 (CST)o From: sms@antinode.org2 Subject: Re: Hobbyist license and layered products) Message-ID: <03112015051805@antinode.org>>  " From: Didier Morandi <no@spam.com>  I > >    So, now, as soon as I become a Software Products Library customer,  > > I'll be all set? y > B > What do you wish me to say, sms? I filled in the form, I got theE > newsletter with  my username and my password and I did successfullyu > download the stuff I needed.  H    Exactly that.  The HP info suggests that it would be a waste of time,: but if you say it's not, that makes a difference.  Thanks.  D    People suggest that I do pointless things all the time.  I try to  filter them out, where possible.  H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-orgu    Saint Paul  MN  55105-2547X   ------------------------------  % Date: Thu, 20 Nov 2003 22:42:34 +0100c" From: Didier Morandi <no@spam.com>2 Subject: Re: Hobbyist license and layered products4 Message-ID: <3fbd354d$0$17086$626a54ce@news.free.fr>   sms@antinode.org wrote:2  J >    Exactly that.  The HP info suggests that it would be a waste of time,< > but if you say it's not, that makes a difference.  Thanks.   We aim to please.i  F >    People suggest that I do pointless things all the time.  I try to" > filter them out, where possible.  I If "People suggest that you do pointless things all the time" and you do u= not agree at all, may I suggest that you pass the MENSA test  F (http://www.mensa.org/workout2.html) to see if you aren't "different"?   :-)a   D. Pdt MENSA Toulouse http://mensa.midipy.free.fr/   ------------------------------  # Date: Fri, 21 Nov 2003 02:32:01 GMT ! From: slaX0r <slaX0r@example.com>p2 Subject: Re: Hobbyist license and layered products; Message-ID: <BOevb.17231$86.338955@twister.tampabay.rr.com>    --____LPHMXLZMXOMRLFKSEJCW____, Content-Type: text/plain; charset=iso-8859-1+ Content-Transfer-Encoding: quoted-printablet         ----------------F "If you don't like the road you're walking, start paving another one."          -Dolly Parton  J >>> Phillip D. Williams<dwilliams296@comcast.net> 11/20/2003 12:14:42 AM = >>><  I You can ftp 5.5 BASIC/COBOL/FORTRAN/VAXC and others from my site. Right =e nowF its straight ftp fromnC 68.35.167.136 username VAXVMS password VAXVMS. The system is a 7.3.- Hope this helps-   Yes! Thank you.p   --____LPHMXLZMXOMRLFKSEJCW____H Content-Type: multipart/related; boundary="____WHPEPQYSAQXEHDGESJXG____"     --____WHPEPQYSAQXEHDGESJXG____- Content-Type: text/html; charset=windows-1252R+ Content-Transfer-Encoding: quoted-printable-   <HTML><HEAD>L <META http-equiv=3DContent-Type content=3D"text/html; charset=3Diso-8859-1"= >1@ <META content=3D"MSHTML 6.00.2800.1276" name=3DGENERATOR></HEAD>7 <BODY style=3D"MARGIN: 4px 4px 1px; FONT: 10pt Tahoma">  <DIV>&nbsp;</DIV>a <DIV>&nbsp;</DIV>eJ <DIV><BR>----------------<BR>"If you don't like the road you're walking, =L start paving another one."<BR>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=L sp; -Dolly Parton<BR><BR>&gt;&gt;&gt; Phillip D. Williams&lt;dwilliams296@c=< omcast.net&gt; 11/20/2003 12:14:42 AM &gt;&gt;&gt;<BR></DIV>L <DIV style=3D"COLOR: #000000">You can ftp 5.5 BASIC/COBOL/FORTRAN/VAXC and =J others from my site. Right now<BR>its straight ftp from<BR>68.35.167.136 =D username VAXVMS password VAXVMS. The system is a 7.3.<BR>Hope this = helps<BR></DIV>.A <DIV style=3D"COLOR: #000000">Yes! Thank you.</DIV></BODY></HTML>     --____WHPEPQYSAQXEHDGESJXG____--    --____LPHMXLZMXOMRLFKSEJCW____--   ------------------------------  % Date: Fri, 21 Nov 2003 07:43:32 +0100 * From: Paul Sture <nospam@sture.homeip.net>2 Subject: Re: Hobbyist license and layered products0 Message-ID: <3FBDC224.5C03494F@sture.homeip.net>   Didier Morandi wrote:k >  > sms@antinode.org wrote:n > I > >    So, now, as soon as I become a Software Products Library customer,e > > I'll be all set? > R > What do you wish me to say, sms? I filled in the form, I got the newsletter withQ > my username and my password and I did successfully download the stuff I needed.t >  > My 2 euros/ > (sorry for the $/euro rate these days, folks)i >   G In actual fact, that rate probably had a positive effect on HP's latestf profit statement.R   -- h
 Paul Sture   ------------------------------    Date: 20 Nov 2003 11:02:58 -0800. From: al5vf03p02@sneakemail.com (William Webb)% Subject: HP 4th qtr earnings releasedr= Message-ID: <d5ce4b06.0311201102.1a17a898@posting.google.com>   B http://www.hp.com/hpinfo/investor/financials/quarters/2003/q4.html    I noticed something interesting:  6 (( Change font to monospaced and this will line up. ))  4 Personal Systems Division: revenues-  $6,000,000,0004 Personal Systems Division: profits-      $21,000,000  4 Enterprise Division: revenues-        $4,100,000,0004 Enterprise Division: profits-           $106,000,000  E Every dollar of Personal Systems Division Revenue contained .35 centsn of profit (that's .0035)  C Every dollar of Enterprise Systems Division Revenue contained 2.585H  cents of revenue (that's .02585)  A Assuming current ratio, (and does *anyone* really believe that PCuB margins are going to INCREASE in the future?) the Personal SystemsA Division will have to bring in over $30,000,000,000 of revenue to # match what Enterprise is doing now.o  D Does anybody really believe that possibility falls anywhere close to reality?  C Contradiction from those who took financial statement analysis moren1 recently than I did is wholeheartedly encouraged.B   WWWebb   ------------------------------  # Date: Thu, 20 Nov 2003 20:45:29 GMTi# From: "John Smith" <a@nonymous.com>i) Subject: Re: HP 4th qtr earnings released I Message-ID: <JJ9vb.128254$HoK.78666@news01.bloor.is.net.cable.rogers.com>    William Webb wrote:aD > http://www.hp.com/hpinfo/investor/financials/quarters/2003/q4.html > " > I noticed something interesting: >o8 > (( Change font to monospaced and this will line up. )) >.6 > Personal Systems Division: revenues-  $6,000,000,0006 > Personal Systems Division: profits-      $21,000,000 > 6 > Enterprise Division: revenues-        $4,100,000,0006 > Enterprise Division: profits-           $106,000,000 >aG > Every dollar of Personal Systems Division Revenue contained .35 centsf > of profit (that's .0035) >VE > Every dollar of Enterprise Systems Division Revenue contained 2.585u" > cents of revenue (that's .02585) >0C > Assuming current ratio, (and does *anyone* really believe that PCaD > margins are going to INCREASE in the future?) the Personal SystemsC > Division will have to bring in over $30,000,000,000 of revenue to-% > match what Enterprise is doing now.0 >QF > Does anybody really believe that possibility falls anywhere close to
 > reality? >eE > Contradiction from those who took financial statement analysis moreg3 > recently than I did is wholeheartedly encouraged.r    J PeeCee's are loss leaders for printer ink and paper. HP is really a forest products company.  :-)   ------------------------------  % Date: Thu, 20 Nov 2003 14:00:11 -0800T# From: "Tom Linden" <tom@kednos.com>E) Subject: RE: HP 4th qtr earnings released.9 Message-ID: <CIEJLCMNHNNDLLOOGNJIGECIIIAA.tom@kednos.com>e   >-----Original Message-----n) >From: John Smith [mailto:a@nonymous.com]$+ >Sent: Thursday, November 20, 2003 12:45 PM= >To: Info-VAX@Mvb.Saic.Com* >Subject: Re: HP 4th qtr earnings released >a >  >William Webb wrote:E >> http://www.hp.com/hpinfo/investor/financials/quarters/2003/q4.html  >># >> I noticed something interesting:L >>9 >> (( Change font to monospaced and this will line up. ))E >>7 >> Personal Systems Division: revenues-  $6,000,000,000p7 >> Personal Systems Division: profits-      $21,000,000t >>7 >> Enterprise Division: revenues-        $4,100,000,000 7 >> Enterprise Division: profits-           $106,000,000>  H Is a breakdown by main product lines available?  VMS, Tru64 HPUX W2K NSK MPIX Storage   >>H >> Every dollar of Personal Systems Division Revenue contained .35 cents >> of profit (that's .0035)s >>F >> Every dollar of Enterprise Systems Division Revenue contained 2.585# >> cents of revenue (that's .02585)t >>D >> Assuming current ratio, (and does *anyone* really believe that PCE >> margins are going to INCREASE in the future?) the Personal Systems D >> Division will have to bring in over $30,000,000,000 of revenue to& >> match what Enterprise is doing now. >>G >> Does anybody really believe that possibility falls anywhere close toe >> reality?0 >>F >> Contradiction from those who took financial statement analysis more4 >> recently than I did is wholeheartedly encouraged. >3 > K >PeeCee's are loss leaders for printer ink and paper. HP is really a forestr >products company.  :-)e >y >G >---' >Incoming mail is certified Virus Free.e; >Checked by AVG anti-virus system (http://www.grisoft.com).3A >Version: 6.0.536 / Virus Database: 331 - Release Date: 11/3/20034 >M ---r& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.536 / Virus Database: 331 - Release Date: 11/3/2003   ------------------------------    Date: 20 Nov 2003 21:41:42 -0800# From: dooleys@snowy.net.au (dooley)o) Subject: Re: HP 4th qtr earnings released = Message-ID: <1ca82fc6.0311202141.652fd3f6@posting.google.com>e  s al5vf03p02@sneakemail.com (William Webb) wrote in message news:<d5ce4b06.0311201102.1a17a898@posting.google.com>...sD > http://www.hp.com/hpinfo/investor/financials/quarters/2003/q4.html > " > I noticed something interesting: > 8 > (( Change font to monospaced and this will line up. )) > 6 > Personal Systems Division: revenues-  $6,000,000,0006 > Personal Systems Division: profits-      $21,000,000 > 6 > Enterprise Division: revenues-        $4,100,000,0006 > Enterprise Division: profits-           $106,000,000 > G > Every dollar of Personal Systems Division Revenue contained .35 centsn > of profit (that's .0035) > E > Every dollar of Enterprise Systems Division Revenue contained 2.585h" > cents of revenue (that's .02585)D Which is the reason Compaq bought Tandem and DEC in the first place. Phil   ------------------------------    Date: 20 Nov 2003 22:23:42 -0800# From: dooleys@snowy.net.au (dooley) ) Subject: Re: HP 4th qtr earnings releasede= Message-ID: <1ca82fc6.0311202223.4e937418@posting.google.com>i  d "Tom Linden" <tom@kednos.com> wrote in message news:<CIEJLCMNHNNDLLOOGNJIGECIIIAA.tom@kednos.com>... > >-----Original Message-----.+ > >From: John Smith [mailto:a@nonymous.com] - > >Sent: Thursday, November 20, 2003 12:45 PMn > >To: Info-VAX@Mvb.Saic.Com, > >Subject: Re: HP 4th qtr earnings released > >  > >p > >William Webb wrote:G > >> http://www.hp.com/hpinfo/investor/financials/quarters/2003/q4.html  > >>% > >> I noticed something interesting:S > >>; > >> (( Change font to monospaced and this will line up. ))r > >>9 > >> Personal Systems Division: revenues-  $6,000,000,000u9 > >> Personal Systems Division: profits-      $21,000,000w > >>9 > >> Enterprise Division: revenues-        $4,100,000,000I9 > >> Enterprise Division: profits-           $106,000,000  > J > Is a breakdown by main product lines available?  VMS, Tru64 HPUX W2K NSK > MPIX Storage > E They seem to separate the server lines but lump all software togethere Alpha    - down 5% NonStop  - down 5%7 PA       - down 10% at high-end, but flat in middle/low. X86      - up 9%3 storage  - up 14% at high-end, and up 109% midrange, software - up 20%  tape     - down 14% (and out)   1 Proliants and SANs seem to be where the money is.i Phil   ------------------------------  % Date: Thu, 20 Nov 2003 21:16:49 -0600,1 From: "David J. Dachtera" <djesys.nospam@fsi.net>e Subject: Re: IA64 TPC results'' Message-ID: <3FBD83A1.67F0EC71@fsi.net>w   John Smith wrote:e >  > David J. Dachtera wrote: > > John Smith wrote:n > >> > >> David J. Dachtera wrote:t > >>G > >>> JL is indeed an HP employee, AFAIK. Unfortunately, he suffers the I > >>> same "PollyAnna" syndrome we see from HP'ers all to frequently here9F > >>> in c.o.v.: Rosy outlook with no foundation in tangible evidence. > >>>eH > >>> I wish one - JUST *ONE* - HP'er would have the balls + backbone toD > >>> explain the bullshit (why one thing is said and another done),0 > >>> instead of just spouting more of the same. > >> > >> David,I > >>  > >> Possible reasons are that : > >> > >> a) They simply don't know.H > >> b) It's NDA material. > >> c) 'Not my department'.F > >> d) There is no logical reason behind the 'bullshit' and they'd be% > >> embarassed trying to explain it.s+ > >> e) It would be a career limiting move.c > >>I > >> Same reasons why there is no response from any HP employee in c.o.v. F > >> when the question is asked - "Why isn't there VMS advertising andI > >> marketing beyond booths at tradeshows devoted to VMS?" or "Are theren? > >> still really 411,000 VMS systems out there in active use?"  > >l$ > > Like I said: "balls + backbone". > L > Most likely it's e). I'm sure that there are some who 'know' but e) weighsK > heavily on their minds. When you live in Nashua and your family has rootsoM > there, where would you pick-up and move to after the witch hunt for leakersu > is through with you?  E From a different angle: How long could you live with yourself knowing ( what what you know and not speaking out?  A Like I said: "balls + backbone". Rarer qualities than honesty and 
 integrity.   -- o David J. Dachterad dba DJE Systems) http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/a   ------------------------------  % Date: Thu, 20 Nov 2003 21:20:24 -0600d1 From: "David J. Dachtera" <djesys.nospam@fsi.net>o Subject: Re: IA64 TPC results ' Message-ID: <3FBD8478.C14FC4E2@fsi.net>t   jlsue wrote: > 9 > On Wed, 19 Nov 2003 20:10:10 -0600, "David J. Dachtera"m  > <djesys.nospam@fsi.net> wrote: >  > >jlsue wrote:s > >> >  > >>P > >> So what you're saying is that the CEO and BOD of your organization actually+ > >> care whether it's an EV79 or and EV7z?a > > E > >I'm saying, in the broader context, that CEOs and BODs will not do B > >business with companies that have proven untrustworthy. Period. > K > To the extent, that they view this "technical" change as backing out of aE > commitment  
 (one of many)E  8 > that their business relied upon, then you are correct. > K > I haven't known too many IT management types (say 2nd level on up) who'vefI > cared all that much about the esoteric details of chip design.  Service8@ > levels to the business is where most of them spend their time.  E Um, HP's service levels are rivalling Compaq for new lows - don't kidoG yourself. Then again, service levels don't mean much when the vendor isaE known for casting commitments aside anytime they become inconvenient.e   --   David J. Dachterau dba DJE Systemsd http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/2   ------------------------------  + Date: Thu, 20 Nov 2003 19:53:54 +0000 (UTC)e. From: John Forkosh <john@SeeSigForAddress.com>" Subject: Re: lat for linux problem, Message-ID: <bpj64i$olj$1@reader2.panix.com>  ' H Vlems <hvlems.nieuw@zonnet.nl> wrote:, : Bart,fM : LAT is indeed completely different from DECnet. Now IIRC the LAT and DECnet3J : implementations for linux are in someway related. An invalid MAC address@ : for DECnet could perhaps affect LAT behaviour. Hence the post. : Hans  % Hi Hans (and Bart and Roland, et al),eF      Thanks for all the help.  Hans turns out to be right -- I changedH the MAC address to AA:00:04:00:14:04 corresponding to valid address 1.20B (or SCSSYSTEMID 1044) according to Hans' instructions, below.  The= DECserver 90L+ then connected immediately.  Problem all gone.lE      If you look back carefully at Caulfield's post that I reproducedlH earlier, you'll see that his last two hex digits were 04 (which xlate toI a valid DECnet address by Hans' instructions).  I'd intentionally changedsE that to 03 for my own reasons, not realizing it would cause problems.tG The posts I'd googled led me to believe that only the leading AA:00:04,p< which designate Digital, mattered.  That's apparently wrong.G      I understand Bart's point that "LAT has noting to do with DECnet". D But the 90L+'s firmware apparently checks more than it theoreticallyH needs to check, leading to Hans' remark that "an invalid MAC address forK DECnet could perhaps affect LAT behavior".  That's apparently what happenedaM with my 90L+, though Roland's DECservers apparently don't exhibit this little  glitch. L      And I guess if I edit sys$system:modparams.dat, change the SCSSYSTEMID,K and rerun autogen, then that will change the VAXstation's MAC address, too,h0 as per Hans' formula.  Haven't tried it, though.3 Thanks again, Hans and everybody, for all the help,r -- p> John Forkosh  ( mailto:  j@f.com  where j=john and f=forkosh )  + : "Bart Zorn" <Bart.Zorn@xs4all.nl> schreef-H : > LAT has nothing to do with DECnet (and vice versa). If the Linux boxJ : > is not running DECnet, there is no need to change it's MAC address. If> : > the Linux box does have a DECnet implementation, then thatD : > implementation will take care of the proper MAC address setting. : >lG : > This does not explain your problem, however, and I am afraid that IG : > cannot help you there.
 : > Bart Zorng : >V, : > "H Vlems" <hvlems.nieuw@zonnet.nl> wrote : > > John,,K : > > not sure whether it is cause of your problem but is aa-00-04-00-0a-03d : > > a valid DECnet address? L : > > IIRC one takes tha last two bytes and reverse them. The leading 6 bits9 : > > designate area, the trailing 10 designate the node.a- : > > 0a-03 -> 03-0A -> 0000 00 11 0000 10 10i9 : > >                                --------- ==========r, : > > That's 0.778 which is invalid, right ? : > > N : > > BTW I use LAT for linux (RedHat 9) without any problems, like Roland the& : > > server is started with defaults.
 : > > Hans : > > 8 : > > "John Forkosh" <john@SeeSigForAddress.com> schreefL : > > > Not sure if this is right group, but anybody here use Lat for Linux,> : > > >           http://linux-decnet.sourceforge.net/lat.htmlE : > > > I have a small 10base2 lan with some Linux PC's, some genuine H : > > > VMS VAXstations, and a DECserver 90L+.  The wires are okay sinceH : > > > I can ping/telnet/ftp between the VS's and PC's, and can connect" : > > > to the VS's from the 90L+. : > > >cK : > > > I also found out about the MAC address problem with the 90L+, i.e.,tH : > > > it barfs on packets from non-Digital addresses, and have applied : > > > the linux fixe5 : > > >           ifconfig hw ether AA:00:04:00:0A:03eF : > > > so the PC looks like a Digital address.  And I'm also applying( : > > > the extra lat-for-linux commands : > > >           latcp -j, : > > >           latcp -x 100 -s -a psistarJ : > > > (where psistar is the name of the PC), though I'm not sure whether% : > > > either is helpful in any way.  : > > >.J : > > > And all this kind of works -- I can sometimes  c psistar  from theD : > > > 90L+ and log on to the PC.  Works fine when it works at all. : > > >eB : > > > Problem is it always takes a _long_ time (90L+ prints lotsG : > > > of ...'s before I ever get a prompt) before connecting, and thezA : > > > 90L+ frequently just gives up with "service unavailable".oG : > > > But if I do manage to connect and then later logout/exit, I cans> : > > > always establish a new PC connection almost instantly.1 : > > > The VS's always connect almost instantly.a : > > >yC : > > > What might be wrong, and what can I do to fix it or tune ite : > > > or whatever?  Thanks,c : > > > --  F : > > > John Forkosh  ( mailto:  j@f.com  where j=john and f=forkosh ) -- n> John Forkosh  ( mailto:  j@f.com  where j=john and f=forkosh )   ------------------------------   Date: 21 Nov 2003 06:10:00 GMT) From: rankin@pactechdata.com (Pat Rankin)c Subject: Re: ODS2 file headerr/ Message-ID: <20NOV200322095918@pactechdata.com>   = In article <d28306e.0311191237.4bd4791e@posting.google.com>,\c/  denny.rich@swagelok.com (Denny Rich) writes...s [ fh2def.h ]C > And I did not find offsets defined for the "allocated blocks" andeF > "used blocks" values in the ODS2 file header.  Is this an oversight?( > or am I not looking in the right area?  F      The 32 byte "record attributes" field in the file header has whatE you're looking for.  You need <fatdef.h> (similar status to fh2def.h)hF in order to interpret it as individual fields.  The HIBLK (high block,D the allocated amount) and EFBLK (end of file block, the used amount)G fields are also vestiges of unreasonable PDP11 compatability; they hold G 32 bits of data by combining two 16 bit fields "backwards" (two little-eC endian words pasted together into a mixed-endian thunk instead of a>E 32 bit longword).  You'll also probably want the FFB (first free byteV of the last used block) field.  2                 Pat Rankin, rankin@pactechdata.com   ------------------------------  # Date: Thu, 20 Nov 2003 23:33:31 GMTE# From: "John Smith" <a@nonymous.com>-. Subject: Re: Peoplesoft Financials and OpenVMSJ Message-ID: <fbcvb.128840$HoK.125592@news01.bloor.is.net.cable.rogers.com>  
 renato wrote:i# > Is anyone else in this situation?  >hF > I work for a small public agency and we are on Peoplesoft FinancialsG > 7.5 SP2 (E&G) with our Oracle database server running on OpenVMS.  We4G > are in internal discussions of which version of Peoplesoft to upgradesG > to.  We would like to go the latest version (FM 8.4) but this versionwH > is not OpenVMS certified and will not be supported.  Currently we have> > no plans on moving away from OpenVMS since it will be a hugeD > undertaking for us.  We have even written a letter to Craig ConwayH > pleading our case.  How much of a change is there between 8.0 and 8.4? >eD > We believe our two options are to 1) upgrade to 8.0 which is stillD > OpenVMS supported, or 2) just go ahead with the 8.4 (8.8?) upgradeG > since for us OpenVMS is only the Oracle database server and have beenx1 > told that there probably won't be any problems.) >y8 > Any advise on this matter will be greatly appreciated. >q > Thank you...    B You should write Carly Fiorina, Rich Marchello, Mark Gorham, PeterK Blackmore, Scott Stallard, and any other HP executives you can think of andr@ tell them that due to lack of VMS advertising and marketing, andH consequential dwindling market share for VMS, your application vendor isF about to screw you over, and ask them what is HP going to do about it.   ------------------------------  + Date: Thu, 20 Nov 2003 20:07:20 +0000 (UTC)o. From: "Insomnee" <robert.heyes@btinternet.com> Subject: Re: Printer problem2 Message-ID: <bpj6tn$3so$1@hercules.btinternet.com>  L "Graham Burley" <burley.not-this@encompasserve-or-this.org> wrote in message3 news:3FBBD51F.65922EAE@encompasserve-or-this.org...p > Insomnee wrote:t >  > >lI > > One however has started wanting the Go button to be pressed before itr willK > > print anything. Example: It says load paper in tray 1, but if you presso GoI > > it overrides and takes from the bulk paper casette. Is there a way toa stop@ > > it from happening and take directly from Tray 3. I did add aK > > /separate=(reset=reset)) on a set queue command but it didnt work. I amdL > > sending it a PCL statement through a device library / form type and dontL > > really want to mess about with that as other printers are using the sameG > > module and working OK. We can print OK from Word and Windows etc...  > >d >tG > A wild guess would be it's an argument about stationery types and thelC > printer no longer thinks that tray N contains what you asked for.s> > Anyway, I'd start by printing out the printer configuration. >e' > Random related PCL control sequences:  >d+ > <ESC>&l26A              ! Paper size = A4t1 > <ESC>&l5H               ! Paper source = Tray 3y. > <ESC>&n6WdPlain         ! Paper type = plain >h >tJ > > Another problem I have is where I have a Square Root symbol at the top leftI > > of the first page, either when I have the banner page or as the firstr* > > character on the page if I go without. > >e >nF > I've seen this with the TCPIP telnet symbiont, in which case I think1 > you need to do this before you start the queue:  >e+ > $ define/system TCPIP$TELNETSYM_RAW_TCP 1i >p >e > Graham >i    I You're all stars shining brightly in the vms sky. i will give these a go.l   ------------------------------  # Date: Thu, 20 Nov 2003 19:08:01 GMTt& From: jlsue <jefflsxxxz@sbcglobal.net>E Subject: Re: Problem with performance while running FTP in batch mode 8 Message-ID: <i84qrvs62nbvju0mpu1lt5vk8l7oru8fhb@4ax.com>  : On Thu, 20 Nov 2003 11:56:26 +0100, "Michal von Grotthuss" <nospam@nospam.pl> wrote:m   >Hi all,: >I have already observed strage behaviour of FTP transfer.M >Session is established on VMS machine to Win2K server running MS ftp server.H   >fL >But when I run this in batch mode, transfer is terrible: about 40KB/sec!!!! >see result:B >1315099648 bytes received in 08:35:47.32 seconds (41.50 Kbytes/s)  H Just out of curiosity.... is the batch queue at a lower priority than an/ interactive user?  That could have some effect.a  I Also, did the batch job run at a different time of day, say the same timet that backups were running?     --- jlse0 The preceding message was personal opinion only.6 I do not speak in any authorized capacity for anyone,  and certainly not my employer.- (get rid of the xxxz in my address to e-mail)o   ------------------------------  % Date: Thu, 20 Nov 2003 21:31:11 -0600u1 From: "David J. Dachtera" <djesys.nospam@fsi.net>eE Subject: Re: Problem with performance while running FTP in batch modeu' Message-ID: <3FBD86FF.D1ACDAAE@fsi.net>    jlsue wrote: > < > On Thu, 20 Nov 2003 11:56:26 +0100, "Michal von Grotthuss" > <nospam@nospam.pl> wrote:v > 
 > >Hi all,< > >I have already observed strage behaviour of FTP transfer.O > >Session is established on VMS machine to Win2K server running MS ftp server.o >  > > N > >But when I run this in batch mode, transfer is terrible: about 40KB/sec!!!! > >see result:D > >1315099648 bytes received in 08:35:47.32 seconds (41.50 Kbytes/s) > J > Just out of curiosity.... is the batch queue at a lower priority than an1 > interactive user?  That could have some effect.c > K > Also, did the batch job run at a different time of day, say the same time  > that backups were running?  C ...and following onto Jeff's train of thought, does anything happen C interactively (for example, LOGIN.COM, etc.) that doesn't happen in0 batch?   Lots o' stuff to look into...u   -- R David J. Dachtera! dba DJE SystemsD http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/"   ------------------------------    Date: 20 Nov 2003 17:16:30 -0800# From: dooleys@snowy.net.au (dooley) " Subject: Re: SQLSRV V7.1-5 Problem= Message-ID: <1ca82fc6.0311201716.4b0a6eef@posting.google.com>f  x paul.j.whapshott@lineone.net (Paul Whapshott) wrote in message news:<51f54c31.0311200135.63f168c0@posting.google.com>... > Hi,e > G > Looking for help. I have an ES45 running VMS 7.3-1 and SQL V7.1-5 andi > RDB V7.1-10. > E > Since upgrading I am getting a large SQS_PROD03_SQLSRV_DIS00371.LOGm. > and am having to stop and start SQL servicesB > (@sys$manager:sqlsrv$shutdown71) and restart to close this large
 > logfile. > C > Every event seems to be logged into the file. The question is, IsaC > there a setting that can limit what gets logged into the logfile? 7 We are using 7.0 and logfiles are usually 3 blocks onlyeH Have a look at the script used to create the service (sqlsrv_create.sqs)F You should be able to change it using sqlsrv_manage using set or alter. If you do this, does it show verify on or off? $ mcr sqlsrv_managet SQLSRV> show settings;	 Settings:t     version:    v7.0     verify:     offl     output:     onD     config-file:        SYS$SYSROOT:[SYSMGR]SQLSRV_CONFIG_FILE.DAT;1     confirm:    on   ------------------------------  % Date: Fri, 21 Nov 2003 16:16:18 +1030n: From: "Barratt, Chris (FMC)" <Chris.Barratt@fmc.sa.gov.au>" Subject: RE: SQLSRV V7.1-5 ProblemN Message-ID: <E829CF9B8F94014887EBC61E1951B0CE9E75BA@sagemshs001.fmc.sa.gov.au>  # What version did you upgrade from ?.  L We noticed a large increase in dispatcher logs once we changed some databaseJ services to be transaction reusable instead of session reusable. To manageL this logging you can get a really big system disk, or restart the dispatcherI on a regular basis (how regular is dependent on your workload).  The only L other alternative is to turn logging off for the dispatcher and just turn itJ back on if you start experiencing a problem (although the logging seems to be a bit useless anyway).d   Cheers,  Chris    > -----Original Message-----< > From: Paul Whapshott [mailto:paul.j.whapshott@lineone.net]( > Sent: Thursday, 20 November 2003 20:06 > To: Info-VAX@Mvb.Saic.Coms  > Subject: SQLSRV V7.1-5 Problem >  >  > Hi,l > G > Looking for help. I have an ES45 running VMS 7.3-1 and SQL V7.1-5 andg > RDB V7.1-10. > E > Since upgrading I am getting a large SQS_PROD03_SQLSRV_DIS00371.LOG>. > and am having to stop and start SQL servicesB > (@sys$manager:sqlsrv$shutdown71) and restart to close this large
 > logfile. > C > Every event seems to be logged into the file. The question is, IsoC > there a setting that can limit what gets logged into the logfile?a >  >  > Thanks for any help, >  >  > Paul >    ------------------------------  % Date: Thu, 20 Nov 2003 13:40:34 -0500H* From: "Bill Todd" <billtodd@metrocast.net>B Subject: Re: Sun to use AMD Opteron - announcement expected Monday2 Message-ID: <Er2dnXhXJ7AClyCiRVn-hg@metrocast.net>  3 "jlsue" <jefflsxxxz@sbcglobal.net> wrote in messagea2 news:l2kprvcm3bajsrijfccd773gk820aol13a@4ax.com...J > On Thu, 20 Nov 2003 02:34:38 -0500, "Bill Todd" <billtodd@metrocast.net> > wrote: >  >f > > L > >Nope.  It's because while SPARC performance satisfies a goodly percentage ofJ > >Sun's customers (and SPARC64 will satisfy even more), it does not offer thesB > >leading-edge performance that *some* Sun customers may require. >iG > I'm curious, why was it OKAY for Sun to allow SPARC to drop behind in J > performance because it "satisfies a goodly percentage of" customers, but: > somehow it's not valid for HP to do the same with Alpha? >tG > Alpha may have been in a leadership position, but that really doesn'toG > negate that the same business principles can apply to the decision tos# > forego more long-term investment.   L Sun didn't 'allow' its SPARC implementation to fall behind:  it simply foundL itself unable to prevent it (possibly in part because of its in-order designD basis - which just might ring a bell in another context).  Alpha, byL contrast, not only retained its performance leadership but had very credibleK plans to extend it in the near future - and at a bargain cost compared with - the reported cost of Sun's SPARC development.>  H So the comparison you're attempting to draw is completely bogus.  Not toD mention your erroneous assumption that Sun is winding down its SPARCH investment:  rather, it's replacing continued development of its currentA lack-luster SPARC product with the combination of a superior (and A wholly-compatible) single-core product from Fujitsu and the (alsoeL compatible) Niagara many-core product that it's developing for about 2 years	 from now.e   >e > > K > >Opteron does offer that performance leadership, and extends (with binary J > >compatibility) a hardware line (x86) that Sun already sells, and is the mostL > >cost/effective general-purpose processor on the planet, and (not entirelyI > >coincidentally) gives Sun a chance to poke Intel in the eye with a bige stickmK > >(the embrace of Opteron by both IBM and Sun removes the last obstacle to>F > >penetrating business customers that Intel fanboys kept harping on). >SD > One problem, though, is that to date there are no mid- to high-end > AMD-based systems.  G Not a problem at all - because (as already noted) Sun isn't *replacing* H SPARC but using Opteron as a supplement in the areas where SPARC is (andH will probably remain) somewhat weak (i.e., low-thread-count applicationsH where many cores can't make up for leisurely single-thread performance).  K I'm sure that Sun would prefer that SPARC could cover such areas as well as-J Opteron can, because it would provide a completely seamless product range.L But since that goal has proven unrealistic, and since they have a very largeF customer base many of whom aren't likely to want to migrate off SPARC,E supplementing it with Opteron (running the same OSs) is the next bestr choice.m  G Whether Sun will also develop high-processor-count Opteron systems willeE likely depend in part upon their success with first SPARC64 and laterdC Niagara (and also on whether their new Opteron customer base startsyK requesting such product extension to give them the same hardware-compatibley# range as the SPARC range provides).T  9   There is no track record for these systems to determinee% > how they perform in the real world.e  J You mean the same as is true of Itanic systems?  Indeed, there seems to beF mounting anecdotal evidence that Itanic's real-world performance fallsE somewhat short of what its benchmark scores would suggest - but since5I Opteron can't depend upon the extensive compiler tweaks that benefit such K Itanic scores its real-world performance is likely to be closer to what its + benchmark results would lead one to expect.8  &   There's nothing to really compare to; > except workstation and hypothetical server-class systems.t >u > > L > >You don't seem to grasp something that even Intel apparently has:  in theK > >large-system space, single-thread performance is plateauing (or, to lookt atK > >it another way, CPU power has gotten so far ahead of memory latency that I > >increasing it farther makes little sense) and the future is throughput  (like H > >Tanglewood and Niagara - or EV8 with many-way SMT).  Niagara is Sun'sL > >large-system throughput solution, but in the 8-processor-and-under serverK > >space it could use the help of a high-performance engine like Opteron to:F > >handle situations where raw performance or cost/performance is more critical > >than SPARC compatibility. >eJ > And, to date, it's not clear how well Opteron will be able to perform inJ > the multi-processor space that competes with other server class systems.  L Actually, it's quite clear - at least up to the point where Sun doesn't needL Opteron (because for large-core-count systems Niagara will be ready about asK soon as any large Opteron system design is likely to be).  At 4 processors, F the current top-of-the-line Opteron currently out-performs the currentH top-of-the-line Itanic on benchmarks like SPECweb99_SSL and SPECint_rateL (well, AMD had only posted a superior 2-processor score for that last time IH looked) - and Opteron has speed upgrades scheduled at a faster rate than/ Itanic does, so this lead should only increase._  J > I'm not saying that it will fail, but there's no way to discuss relative3 > capabilities because they don't much exist today.C  K And there's no need to discuss them, either - because the capabilities thatM' Sun wants to leverage *do* exist today.H     They may have to make:I > trade-offs in the system design that inhibit it's competing well in the : > over-4-processor market.  We just really don't know yet.  H And don't care, as noted above (as least from Sun's viewpoint:  I'm sureH that Intel is nervous that someone will create a large Opteron system to' threaten Itanic across its full range).0   - bill   ------------------------------  % Date: Thu, 20 Nov 2003 13:56:15 -0500.* From: "Bill Todd" <billtodd@metrocast.net>B Subject: Re: Sun to use AMD Opteron - announcement expected Monday2 Message-ID: <QOydnUVn_Y7RkyCi4p2dnA@metrocast.net>  3 "Rick Jones" <foo@bar.baz.invalid> wrote in message#+ news:0o7vb.9422$z8.5159@news.cpqcorp.net... + > Bill Todd <billtodd@metrocast.net> wrote:PJ > > 2.  SPARC would not have a well-defined future product well on the way toK > > completion that offered even greater performance leadership; Alpha did.: >b= > So long before SPARC was terminated, Sun would have stopped9+ > development of future SPARC CPUs/systems?k  H Don't keep trying to twist things, Rick:  the point (in the hypotheticalG discussion of the conditions under which Sun might eventually terminate D future SPARC development) was that as well as not having had currentK leadership performance, Sun would have had no leadership to look forward to A in the future either - in marked contrast to Alpha in both cases.,  B Whether Sun eventually actually *does* decide to discontinue SPARCH development will presumably depend upon its estimation of the continuingJ demand for it (again unlike the case with Alpha, where demand was not onlyI more than sufficient to justify continued development but could have been-I significantly increased with any effort).  And whether at some later date/E Sun then terminates SPARC availability (and even later, support) will F presumably depend upon demand as well.  And since (unlike cHumPaq) SunJ actually listens to its customers (witness its reversal of the decision toG drop x86 Solaris), these decisions (if they occur) will likely prove at L least acceptable even if some customers might wish the need for them did not exist.   - bill   ------------------------------  % Date: Thu, 20 Nov 2003 15:17:19 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>B Subject: Re: Sun to use AMD Opteron - announcement expected Monday) Message-ID: <3FBD214D.A2234F92@istop.com>    jlsue wrote:L > in ".com" my arse).  IT managers have, unfortunately, bought into the ideaL > that Intel spells "industry standard".  I've never liked it, but I stopped2 > letting it get my blood boiling a long time ago.  K IT managers are under the impression that Intel makes only one product, theoF 8086. It is the 8086, made by many manufacturers which is the industryH standard. It is the 8086 which is ubiquitous, runs windows, available in/ commodity PCs, which is heavily advertised etc."  & It is the 8086 which is "compatible".     H IA64 is nothing more than an incompatible, expensive proprietary, singleH source product. It matches NONE of the attributes of a industry standard commodity product.   ------------------------------  % Date: Fri, 21 Nov 2003 01:21:21 -0500 * From: "Bill Todd" <billtodd@metrocast.net>B Subject: Re: Sun to use AMD Opteron - announcement expected Monday2 Message-ID: <a6mdnc3f6PtCMyCiRVn-jg@metrocast.net>  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:5qXmWRWqACW9@eisner.encompasserve.org...0@ > In article <sYSdnaUwDrD5jyaiRVn-tA@metrocast.net>, "Bill Todd"  <billtodd@metrocast.net> writes:< > > "Rob Young" <young_r@encompasserve.org> wrote in message >e > >o/ > >   Your "facts" have us believe Montecito is)I > >> little more than a shrink to 90 nm , 12 MByte cache , with less than % > >> stellar performance improvement.f > >?I > > No, what I described was the product that Intel had planned for 2004.0 Ther> > > details that Intel had made public are a matter of record. > >y >a< > No - just your view.  In February Intel execs were stating > 18 MByte L3 for Montecito.  2 Look at the date on the post you quoted, you dolt.   ...t  0 > Well, you would be wrong about Montecito then.  L Not according to the statements Intel was making up to the time I posted theE material you quoted (i.e., a month before the 18 MB figure surfaced).3     It is obviouslymF > a new core, right?  How else to explain the multi-threading features; > and that Intel states it will appear as 4 CPUs to the OS?m  I The degree to which it can be considered a 'new' core depends (as I noted$D previously) on the nature of that multi-threading support.  EV7, forD example, doesn't really qualify as a 'new' Alpha core, despite minor changes.     But/5 > yet, you would have us believe (your futures) that:h > L > "This article is as good an example of spin as I've ever seen.  Instead ofJ > questioning what caused Intel to slip Montecito (the Itanic shrink to 90 nm,,2 > with 12 MB of on-chip cache) from 2004 to 2005," >2. > Now obviously your sourcing isn't very good.  D As already noted, my sourcing was just fine as of the date I posted.     But you aren't alone@ > as Montecito is more than "a shrink to 90nm with 12 MB on-chip	 > cache."   B The fact that Intel changed its mind doesn't alter what its public7 statements had been up until January 16th of this year.   6   But Intel certainly wasn't/isn't going to let you inA > on the details and like many in the industry play cat and mouse( > with the competition.b  D The fact that Intel announced changes not only to Montecito's publicJ schedule (and had to rustle up an interim Madison product as a result) butG to significant aspects of its design (number of cores, cache size) lastlL January suggests rather that Intel ran into problems - related to its 90 nm.K process (and indeed we're seeing some problems there right now, though they J may be getting ironed out) and/or related to the products it felt it wouldI need (e.g., HP's announced intention to create a dual-chip module in 2004eL may have pointed out a need for a multi-core chip that it had not foreseen).  )   We still remember how great a performerr+ > P6 was when it came out, surprising many.   L Let's also remember your glowing predictions for McKinley (as of a post here on July 13, 2001):  F "A senior circuit designer put it to me that he was totally shocked inJ February when Intel pointed out that McKinley was/is at 1.4 GHz.  Don't be' surprised if McKinley comes in higher."(  J Well, we weren't surprised, but it wasn't because McKinley came in higher.   >1> > Any predictions on Montecito performance?  SpecInt?  SpecFP? > tpmC?b  I That really will depend first upon the degree to which Intel tames its 90sK nm. process, as I already noted.  If it doesn't, then Montecito won't offer L much of any performance boost over next year's MadisonII (which isn't likely; to be more than 10% - 20% faster than the current Madison).   L Assuming that the process *is* tamed, it will then depend upon whether IntelJ is willing to exceed its stated power envelope for Itanic2 (130W).  If notF (and TTBOMK Intel has said it won't change that envelope right throughI Montecito), then Montecito won't noticeably out-perform Madison II unlessr< you disable one of its two cores - kind of a waste, I'd say.  K So (if the process works out OK) I'd put Montecito's SPECint performance atmJ introduction (with one core disabled to limit power dissipation) at aroundJ 2000 (about what Opteron should be hitting in the same time-frame) and itsI SPECfp performance at around 3000 (yes, it should lead there).  Though if1H some critical elements of the SPEC suite fit into the larger cache which, didn't before those numbers could be higher.  I TPC-C performance is a lot harder to predict - especially if there's somesJ form of SMT in Montecito (though if it really is only two threads per coreI it sounds a lot more like the toy SMT in P4 than the serious SMT in EV8).iK My guess is that you'd get better tpmC by running both cores throttled back0F (to stay within the power envelope) than by running one at full speed,  regardless of the SMT situation.   >  > >>C > >> "(Chip) frequency will play a role, but it will be eclipsed by0 > > improvements int, > >> multicore and multithreading," he said. > >> > >> --- > >>A > >> Seven times the performance of Madison.  7 times the integerr, > >> performance, or 7 times floating point?  H Pretty clearly not, given the Intel quote immediately above.  Unless you* were referring to SPECint/fp_rates scores.     Or maybe 7 times tpmC?   That would be believable.    > >>A > >> I suppose Opteron is slated to double in performance in that.& > >> timeframe, or maybe triple?  LOL. > >-L > > It's seldom wise to speak from complete ignorance, Rob.  Opteron will beH > > dual-core itself in 2006 (if not earlier:  I'm not sure that has yet beenH > > clarified), in the same 65 nm. process that Tanglewood should be in, with aJ > > clock rate around 4 GHz and additional on-chip cache.  That translates to aB > > 4x performance improvement right there, leaving aside whatever enhancementsG > > the K9 core brings to the table.  So on a per-core basis it'll blownI > > Tanglewood (and any left-over Montecitos) completely away (TanglewoodlI > > offering a bit less per-core performance than Madison, from the above:F > > description), though Tanglewood may have some advantage in heavilyJ > > multi-threaded throughput depending (as noted) on exactly how major an > > advance the K9 core is.n >s >i > Opteron blow Tanglewood away?a  K Read the words, Rob:  *on a per-core basis* - i.e., for single-threaded (ore low-thread-count) execution.  #   We don't have substantial detailssC > yet.  We know now that Montecito is considerably more substantialiA > than it appeared at the beginning of the year.  Here's a guess.iC > I'd guess that the 7 times performance improvement would apply to>? > database/tpmC numbers.  After all, that is a heavily threaded ? > environment.  If tpmC and friends jump a factor of 7, that isSB > a sweet niche for IA64.  Personally, focusing on databases makes: > the most sense as that is where most of the money is at. > @ > Recent speculation has Tanglewood with SMT.  If they pull thatB > off, the game is over in my opinion.  8 cores with SMT?  Unreal.  3 Not really:  Niagara offers exactly the same thing.l   - bill   ------------------------------  # Date: Thu, 20 Nov 2003 19:27:54 GMT # From: "John N." <JNixon@cfl.rr.com> = Subject: SYS$LIBRARY:TRACE_V74.EXE on VMS 7.3-1 causing errord; Message-ID: <_A8vb.14025$86.286496@twister.tampabay.rr.com>s  K When rebuilding our application on our VMS 7.3-1 node, a programmer reportsn the following error:  0 ************************************************0 %LIBRAR-S-REPLACED, module TERMINATE replaced in- TECHDISK:[TECHSUP.PRODUCTS.BIN.AXP]PSAP.OLB;1- compiling TRACE.C-7 %LIBRAR-E-SEQNCE, illegal record sequence in module  inv" SYS$COMMON:[SYSLIB]TRACE_V74.EXE;1/ *********************************************** J This rebuild has worked well since VMS 4.X but seems to have broken on VMS 7.3-163 Is anyone familiar with  SYS$LIBRARY:TRACE_V74.EXE?r  6 Compiler is "Compaq C V6.5-001 on OpenVMS Alpha V7.3-1   ------------------------------  # Date: Thu, 20 Nov 2003 20:16:48 GMTc# From: hoff@hp.nospam (Hoff Hoffman)aA Subject: Re: SYS$LIBRARY:TRACE_V74.EXE on VMS 7.3-1 causing errori1 Message-ID: <Qi9vb.9437$fk.1049@news.cpqcorp.net>o  a In article <_A8vb.14025$86.286496@twister.tampabay.rr.com>, "John N." <JNixon@cfl.rr.com> writes:tL :When rebuilding our application on our VMS 7.3-1 node, a programmer reports :the following error:   F   Try this test again, but using routine names and/or module filenamesG   other than TRACE or DEBUG -- if I were guessing, your system probablye#   has a TRACE logical name defined.h  E   DEBUG is one of the more common ways folks get into trouble in thisn$   area, but TRACE works similarly.    E   There is a mechanism -- I do not recall if it has become documentednC   anywhere off-hand, but the scheme has been around for eons -- for F   substituting the standard OpenVMS debugger or the traceback handler,E   using definitions of logical names such as LIB$DEBUG and DEBUG, ando   LIB$TRACE and TRACE.    B   I suspect you are encountering a definition of TRACE, obviously.A   (Possibly from a debugger-related software kit, or an ECO kit?)o    N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faqIN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comn   ------------------------------  # Date: Thu, 20 Nov 2003 20:35:00 GMTk# From: "John N." <JNixon@cfl.rr.com>oA Subject: Re: SYS$LIBRARY:TRACE_V74.EXE on VMS 7.3-1 causing errorp; Message-ID: <Uz9vb.14793$86.292814@twister.tampabay.rr.com>    Thanks,s  9 There is a system logical  Trace =SYS$SHARE:TRACE_V74.EXEt  L I have asked the programmer to check his build procedures and not use TRACE.K This logical is defined on our only VMS 7.2-1 node and on 4 out of 5 of ourhL VMS 7.3-1 nodes.  I don't know where it gets defined or if I can prevent it.  0 "Hoff Hoffman" <hoff@hp.nospam> wrote in message+ news:Qi9vb.9437$fk.1049@news.cpqcorp.net... G > In article <_A8vb.14025$86.286496@twister.tampabay.rr.com>, "John N."t <JNixon@cfl.rr.com> writes:aF > :When rebuilding our application on our VMS 7.3-1 node, a programmer reportsD > :the following error:f >wH >   Try this test again, but using routine names and/or module filenamesI >   other than TRACE or DEBUG -- if I were guessing, your system probablyc% >   has a TRACE logical name defined.t >pG >   DEBUG is one of the more common ways folks get into trouble in this $ >   area, but TRACE works similarly. >iG >   There is a mechanism -- I do not recall if it has become documentedoE >   anywhere off-hand, but the scheme has been around for eons -- forHH >   substituting the standard OpenVMS debugger or the traceback handler,G >   using definitions of logical names such as LIB$DEBUG and DEBUG, andn >   LIB$TRACE and TRACE. >oD >   I suspect you are encountering a definition of TRACE, obviously.C >   (Possibly from a debugger-related software kit, or an ECO kit?)e >i >o( >  ---------------------------- #include' <rtfaq.h> -----------------------------t4 >     For additional, please see the OpenVMS FAQ --  www.hp.com/go/openvms/faqt, >  --------------------------- pure personal# opinion ---------------------------GG >         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com. >u >o   ------------------------------  % Date: Thu, 20 Nov 2003 21:47:57 +0000o- From: John Laird <nospam@laird-towers.org.uk>EA Subject: Re: SYS$LIBRARY:TRACE_V74.EXE on VMS 7.3-1 causing error 8 Message-ID: <tcdqrvc4sc0l3ohfg0f4ad290bc88i5j35@4ax.com>  F On Thu, 20 Nov 2003 19:27:54 GMT, "John N." <JNixon@cfl.rr.com> wrote:  L >When rebuilding our application on our VMS 7.3-1 node, a programmer reports >the following error:t > 1 >************************************************t1 >%LIBRAR-S-REPLACED, module TERMINATE replaced in . >TECHDISK:[TECHSUP.PRODUCTS.BIN.AXP]PSAP.OLB;1 >compiling TRACE.C8 >%LIBRAR-E-SEQNCE, illegal record sequence in module  in# >SYS$COMMON:[SYSLIB]TRACE_V74.EXE;1 0 >***********************************************K >This rebuild has worked well since VMS 4.X but seems to have broken on VMSn >7.3-14 >Is anyone familiar with  SYS$LIBRARY:TRACE_V74.EXE? >97 >Compiler is "Compaq C V6.5-001 on OpenVMS Alpha V7.3-1/  J Has the look of a $ LIBRARY/REPLACE PSAP.OLB TRACE command, where there isJ now a logical name TRACE pointing to SYS$LIBRARY:TRACE_V74.EXE.  The errorF message is consistent with the librarian reading that executable as an" object module and objecting to it.  J If you explicitly specify TRACE.OBJ on the $ LIBRARY command then the fileJ name won't be subject to any logical name translation.  In general this isL always a good idea in build environments - you can be lazier when typing DCL( as you will soon notice the unexpected !   --  ? You're growing old when your knees buckle and your belt won't. e   Mail john rather than nospam...    ------------------------------  + Date: Thu, 20 Nov 2003 20:28:29 +0000 (UTC)+P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) Subject: Re: TELNET/PORT$ Message-ID: <bpj85d$hq4$1@online.de>  1 In article <3FBBE688.D4E6551@istop.com>, JF Mezei $ <jfmezei.spamnot@istop.com> writes:   L > Which router allows you to remap a port to another port ? I'd like to have4 > that. I have a Netgear RT314 and it won't do that.  E It's a Linksys.  It's 500 km away right now and I don't remember the c> model number.  Email me Sunday if I haven't posted it by then.   ------------------------------  + Date: Thu, 20 Nov 2003 20:30:00 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) Subject: Re: TELNET/PORT$ Message-ID: <bpj888$hq4$2@online.de>  A In article <1031119200814.3301B-100000@Ives.egh.com>, John Santosm <JOHN@egh.com> writes: (  = > Maybe the mapping is only happening on inbound connections?0> > In otherwords, the packet has to come from the WAN interface; > and be destined for the LAN interface (or vice versa) forn > the mapping to occur?d  E That's what I'm doing.  Sometimes it DOES work.  It just gets "hung" nH sometimes, but only for the machine from which the previous port-mapped D connection had come.  Since one can get through from elsewhere, the  router itself is not hung.   ------------------------------  + Date: Thu, 20 Nov 2003 20:33:16 +0000 (UTC)hP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) Subject: Re: TELNET/PORT$ Message-ID: <bpj8ec$hq4$3@online.de>  A In article <3fbc9757$0$254$636a55ce@news.free.fr>, Didier Morandih <no@spam.com> writes:   K > >>There is a menu where one can forward incoming connections to port X to A > >>a specific port and a specific IP address behind the router.   > > N > > Which router allows you to remap a port to another port ? I'd like to have6 > > that. I have a Netgear RT314 and it won't do that. > Q > It's called NAT (as you know). You don't have a NAT submenu in your router WEB t > interface?  H Actually, NAT just substitutes one IP address for another.  This is PAT. See   7    http://www.garykessler.net/library/tcpip.html#IPcidrw   ------------------------------  + Date: Thu, 20 Nov 2003 20:37:52 +0000 (UTC)nP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) Subject: Re: TELNET/PORT$ Message-ID: <bpj8n0$hq4$4@online.de>  9 In article <00A292BB.5D44C34E@SendSpamHere.ORG>, VAXman- e @SendSpamHere.ORG writes:   F > I have a Netopia for T1 routing.  I *had* a Linksys for cable modem E > routing -- a piece of junk IMO (Linksys and the cable modem).  JustiE > recently I replaced the Linksys with a Netopia ethernet router.  IttC > does allow NAT and PAT assignments whereas the Linksys would not.e  I My Linksys is the one which does PAT (the rebadged Zyxel doesn't).  It's  I more or less OK.  Apart from the problem mentioned, it has "gotten hung" )% once or twice in the past year or so.y  E The main web page (there is no telnet interface to bring up an ASCII nD menu like on the rebadged Zyxel, though this is undocumented there) G needs a fancy browser to configure it, but fortunately I don't have to h do that often.  I I use it for a connection with a dynamic IP address which is updated via nC a DCL procedure which uses lynx -dump to query the router every 30 kF seconds or so.  (The Zyxel has a built-in update client, but it's not E reliable.  In any case, I update my connection quite often, not just  C when the IP address changes, so that the world will notice if I go 0E offline (instead of the IP address getting assigned to someone else, >, while the DNS record for me is still valid).   ------------------------------  % Date: Thu, 20 Nov 2003 22:44:00 +0100c" From: Didier Morandi <no@spam.com> Subject: Re: TELNET/PORT4 Message-ID: <3fbd35a1$0$17086$626a54ce@news.free.fr>  / Phillip Helbig---remove CLOTHES to reply wrote:o  C > In article <3fbc9757$0$254$636a55ce@news.free.fr>, Didier Morandi  > <no@spam.com> writes:  >  > K >>>>There is a menu where one can forward incoming connections to port X toeA >>>>a specific port and a specific IP address behind the router. a >>>IM >>>Which router allows you to remap a port to another port ? I'd like to haver5 >>>that. I have a Netgear RT314 and it won't do that.c >>Q >>It's called NAT (as you know). You don't have a NAT submenu in your router WEB r >>interface? >  > J > Actually, NAT just substitutes one IP address for another.  This is PAT.  I Yeah. I know. NAT = Network Address Translation. But my ADSL router does h0 offer a NAT menu, in which I can remap my ports.   D.   ------------------------------  % Date: Thu, 20 Nov 2003 11:07:34 -0800t# From: "Tom Linden" <tom@kednos.com>lY Subject: RE: The Alpha Chip Engineering Group Revival (aka: today's HP/Intel         pres 9 Message-ID: <CIEJLCMNHNNDLLOOGNJICEBPIIAA.tom@kednos.com>y   >-----Original Message----- * >From: Didier Morandi [mailto:no@spam.com]+ >Sent: Thursday, November 20, 2003 10:27 AMg >To: Info-VAX@Mvb.Saic.ComI >Subject: The Alpha Chip Engineering Group Revival (aka: today's HP/Intele4 >presentation in Toulouse) in Toulouse) in Toulouse) >  >c@ >I just attended an (excellent) HP/Intel co-presentation on the  >following subject:e >C  >"HP integrity Servers Strategy" >r% >Good news. Hear (I signed *no* NDA):b >h >[start of Intel speaker quote]h? >1. "The Intel Engineering Group is fully busy on the Itanium2   >chip, so Intel has = >given to the former Compaq Alpha Engineering Group that was   >purchased two years n+ >ago the mission to produce the Montecito."b >0F >Good news. If this is true, we may see a *really good* Itanium3 chip. > B >2. "We have invented with the Itanium family a revolutionary new  >technology. A gC >CPU has three functions: load/compute/store. The Itanium chip can t >load the next n5 >instruction while it is computing the previous one."c >o >no comment. >eB >3. "There are 75'000 employees at Intel today, among them 10'000 
 >are working  ) >only on software, and 700 on compilers".i  : I don't there are that many compiler engineers in the USA.   >v' >10'000 on software? Nice to hear that.  >o >[start of HP speaker quote]< >4. "When we had to merge HP and COMPAQ, there were 215'000  >workstations and   >227'000 mailboxes to federate". >o, >%DCL-W-MBXDUP, duplicate mailboxes detected >p >Now, the best one : >eD >5. "The HP-UX/11 PA-RISC to Itanium migration is a non-destructive 
 >upgrade". >-6 >??? HP Engineering has (had?) "destructive" upgrades? >:-) >r >D.D >d >---' >Incoming mail is certified Virus Free.:; >Checked by AVG anti-virus system (http://www.grisoft.com).uA >Version: 6.0.536 / Virus Database: 331 - Release Date: 11/3/2003n >u ---N& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.536 / Virus Database: 331 - Release Date: 11/3/2003   ------------------------------  % Date: Thu, 20 Nov 2003 14:21:17 -0500 * From: "Bill Todd" <billtodd@metrocast.net>Y Subject: Re: The Alpha Chip Engineering Group Revival (aka: today's HP/Intel presentation82 Message-ID: <XKudnVht3au3iSCiRVn-iw@metrocast.net>  / "Didier Morandi" <no@spam.com> wrote in messagei- news:3fbd0786$0$2773$626a54ce@news.free.fr...eJ > I just attended an (excellent) HP/Intel co-presentation on the following subject:  I Some of the quotes that you present below seem to call into some questionn; the 'excellence' (or even competence) of this presentation.e   >I! > "HP integrity Servers Strategy"0 >a& > Good news. Hear (I signed *no* NDA): >-  > [start of Intel speaker quote]H > 1. "The Intel Engineering Group is fully busy on the Itanium2 chip, so	 Intel haswK > given to the former Compaq Alpha Engineering Group that was purchased two  years0, > ago the mission to produce the Montecito."  K Now, that's interesting phrasing (assuming there's any truth in it at all):o  H Did the Montecito design team (which clearly didn't start out with AlphaI people) just throw the design over the wall for Alpha people to implement I (boy, I'll bet they would have been happy with that!)?  Or did an earlieriL Montecito design (the one that started out as a single-core, 12-MB-cache, 90C nm. 2004 product) get scrapped and replaced by a from-the-ground-upyE Alpha-team-designed product (I find that hard to believe as well, I'm.K afraid:  among other things, Alpha people likely wouldn't be farting around J with mere dual-thread SMT - and in any event if they only took over a yearL or so ago with a redesign - even one based on the already-started TanglewoodG work - shipping the product any time before 2006 would be a challenge).e   >sG > Good news. If this is true, we may see a *really good* Itanium3 chip.    Another major 'if' there.g   >QB > 2. "We have invented with the Itanium family a revolutionary new
 technology. AuL > CPU has three functions: load/compute/store. The Itanium chip can load the next6 > instruction while it is computing the previous one." >e
 > no comment.0   You're too kind.   ...B   > Now, the best one :i >aD > 5. "The HP-UX/11 PA-RISC to Itanium migration is a non-destructive	 upgrade".9 >47 > ??? HP Engineering has (had?) "destructive" upgrades?t  L Of course:  the only way one could characterize the transition from Alpha to5 Itanic as an 'upgrade' would be as a destructive one.c   - bill   ------------------------------    Date: 20 Nov 2003 17:38:09 -06004 From: kuhrt@nospammy.encompasserve.org (Marty Kuhrt) Subject: The Uptimes Project3 Message-ID: <4bg$QxTIG81R@eisner.encompasserve.org>e  ? It might be a meaningless milestone, but my production cluster  > machines in Ireland are now number two and three in the Active= Hosts list at Uptimes.  The cluster uptime exceeded 1000 days  today.  3 http://uptimes.hostingwired.com/stats.php?op=active.   ------------------------------  % Date: Thu, 20 Nov 2003 18:19:43 -0800n3 From: Alan Frisbie <Usenet01REMOVE@Flying-Disk.com>m  Subject: Re: The Uptimes Project. Message-ID: <3FBD763F.6050406@Flying-Disk.com>   Marty Kuhrt wrote:A > It might be a meaningless milestone, but my production cluster V@ > machines in Ireland are now number two and three in the Active? > Hosts list at Uptimes.  The cluster uptime exceeded 1000 days  > today. > 5 > http://uptimes.hostingwired.com/stats.php?op=activeu  < My production system (non-clustered) at Nelson Nameplate was> doing great at 702 days until we had a long power failure last> week which dropped us back to zero.   We were only a few weeks? short of two years without a reboot.   Sigh.   I'll have to get 
 a bigger UPS.n  ? Were it not for that, we would have had VMS systems in three of28 the top four slots.   And all AlphaServer 1000A systems!   Alan   ------------------------------  % Date: Thu, 20 Nov 2003 19:54:49 -0600., From: "Dave Gudewicz" <k9jdk@NOSPAMarrl.net>& Subject: Re: VMS 2003 bootcamp summary/ Message-ID: <vrqs3bjv7geube@corp.supernews.com>C  L Dunno how many newbies there were, but that might be something to add to theH registration info. for next time.  And while it was good to see many old! faces, new faces are a good sign.e   Dave...   @ "Brad McCusker" <brad.mccuskerNosp@Mcompaq.com> wrote in message) news:vrll2c5isj6hcb@corp.supernews.com...  >dB > "Alan Frisbie" <Usenet01REMOVE@Flying-Disk.com> wrote in message* > news:3FBAC38C.5070103@Flying-Disk.com... >bG > >    o  Since the event is called "bootcamp", perhaps you should have + > >       a track dedicated to VMS newbies.i >nL > This is an interesting point.  I had one or two VMS-newbies in my hands-onG > Open Source Porting seminar each of the three days it was offered.  I  wonderG > how many newbies there actually was?   To provide a newbie track that L > included hands-on might be a big draw, if we are sure to mention it in the > program descriptions.r >rI > (The really good news is, the "newbies" were actually UNIX/Linux peopler andtF > they were impressed with BASH on VMS.  Had plenty of suggestions and@ > requests for missing utilities, but, impressed none the less!) >eJ > Any of the other attendees have a sense for how many newbies there were? >' > Brad >  >t   ------------------------------  % Date: Thu, 20 Nov 2003 22:02:21 -0700a+ From: "Barry Treahy, Jr." <Treahy@MMaz.com>s) Subject: Re: VMS Defragmentation Softwares' Message-ID: <3FBD9C5D.1060506@MMaz.com>    Bob Koehler wrote:  T >In article <3FB468F5.10708@MMaz.com>, "Barry Treahy, Jr." <Treahy@MMaz.com> writes: >, >  e >nH >>Yes, if expect that if you recall, Digital years back would state how G >>dangerous on-line defraging was, so their only 'approved' method was t >>backup/restore.  >>     >> >iG >   You're over a decade out of date.  Digital fixed this by adding thecH >   move file primitive to the OS.  Move file is approved, used by DEC's: >   (HP's) product, and used by most third party products. >    >pG You took what I said out of context.  Yes, I know about the primative,  G as does Raxco...  My reference to Digital was for the time frame prior aI to the primative, and how even then Raxco had a quality product that was n safe...s   BarryA   -- a  > Barry Treahy, Jr                       E-mail: Treahy@MMaz.com> Midwest Microwave                          Phone: 480/314-1320> Vice President & CIO                         FAX: 480/661-7028                        a   ------------------------------  % Date: Thu, 20 Nov 2003 22:05:42 -0700d+ From: "Barry Treahy, Jr." <Treahy@MMaz.com>r) Subject: Re: VMS Defragmentation Software ' Message-ID: <3FBD9D26.8090405@MMaz.com>t    Peter 'EPLAN' LANGSTOEGER wrote:  T >In article <3FB468F5.10708@MMaz.com>, "Barry Treahy, Jr." <Treahy@MMaz.com> writes: >  l >s" >>Peter 'EPLAN' LANGSTOEGER wrote: >>     >>m >>>In article <vr7ef11drh320f@corp.supernews.com>, Nathan Hartley <nathan@removethis.ilothlorien.com> writes:r	 >>>      n >>>oE >>>>Anyone have an opinion on VMS open-file defragmentation software?m >>>>         >>>>N >>>Yes, I use DEC's DFO/DFG/DEFRAG tool for many years now (running it daily).I >>>It still has some bugs, but I haven't had a data corruption so far ;-)tA >>>And it is also part of the OpenVMS hobbyist license program !!o >>>g	 >>>      e >>>cM >>>>I am currently looking at Raxco's Performance Suite and Executive's Disk a >>>>Keeper.l >>>>         >>>>I >>>I tend to prefer products of the same company as the underlying opsys.c >>> 	 >>>        >>>eH >>Yes, if expect that if you recall, Digital years back would state how G >>dangerous on-line defraging was, so their only 'approved' method was  I >>backup/restore.  I have to believe that DFO was only developed because uH >>the likes of Raxco proved otherwise and enabled 24/7 shops to stay up  >>and running... >>     >> >dJ >But if you recall also, DEC changed the ODS and the XQP to get a movefile6 >primitive to make a non-dangerous defragger happen... >n >  f >hF Yes, at which point the existing defrag products started to use those F primatives.   DEC added them because they could, and Raxco could no,t G but obviously DEC saw the need to compete in the defrag market because s8 backup/restore was not a viable option for 24x7 shops...   Barryn   --    > Barry Treahy, Jr                       E-mail: Treahy@MMaz.com> Midwest Microwave                          Phone: 480/314-1320> Vice President & CIO                         FAX: 480/661-7028                        h   ------------------------------  % Date: Thu, 20 Nov 2003 22:09:01 -0700-+ From: "Barry Treahy, Jr." <Treahy@MMaz.com>e) Subject: Re: VMS Defragmentation Softwarey' Message-ID: <3FBD9DED.5090506@MMaz.com>n   jlsue wrote:  3 >On Thu, 13 Nov 2003 12:05:51 -0500, Nathan Hartley + ><nathan@removethis.ilothlorien.com> wrote:l >u >  f >sC >>Anyone have an opinion on VMS open-file defragmentation software?  >>K >>I am currently looking at Raxco's Performance Suite and Executive's Disk t	 >>Keeper.I >> >>     >> >eG >Why do you think you need disk defragmentation?  What problems are youh- >seeing that point to fragmentation problems?i >tE >What kind of storage configuration do you have?  Note that RAID-typesL >configurations actually "defragment" your files at the hardware end anyway. >  e > C I've never heard of this!  How can you make such a claim when most yG external RAID arrays are file system neutral?  If the RAID array knows h6 nothing of the file system, it cannot defrag a file...   Barryj   -- l  > Barry Treahy, Jr                       E-mail: Treahy@MMaz.com> Midwest Microwave                          Phone: 480/314-1320> Vice President & CIO                         FAX: 480/661-7028                        l   ------------------------------  % Date: Fri, 21 Nov 2003 07:50:26 +0100.* From: Paul Sture <nospam@sture.homeip.net>3 Subject: Re: [OT] Mac OS X: system folder not founde0 Message-ID: <3FBDC3C2.6FD678DC@sture.homeip.net>   Roland Barmettler wrote: >  > Hello Paul > 8 > > Hi Roland, can you tell me where you got it, please?6 > > I haven't managed to locate a copy here in CH yet. > 2 > Believe it or not, in Mediamarkt (Dietikon) :-))H > I got it on the 25th of October where most Mac shops were already sold: > out, but Mediamarkt had still a whole shelf of Panthers. >  > Gressli, Roland >    Amazed! I'll try them.   Merci vielmal.  s > -- > 3rd Law of Computing:t! >         Anything that can go wrh" > Segmentation fault (core dumped)   -- s   -- w
 Paul Sture   ------------------------------   End of INFO-VAX 2003.645 ************************