1 INFO-VAX	Fri, 01 Aug 2003	Volume 2003 : Issue 421       Contents:% Re: .com file to zero out disk errors % Re: .com file to zero out disk errors % AUTO_DLIGHT_SAV and Multinet settings  Corrupted Indexed RMS file Re: Corrupted Indexed RMS file# Disk Special 36GB 15KRPM For USD379  Re: DSSI problem Re: duplicating system disks Re: duplicating system disks FA: VMS 5.2 Books  Free VMS course for new users!" Re: Free VMS course for new users! Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news...  Re: Itanium performance news... + Re: Moving cluster whilst maintaing quorum. # Re: Mozilla... the new NULL process  NCP question Re: NCP question" Re: Packed decimal arithmetic in C Re: Quorom Disk/Node Overhead ! Re: Running Java program detached ! Re: Running Java program detached 4 Re: SFF commands for TCPIP$SYSTEM:TCPIP$SMTP_SFF.EXE4 Re: SFF commands for TCPIP$SYSTEM:TCPIP$SMTP_SFF.EXE4 Re: SFF commands for TCPIP$SYSTEM:TCPIP$SMTP_SFF.EXE4 Re: SFF commands for TCPIP$SYSTEM:TCPIP$SMTP_SFF.EXE4 Re: SFF commands for TCPIP$SYSTEM:TCPIP$SMTP_SFF.EXE4 Re: SFF commands for TCPIP$SYSTEM:TCPIP$SMTP_SFF.EXEG SPARC/Solaris Question (was Re: Yamahill? Prescott? Wot's it all mean?) K Re: SPARC/Solaris Question (was Re: Yamahill? Prescott? Wot's it all mean?) K Re: SPARC/Solaris Question (was Re: Yamahill? Prescott? Wot's it all mean?)  Re: Sun and SCO  Re: Sun and SCO  Re: Sun and SCO 3 Re: Sun tries to woo AlphaServer users away from HP 3 Re: Sun tries to woo AlphaServer users away from HP  TCP/IP feature request Re: Web Apps for VMS Re: Web Apps for VMS* Re: Yamahill? Prescott? Wot's it all mean?& Re: [Change topic -> OT] Pence/Pennies& Re: [Change topic -> OT] Pence/Pennies& Re: [Change topic -> OT] Pence/Pennies  F ----------------------------------------------------------------------  % Date: Thu, 31 Jul 2003 13:03:17 -0500 1 From: Bob Blunt <robert.blunt@hp.donotspamme.com> . Subject: Re: .com file to zero out disk errors, Message-ID: <3f295f94$1@usenet01.boi.hp.com>   Chuck Aaron wrote: > anyone have one. >   I Not in DCL, but a solution does exist in Macro for VAX and Alpha flavors  I of OpenVMS.  If you're of a non-VMS mindset, a '.com' might equate to an   "executable image" on VMS.  I The tool resets counters in the O/S in inner access modes, so a solution  3 that uses DCL only won't be available until V7.3-2.   K The Macro solution is, IIRC, called ZDEC.  Use at your own risk, YMMV, etc.    bob    ------------------------------  % Date: Thu, 31 Jul 2003 15:39:26 -0700 + From: "Barry Treahy, Jr." <Treahy@MMaz.com> . Subject: Re: .com file to zero out disk errors' Message-ID: <3F299A9E.6070705@MMaz.com>    Bob Blunt wrote:   > Chuck Aaron wrote: >  >> anyone have one.  >> > C > Not in DCL, but a solution does exist in Macro for VAX and Alpha  F > flavors of OpenVMS.  If you're of a non-VMS mindset, a '.com' might ) > equate to an "executable image" on VMS.  > B > The tool resets counters in the O/S in inner access modes, so a ? > solution that uses DCL only won't be available until V7.3-2.    G Eigen, a small company from NY I believe, has an inexpensive tools set  = for VMS which includes among them, CLRERR for such a thing...    Barry    --    > Barry Treahy, Jr                       E-mail: Treahy@MMaz.com> Midwest Microwave                          Phone: 480/314-1320> Vice President & CIO                         FAX: 480/661-7028                            ------------------------------    Date: 31 Jul 2003 15:03:32 -0700+ From: spamdump@mccready.com (Gary McCready) . Subject: AUTO_DLIGHT_SAV and Multinet settings= Message-ID: <ffd79a6c.0307311403.1b63921a@posting.google.com>   C I'm running Multinet 4.4, VMS 7.3-1 (and 7.3), and I've defined the D logical NET$DISABLE_DTSS to 1 to turn DTSS off. If I am running XNTPF as a client, how should I set the VMS sysgen parameter AUTO_DLIGHT_SAVE so that daylight savings time will change automatically as it should?   D I've already used both net$configure and utc$time_setup to make sure< the right timezone is set at the VMS level, and the multinetC configure/interface and multinet set/timezone commands to make sure B the timezone is set at the multinet level.  And of course, XNTP is- enabled. Are there any other settings needed?   E And what should be done prior to VMS 7.3, where that sysgen parameter  does not exist?    Thanks in advance,
 Gary McCready    ------------------------------    Date: 31 Jul 2003 12:49:08 -0700$ From: erikbenada@yahoo.ca (ericisko)# Subject: Corrupted Indexed RMS file = Message-ID: <9aa633ca.0307311149.5b121bcc@posting.google.com>    Hi,   F recently we had problem with application hanging. We traced it down toC 1 indexed file. We ran analyze/rms on it. This is output (I deleted  page headers):     FILE HEADER   4         File Spec: USRDSK_1:[B1_BILLING]SERHIS.DAT;1         File ID: (993,45,0)          Owner UIC: [1,4]C         Protection:  System: RWED, Owner: RWED, Group: RWED, World:  RWED0         Creation Date:   25-JUL-2002 04:54:07.46?         Revision Date:   23-JUL-2003 23:50:51.05, Number: 30188 '         Expiration Date: none specified 0         Backup Date:     26-JUL-2003 13:45:23.23!         Contiguity Options:  none !         Performance Options: none !         Reliability Options: none !         Journaling Enabled:  none    RMS FILE ATTRIBUTES   "         File Organization: indexed         Record Format: fixed+         Record Attributes:  carriage-return           Maximum Record Size: 103         Longest Record: 103 <         Blocks Allocated: 2682696, Default Extend Size: 5000         Bucket Size: 15 !         File Monitoring: disabled          Global Buffer Count: 0   FIXED PROLOG  6         Number of Areas: 1, VBN of First Descriptor: 2         Prolog Version: 3   + AREA DESCRIPTOR #0 (VBN 2, offset %X'0000')            Bucket Size: 15          Reclaimed Bucket VBN: 0 F         Current Extent Start: 2677671, Blocks: 5015, Used: 2805, Next: 2680476 %         Default Extend Quantity: 5000 !         Total Allocation: 2682623   * KEY DESCRIPTOR #0 (VBN 1, offset %X'0000')  :         Index Area: 0, Level 1 Index Area: 0, Data Area: 0         Root Level: 3 3         Index Bucket Size: 15, Data Bucket Size: 15          Root VBN: 750250         Key Flags:'                 (0)  KEY$V_DUPKEYS    0 '                 (3)  KEY$V_IDX_COMPR  1 '                 (4)  KEY$V_INITIDX    0 '                 (6)  KEY$V_KEY_COMPR  1 '                 (7)  KEY$V_REC_COMPR  1          Key Segments: 1          Key Size: 26         Minimum Record Size: 26 ;         Index Fill Quantity: 7680, Data Fill Quantity: 7680 "         Segment Positions:       0"         Segment Sizes:          26         Data Type: string %         Name: "service id-date-order"           First Data Bucket VBN: 3? ***  VBN 1848100:  The bucket chains for key #0 contain a loop. 5 Unrecoverable error encountered in structure of file.      The analysis uncovered 1 error.      ANAL/RMS BIL$FILES:SERHIS.DAT     E We checked also disk for errors (analyze/disk/read) and error log for D some errors that could be related to this file corruption. We didn't find any problems.  8 Has anyone had similar problems? I found ECO Kit for RMSD (VMS721_RMS-V0500) that mentions RMS loop during attempted duplicateF key insert for compressed keys - this might be related. But we applied that patch long time ago. F Could this corruption be caused by application crash? This seems to beE quite serious corruption to the file to be caused by mere application A crash (fortunately it's test machine - but we would like to avoid # similar problems in production...).   A Our system is OpenVMS version 7.2-1, machine is Alpha Server 2100  5/300.0 Please let me know if you need more information.   Thanks,       Erik    ------------------------------  % Date: Thu, 31 Jul 2003 13:02:53 -0700 * From: "Jack Peacock" <peacock@simconv.com>' Subject: Re: Corrupted Indexed RMS file 2 Message-ID: <y4Sdne0JOstz6LSiXTWJgA@mpowercom.net>  1 "ericisko" <erikbenada@yahoo.ca> wrote in message 7 news:9aa633ca.0307311149.5b121bcc@posting.google.com...  > Hi,  > H > recently we had problem with application hanging. We traced it down toE > 1 indexed file. We ran analyze/rms on it. This is output (I deleted  > page headers): > J You might run analyze for a large number of files and see what you get.  IL had an AlphaServer 2100 internal SCSI bus with an intermittent write failureL which produced the same results.  The hardware failure did not get logged soK the customer was unaware of it until every SCSI device on the bus (disk and H tape) had errors on it.  Do not discount a hardware failure just because  there isn't anything in the log.  E Do you verify a tape backup?  That was how I determined it was a SCSI K bus/controller failure, since every disk (even known good ones installed as H replacements) eventually wound up with corrupt files when restoring from tape.     Jack Peacock    ------------------------------  % Date: Thu, 31 Jul 2003 18:33:06 -0400 ) From: "David @ Island" <david-@-hpaq.net> , Subject: Disk Special 36GB 15KRPM For USD379/ Message-ID: <vij67qicebtld5@news.supernews.com>   ! Qty in stock - Brand NEW HP Disks  68Pin Low Profile U160 Disk    $379 !!!       -- Island Computers US Corp 2700 Gregory St, Ste 180 Savannah GA 31404  Tel; (00) 1 912 447 6622 Fax; (00) 1 912 201 0402 dbturner@hpaq.net  www.hpaq.net  - Please read our warranty terms and conditions  before placing your order ! http://www.hpaq.net/warranty.html ! Support: http://www.hpaq.net:8004 $ Support  FTP: ftp://www.islandco.com (User name & Password required)    ------------------------------  % Date: Fri, 01 Aug 2003 08:51:53 +0800 , From: Paul Repacholi <prep@prep.synonet.com> Subject: Re: DSSI problem - Message-ID: <87el06nqs6.fsf@prep.synonet.com>   7 You didn't knock either of the terminators off did you?    --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Fri, 01 Aug 2003 08:14:38 +0800 , From: Paul Repacholi <prep@prep.synonet.com>% Subject: Re: duplicating system disks - Message-ID: <87n0eunsi9.fsf@prep.synonet.com>   ( jlsue <jefflsxxxz@sbcglobal.net> writes:  B > If you move all of the critical, modifiable files (sysuaf, queueF > database files, rightslist, etc) off the system disk, then a on-line) > image backup will probably work well...   C > Of course, you'll still need to copy those files in some reliable @ > way.  Convert/share to a "backup" directory, which then gets aD > backup to another saveset can work.  It just adds an extra step on& > the restore before you can use them.  E Why not convert them back to the *original* place on the system disk? G You are then able to boot instantly with everything on the system disk, ? and can then convert them out to some other `common file' disk.    --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Fri, 01 Aug 2003 05:55:08 GMT % From: "Mike Naime" <mnaime@kc.rr.com> % Subject: Re: duplicating system disks ; Message-ID: <0hnWa.50162$7O4.1170472@twister.rdc-kc.rr.com>   1 jlsue <jefflsxxxz@sbcglobal.net> wrote in message 2 news:86qfiv4hhlfhmt98t4rg54qmgvaei4ju23@4ax.com...J > On Tue, 29 Jul 2003 01:53:04 GMT, "Mike Naime" <mnaime@kc.rr.com> wrote: >  > >  > K > If you move all of the critical, modifiable files (sysuaf, queue database J > files, rightslist, etc) off the system disk, then a on-line image backup > will probably work well...  K But those are the files that I really do not care about!  When a new client G build is done, all of the above get modified/replaced.  Each client has / different printers.  One of the things we do is I START/MANAGER/QUE/NEW/CLUSTER to wipe the old que manager and start a new K database.   The logons are all client specific and created for the specific L client during the new build phase that lasts a couple of months (Sometimes a year) prior to go-live.   J I will agree that the Standalone backup is the safest way to make a backup; copy of your OS disk.  But it is not the only way to do it.   L Maybe I am misunderstanding the original intent of the question that started this thread ? > > There must be many folks who have multiple system disks for B > > redundancy, which should be identical except for NODE-SPECIFICE > > stuff.  Rather than upgrading, installing layered products on etc D > > ALL disks, it would make more sense to do it just on one "master@ > > disk" then make copies of this for other system disks (quite2 > > comfortably if all system disks are shadowed).  J I thought that it was asking about Duplicating disks for new clusters.  IfL that was not what they where asking about then I obviously misunderstood the	 question.   G We have OS disks that are mirror on the HSG controllers for redundancy. I Duplicating OS disks is what I am doing, and have been doing for the last G three years to make 70+ OS disks.  The only problems that I have had in I doing this have been proven to be caused by new hardware failures.  I had L one new DS20E motherboard that trashed any OS disk that you tried to boot itJ on.  HP could not explain it, they just replaced the mother board, and the problem went away.   > H > Of course, you'll still need to copy those files in some reliable way.L > Convert/share to a "backup" directory, which then gets a backup to anotherI > saveset can work.  It just adds an extra step on the restore before you  can  > use them.  >   
 Mike Naime   ------------------------------  % Date: Thu, 31 Jul 2003 21:27:29 -0400 0 From: "Homer Simpson" <hsimpson@burnsenergy.com> Subject: FA: VMS 5.2 Books: Message-ID: <DejWa.10473$f%2.7601@fe05.atl2.webusenet.com>  
 User's Manual K http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=3524&item=2744461581    System Manager's Manual K http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=3524&item=2744461865    DCL DictionaryK http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=3524&item=2744463433    ------------------------------    Date: 31 Jul 2003 20:20:17 -0700( From: bob@instantwhip.com (Bob Ceculski)' Subject: Free VMS course for new users! = Message-ID: <d7791aa1.0307311920.31fe89a5@posting.google.com>   4 for those new to vms, here is a link to an excellent introductory vms course ...   , http://h71000.www7.hp.com/wbt/pc/welcome.htm   ------------------------------  * Date: Fri, 1 Aug 2003 00:33:14 -0500 (CDT)0 From: Kevin Monceaux <OwnedByDogs@grandecom.net>+ Subject: Re: Free VMS course for new users! H Message-ID: <Pine.LNX.4.50.0308010028370.6692-100000@Linux.monceaux.com>  J Humm, an introductory course on VMS that requires Windows 95/98/NT to viewI the course.  It would seem that a web based course on VMS would be hosted @ on a VMS server and be accessible from any os with VMS being the preferable os.  ( On Thu, 31 Jul 2003, Bob Ceculski wrote:  6 > for those new to vms, here is a link to an excellent > introductory vms course ...  > . > http://h71000.www7.hp.com/wbt/pc/welcome.htm >  >    ------------------------------  % Date: Thu, 31 Jul 2003 13:47:33 -0400 * From: "Bill Todd" <billtodd@metrocast.net>( Subject: Re: Itanium performance news...2 Message-ID: <fB6cnRQBTdXey7SiXTWJhQ@metrocast.net>  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:M88bC9W1s+$t@eisner.encompasserve.org... @ > In article <JqecnVBm8Nhr27SiXTWJhw@metrocast.net>, "Bill Todd"  <billtodd@metrocast.net> writes: > > H > > "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> wrote in message/ > > news:Dk8Wa.1177$cE4.319@news.cpqcorp.net... E > >> Typical.  The big three JS, JFM, BT all weigh in with criticism.  > > H > > Typical - and fully warranted.  As long as people like you and Keith take it I > > upon yourselves to introduce marketeering into c.o.v., people like us  willG > > take steps to bring back a modicum of reality to your more fanciful J > > excursions.  If you don't like that, I suggest that you return to yourK > > engineering day jobs and leave the spinning to those who are paid to do  it.  > >  > D > It is their version of reality versus the JBJ troika.  One must do7 > something to counter the affable Bill Todd's spinola:  >  > L http://groups.google.com/groups?selm=hNKcnXrKKfx0YXKjXTWJig%40metrocast.net& oe=UTF-8&output=gplain > , > From: "Bill Todd" <billtodd@metrocast.net> > Newsgroups: comp.os.vms & > Subject: Re: Sun to be the next DEC!' > Date: Wed, 18 Jun 2003 01:47:16 -0400  > L > Sun should do fine as soon as the market starts to recover:  its customersF > seem happy to pay a reasonable price for the hardware, software, and support D > that Sun gives them, and at least many will continue to prefer the	 perceived < > solidity of Solaris to the more adventurous move to Linux. >  > ---  > " > The server market has recovered.  L In your dreams.  But I agree that it has *started* to recover - so now we'll1 see what Sun's results look like in this quarter.    >  Sun isn't doing fine,  G Sun *wasn't* doing fine last quarter.  But if you look at the full-year C results for 2002 over 2001, Sun grew significantly while HP shrank: 5 different companies have problems at different times.     and Sun won't > be doing fine.  I Another fearless Rob Young prognostication.  Since it'll take you a while L yet to live down your Wildfire predictions, I think I prefer to wait and see what actually happens.   ...     with Itanium thatE > day will arrive too.  Look back and show such opinion was warranted $ > (i.e. Itanium's eventual success).  L *What* success, Rob?  In the real world success is measured by sales, and so! far Itanic ain't got much of any.       After all, there are some veryA > fine future  processes coming and Itanium 2 is doing quite well  > performance-wise  I As I've noted elsewhere, Itanic2 is at best POWER4+'s equal - and now has K nowhere to go but down in performance and power consumption relative to its K POWERx competition until Tanglewood appears in 2006-7.  It has no scheduled H core enhancements before Tanglewood (just process shrinks - on about theJ same schedule as POWERx - plus cache-size increases and the dual-core chipI in 2005), whereas POWER5 gets SMT, on-chip offload engines, and God knows E what else next year (and then whatever POWER6 has in store for 2006).   L Add in Opteron's (and Xeon's) far superior price/performance on the low end,/ and Itanic has uphill battles across the board.    - bill   ------------------------------  % Date: Thu, 31 Jul 2003 13:51:35 -0400 ) From: JF Mezei <jfmezei.nospam@istop.com> ( Subject: Re: Itanium performance news...) Message-ID: <3F295725.4AEC01D5@istop.com>    Fred Kleinsorge wrote: > B > Typical.  The big three JS, JFM, BT all weigh in with criticism.  I I asked a simple question: if Alpha were truly tested to its maximum, how N would it compare to the "record breaking" performance you pointed to for IA64. I have received no answer.   ------------------------------    Date: 31 Jul 2003 11:05:03 -07001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) ( Subject: Re: Itanium performance news...= Message-ID: <cf15391e.0307311005.35944e9c@posting.google.com>   Y JF Mezei <jfmezei.nospam@istop.com> wrote in message news:<3F288452.1543611@istop.com>... L > Can any objective person comment on whether Alpha EV7 could have beat thatL > record had Digital still been a competitor of HP with a vested interest in= > marketing and making Alpha compete head on against intel ?%   C People could argue forever about whether DEC/Compaq didn't have the E WILL to compete head-to-head with Intel, or didn't have the MONEY (as F they claimed).  In either case, they DIDN'T spend the money to keep up? in semiconductor process technology, and Intel has used leading ? process technology to more than make up for any shortcomings in 
 architecture.   " A recent graph by Paul DeMone (seeD http://www.realworldtech.com/page.cfm?ArticleID=RWT060503232439&p=3)9 of RISC vs. Intel speed with a parallel graph showing the B corresponding semiconductor process generations really puts thingsB into perspective.  Alpha was able to retain about a 2X performanceE lead over Intel in general until the lead started to fade about 1999, E and the chart indicates the performance lead has now evaporated.  One F big factor apparent from Paul's illustration is that Intel is TWO FULL? PROCESS GENERATIONS ahead of Alpha now (fully 1/2 feature size, F resulting in 4X density, and smaller capacitance and thus faster clock$ speeds and lower power consumption).  B In addition to process improvements, the graphs show that a little= Alpha technology in the Pentium Pro timeframe and more clever @ engineering in the Pentium 4 timeframe also helped.  By the time> mainstream VMS customers transition to Itanium and need higherE performance, the Alpha engineers' (now-legal) contributions will kick  in.    ------------------------------  % Date: Thu, 31 Jul 2003 14:30:24 -0400 ) From: JF Mezei <jfmezei.nospam@istop.com> ( Subject: Re: Itanium performance news...) Message-ID: <3F29603B.F5C82D6E@istop.com>    Bill Todd wrote:K > To be fair, you're not as much of a pest in this regard as Keith has been L > recently - but the qualitative resemblance was sufficient to lump you both > together.   F I have to disagree. Mr Parris (undoubtedly the great great great greatJ grandfather of Tom Parris), has posted documentrs that are closer to press# releases than to personal opinions.   K Of course, some of us feel the need to debate the propaganda released by HP I because it doesn't represent the reality when it comes to discussing VMS.    ------------------------------  # Date: Thu, 31 Jul 2003 18:51:45 GMT # From: "John Smith" <a@nonymous.com> ( Subject: Re: Itanium performance news...H Message-ID: <5zdWa.10757$4UE.10190@news01.bloor.is.net.cable.rogers.com>  6 "JF Mezei" <jfmezei.nospam@istop.com> wrote in message# news:3F295725.4AEC01D5@istop.com...  > Fred Kleinsorge wrote: > > D > > Typical.  The big three JS, JFM, BT all weigh in with criticism. > > > I asked a simple question: if Alpha were truly tested to its maximum, howF > would it compare to the "record breaking" performance you pointed to	 for IA64.  > I have received no answer.   It's a fair question AFAIK.   > Since there are no customer purchaseable IG64/VMS systems withE production quality VMS on them, it would seem to me that for the next F 2 years HP has to show some decent performance numbers on Alpha/VMS ifD they expect to continue to sell Alphas and VMS until IA64/VMS is GA.   ------------------------------    Date: 31 Jul 2003 14:04:08 -0500+ From: young_r@encompasserve.org (Rob Young) ( Subject: Re: Itanium performance news...3 Message-ID: <bybJ6pVnhmln@eisner.encompasserve.org>   _ In article <fB6cnRQBTdXey7SiXTWJhQ@metrocast.net>, "Bill Todd" <billtodd@metrocast.net> writes:  > : > "Rob Young" <young_r@encompasserve.org> wrote in message/ > news:M88bC9W1s+$t@eisner.encompasserve.org...    >># >> The server market has recovered.  > N > In your dreams.  But I agree that it has *started* to recover - so now we'll3 > see what Sun's results look like in this quarter.  >   B 	Dreams?  "Started" to recover?  Ahh... vintage Andrew.  "Started". 	means anything and nothing.  Talk about spin.   http://www.marketwatch.com/news/yhoo/story.asp?source=blq/yhoo&siteid=yhoo&dist=yhoo&guid=%7BCAD48FC8%2D1EBF%2D4496%2DAC0F%2D4FB0FC869E24%7D  ! H-P retains server shipment lead    By Rex Crum, CBS.MarketWatch.com& Last Update: 6:18 PM ET July 28, 2003    J H-P (HPQ: news, chart, profile) remained No. 1 in worldwide shipments withL 376,100 units and a 29.5 percent market share, up almost 16 percent from itsM year-ago totals. The Gartner figures also put H-P in first place in shipments G in the Europe, Middle East and Africa, Asia Pacific, and Latin American  regions.  M Dell (DELL: news, chart, profile) claimed second-place with 261,600 units and O 20.5 percent of the worldwide market, a 33-percent rise over the second quarter I of 2002. Dell took the top spot in server shipment to the U.S. and Japan.   O IBM (IBM: news, chart, profile) came in third with almost 200,000 shipments and M 15.7 percent of the market, up 33.5 percent from a year ago. Sun Microsystems N (SUNW: news, chart, profile) kept its fourth-place position with 64,000 serverK shipments and a 5-percent stake. However, Sun's shipments fell 19.4 percent  from a year ago.    B 	HP sells 40000+ more servers than last year, Dell sells 33% more,; 	IBM 33.5% more.  Sun - in this same timeframe - manages to ) 	sell 15000+ less servers than last year.   G 	Server sales are up tremendously for large OEMS - except Sun - in the  C 	last year and yet... and yet Sun has shrunk badly.  Spin you must!   @ > so now we'll see what Sun's results look like in this quarter.  G 	That's funny.  It's holiday season.  Hardly anyone buys in the summer  B 	time.  Maybe Sun sells 10 more servers than last year, that would 	be a triumph. 	  >>  Sun isn't doing fine,  > I > Sun *wasn't* doing fine last quarter.  But if you look at the full-year E > results for 2002 over 2001, Sun grew significantly while HP shrank: 7 > different companies have problems at different times.  >   < 	Nonsense.  Sun had a disasterous 2002 (on par with the 2003( 	disaster?  Don't know - have to check).  A 	Sun is shrinking itself out of existence.  Sun did $18.2 billion ; 	in sales year ending June 2001.  Sun did ~$12.5 billion in = 	sales year ending June 2002.  An amazing growth rate of -32% 	 	in revs.   ? 	Sun is doing very poorly as has been rehashed here in the last ? 	2 weeks.  Declining revenues for 9 straight quarters (and yes, D 	we now know what view to take when comparing quarters).  Sun hasn'tB 	grown revenues at all in the last 9 quarters.  Are you looking atA 	the first quarter 2001 and comparing that to first quarter 2000?   B 	Or don't want to talk about how much Sun shrunk in the last year?   > Since it'll take you a whileN > yet to live down your Wildfire predictions, I think I prefer to wait and see > what actually happens.  @ 	Dig deep.  My foul predictions are long buried.  Your's have a   	certain fresh scent about them.   				Rob    ------------------------------  # Date: Thu, 31 Jul 2003 19:49:56 GMT 9 From: "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> ( Subject: Re: Itanium performance news...1 Message-ID: <EpeWa.1222$gY4.106@news.cpqcorp.net>   5 "Bill Todd" <billtodd@metrocast.net> wrote in message , news:wXadnZIs6PXtz7SiXTWJhg@metrocast.net...   > J > I'm paid to create new distributed file system architectures, Fred.  AndI > when I can't find anyone interested in developing an interesting one, I  work > on my own. >   H Must have been a slow time for a long time.  We just can't get enough of you.   > I > Since you appear to have been drooling on your keyboard while composing  the F > balance of your post, that's about all that was worth responding to. >    Ayup.  Guess so.   ------------------------------  # Date: Thu, 31 Jul 2003 20:28:00 GMT 9 From: "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> ( Subject: Re: Itanium performance news...1 Message-ID: <kZeWa.1229$kU4.820@news.cpqcorp.net>   L Yup.  Not - "Great news, maybe VMS on Itanium is going to be great" - just aD another rehash of "Gee, if they'd just kept Alpha - gosh it would be wonderful".     6 "JF Mezei" <jfmezei.nospam@istop.com> wrote in message# news:3F295725.4AEC01D5@istop.com...  > Fred Kleinsorge wrote: > > D > > Typical.  The big three JS, JFM, BT all weigh in with criticism. > K > I asked a simple question: if Alpha were truly tested to its maximum, how J > would it compare to the "record breaking" performance you pointed to for IA64.  > I have received no answer.   ------------------------------  # Date: Thu, 31 Jul 2003 21:07:29 GMT # From: "John Smith" <a@nonymous.com> ( Subject: Re: Itanium performance news...H Message-ID: <lyfWa.33322$hOa.18893@news02.bloor.is.net.cable.rogers.com>  > A proper answer from Fred, would have been along the lines of:  = '....Based on the most recent benchmarks which were done with @ configuration X, and reasonably extrapolating to the most recentD processors, latency, etc... we might expect that the result would be@ approximately Z, but confirmation of that will have to await theC published figures. I'll check around and  see when that information F may be expect to be released and get back to you.' or perhaps a simple 'I don't know.'     F Fred, the bottom line is that 99 and 44/100% of all the bitching (yourE characterization) - criticism to most others - would simply vanish if D HP would advertise VMS and market it effectively to new and existing
 customers.  C I know that you and your colleagues each put forth an honest effort E each day towards making VMS better than it was the day before, and we < all appreciate it. Wouldn't it be nice if the fruits of yourB collective labors were shared amongst a larger and rapidly growingF customer base? What have you seen the HP advertising department do forC VMS lately? Don't tell me that you don't sometimes wonder just what < the hell is going on with advertising and marketing for VMS.  F Let's say that VMS were EOL'd tomorrow and that in 5 years all supportF were to cease. You have marketable, transferable skills and would mostA likely get a decent job in fairly short order. Your personal risk E should VMS be EOL'd is pretty low. Now consider the VMS customer base ? should VMS be EOL'd. They'd be out many hundreds of millions of 8 dollars, or more, in porting costs and possible business. interruptions. That is a somewhat bigger risk.  ? You might counter with 'VMS isn't going to be EOL'd...there's a C published roadmap, etc...'. True enough. But is that enough to stop > the erosion of the customer base or the reduction in installedF systems? The sales cycle is long and without effective advertising andF marketing, HP will not reach enough potential new customers or replaceC defecting customers fast enough since there isn't a large number of E prospects in the sales pipeline. VMS, while excellent, will represent @ less and less of HP's revenues, and be further marginalized in a0 growing market once the recession is truly over.  @ Many excellent products have been roadkill simply because lesserF products garnered greater mindshare and more of the allocated customer dollars.      D "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> wrote in message+ news:kZeWa.1229$kU4.820@news.cpqcorp.net... > > Yup.  Not - "Great news, maybe VMS on Itanium is going to be great" - just a F > another rehash of "Gee, if they'd just kept Alpha - gosh it would be
 > wonderful".  >  > 8 > "JF Mezei" <jfmezei.nospam@istop.com> wrote in message% > news:3F295725.4AEC01D5@istop.com...  > > Fred Kleinsorge wrote: > > > F > > > Typical.  The big three JS, JFM, BT all weigh in with criticism. > > @ > > I asked a simple question: if Alpha were truly tested to its maximum, howE > > would it compare to the "record breaking" performance you pointed  to for > IA64.  > > I have received no answer. >  >    ------------------------------  # Date: Thu, 31 Jul 2003 21:14:03 GMT & From: Rick Jones <foo@bar.baz.invalid>( Subject: Re: Itanium performance news...2 Message-ID: <vEfWa.1240$C15.1043@news.cpqcorp.net>  " John Smith <a@nonymous.com> wrote:@ > A proper answer from Fred, would have been along the lines of:  ? > '....Based on the most recent benchmarks which were done with B > configuration X, and reasonably extrapolating to the most recentF > processors, latency, etc... we might expect that the result would beB > approximately Z, but confirmation of that will have to await theD > published figures. I'll check around and see when that informationA > may be expect to be released and get back to you.' or perhaps at > simple 'I don't know.'  ; Actually, if we are talking about TPC, public discussion of!C estimates/extrapolations/etc are prohibited.  I'm not sure what the B rules are for Oracle Apps or SAP SD, but would not be surprised to hear that they were the same..  
 rick jones -- fG oxymoron n, commuter in a gas-guzzling luxury SUV with an American flagcF these opinions are mine, all mine; HP might not want them anyway... :)A feel free to post, OR email to raj in cup.hp.com  but NOT BOTH...e   ------------------------------  % Date: Thu, 31 Jul 2003 17:20:21 -0400 ( From: David Froble <davef@tsoft-inc.com>( Subject: Re: Itanium performance news..., Message-ID: <3F298815.6050803@tsoft-inc.com>   JF Mezei wrote:e   > Bill Todd wrote: > K >>To be fair, you're not as much of a pest in this regard as Keith has been L >>recently - but the qualitative resemblance was sufficient to lump you both >>together.e >> > H > I have to disagree. Mr Parris (undoubtedly the great great great greatL > grandfather of Tom Parris), has posted documentrs that are closer to press% > releases than to personal opinions.m    K Gee, isn't that what many have asked for?  Some press releases?  A form of $Q marketing.  Are you then against marketing of VMS in this manner?  In any manner?s    M > Of course, some of us feel the need to debate the propaganda released by HPsK > because it doesn't represent the reality when it comes to discussing VMS.s  ^ If there is any untruth in the 'press releases', that would be good to expose.  Is there such?    P A suggestion.  Let Fred and Keith do their jobs, while avoiding propaganda.  If P you're still angry with Palmer/Curly/Carly, direct your anger in that direction.  N The reality you refer to is that for now we're stuck with IA-64.  If it hangs M around long enough, and actually works to some degree, it should satisfy the a needs of most.  * Is it 'industry standard'?  That's a joke!  > Is it more important to Intel than the desktop?  Another joke!  P Will it get buried by desktop quantity just like every other enterprise CPU?  I 5 know which side of that bet I won't waste even $1 on.e     Dave   -- s4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com6 T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486   ------------------------------  % Date: Thu, 31 Jul 2003 17:36:34 -0400i* From: "Bill Todd" <billtodd@metrocast.net>( Subject: Re: Itanium performance news...2 Message-ID: <c6GdnRLFxbttFrSiXTWJig@metrocast.net>  5 "David Froble" <davef@tsoft-inc.com> wrote in message-& news:3F298815.6050803@tsoft-inc.com... > JF Mezei wrote:9 >  > > Bill Todd wrote: > >,H > >>To be fair, you're not as much of a pest in this regard as Keith has beenI > >>recently - but the qualitative resemblance was sufficient to lump yout both
 > >>together.e > >> > > J > > I have to disagree. Mr Parris (undoubtedly the great great great greatH > > grandfather of Tom Parris), has posted documentrs that are closer to press ' > > releases than to personal opinions.   I And are therefore less appropriate in this venue.  While I often disagree L with Fred's analysis, at least it is most of the time (with the exception ofF the Itanic press release that started this thread) it's a personal one$ rather than pure regurgitated HP PR.   >a > L > Gee, isn't that what many have asked for?  Some press releases?  A form ofK > marketing.  Are you then against marketing of VMS in this manner?  In any& manner?   H I think you're confused.  The press release that started this thread hadA *nothing* to do with VMS:  it was just more of HP's "Isn't ItanicmL wonderful!" propaganda, and had almost as little relevance to c.o.v. as someH of Keith's "Gee:  Sun seems to be on the rocks..." and "Wow!  HP sure is doing well!" recent spam.   L I suspect that most people (including yourself) who advocate more active VMSK marketing want to see it *outside* the existing customer base, not confinedrG to the faithful here.  If such external VMS-specific marketing activityiK actually occurred, it would be entirely appropriate to call attention to itu> here ("Hey, look:  HP actually *is* starting to market VMS!").   ...t  1 > A suggestion.  Let Fred and Keith do their jobso  I I thought they were engineers, not marketeers.  Several of Keith's recentsB posts here have *nothing* to do with engineering (or with c.o.v.).   - bill   ------------------------------  % Date: Thu, 31 Jul 2003 17:45:05 -0400 * From: "Bill Todd" <billtodd@metrocast.net>( Subject: Re: Itanium performance news...2 Message-ID: <E4ednbjUS71yELSiXTWJkg@metrocast.net>  > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message7 news:cf15391e.0307311005.35944e9c@posting.google.com...t6 > JF Mezei <jfmezei.nospam@istop.com> wrote in message$ news:<3F288452.1543611@istop.com>...I > > Can any objective person comment on whether Alpha EV7 could have beat  thatK > > record had Digital still been a competitor of HP with a vested interest  in? > > marketing and making Alpha compete head on against intel ?%  >DE > People could argue forever about whether DEC/Compaq didn't have the4G > WILL to compete head-to-head with Intel, or didn't have the MONEY (as H > they claimed).  In either case, they DIDN'T spend the money to keep upA > in semiconductor process technology, and Intel has used leadingrA > process technology to more than make up for any shortcomings ino > architecture.   C A regrettably common misconception resulting from the acceptance of K cHumPaq's lies by the gullible:  one might have expected an engineer not to J parrot it back without at least a *bit* of thought, but I guess you're notB very critical of statements that match your employer's party line.  H Alpha hasn't been in the process-technology biz since the Hudson fab wasK sold to Intel:  Alpha's process technology *was* Intel's process technology G for a while thereafter, and is now the equally-estimable IBM's.  That's G likely why Alpha was able to move relatively quickly through its 250 nmxL generation to its 180 nm generation:  it only had to tweak the existing coreF to the new process, not pioneer the process technology itself as well.  K So whatever lead has evaporated has not gone away because of some intrinsic:L Intel process technology advantage that Compaq couldn't have matched even ifJ it had tried to (because Compaq didn't need to match it:  it just needed aJ partner like IBM who matched it):  it has evaporated solely because CompaqF started stealthily slowing down *Alpha* development about 4 years ago.   > $ > A recent graph by Paul DeMone (seeF > http://www.realworldtech.com/page.cfm?ArticleID=RWT060503232439&p=3); > of RISC vs. Intel speed with a parallel graph showing theQD > corresponding semiconductor process generations really puts thingsD > into perspective.  Alpha was able to retain about a 2X performanceG > lead over Intel in general until the lead started to fade about 1999,b  C What an *amazing* coincidence:  just when Curly took over and AlphaeI clock-rate increases suddenly ground to a previously-unscheduled halt forS over a year.  B > and the chart indicates the performance lead has now evaporated.  L Er, no:  it only shows such evaporation for *FP* performance - IA32 achievedF rough *integer* performance parity with Alpha all the way back in 1995I (though it sagged back a bit around 1997).  But had Alpha development notpG slowed down, Alpha could have moved back into the integer lead as well.p  H Of course, as Wildfire's (and now SuperDome's) lackadaisical performanceF demonstrated there's a lot more to system performance than SPECint andL SPECfp, and even with its older, slower processors Marvel systems still haveL substantial performance leads over of top-of-the-line Intel systems (whetherL 32- or 64-bit, with the possible exception of the apparently faltering SGI's& new Altix platforms) in several areas.     OneiH > big factor apparent from Paul's illustration is that Intel is TWO FULLA > PROCESS GENERATIONS ahead of Alpha now (fully 1/2 feature size,uH > resulting in 4X density, and smaller capacitance and thus faster clock& > speeds and lower power consumption).  G Someone who purports to be an engineer should at least be able to count  accurately to 2.  I a)  To be even *one* 'full' process generation ahead of Alpha, IA32 would G have had to have entered its current 130 nm process no later than AlphasF entered its current 180 nm process (in early 2001, according to Paul's chart).  It did not.  L b)  To be *two* 'full' process generations ahead IA32 would have had to haveG entered its 90 nm process before Alpha entered its 180 process in earlytK 2001.  IA32 has not yet shipped a 90 nm part today (more than 2 years aftera, that date), though it's reportedly about to.  L c)  And both the above comparisons apply to IA32 processors, not the ItanicsG about which JF asked his question.  Itanic2 only moved into the processiH generation beyond Alpha's at the end of *last month*:  before that, theyL were in the same process generation.  And if Compaq hadn't been dragging itsD feet since 1999, EV8 would have appeared in that 130 nm next processH generation already:  in late 1998 EV8 was scheduled to ship in December,H 2001, and while that was unquestionably an aggressive date it could haveE slipped over a year and a half and still have shipped by now (and, of9E course, would now be beating the daylights out of *all* competition).a  K (I suppose by 'full process generation' above you might have merely meant aaK 'full' 29.3% linear shrink rather than some intermediate shrink as recentlyeI occurred when SPARC shrank from a 150 nm process to a 130 nm process.  BywI that definition, Itanic and IA32 would be considered *one* 'full' processrL generation ahead of Alpha, and while IA32 would shortly be 'two full processL generations' ahead of Alpha Itanic isn't scheduled for its next shrink untilJ 2005.  But that's a fairly silly definition to use when you're talking, asC you were, about supposed advantages *over time* rather than about a I point-in-time performance comparison and what differences enter into it.)    > D > In addition to process improvements, the graphs show that a little? > Alpha technology in the Pentium Pro timeframe and more clevereB > engineering in the Pentium 4 timeframe also helped.  By the time@ > mainstream VMS customers transition to Itanium and need higherG > performance, the Alpha engineers' (now-legal) contributions will kick< > in.e  I The problems, of course, are  a) that many customers would have preferrediK not to have to migrate *at all* to achieve improved performance (and CompaqyK had assured them that they would not have to) and  b) that they'd have thatsL vastly-improved performance with EV8 *today* (not in 2006-7) had Compaq kept to its stated commitments.   - bill   ------------------------------  % Date: Thu, 31 Jul 2003 18:43:48 -0400r* From: "Bill Todd" <billtodd@metrocast.net>( Subject: Re: Itanium performance news...2 Message-ID: <kamdnZmlr_4vBrSiXTWJiQ@metrocast.net>  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:bybJ6pVnhmln@eisner.encompasserve.org...l@ > In article <fB6cnRQBTdXey7SiXTWJhQ@metrocast.net>, "Bill Todd"  <billtodd@metrocast.net> writes: > >m< > > "Rob Young" <young_r@encompasserve.org> wrote in message1 > > news:M88bC9W1s+$t@eisner.encompasserve.org...  >  > >>% > >> The server market has recovered.  > > J > > In your dreams.  But I agree that it has *started* to recover - so now we'llt5 > > see what Sun's results look like in this quarter.  > >r >tC > Dreams?  "Started" to recover?  Ahh... vintage Andrew.  "Started"-/ > means anything and nothing.  Talk about spin.5  K You really are a disgusting little turd, Rob - much like your friend in theeJ White House.  You just blithely trot out your spin with complete disregardJ for the facts, presumably because you assume that a significant portion of5 your audience is too lazy and/or too stupid to check.4   >  >kL http://www.marketwatch.com/news/yhoo/story.asp?source=blq/yhoo&siteid=yhoo&d@ ist=yhoo&guid=%7BCAD48FC8%2D1EBF%2D4496%2DAC0F%2D4FB0FC869E24%7D   ...c  > > Server sales are up tremendously for large OEMS - except Sun  G Worldwide server sales in 2003Q2 were up 17.6% over 2002Q2:  as I said, K that's the *start* of the recovery.  There's a hell of a long way to go yeta before it's complete.n  > In Q1, worldwide server sales were up only 10.4% over Q1, 2002K  http://www3.gartner.com/5_about/press_releases/pr28apr2003b.jsp ):  that'seE more like the *suggestion* of a recovery, a suggestion which just gote" confirmed by the new 17.6% figure.  G If you check earlier Gartner press releases, you'll find that full-yearn0 server growth from 2001 to 2002 was a mere 4.2%.   ...    > >>  Sun isn't doing fine,e > > K > > Sun *wasn't* doing fine last quarter.  But if you look at the full-yeartG > > results for 2002 over 2001, Sun grew significantly while HP shrank:t9 > > different companies have problems at different times.n > >  >e= > Nonsense.  Sun had a disasterous 2002 (on par with the 2003l) > disaster?  Don't know - have to check).,  H Of course you don't know - because you're wrong.  Unlike you, however, I$ knew what I was talking about above:  ? http://www3.gartner.com/5_about/press_releases/pr24jan2003a.jspu  J If you will examine the Worldwide Server Unit Shipment chart, you will seeG that (the combined) cHumPaq sold 4.6% fewer server units and held lowersL market share in 2002 than in 2001.  And that Sun sold 6.7% more server units> and held (slightly) greater market share in 2002 than in 2001.   - bill   ------------------------------  % Date: Thu, 31 Jul 2003 19:31:22 -0400d* From: JF Mezei <jfmezei.spamnot@istop.com>( Subject: Re: Itanium performance news...) Message-ID: <3F29A6B4.86B4A01C@istop.com>i   David Froble wrote:oL > Gee, isn't that what many have asked for?  Some press releases?  A form ofS > marketing.  Are you then against marketing of VMS in this manner?  In any manner?o  G These press releases/propaganda don't mention VMS. They mention anothera company's products (Itanium).a   ------------------------------  % Date: Thu, 31 Jul 2003 21:56:29 -0400 ( From: David Froble <davef@tsoft-inc.com>( Subject: Re: Itanium performance news...* Message-ID: <3F29C8CD.90500@tsoft-inc.com>   Bill Todd wrote:  7 > "David Froble" <davef@tsoft-inc.com> wrote in message2( > news:3F298815.6050803@tsoft-inc.com... >  >>JF Mezei wrote:. >> >> >>>Bill Todd wrote:& >>>o >>>2H >>>>To be fair, you're not as much of a pest in this regard as Keith has >>>> > been > I >>>>recently - but the qualitative resemblance was sufficient to lump youe >>>> > both > 
 >>>>together.2 >>>> >>>>I >>>I have to disagree. Mr Parris (undoubtedly the great great great great0G >>>grandfather of Tom Parris), has posted documentrs that are closer ton >>>e > presss > & >>>releases than to personal opinions. >>>g > K > And are therefore less appropriate in this venue.  While I often disagreesN > with Fred's analysis, at least it is most of the time (with the exception ofH > the Itanic press release that started this thread) it's a personal one& > rather than pure regurgitated HP PR. >  >  >>L >>Gee, isn't that what many have asked for?  Some press releases?  A form ofK >>marketing.  Are you then against marketing of VMS in this manner?  In anye >>	 > manner?/ > J > I think you're confused.  The press release that started this thread hadC > *nothing* to do with VMS:  it was just more of HP's "Isn't ItanictN > wonderful!" propaganda, and had almost as little relevance to c.o.v. as someJ > of Keith's "Gee:  Sun seems to be on the rocks..." and "Wow!  HP sure is > doing well!" recent spam./    = Yeah, wonderful.  When was it going to actually be available?e    N > I suspect that most people (including yourself) who advocate more active VMSM > marketing want to see it *outside* the existing customer base, not confinedc > to the faithful here.y     Agreed.a  2 > If such external VMS-specific marketing activityM > actually occurred, it would be entirely appropriate to call attention to iti@ > here ("Hey, look:  HP actually *is* starting to market VMS!"). > ^ Agreed.  It's just that JF's so off balance he'd fall if he wasn't already flat on the ground. >  > 1 >>A suggestion.  Let Fred and Keith do their jobsi >> > K > I thought they were engineers, not marketeers.  Several of Keith's recentyD > posts here have *nothing* to do with engineering (or with c.o.v.). >  > - bill    F Yeah, but refer again to above about JF, and it goes for Mr anon also.     Dave     --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com6 T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486   ------------------------------  % Date: Fri, 01 Aug 2003 10:06:31 +0800 , From: Paul Repacholi <prep@prep.synonet.com>( Subject: Re: Itanium performance news...- Message-ID: <8765linnbs.fsf@prep.synonet.com>d  ; "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> writes:n  B > Typical.  The big three JS, JFM, BT all weigh in with criticism.  > So when are you posting the EV7, VMS or RDB benchmark results?   -- s< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------    Date: 31 Jul 2003 23:02:53 -0500+ From: young_r@encompasserve.org (Rob Young) ( Subject: Re: Itanium performance news...3 Message-ID: <gzCra0GHiUHR@eisner.encompasserve.org>a  _ In article <kamdnZmlr_4vBrSiXTWJiQ@metrocast.net>, "Bill Todd" <billtodd@metrocast.net> writes:  > : > "Rob Young" <young_r@encompasserve.org> wrote in message   >>> >> Nonsense.  Sun had a disasterous 2002 (on par with the 2003* >> disaster?  Don't know - have to check). > J > Of course you don't know - because you're wrong.  Unlike you, however, I& > knew what I was talking about above: >   : 	Wrong about what?  That Sun had a disasterous 2002 versus7 	2001?  They went from a 18 billion dollar company witht= 	associated earnings and slid to a 12 billion dollar company.y  A > http://www3.gartner.com/5_about/press_releases/pr24jan2003a.jsp  > L > If you will examine the Worldwide Server Unit Shipment chart, you will seeI > that (the combined) cHumPaq sold 4.6% fewer server units and held lower N > market share in 2002 than in 2001.  And that Sun sold 6.7% more server units@ > and held (slightly) greater market share in 2002 than in 2001.  ; 	So?  They sold more kit.  Big deal.  That obviously wasn'tz< 	one of my supporting points claiming they had a disasterous( 	2002 compared to 2001.  This might make: 	things even more clearer (as if they weren't before).  In 	Sun's own words:   F http://www.sun.com/smi/Press/sunflash/2002-07/sunflash.20020718.1.html  O "For the full 2002 fiscal year, Sun reported revenues of $12.5 billion, down 32-J percent from record high revenues of the prior year. Net loss for the 2002G fiscal year was $255 million and the net loss per common share was $.08gL (excluding a $517 million restructuring charge, a $99 million loss on equityO investments, a $3 million charge for in-process research and development, and atK $246 million benefit for the related tax effects). Including these amounts, A GAAP net loss was $628 million and GAAP loss per share was $.19."i  ' 	They lost $255 million for the year.  -   	2001:  F http://www.sun.com/smi/Press/sunflash/2001-07/sunflash.20010719.1.html  L "For fiscal year 2001, Sun reported revenues of $18.25 billion, up from lastM year's revenues of $15.72 billion. Pro forma net income was $1.45 billion for H the fiscal year, compared with last year's pro forma net income of $1.86M billion. Pro forma earnings per share for the fiscal year was $0.42, comparedl8 with last year's pro forma earnings per share of $0.55."  & 	They made $1.45 billion for the year.  A 	Trotting out Gartner to show they sold more servers is admirable  	but misses the point entirely.   ? 	Here is a hard fact.  HP, IBM and Dell sold substantially mored@ 	servers, made quite a bit more money than Sun.  Sun barely made> 	a profit and sold quite a percentage less in kit - nearly 20%? 	less.  Sun is in big trouble.  The fact that Scott is avoidingh 	the very hard questions:h  < http://www.thestreet.com/_yahoo/tech/kcswanson/10102503.html  H "The cost of components didn't decline as fast as it did in the previousL quarter," said Sun chief Scott McNealy, explaining why gross margins fell in the June quarter.   O But he sought to evade a question about how Sun's margins are likely to fare intF the longer term, given the company's push into lower-margin hardware.     > 	Isn't surprising at all as he can't answer that without being 	pinned down.     > 	So sure Bill.. maybe when Sun's server sales increase and you; 	can point out the increase for us.  But it will be hard toc8 	hide the loss in revenue and profits something I'm sure. 	most of us will notice.  Wall Street is on to= 	Sun - check out their stock price.   There aren't any brightt? 	spots for Sun anymore (sliding market share, sliding revenues,t; 	zero profit, declining stock price - you get the picture).t   				Robo   ------------------------------  + Date: Thu, 31 Jul 2003 20:31:22 +0000 (UTC)i7 From: moroney@world.std.spaamtrap.com (Michael Moroney) 4 Subject: Re: Moving cluster whilst maintaing quorum.( Message-ID: <bgbuaq$nif$1@pcls4.std.com>  & issinoho@slayme.com (issinoho) writes:  F >OK, there's lots of good stuff in all these replies, however if I canF >draw the subject back to the original question... if you were gettingF >PAID to do this job and didn't want to stuff things up, what strategy >would you employ?  E Perhaps you should tell us (or ask your employer) whether they reallyTE need continuous coverage, can they accept some outage at 3:00AM, on aaD weekend or holiday?  Vs. the chance of some outage if "hot-swapping"H DSSI causes problems.  You are at risk 4 times - two disconnects and two# connects with part of the bus live.   E >And - sorry I didn't mention this originally - the two storage units E >have shadowed volume sets, except the quorum disk of course; if thatn >makes any difference.  G It _does_ make a difference, unless your database or whatever is "read eF only" at the time.  Remember to factor in the time for two full shadowH copies of all your mounted shadowsets, once when the first node is movedD and brought back into the cluster, once when the second is moved andJ readded.  The first copy will go over the inter-site link, if it's slower., What version, is the new minicopy available?  K Dismount any unnecessary shadowsets entirely before the first node is shut hJ down and don't remount them until everything is in place and OK.  This way& they won't have to undergo two copies. -- n -Mike    ------------------------------  % Date: Thu, 31 Jul 2003 16:24:15 -0400e5 From: "Zeni Schleter" <SchleterZb@y12notreal.doe.gov> , Subject: Re: Mozilla... the new NULL process( Message-ID: <bgbtro$9l3$1@sws1.ornl.gov>  H "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in message- news:878Ja$CiHx2a@eisner.encompasserve.org...e< > In article <00A230CF.38A10C27@SendSpamHere.ORG>,   VAXman- @SendSpamHere.ORG writes:a > > OK,r > > K > > Why does Mozilla eatup CPU cycles when it is sitting on a web page withCF > > nothing actively changing?  No Javascript, Java, etc.  As a test I created  > > a page with: > E >    Mozilla has to poll between VMS event flags and socket select(),tG >    because the authors of the VMS port didn't want to rewrite all ther >    socket select() stuff.w >"I >    COE work to expand the coverage of select() to non-socket interfacese* >    should at least partially solve this. >sI >    IIRC there is a logical name than can be used to control the pollingD >    speed.n  L I am way behind on this thread but last fall after installing CSWB v1.0 on aH production system,  I saw the CPU utilization elevated with no effectiveL process activity on two Alphaserver 4100 each with  2 processors .  On otherE 4100s (development) that had only one processor each, CSWB idled with I minimal activity.   I called HP support.  The final solution was droppingiJ the polling time down (which I think was mentioned in the thread already).F Since the pages don't change , the refreshing interval did not matter.  I We have some in house software that calculates idle time.  When we portedtE from Vax to Alphas the idle time was off by a factor of the number ofsL processors.  I had wondered at the time if this was related.  I had not seenK any reports about others experiencing the same kind of cpu utilization withnH CSWB when it is idle.  I wonder if this will disappear when we install aI newer release.  We do not have any dual processor development machines to 
 test this on..  I FWIW, this is old news.  Is there any cases of similar non-idle CSWB v1.0M9 behavior that has been corrected with the newer releases?s   -- Z.+   ------------------------------  % Date: Thu, 31 Jul 2003 14:14:22 -0700g1 From: "Wolf, Gerald J" <gerald.j.wolf@boeing.com>. Subject: NCP questionkR Message-ID: <3BFEACE361F5BF429DD1DA593E3A7C090178DBA1@xch-nw-28.nw.nos.boeing.com>  ) Currently the NCP database does not show:o   Designated routers
 Adjacent node & and the hardware address are all zeros  / How can I set/define these in the NCP database?>   Regards,   Gerald Wolf    ------------------------------  % Date: Thu, 31 Jul 2003 17:18:59 -0500 1 From: Bob Blunt <robert.blunt@hp.donotspamme.com>t Subject: Re: NCP questionf, Message-ID: <3f299b83$1@usenet01.boi.hp.com>   Wolf, Gerald J wrote:u+ > Currently the NCP database does not show:3 >  > Designated router- > Adjacent node ( > and the hardware address are all zeros > 1 > How can I set/define these in the NCP database?2 > 
 > Regards, > 
 > Gerald Wolfs  I Gerald, in what order are your networking stacks started on your system? SF   It looks like you're using Phase IV, but this can be checked easily:  - $ WRITE SYS$OUTPUT F$GETSYI("DECNET_VERSION")d  G If the response is "00040000" you have Phase IV.  DECnet Phase IV MUST tH be started before all other network protocols because it will reset the F MAC address on the Ethernet interface.  If another protocol that uses G the MAC address starts first, it can prevent DECnet from resetting the t> address and DECnet won't start.  From NCP check the line with:   MC NCP SHO LINE xyz-0 CHAR 	and MC NCP SHOW LINE xyz-0 STAT   H IF the line status shows anything but a state of "ON" post the results,  example:   NCP> show line sva-0 stat-  / Line Volatile Status as of 31-Jul-2003 17:07:180   Line = SVA-0   State                   = on  / You may see "on-synchronizing" or "on-starting"p  G If you have a Alpha running OpenVMS, you should ONLY have one line and BB one circuit defined in the database (and those should be the SAME F devices, ie SVA-0, EIA-0, ERA-0 or others), REGARDLESS if your DECnet G license is DVNETEND or DVNETRTG.  DECnet Phase IV won't support a full rG routing node on Alpha.  You might also check to verify that you have a tB valid license for DECnet.  Besides the DVNET%%% licenses, certain A NET-APP-SUP-nnn licenses also enable DECnet, there may be others.o  % So the things to check are, at least:y  # 1)  Did DECnet Phase IV start firstt* 2)  Do you have a valid, unexpired license4 3)  Only one line and one circuit defined in the D/B6 4)  The device types for line and circuit are the same0 5)  The state of the line (on, on-starting, etc)  I Those are the software things that I've found to be problems most often.  E   Once you get past those, you may have a hardware problem, but lets t eliminate S/W first...     bobt   ------------------------------  % Date: Fri, 01 Aug 2003 08:01:51 +0800m, From: Paul Repacholi <prep@prep.synonet.com>+ Subject: Re: Packed decimal arithmetic in C:- Message-ID: <87vftint3k.fsf@prep.synonet.com>   , JF Mezei <jfmezei.spamnot@istop.com> writes:  E > SANTA CLARA, Calif., July 23, 2003 -- Intel Corporation announced at? > breakthrough in computing with the introduction of compressed : > decimal arithmetic on its Itanium II architecture. OlderC > technologies such as packed decimal used by competitors have been"B > unable to cope with the need to handle ever increasing values by > busineses.  C Is this the wonderfull Sinking Point Processor we have been hearing % rumours about since the last century?n   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.m@                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  + Date: Thu, 31 Jul 2003 18:02:50 +0000 (UTC) 7 From: moroney@world.std.spaamtrap.com (Michael Moroney)i& Subject: Re: Quorom Disk/Node Overhead( Message-ID: <bgblka$5ni$2@pcls4.std.com>  * robert_kersey@bat.com (Rob Kersey) writes:  N >Does a quorom disk produce much more overhead than an individual quorum node?  O >What are the disadvantages of using a quorom disk as opposed to a quorom node?y  D A cluster transition involving a quorum disk takes 4*QDSKINTERVAL toG happen.  The default is now 3 seconds (I shortened it from 10 seconds agD few years back) so this means transition times of around 12 (or 40) = seconds.  Quorum I/Os on a busy disk often get timed out/failuJ for some other reason and this will cause unnecessary cluster transitions.  E A quorum node of something like an old Alphastation 2xx with its own eG system disk can be had for cheap.  Just make sure the sysgen parametersyG are set so it doesn't try to master lock trees or whatever if the otherb% cluster members are powerful systems.  -- u -Mikeb   ------------------------------  % Date: Thu, 31 Jul 2003 15:16:17 -0400l0 From: "Brian Tillman" <Tillman@sparkingwire.com>* Subject: Re: Running Java program detached$ Message-ID: <3f296b07$1@news.si.com>  8 >Your LOGIN.COM is not executed in the detached process.J >User the /AUTHORIZE qualifier in your RUN command, that will take care of >this.  F Since he is running loginout.exe as the image, I don't believe you are correct. -- iI Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot comp5 Smiths Aerospace  Addresses modified to prevent SPAM. D 3290 Patterson Ave. SE, MS 1B3 Replace "at" with "@", "dot" with "." Grand Rapids, MI 49512-1991d8        This opinion doesn't represent that of my company   ------------------------------  + Date: Thu, 31 Jul 2003 20:10:11 +0000 (UTC)b- From: lewis@spyder.mitre.org (Keith A. Lewis)r* Subject: Re: Running Java program detached. Message-ID: <bgbt33$ksn$1@newslocal.mitre.org>   thomas.krebs@deutsche-boerse.com (Thomas Krebs) writes in article <2387c644.0307300748.42ff88c@posting.google.com> dated 30 Jul 2003 08:48:45 -0700:G >I am using JRE 1.4.0-1 on OpenVMS 7.3, trying to run a Java program aspE >a detached process. Therefore I have written a DCL, which looks likee >this: >t' >$ @sys$manager:java$140_setup.com FAST." >$ define JAVA$CLASSPATH .........
 >$ java "..."  > G >Running this DCL interactive works like a charm. However, when I start. >it with > 0 >$run /detached/input=dcl-above.com/output=t.log >sys$system:loginout.exe >o! >It fails with the error message:i? >Load class: sun/reflect/MagicAccessorImpl: could not find filee >oF >By switching on verbose mode in the java command, I figured that thisF >a class from  rt.jar which is the first class java tries to find. So,F >obviously in the detached process it isn't able to locate its runtime
 >environment.n >i >Any ideas?s  L I second what the other guy said about trying /AUTHORIZE.  When you use thatH qualifier, you get your job quotas from the UAF.  It's possible that the5 PQL_ default quotas aren't enough to run Java at all.e  L Another experiment which might be enlightening would be to run it as a batchK job, moving your LOGIN.COM somewhere else to make sure it isn't a factor.     J BTW, is there a reason you're doing this as detached rather than batch?  IH do that myself with apps that I intend to keep running indefinitely, butJ jobs I expect to finish get sent to the batch queues; they are much easier
 to deal with.c  + --Keith Lewis              klewis$mitre.org > The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Thu, 31 Jul 2003 14:22:05 -0400 ) From: JF Mezei <jfmezei.nospam@istop.com>a= Subject: Re: SFF commands for TCPIP$SYSTEM:TCPIP$SMTP_SFF.EXEn) Message-ID: <3F295E48.730775B1@istop.com>r   John Brandon wrote:OP > I have been digging around looking for the command/parameter list for use with > the SMTP SFF.   J There are two sources for SFF documentation. One is on the FAQ (or ask theL wizzard) on the vms web site. This one is more complete. The other is in theK TCPIP 5.3 documentation (TCPIP Services management) manual available on thes VMS documentation web site.q  & > ==================================== > MAIL FROM:<someone>r, > RCPT TO:<smtp%"9008882222@mobile.att.net">% > RCPT TO:<smtp%"me.name@domain.com">  > DATA  > SUBJECT:TESTing in progress... > This is great!& > ====================================  N Everyline except the SUBJECT: and "this is great" are part of RFC821 commands.2 Everything below DATA is in the RFC822 definition.  L The MAIL FROM: typically results in the Return Path: line being added at theY top of the DATA contents. It is typically used to send back a non-delivery notifications.a  N In terms of faking the "From", there are logicals that, when set, require your, application has SYSPRV to fake a From: line.  J In terms of commands other than MAIL FROM and RCPT TO: and DATA, there areM others ones for identification such as HELO (eor a more sophisticated EHELO), J a NOOP to just keep the server connection up, and I think QUIT which is toL cancel the connected to the server. However, I do not believe they have much relevance in this case.a   ------------------------------  % Date: Thu, 31 Jul 2003 14:49:53 -0500e( From: brandon@dalsemi.com (John Brandon)= Subject: Re: SFF commands for TCPIP$SYSTEM:TCPIP$SMTP_SFF.EXEe1 Message-ID: <03073114495351@dscis6-0.dalsemi.com>t  # JF Mezei [jfmezei.nospam@istop.com] L > There are two sources for SFF documentation. One is on the FAQ (or ask theN > wizzard) on the vms web site. This one is more complete. The other is in theM > TCPIP 5.3 documentation (TCPIP Services management) manual available on the4 > VMS documentation web site.-  E The SFF script I wrote was using the references to SFF/SMTP from ATW.d  , MAIL, SMTP, MIME, and attachments? (take II). http://h71000.www7.hp.com/wizard/wiz_4492.html  ( > > ==================================== > > MAIL FROM:<someone>g. > > RCPT TO:<smtp%"9008882222@mobile.att.net">' > > RCPT TO:<smtp%"me.name@domain.com">  > > DATA" > > SUBJECT:TESTing in progress... > > This is great!( > > ====================================  O From what I understand the DATA is the header terminator; lines above it define < the source and recepient, lines below it define the message.* However this does not seem to be the case.  K I put a FROM: after the DATA and the e-mail reflected that value.  Also therA SUBJECT: must immediately following DATA for it to work properly.e  P > Everyline except the SUBJECT: and "this is great" are part of RFC821 commands.4 > Everything below DATA is in the RFC822 definition.  O The SUBJECT: actually properly parsed the subject into the e-mail subject.  And  works - I guessed at it.  N > The MAIL FROM: typically results in the Return Path: line being added at theL > top of the DATA contents. It is typically used to send back a non-delivery > notifications.  O The MAIL FROM: will parse the @domain.com to the end of whatever you put there.r9 Adding a FROM: line after the DATA also acts differently.l      MAIL FROM:<one two three>         Results in:s       three@domain.com    e    MAIL FROM:<someone>
    RCPT:<...>     DATAc    FROM:one two three          Results in:l       one.two.three@domain.com  P > In terms of faking the "From", there are logicals that, when set, require your. > application has SYSPRV to fake a From: line.  C I saw that somewhere in ATW or FAQ.  Will need to look at it again.l  L > In terms of commands other than MAIL FROM and RCPT TO: and DATA, there areO > others ones for identification such as HELO (eor a more sophisticated EHELO),bL > a NOOP to just keep the server connection up, and I think QUIT which is toN > cancel the connected to the server. However, I do not believe they have much > relevance in this case.-  . Rather than re-event the wheel (oh why not...)   I found this sitel  $ http://www.faqs.org/rfcs/rfc821.html  = I think I understand now... I also said that about my wife...c     Thanks for the input JF Mezeir       J*o*h*n B*r*a*n*d*o*nn VMS Systems Administrator7* firstname.lastname.spam.me.not@dalsemi.com   ------------------------------  % Date: Thu, 31 Jul 2003 17:30:48 -0400i* From: JF Mezei <jfmezei.spamnot@istop.com>= Subject: Re: SFF commands for TCPIP$SYSTEM:TCPIP$SMTP_SFF.EXEh) Message-ID: <3F298A76.5FBB3AB7@istop.com>t   John Brandon wrote:nQ > From what I understand the DATA is the header terminator; lines above it defineu> > the source and recepient, lines below it define the message., > However this does not seem to be the case.  N Nop, this is correct. MAIL FROM: and RCPT TO: build the real delivery envelopeL of the message (as well. Anything below the DATA is just the contents of theN message. The contents include the RFC 822 headers that you see when you read aU message. They are parsed by clients but not by MTAs (unless you consider virus/spam).m  L However, each MTA will ADD to the contents. For instance, it is the MTA thatN adds the reply-path line, as well as each Received from <xxx> by <xxxx> lines.      M > I put a FROM: after the DATA and the e-mail reflected that value.  Also thepC > SUBJECT: must immediately following DATA for it to work properly.'  I Incorrect. The order of fields in the RFC822 header is not fixed. What isr: fixed is that the header stops after the first blank line.  
 for instance:   ! MAIL FROM:<johndoe@pinkhouse.gov>s, RCPT TO:<Elizabeth.Windsor@buckingham.co.uk> DATA& From: John Doe <johndoe@pinkhouse.gov>) To: Bill CLinton <b.clinton@lewinsky.org>h Date: 1 Apr 2003 15:25 GMT   Dear Queen, K While this message appears to be adressed to your lover, Mr Clinton, it wash infact deliveredH to you. In order to improve the trans atlantic relations between our twoJ countries, I would strongly suggest that you arrange for Prince William toL meet and fall in love with Mrs Hillary Clinton's daughter. This would ensureI that our british media would thrive on all the rumours and innuendos that K would follow the wedding of the future king of england with the daughter of   the future president of the USA. .l    5 Everything below the DATA command is just that: DATA.sH What the above will do is cause your local SMTP server to connect to theI buckingham.co.uk SMTP server and transmit those commands after your localtI server has identified itself to the buckingham server, and will only sendlM those RCPT TO: commands that belong to buckingham.co.uk. Separate connectionssH would be made for each RCPT TO: that are destined to other recipients in different domains.    Q > The SUBJECT: actually properly parsed the subject into the e-mail subject.  Andd > works - I guessed at it.  M It is parsed by the mail client when it opens the contents of the message. It J is not handled by the MTAs when delivering the messages to each other. (InK fact, most anti-spam features are done right at the RCPT TO: commands, welleS before the DATA phase, so the MTA will block messages before getting t the subject.a   >    MAIL FROM:<one two three> >  >       Results in:e >       three@domain.com  I MAIL FROM: <one two three> should normally result in an error, unless thet reciving MTA is really dumb.  	 >    DATAb >    FROM:one two threee >  >       Results in:   >       one.two.three@domain.com  J The "From:" line in the DATA phase isn't parsed for validity. It is simplyL used by the mail client. (the mail client, especially those who pickup theirL messages via POP) have no idea of the true delivery envelope information, soF they rely on the headers that are part of the contents of the message.   ------------------------------  % Date: Thu, 31 Jul 2003 18:40:04 -0500n( From: brandon@dalsemi.com (John Brandon)= Subject: Re: SFF commands for TCPIP$SYSTEM:TCPIP$SMTP_SFF.EXEc1 Message-ID: <03073118400456@dscis6-0.dalsemi.com>m  O No argument - however there is no specific document as to the syntax for SFF onoL VMS - and I believe that what you are stating is coming from the UNIX RFC822B and not $SYS$COMMON:[SYSEXE]TCPIP$SMTP_SFF.EXE  - just my thought.  K I am also TCPIP V5.0a - so that may also be part of the odd behavior I have) seen.h         J*o*h*n B*r*a*n*d*o*n) VMS Systems Administratori* firstname.lastname.spam.me.not@dalsemi.com   ------------------------------    Date: 31 Jul 2003 17:20:08 -0700# From: dooleys@snowy.net.au (dooley)t= Subject: Re: SFF commands for TCPIP$SYSTEM:TCPIP$SMTP_SFF.EXEo= Message-ID: <1ca82fc6.0307311620.4a5f7be2@posting.google.com>s  a brandon@dalsemi.com (John Brandon) wrote in message news:<03073111110795@dscis6-0.dalsemi.com>... P > I have been digging around looking for the command/parameter list for use with > the SMTP SFF.  l > 8 > So far I have found that this works (enclosed in ====) > & > ==================================== > MAIL FROM:<someone> , > RCPT TO:<smtp%"9008882222@mobile.att.net">% > RCPT TO:<smtp%"me.name@domain.com">m > DATA  > SUBJECT:TESTing in progress... > This is great!& > ==================================== >  > N > However, I can not find the documentation that details additional parametersF > (if any), qualifiers, logicals, etc...  Anyone know this or have it? >  there is a summary at V http://h71000.www7.hp.com/doc/tcpip51/documentation/v51_relnotes_001.html#sff_overview Phil   ------------------------------  % Date: Thu, 31 Jul 2003 19:32:58 -0500 ( From: brandon@dalsemi.com (John Brandon)= Subject: Re: SFF commands for TCPIP$SYSTEM:TCPIP$SMTP_SFF.EXEe1 Message-ID: <03073119325897@dscis6-0.dalsemi.com>o   Phil wrote:  > there is a summary at X > http://h71000.www7.hp.com/doc/tcpip51/documentation/v51_relnotes_001.html#sff_overview  2 Thanks... looked everywhere but the release notes.  ? Pretty much the same information as contained in ATW and FAQ.  o         J*o*h*n B*r*a*n*d*o*n  VMS Systems Administratorn* firstname.lastname.spam.me.not@dalsemi.com   ------------------------------  % Date: Thu, 31 Jul 2003 19:11:22 -0400r) From: "Neil Rieck" <n.rieck@sympatico.ca>oP Subject: SPARC/Solaris Question (was Re: Yamahill? Prescott? Wot's it all mean?)7 Message-ID: <umhWa.106$Ji1.40166@news20.bellglobal.com>e   SUN segue...  5 "Bill Todd" <billtodd@metrocast.net> wrote in message , news:Y7icnbZlTKwBV7WiXTWJgA@metrocast.net... >- > [...snip...] >,H > If Itanic ever does become more popular, it will compete directly withI > Opteron (and Xeon, and, if it appears, Yamhill) in the low end and withiG > POWER4+ in the high end (and with SPARC across the board, though withaF > SPARC - as usual - at a major disadvantage in terms of raw processorF > performance).  Unless Sanmina - or IBM - fields large-system OpteronH > chipsets, in which case Opteron will compete across the board as well. >  > [...snip...] >@N Since you've brought up SPARC, we've recently noticed some strange performanceM problems on systems running Solaris on a Sun-Ultra-10 (like a modestly loadedwN DHCP server causing CPU panic messages in high memory) . This has started someN debate in our office as to whether these systems are multi-threaded or not? Do you have any info in this area?5    
 Neil Rieck Kitchener/Waterloo/Cambridge,b Ontario, Canada.! http://www3.sympatico.ca/n.rieck/u   ------------------------------  % Date: Thu, 31 Jul 2003 21:23:03 -0400u* From: "Bill Todd" <billtodd@metrocast.net>T Subject: Re: SPARC/Solaris Question (was Re: Yamahill? Prescott? Wot's it all mean?)2 Message-ID: <7S2dnfXRJ6WYXLSiXTWJkA@metrocast.net>  4 "Neil Rieck" <n.rieck@sympatico.ca> wrote in message1 news:umhWa.106$Ji1.40166@news20.bellglobal.com...s   ...   D > Since you've brought up SPARC, we've recently noticed some strange performance1H > problems on systems running Solaris on a Sun-Ultra-10 (like a modestly loadedK > DHCP server causing CPU panic messages in high memory) . This has started  someH > debate in our office as to whether these systems are multi-threaded or not? Dor! > you have any info in this area?k  L While I'm hardly the person to ask about SPARC details, my strong impressionJ is that no SPARC processor supports any kind of *hardware* multi-threading
 (SMT or CMP).a   - bill   ------------------------------   Date: 1 Aug 2003 01:55:51 GMT'2 From: "Zane H. Healy" <healyzh@shell1.aracnet.com>T Subject: Re: SPARC/Solaris Question (was Re: Yamahill? Prescott? Wot's it all mean?), Message-ID: <bgchb702u61@enews4.newsguy.com>  ) Bill Todd <billtodd@metrocast.net> wrote:sI >> problems on systems running Solaris on a Sun-Ultra-10 (like a modestlyl > loadedL >> DHCP server causing CPU panic messages in high memory) . This has started > someI >> debate in our office as to whether these systems are multi-threaded orw	 > not? Doo" >> you have any info in this area?  N > While I'm hardly the person to ask about SPARC details, my strong impressionL > is that no SPARC processor supports any kind of *hardware* multi-threading > (SMT or CMP).0  J Also, unless the Ultra 10 has been upgraded with a better disk, they stinkL performance wise.  I forget what speed EIDE drive mine came with, but it wasK less than 5400RPM.  Putting a 7200RPM EIDE Drive in it helped, but it still.J only has a ATA33 interface.  I'd imagine that putting a UW-SCSI card in itL would seriously help. Something else to consider is if the system has enoughD RAM (you really don't want to be swapping to the crummy stock disk).   				Zane   ------------------------------  % Date: Fri, 01 Aug 2003 08:28:23 +0800e, From: Paul Repacholi <prep@prep.synonet.com> Subject: Re: Sun and SCO- Message-ID: <87ispinrvc.fsf@prep.synonet.com>-  Q Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> writes:>  < > I say perspective because no one as suggested yet that SCOF > definitely don't have a case against IBM. What happens if the courts7 > find for SCO ruling that IBM have missused SCO's IP ?s  E > Posters from HP who so far have gone bad SCO being nasty to IBM andoD > Linux while attempting to include Sun in the SCO camp as well willE > all be sitting on their hands. HP is just as keen to protect its IPl+ > as any other commercial software company.h  G Well then they will have to dump nasty poluted PHUX and port pure legale unpoluted T64 to the itanic.   -- g< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.o@                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Fri, 01 Aug 2003 01:35:38 GMT ) From: bob smith <sfmc68@bellatlantic.net>  Subject: Re: Sun and SCO6 Message-ID: <KtjWa.3010$M46.2856@nwrddc02.gnilink.net>  ( Andrew Harrison SUNUK Consultancy wrote:   > Eric Bruno wrote:c > K >> "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in messageV0 >> news:4jPXVjhbY4yr@eisner.encompasserve.org... >>7 >>> In article <uqpPmMr5zJvw@eisner.encompasserve.org>,. >> >>@ >> koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes: >>K >>>> In article <bfgvpg$lik$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUKo >>>o >>>a> >> Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> writes: >>D >>>>> You didn't consider that the deal could simply be a technologyC >>>>> licensing deal that has nothing to do with SCO's ongoing casey >>>>> against IBM ???s >>>> >>>> >>>>   Nope. >>>t >>>cH >>>   And the July 21 interview with Scott McNealy in eWeek supports our4 >>>   skepticism on Sun's true feelings about Linux. >>>g >>K >> Two years ago Sun as going to kill it's Solaris x86 platform in favor ofh >> Linux.   This now hasH >> flipped 180%.  The Solaris x86 community fought long and hard change  >> Sun's >> position.L >> Since Sun has a solid x86 Unix with Solaris x86 offering. Linux is not as+ >> critical to Sun in order to get into x86.K >> installations, unlike HP and IBM which do not have x86 versions of theirf >> Unix flavors.K >> Installing Solaris x86 from a sales and marketing position will possiblee >> allow Sun a bettercE >> opportunity to sell Ultrasparc technology as machines come due forv >> replacement. Since then >> OS is the same. >>L >> Due to neglect Sun is way behind in drivers for current hardware for x86. >> Access to< >> SCO's source base provides access to current x86 drivers. >> >> Eric Bruno. >> > ? > We also hired a number of SCO staff who used to work in theirs? > device driver team, they now work in Sun's kernel engineeringhC > team in Watford (NR London) producing Solaris x86 device drivers.o > G > The conspiracy theorists and people who publish unresearched BS about D > Sun on this newsgroup will probably still think that the whole SCO@ > case is just a put up job by Sun. As usual the truth is rather > more mundane.t > 	 > Regardsp > Andrew Harrison, Andrew, B I have given you a hard time in the past, but this post is pretty  decent. Goo post.  bobe   ------------------------------  % Date: Thu, 31 Jul 2003 21:49:16 -0400e( From: David Froble <davef@tsoft-inc.com> Subject: Re: Sun and SCO, Message-ID: <3F29C71C.3020505@tsoft-inc.com>  ( Andrew Harrison SUNUK Consultancy wrote:   > Bob Koehler wrote: > J >> In article <bg86nt$1gq$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK > >> Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> writes: >> >>B >>> In a previous discussion the only thing of note that an of the@ >>> HP appologists to come up with as leading the Linux, ItaniumC >>> port. Not that noteworthy when you remember what Linus Torvaldsw$ >>> thinks of Itanium as a platform. >> >> >>I >>    Just because Torvalds wrote a kernel for an 80386, doesn't make him J >>    a great judge of technology.  He actually dislikes VMS so he doesn't >>    get any credit here. >>     >  > ? > I havn't detected any like for or much understanding of Linuxm0 > on the part of Fred, doesn't stop him posting. >  > 	 > Regardsh > Andrew Harrison  >   
 Hey andy boy,r  C You running out of FUD?  Back to picking on Fred?  What's that pay?S   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com6 T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486   ------------------------------  % Date: Thu, 31 Jul 2003 14:23:35 -0400 ) From: JF Mezei <jfmezei.nospam@istop.com>w< Subject: Re: Sun tries to woo AlphaServer users away from HP) Message-ID: <3F295EA3.892CA085@istop.com>o   Rob Young wrote:P > "You're going to see Linux and Windows absolutely eviscerate the midrange Unix > market," said Capellas.  > G >         Capellas was/is the source and we beat this about quite a bit1 >         1.2 years ago.  L Sorry, Winkler was the original source. Curly was merely continuing what his& handlers such as Winkler had told him.   ------------------------------  % Date: Fri, 01 Aug 2003 10:08:24 +0800 , From: Paul Repacholi <prep@prep.synonet.com>< Subject: Re: Sun tries to woo AlphaServer users away from HP- Message-ID: <871xw6nn8n.fsf@prep.synonet.com>n  % "Tom Linden" <tom@kednos.com> writes:   F > I wish we had better news to report on this issue of porting PL/I toD > Itanium.  As you know on Alpha we interface to the GEM backend andE > the effort to hook it up to newer version which supports Itanium isiA > really no big deal.  We have made the request to HP, but at thelD > moment they don't seem to understand the business opportunity that: > is at risk.  As the recent discussions on packed decimalE > calculations pointed out, these can not be performed with C or C++,t@ > which is, of course, one reason that all the largest financialE > institutions stick with PL/I.  This also ties in with the thread onhF > Disaster Recovery, i.e.  you can have the best scheme out there, but5 > if your applications don't run there it is useless.a  @ > I think this question would be appropriately addressed to Mark	 > Gorham.c  ' And that means no SDL on the good ship.T   Another log on the fire...   -- a< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.e@                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Thu, 31 Jul 2003 19:37:00 -0400r) From: "Neil Rieck" <n.rieck@sympatico.ca>o Subject: TCP/IP feature requesth7 Message-ID: <wKhWa.111$Ji1.44454@news20.bellglobal.com>m  L At a recent OpenVMS symposium in Ottawa, I approached Gaitan D'Antoni with aK suggestion to add an "FTP API" and a "Telnet API" into "TCP/IP Services forlL OpenVMS". After all, every major OS (including Windows) provides this so whyG not OpenVMS? (BTW, these modules have been available forever in Process2% Software Corps "TCPware for OpenVMS")l  ; I followed up with an e-mail and this is Gaitan's response.s  	 * * * * *i   -----Original Message-----8 From: D'Antoni, Gaitan [mailto:gaitan.dantoni[at]hp.com] Sent: 2003-07-30 21:21 To: RIECK, NEILt; Cc: Woodcome, Lawrence [mailto:lawrence.woodcome[at]hp.com] % Subject: RE: OpenVMS TCP/IP Wish List    Hi Neil,  M I heard back from the Product Manager and the Technical Leader regarding youryD suggestion. This is one of the items on the list of things to do butM unfortunately it never seems to make it to the top of the list due to limited,J resources and other requirements. I'd say the soonest we could get to this( would be post V5.5, or sometime in 2005.  K If you need any other information or have other requirements please contact-G Larry Woodcome, [mailto:lawrence.woodcome[at]hp.com] the TCP/IP Serviceh Product Manager.   Gaitan  	 * * * * *k  N It is my belief that if there is a demand for this then they will build it (orJ at least build it sooner). If anyone out there thinks this is a good idea,7 then please send a brief polite note to Larry Woodcome.u  N p.s. I've taken the liberty of de-spamming their e-mail addresses; You'll know what to do to make them usable.d    
 Neil Rieck Kitchener/Waterloo/Cambridge,  Ontario, Canada.! http://www3.sympatico.ca/n.rieck/e   ------------------------------  % Date: Thu, 31 Jul 2003 14:19:32 -0400f< From: "Carlc Internet Services" <cc@nospamming-to-carlc.com> Subject: Re: Web Apps for VMS.1 Message-ID: <U4dWa.416$kl5.23885@news.uswest.net>   = "Kevin Monceaux" <OwnedByDogs@grandecom.net> wrote in messageuC news:Pine.LNX.4.50.0307302107470.19634-100000@Linux.monceaux.com...d >gL > I'm a little behind on reading my Info-VAX mail.  Oddly enough today I wasL > thinking about installing a web server on my VAX and then came across thisJ > thread.  I have Apache/PHP/MySQL running on a couple of Linux boxes.  ItJ > would be nice to have Apache available for the VAX architecture.  One ofI > these days I really should locate a previously loved Alpha so I can try I > out VMS for the Alpha paltform.  Of the two web servers available for a L > VAX what are the differences and which would be a good one as a first try?    J I've been using OSU on both VAX then on Alpha as the business grew. As oneD of the FEW OpenVMS ISP Webhosting companies left in the USA, OSU hasH been completely 100% rock solid (yes, OpenVMS helps <GRIN>). It has it'sK own config file formats, but once your use to it, it's the energizer bunny.n+ Just keeps on going and going and going....   G I have PHP working on it today! Only took downloading the file from thenD OSU site and viola, PHP is working. OSU handles PHP as a cgi-bin, soH you can do some really cool security with it (like run PHP programs from other user accounts).r  H I've messed with MySQL a bit on OpenVMS. I can't get it to compile causeB I don't have the cxx (c++ I think its called) compiler. There is a pre-compiledG version, but requires a few other things I didn't want to preload (like  openssl J older version). Eventually, I'll get it running for another project I have in mind.  F There is a users group with a low volume email talk available for OSU.G OSU is found at http://kcgl1.eng.ohio-state.edu/www/doc/serverinfo.html    Carl   ==================== http://www.carlc.com/e  =       "Price, Performance, Quality. Choose any two you like."w   ------------------------------   Date: 1 Aug 2003 02:49:12 GMTe2 From: "Zane H. Healy" <healyzh@shell1.aracnet.com> Subject: Re: Web Apps for VMSt* Message-ID: <bgckf805t@enews4.newsguy.com>  ) Bob Ceculski <bob@instantwhip.com> wrote:B< > Kevin, if you want the best ... both for alpha and vax ...  5 > http://www.sss.co.nz/software/purveyor/purveyor.htm   L ?!?!?!  Purveyor *was* a nice product, but it's been dead for *how* long?  II think Process will provide Hobbyist Licenses for it, but that's about it.0   		Zane   ------------------------------  % Date: Fri, 01 Aug 2003 09:17:45 +0800 , From: Paul Repacholi <prep@prep.synonet.com>3 Subject: Re: Yamahill? Prescott? Wot's it all mean?t- Message-ID: <87adaunpl2.fsf@prep.synonet.com>d  + JF Mezei <jfmezei.nospam@istop.com> writes:e  C > since it will be available from multiple sources, whereas IA64 isn* > proprietary, available from Intel only).  > Well, to be pedantic, it is not from Intel. It is from Aliance- for Advanced Computing or some such rubish...w  F The entire idea is to cut adrift all those who have deals and licencesD with intel so that *ONLY* intel and hp can play in that pond. If youE think this will be good for you, then there is a long queue of peoplen0 with wonderfull bridges who wish to talk to you.   -- k< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.s@                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  + Date: Thu, 31 Jul 2003 21:21:22 +0000 (UTC)a) From: Dan Foster <dsf@globalcrossing.net>P/ Subject: Re: [Change topic -> OT] Pence/Pennies03 Message-ID: <slrnbij22p.13v.dsf@gaia.roc2.gblx.net>t  ] In article <Xns93C974F248D0Afalkarcabca@205.233.108.180>, Alfred Falk <falk@arc.ab.ca> wrote:fI > Also, in U.S. and Canada, "penny" is not the official name of any coin oJ > or amount of money.  It is "one cent", not "one penny", and has been so  > for a very long time.Y  I That's correct. The U.S. defined the coin units and other parameters wheneG the Congress passed The Coinage Act on April 2, 1792 -- the cent as therJ proper name of the basic unit was defined as being the official 1/100th ofH a dollar unit. (We also had half cent coins then, as well...and also theD two and three cent coins but was discontinued by the end of the 19th	 century.)P  8 The full text of The Coinage Act (U.S.) can be found at:  2 http://landru.i-link-2.net/monques/coinageact.html  I Sections 9 and 20 defined the units -- and spelled a dime as a disme, and K also defined a mille as being a thousandth part of a dollar, or 1/10th of an cent.i  E Popular usage dies hard, though, so you still hear people in the U.S.h refer to cents as 'pennies'. :)I   -Dan   ------------------------------  # Date: Fri, 01 Aug 2003 00:44:19 GMT  From: Rob.Buxton@wcc.govt.nz/ Subject: Re: [Change topic -> OT] Pence/Pennies-$ Message-ID: <3f29b67e.16927968@news>  / On Thu, 31 Jul 2003 06:42:26 GMT, Mike Rechtmane  <michael.rechtman@hp.com> wrote:   >Roy Omond wrote:i >>   >> Bob Koehler wrote:. >> ri >> > In article <bg6c0q$jrogm$2@ID-120847.news.uni-berlin.de>, "John Travell" <john@jomatech.com> writes:  >> >& >> >>Plural - Pence, Singular - Penny. >> >E >> >    We must have been de-anglicized somewhere.  Plural - pennies.t >> n4 >> I'd say it's slightly more complicated in the UK. >> e; >> The "pence" usage is (was?) usually for a monetary valueh >> e.g. fourpence, sixpence. >> l5 >> I think that it would have been common to say e.g.P; >> "there are four pennies on the table" (i.e. 4 individual ; >> penny coins), in contrast to "there are fourpence on the 9 >> table" (possibly made up of one threepenny bit and one 
 >> penny). >> a > > >From faraway memories of growing up in one of "the colonies": >oA >"there *is* fourpence on the table" (as one consolidated amount. > >also "thrup'ny" not threepenny, and "tuppence-three-farthing"( >(price of a loaf of bread back when...) >-, And there were probably regional variations.= I recall three ha'penny being used as one and a half pennies.aF And from above, 3d was generally called a thrup'ny bit. The d became a p at decimilisation.- Not to mention tanners, bobs and  half-crowns    >e9 >> Ah, the threepenny bit ... yeah, I know, I can hear ita; >> coming - "you had a threepenny bit ? you were lucky ..."o >> :-) >> o >> Showing my age ( >L ),  >> e >> Roy Omond >> Blue Bubble Ltd.k >k >4 >( approaching LX )p >-- F >---------------------------------------------------------------------F >Usual disclaimer: All opinions are mine alone, perhaps not even that.@ >Mike Rechtman                            *rechtman@tzora.co.il*G >Kibbutz Tzor'a.                          Voice (home): 972-2-9908337  mC >  "20% of a job takes 80% of the time, the rest takes another 80%"mF >---------------------------------------------------------------------   ------------------------------  % Date: Fri, 01 Aug 2003 08:07:42 +0800e, From: Paul Repacholi <prep@prep.synonet.com>/ Subject: Re: [Change topic -> OT] Pence/Pennies3- Message-ID: <87r846nstt.fsf@prep.synonet.com>n  / Roy Omond <Roy.Omond@BlueBubble.UK.Com> writes:J  F > Ah, the threepenny bit ... yeah, I know, I can hear it coming - "you1 > had a threepenny bit ? you were lucky ..."  :-)   D Thrupence! You could buy a bottle of Coke with that! (and 4p for the- mid size, and a whole 6p for the 26oz bottle)    -- I< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.t@                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------   End of INFO-VAX 2003.421 ************************ (Keith A. Lewis)r* Subject: Re: Running Java program detached. Message-ID: <bgbt33$ksn$1@newslocal.mitre.org>   thomas.krebs@deutsche-boerse.com (Thomas Krebs) writes in article <238 @4 ݱ `  %+oA +@, }( ] @  *OA
 
B( ]    (A @  $ ]G  .  2.U J 2J+FԔ FvJF  >
0A  .$ = J0 AFF$ =h =   E;A0`CzC. ` $@G  ZC E@CG @ TGBGD % bv~Zk  C   D?4 ݡ, }-A
 
@.A @4 ݱ `  %+fA +B, }( ] @  E*FA
 JB( ]    &A @    .GvmJhBGU JF  >
0A
#.GzZk J = !F	 !K   
, H@ H Da/H D 
<G ݠ$ =&E ݠ$ X } =F  /	0 B$ = K `8K8B 1"Q.QQJ'J1!J C C$ = = F `&E@&E@  i/ iK	0 A$ = `8K8C #,@HH H C4G h G$ =  'A'D# xGF $ =G F   I.&E IJ	0 AXGF@g&E$ =x`&E# `$ = =  . J	0 A$ = `7J7C ",@HH H C$ =@G &E  XF F
  1AK "BtJ  Q.&E0@CR QJTF  Q>? G;AK }"{BXF  ;/ CK@Y ;K  ;?$ !` #  , H D ( BxZk( B =# @ C