1 INFO-VAX	Sun, 01 Jul 2001	Volume 2001 : Issue 362       Contents:' A black week for the computer industry? + Re: A black week for the computer industry? + Re: A black week for the computer industry? + Re: A black week for the computer industry? / a CPQ Intel/Alpha Roadmap,  Hess-Neckels (ISSG)  Re: a letter from Rich Marcello  Re: Compaq switches to IA-64D Re: Compilers go to Intel, hobbyist, CSLG, etc. goes out the window?9 Free text-to-HTML converter AsctoHTM 4.0 for VMS released   Re: Full port of VMS to Itanium.* Re: Hobby LIcense: One USER, or one LOGIN?; Re: I  (of all people) just got a letter from Rich Marcello ; Re: I  (of all people) just got a letter from Rich Marcello ; Re: I  (of all people) just got a letter from Rich Marcello  Re: IA64 Rocks My World  Re: IA64 Rocks My World ) Re: One more dreadful thought to consider 6 Re: Question to Charlie Matco - reply to Terry Shannon6 Re: Question to Charlie Matco - reply to Terry Shannon RMS file performance' Re: Thanks Compaq for the new business!  The Alpha/IA64 Hybrid  Re: The Alpha/IA64 Hybrid  Re: The Alpha/IA64 Hybrid  Re: The Alpha/IA64 Hybrid  Re: The Alpha/IA64 Hybrid  Re: The Alpha/IA64 Hybrid 1 Re: The death of VMS has been greatly exaggerated $ Using SCSI controller on QBUS system VMS 7.3 experiences? RE: VMS 7.3 experiences? Re: VMS 7.3 experiences? Re: VMS V7.3 SPD Error Re: VMS V7.3 SPD Error< Re: Wailing and moaning... (was: Question to Charlie Matco.)< Re: Wailing and moaning... (was: Question to Charlie Matco.) Re: What is lockstep ?! Re: where should SYSUAF.DAT live? + Re: Why Compaq won't succeed on IA64 either + Re: Why Compaq won't succeed on IA64 either / Re: Yet Another Reason Why Windoze .ne. OpenVMS / RE: Yet Another Reason Why Windoze .ne. OpenVMS / Re: Yet Another Reason Why Windoze .ne. OpenVMS   F ----------------------------------------------------------------------  % Date: Sun, 01 Jul 2001 10:43:41 -0400 ) From: "Joseph B. Gurman" <gurman@ari.net> 0 Subject: A black week for the computer industry?= Message-ID: <gurman-6B244D.10434101072001@news.crosslink.net>   H     Looking back on the past week, I still get a little short of breath ) realizing that in the span of a few days:   9 1. Compaq has sold the farm (as we know it) to Intel, and   E 2. the breakup of Microsoft has been rescinded by an appellate court  C full of ostensible former local bar associates of Bill Gates's dad.   I     Either of these could end up being for the good, I suppose (e.g. all  B the comments here about OpenVMS on a cheaper platform, ability to F compete directly with MS OS's, &c.), but I think good for us (OpenVMS  users) depends critically on:   H 1. Compaq's really staying the course (lots of resources. lots) for the I big software engineering job of porting OpenVMS to the IPF architecture,   and   G 2. Intel actually producing something, in OEM quantities, that OpenVMS   can run on by 2004.   G     With people like Hoff on the job, 1 should be, as he says, "hard,"  H but doable --- if Compaq doesn't run into enough bad quarters that they I decide to ditch the effort. Perhaps they already have (sorry, conspiracy  % theory). It certainly won't be cheap.   D     It's number 2 that really scares me, with or without conspiracy G theory. I really find myself wanting to say that Alpha technology will  C help them come up with a viable product (in the Madison timeframe,  H whenver that is), but Intel has really good engineers and they're still ? having lots of problems producing their first family of 64-bit   processors.   G     I feel a lot of unease. The OpenVMS porting group is going to live   in interesting times.   G     And the Microsoft news? Clearing the way for the continued triumph  I of the shoddy? All other OS's bascially irrelevant, as the Borg head for   universal domination?   F     I could ramble on, but I gotta go finish my 7.3 upgrades and play . around some more with our Mac OS X systems....                     Joe Gurman   ------------------------------  % Date: Sun, 01 Jul 2001 10:13:57 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> 4 Subject: Re: A black week for the computer industry?' Message-ID: <3B3F3E35.786BCF28@fsi.net>    "Joseph B. Gurman" wrote:  > [snip]H >     With people like Hoff on the job, 1 should be, as he says, "hard,"I > but doable --- if Compaq doesn't run into enough bad quarters that they J > decide to ditch the effort. Perhaps they already have (sorry, conspiracy' > theory). It certainly won't be cheap.    Ever heard of "Emerald"? See:   ' http://www.djesys.com/vms/vmspolls.html    ...especially...  / http://www.djesys.com/vms/vmspolls.html#emerald    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/   F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.    ------------------------------  # Date: Sun, 01 Jul 2001 15:36:30 GMT 4 From: "Terry C. Shannon" <terryshannon@mediaone.net>4 Subject: Re: A black week for the computer industry?; Message-ID: <2sH%6.833$9r6.1508386@typhoon.ne.mediaone.net>   4 "Joseph B. Gurman" <gurman@ari.net> wrote in message7 news:gurman-6B244D.10434101072001@news.crosslink.net... I >     Looking back on the past week, I still get a little short of breath + > realizing that in the span of a few days:  > ; > 1. Compaq has sold the farm (as we know it) to Intel, and  > F > 2. the breakup of Microsoft has been rescinded by an appellate courtE > full of ostensible former local bar associates of Bill Gates's dad.  > J >     Either of these could end up being for the good, I suppose (e.g. allC > the comments here about OpenVMS on a cheaper platform, ability to G > compete directly with MS OS's, &c.), but I think good for us (OpenVMS  > users) depends critically on:  > I > 1. Compaq's really staying the course (lots of resources. lots) for the J > big software engineering job of porting OpenVMS to the IPF architecture,  F Kent Ferson in Tru64-land last Tuesday claimed that CPQ has authorized; additional budget for the porting efforts on Tru64 and VMS.    > and  > H > 2. Intel actually producing something, in OEM quantities, that OpenVMS > can run on by 2004.    That remains an issue.   > H >     With people like Hoff on the job, 1 should be, as he says, "hard,"I > but doable --- if Compaq doesn't run into enough bad quarters that they J > decide to ditch the effort. Perhaps they already have (sorry, conspiracy' > theory). It certainly won't be cheap.   
 No, it won't.    ------------------------------  $ Date: Sun, 1 Jul 2001 14:00:51 -0400' From: "Bill Todd" <billtodd@foo.mv.com> 4 Subject: Re: A black week for the computer industry?( Message-ID: <9hno9r$2p9$1@pyrite.mv.net>  4 "Joseph B. Gurman" <gurman@ari.net> wrote in message7 news:gurman-6B244D.10434101072001@news.crosslink.net...    ...   J >     Either of these could end up being for the good, I suppose (e.g. allC > the comments here about OpenVMS on a cheaper platform, ability to G > compete directly with MS OS's, &c.), but I think good for us (OpenVMS  > users) depends critically on:  > I > 1. Compaq's really staying the course (lots of resources. lots) for the J > big software engineering job of porting OpenVMS to the IPF architecture, > and  > H > 2. Intel actually producing something, in OEM quantities, that OpenVMS > can run on by 2004.   I I suggest that you left out the most important point of all:  the port of H VMS is at best pretty much goodness-neutral (compared with simply havingL continued to support - sic - Alpha) in and of itself.  VMS sales will take aH hit from customers who see dropping Alpha as reason to mistrust Compaq'sJ commitment to all things non-Wintel, and may get a compensating boost fromJ customers enthused about the possibility of running it on Intel hardware 3 years down the road.  L So for VMS's fortunes to *rise*, Compaq will need to do what it has shown noK interest in doing for 3 years now:  actually *market* and more aggressively D develop VMS to indicate to potential new customers that Compaq is as6 committed to the platform as it would like them to be.   - bill   ------------------------------  # Date: Sun, 01 Jul 2001 02:27:47 GMT  From: LBohan@dbc.spam_less..com 8 Subject: a CPQ Intel/Alpha Roadmap,  Hess-Neckels (ISSG)8 Message-ID: <s72tjt0qfc1sjfdo819l493a50lvag0tri@4ax.com>  " a roadmap from a ISSG pt of view.   5 While I know I could care less about CPQ's ISSG group 8 and their products, the asides/remarks here inre Alpha, 0 and also their 32-cpu intel box are of interest.    ? http://www.siliconinvestor.com/stocktalk/msg.gsp?msgid=16013998   ! To:Night Writer who wrote (91949)  From: Night Writer Friday, Jun 29, 2001  2:25 PM  Respond to of 91995   + Compaq Fleshes Out Intel and Alpha Roadmaps   C Jun 29, 2001 (ComputerWire via COMTEX) -- Compaq Computer Corp will ? launch a 32-way implementation on Intel's Foster Pentium 4 Xeon C processors by, at the latest, the end of the first quarter of 2001, D then a 32-way IA64 box in the 2002/2003 timeframe, whenever McKinley ships.  E It is in the Madison iteration of the Itanium family in 2004 that the C Alpha technology from Compaq will be integrated into the processor, E explained Carol Hess-Nickels, director of the computer manufacturer's ; industry standard server business unit for the EMEA region.   B Hess-Nickels laid out the roadmap for Compaq's NT and Alpha serverC lines at the launch of two more ProLiant and two appliance servers, F including the Houston-based company's first product to ship with Linux from the factory.   C The appliances are the TaskSmart W2200 and C4000, both based on the D ProLiant DL320 server platform. The former is a web server, bundlingC RedHat Linux and Apache web server software, and it is this product F that marks Compaq's entry into the shipment of Linux-based appliances.E Previously it would have been the channel or the customers themselves C that would have loaded the open source OS onto Compaq boxes for web D serving, Hess-Nickels explained. Indeed, it was after observing that@ this was the normal route for web serving that Compaq decided to; launch a product with the free OS bundled from the factory.   E The C4000 is a caching appliance that marks the beginning of Compaq's B relationship with Inktomi, whose Traffic Server Engine + Media IXTE ships bundled onto the box. As such it replaces the previous C-Series A appliances, which shipped with Novell Inc's caching software. The C switch to Inktomi Corp, Hess-Nickels explained, was due to the fact @ that its product supports all three of the major streaming mediaC formats, i.e. Real Media, Windows Media and QuickTime. The C-Series 8 box actually ships in mid-July and no price has yet been= announced. The W-Series has actually been shipping since May.   A These products complement the N-Series, Compaq's network attached B storage (NAS) product line launched last year. That line currentlyC runs a cut-down NT, but the advent of the C4000 inevitably begs the ? question whether a Linux-based NAS box might not also be in the F offing, particularly as there are other competitors such as the NASRaQ9 from Sun Microsystems Inc's Cobalt division and the Snap! ? Server line from Quantum Corp's Snap! Division that are already @ running on the free OS. "Directionally that's a good way to go," commented Hess-Nickels.   C The two ProLiant unveiled were the DL380, a 2U rack product and the ; ML370, a 5U tower. What is new about these boxes, explained D Hess-Nickels, is the addition of an online space memory module. ThisD Online Redundant Space Memory, as it is called, is the result of the1 company's Advanced Memory Protection Architecture @ for higher levels of server uptime by offering DIMM level memoryC failover. No prices are currently available on these boxes, as they & are yet to become generally available.  ? As for the Alpha line of products, Hess-Nickels fleshed out the D comments made earlier in the week by CEO Michael Capellas that thereE would be "two more speed bumps" in the processor roadmap prior to the , merger of the Alpha technology with Itanium.  E Alpha is currently at the Ev6 iteration, she recalled, and there will C actually be what she termed an "Ev6x" with some enhancements to the C current product. Then there is the actual next iteration, Ev7, with B possibly also an Ev7x, and it is at Ev8 that the technology passesE across into Itanium. So technically that should be three speed bumps.   > On the 32-way Intel side, the McKinley-based box coming out inD 2002/2003 will, of course, already be a 64-bit machine. With Madison@ incorporating the Alpha technology, however, and that technologyF forming the basis for further Compaq 32-way machines, the product line1 should gain considerable processing capabilities.   F Hess-Nickel added that Compaq decided to drop the Unisys Corp's 32-wayF ES7000 boxes earlier this year (CI No 4,154) after announcing it wouldE OEM and actually starting selling it because "we were selling it with F [Microsoft's] Data Center Server, but 95% of customer preferred to buy the 8-way and cluster."   F So why continue work on 32-way? "By the time we launch our product theE Data Center software will have become more accepted by customers, who E for the time being are still asking why not just go with the Advanced  server," she explained.   5 Computergram International: Issue 4196, June 29, 2001    ------------------------------   Date: 1 Jul 2001 16:54:24 +0200  From: gazso@ludens.elte.hu( Subject: Re: a letter from Rich Marcello! Message-ID: <LILPgltBUDuQ@ludens>    Rich Marcello wrote in > ----- Original Message -----  ? >> First let me assure you that Compaq will immediately begin a : >> full port of the Tru64 UNIX, OpenVMS and NonStop Kernel9 >> operating systems and development tools to the Itanium : >> processor family. This means that all of the enterprise? >> features and characteristics that you have relied on will be ; >> there on the Itanium platform. Additionally, Compaq will : >> continue development and enhancement of these operating? >> system environments ensuring that we can continue to deliver 3 >> leadership capabilities - such as clustering and : >> availability. Our intention is to have our first Compaq= >> Itanium-based systems for Tru64 UNIX and OpenVMS available ; >> in 2003 for early ISV testing and generally available in  >> 2004.  % Seems here it is what I asked for :-)   9 >>                                           Compaq's top < >> priorities for Alpha-based systems remain unchanged -- to= >> deliver improved performance and faster implementations of 9 >> the Alpha microprocessor to meet committed performance 5 >> enhancements, while continuing to provide the most 2 >> available, scalable systems at the lowest cost.2                                ^^^^^^^^^^^^^^^^^^^A I very hope that it could mean a widening on the hobbyist program B or an OpenVMS/Home use license at a low cost... Then we could tell= "Try OpenVMS from Compaq" to our friends, and show how to use @ it as well, tell about the security and the reliability - CompaqG won't be able to find such an efficient mean of adverisment I guess :-)   5 (let the goat suffice and the cabbage remain as well)   E Smposium: that seems to be a good idea, some people of this newsgroup B living in America could tell Compaq the concerns discussed here... Will be there someone?      Gaspar    ------------------------------   Date: 1 Jul 2001 16:26:07 GMT - From: djweath@attglobal.net (Dave Weatherall) % Subject: Re: Compaq switches to IA-64 5 Message-ID: <DTiotGxQ0bj6-pn2-VoijvijMo5Qv@localhost>   . On Fri, 29 Jun 2001 06:12:24, Vance Haemmerle " <vance@toyvax.Tucson.AZ.US> wrote:   > Duane Sand wrote:  > > 3 > > The chances of Intel adding Vax format is zero. K > > And you don't want systems built out of special-order customized chips. E > > You want systems built from the same high-volume pool of chips as  > > everyone else that year.4 > > So Vax fp will need to be done by some software. > H >   That's gonna suck.  One nice thing about the VAX to Alpha transitionJ > was that data files could stay the same even though the executables wereG > different.  Now, with Itanium not supporting F, D and G floats, it's nG > almost as bad as having to switch endian-ness.  Will VAX fp really be F > emulated in software?  Compaq gave up porting Java to VAX because ofI > the horrible performance emulating IEEE on VAX.  How bad is it going tos! > be emulating VAX floats on IPF?   E Well this is something that would/will complicate the lives of those .F people who gather information with Vax (We still use ELN) and display C it on VMS. It works perfectly now with no IF/ELSE/ENDIFS in sight.  F This implies one reason that iVMS will _not_ be compile and go, as it ! was, for the most part, on Alpha.   E This is another 'well if we have to change maybe we change away from e Compaq/VMS...'  F Yes I know its only a small thing and I know there are ways around it D and I could prepare my Alpha VMS code well in advance but  ... will 8 the people who pay the bills see it the same simple way?  F On the other hand, trying to be +ve, if in addition to iVMS there are F IA64 VME boards that can _boot_ and run iVMS or IA64/ELN (some hope!!)E then I could possibly sell it as it would give us a way forward. The M+ only problem would be Compaq's credibility.   D I have to confess a bias against Intel. It started with the 8080. I A preferred the 6800 and Z80. All my PCs have AMD inside. Even our  = Rainbows used NEC V20s (apart from the one with the Suitable hC Soluttions 286 board). I did have to bite the bullet for my wife's wA machine tho'. AMD didn't do MP cpus then so I had to build her a eC Dual-P3 for her animation work. It is quick though. When she first tE started we considered an Alpha but then the 3d/Max port was dropped. e Probably just as well...  --  Cheers - Dave.  E PS. I just read a couple of other messages and wanted to stress that 4" the above is a real-time system...   ------------------------------   Date: 1 Jul 2001 09:35:54 -0500e9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) M Subject: Re: Compilers go to Intel, hobbyist, CSLG, etc. goes out the window?e3 Message-ID: <t$OY720dA3f9@eisner.encompasserve.org>   a In article <9hlbj6$keb@gap.cco.caltech.edu>, mathog@seqaxp.bio.caltech.edu (David Mathog) writes: H > It would be nice if once Q managements' eyes stop spinning they could I > clarify the repercussions on the hobbyist and education programs of thes, > sale of their compiler divisions to Intel.  B Where does it say that "compiler divisions" will be sold to Intel.@ Everything I have read indicates Intel will be getting GEM, with& Compaq still making the VMS compilers.   ------------------------------  % Date: Sun, 01 Jul 2001 16:44:34 +0100o+ From: John A Fotheringham <jaf@jafsoft.com> B Subject: Free text-to-HTML converter AsctoHTM 4.0 for VMS released8 Message-ID: <p5hujt88qn4r2t3haldeu17hqjrillj7r9@4ax.com>  A JafSoft Limited in collaboration with Yezerski Roper Limited havei= released AscToHTM 4.0 for VMS, a program that converts plain e@ text files (.txt) into good looking HTML web pages in seconds.    = Version 3.0 was on the Compaq freeware disk.  This is a majorhA update to that version, although initially only an AXP version isf	 availablew  - home page : 	http://www.jafsoft.com/asctohtm/t9 documentation :	http://www.jafsoft.com/doco/docindex.html,7 VMS page : 	http://www.jafsoft.com/asctohtm/a2hvms.html  AXP download  : ? http://www.jafsoft.co.uk/download/openvms/a2h400axp.zip (2.2Mb) ( YRL freebies : http://www.yrl.co.uk/free  A This software is free to VMS users, and available commercially tosD Windows ($40) and (in the near future) Linux users.  The software isA developed in C++ under VMS and then ported to the lesser OS's :-)e   Cheers, Jaf      AscToHTM 4.0 Press release --------------------------= AscToHTM converts text files using advanced text-recognition u@ algorithms that detect the document's structure.  It recognizes C headings, bullets, lists, indentation, URLs, tables, ASCII art and mD much more.  The resulting HTML faithfully reflects the structure of  the original document.  E This easy-to-use program adds value to the HTML e.g. by turning URLs oC into hyperlinks and adding color to your pages.  It can generate a aD hyperlinked contents list from your headings, split large files intoE a set of linked pages, or place your HTML in a set of frames.  These r? extra features make the software ideal for use as an authoring e" tool as well as a conversion tool.  E Novice users will appreciate how easy it is to quickly generate HTML pC - just choose the files to be converted, press a single button and r, the results are displayed in your browser.    C Power users will love the control that they can exercise over such wC things as title, colors, <META> tags, style sheets and a myriad of  A other aspects of HTML production.  For efficiency, your selected cC options may be saved in "policy files" for reuse next time you run a the program.  F Whether you are a business user that wants to place key documents and D reports on your Intranet, or a budding author who wishes to publish E your prose as good-looking e-books, this software will save you many  E hours of effort.  The extensive documentation will guide you through n the program's many features.  B Version 4.0 introduces support for frames generation; copying the ? HTML generated directly to the windows clipboard; and allowing e> user-supplied HTML fragments to be used to customize the HTML  generated by the program.      -- John A Fotheringham + Visit my search engine and web robots pagesb% http://www.jafsoft.com/searchengines/b   ------------------------------  $ Date: Sun, 1 Jul 2001 12:25:49 -0500) From: "T. S. Murphy" <murphts@swbell.net>s) Subject: Re: Full port of VMS to Itanium.e. Message-ID: <z2J%6.14$Y36.27252@nnrp3.sbc.net>  2 "Bill Todd" <billtodd@foo.mv.com> wrote in message" news:9hgb60$4dm$1@pyrite.mv.net...  L > I'm no expert, but I'd have to see explicit examples (e.g., non-privilegedF > instructions in some generation that aren't supported in some futureJ > generation) before I believed this.  I've got XT DOS software that still- > runs under Pentiums and Win9x, for example.r  G Intel has retired one instruction ever, which was the move to/from test K registers instruction. It was present in the 486 and retired in the PentiumSL processor. Chapter 17 of the IA-32 manual volume 3 details the architecturalJ differences between the different generations, and there really isn't very much.e  C > New instructions really do constitute features, not objectionableoK > incompatibilities:  it's *upward* compatibility that's usually important,'J > and AFAICT Intel has always done an excellent job of maintaining it.  IfJ > *downward* compatibility was important to any application, it could haveJ > achieved it simply by not using the newer instructions (and watching itsE > memory requirements) - but that's not an Intel responsibility but a  software > one.  L I agree. If Intel had to maintain downward compatibility, no processor couldI ever support 32 bit addressing, protected mode, floating point, or any ofnI the number of features which Intel has added since the 8086. All of these.L features unquestionably make the processor better (unless you only need 640KG of memory, enjoy programming with 64k segments, and don't need floating I point), and Intel would have promptly gone out of business if they didn't " keep improving their architecture.  F Similarly, Microsoft can't really support old hardware if they want toG improve their products. Microsoft gets a lot of flack since they didn'toH support protected mode until Windows 95, about ten years after the firstI processor which supported it shipped. But they couldn't have supported itaH until the majority of consumers had switched to the new architecture, orC else they would alienate their potential customers. Windows 2000 is I inarguably many, many orders of magnitude superior to Windows 3.0, but it J couldn't have been built if it also had to run on an old XT (which was one of the requirements of 3.0)h  D A lof of vendors are worse: you can't run MacOS X even on one of the4 original iMacs which are only a couple of years old.   ------------------------------  $ Date: Sun, 1 Jul 2001 12:46:25 +0200" From: "Hans Vlems" <hvlems@iae.nl>3 Subject: Re: Hobby LIcense: One USER, or one LOGIN? ( Message-ID: <9hmut3$4l7$1@news.IAEhv.nl>  D Assuming that the hobbyist license uses LMF, it depends on what type of VMS license you have. Is it an A, B, C or D license?   Hans. Mark Vance <mvance@iglou.com> wrote in message# news:3B3DF757.FA66C5EE@iglou.com...dE > Does the hobbyist license confine you to ONE USER at a time, or one F > LOGIN of one user at one time.  Id sure like to be able to login andD > stop/id an interactive proc, instead   of turning the damn machine > off...Thanks...  >  >1 > Mark.  >e >n   ------------------------------  $ Date: Sun, 1 Jul 2001 12:26:51 +0200" From: "Hans Vlems" <hvlems@iae.nl>D Subject: Re: I  (of all people) just got a letter from Rich Marcello( Message-ID: <9hmtoe$2rd$1@news.IAEhv.nl>  * Correct. In Europe we got it by snailmail.  = Terry C. Shannon <terryshannon@mediaone.net> wrote in messagec5 news:6It%6.745$9r6.1232620@typhoon.ne.mediaone.net...  >n< > "Joe Heimann" <heimann@nog.ecs.umass.edu> wrote in message' > news:9hlkh7$i0$1@odo.ecs.umass.edu...l' > > Joe <cstranslations@msn.com> wrote: $ > > > I just got an email from Rich. >r	 > > > Joes > > J > > I got the same message, looks like it may have been sent to all DECUS/$ > > Encompass registered membership. > I > Yep, this appears to be the case. Given the timing of the announcement,eG > CETS2001 will likely be the earliest opportunity to get all the luridp" > details on the Alpha-->IPF move. >0J > Of course, this throws the CETS2001 content development team in overtimeC > mode, as the existing session schedule will have to be overhauled 	 big-time!d >  >h   ------------------------------   Date: 1 Jul 2001 09:39:33 -0500 9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)tD Subject: Re: I  (of all people) just got a letter from Rich Marcello3 Message-ID: <xEgeewLGMBDj@eisner.encompasserve.org>i  [ In article <9hlkh7$i0$1@odo.ecs.umass.edu>, Joe Heimann <heimann@nog.ecs.umass.edu> writes:a  H > I got the same message, looks like it may have been sent to all DECUS/" > Encompass registered membership.  F I will keep an eye on the mailbox, but so far we just know it has beenE sent to registered members who have allowed their email address to bev used.S  N ==============================================================================N Great Inventors of our time: Al Gore -> Internet; Sun Microsystems -> ClustersN ==============================================================================   ------------------------------   Date: 1 Jul 2001 09:40:37 -0500y9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)uD Subject: Re: I  (of all people) just got a letter from Rich Marcello3 Message-ID: <SJ9d5sx5aGmV@eisner.encompasserve.org>   M In article <9hmtoe$2rd$1@news.IAEhv.nl>, "Hans Vlems" <hvlems@iae.nl> writes:o, > Correct. In Europe we got it by snailmail.  < Perhaps next week it will make it to Massachusetts via mail.   ------------------------------  % Date: Sun, 01 Jul 2001 03:29:20 -0400o( From: Hamlyn Mootoo <univms@bigfoot.com>  Subject: Re: IA64 Rocks My World+ Message-ID: <3B3ED150.FEC94D03@bigfoot.com>n  D Digital, in the mid- 80's had no aversion whatsoever to "obsoleting"D their own systems.  And I can only assume that it WAS marketing, notG engineering who made these decisions.  One model would come out, then 5 G months later, something far faster and cheaper would arrive killing theiD revenue stream of the earlier product line. Intel made BILLIONS fromA knowing how to milk the revenue stream from the 286, 386, and 486fE architectures because THEIR marketing people didn't pop quaaludes foruB breakfast, and probably graduated high school.  I suspect the onlyG reason the pentium was released so quickly (too quickly) was that Intel < sensed competition from other processors.  In the absence ofH competition, it is prudent NOT TO KILL your own product lines.  And withF proprietary hardware in the mid 80's, as was the vogue for IBM and theB like, there often was no direct competition because they has theirH customers locked up. For customers, switching to another computer vendorE back then was simply too costly.  As a company, when you've done youraA R+D, and manufactured a product, you want as much sales from thattH product as possible, because there are no more initial costs so you justH get to keep selling the thing.  Ever heard of an operating system calledE DOS by a company called Microsoft?  Yes Virginia, there was in fact ahH version 6 of this product.  For at LEAST 7 years, DOS remained virtuallyE unchanged, while gates just pocketed the cash.  If it weren't for theeH likes of Xerox PARC and Apple, old Billy would be serving up DOS versionA 45.1 as we speak!  You want to be successful? Don't kill your ownaC revenue streams from products unless incipient competition makes iteC imperative.  Anyone want to buy the brooklyn bridge? I'm selling it ( cheap, and I've only got three left.  :)     HM   HM   Bengt Larsson wrote: > D > In comp.arch, Eric Smith <eric-no-spam-for-me@brouhaha.com> wrote: > S > >Casper.Dik@Holland.Sun.Com (Casper H.S. Dik - Network Security Engineer) writes:oC > >> Engineers shouldn't blame everything on marketing; Digital hasm= > >> made serious mistakes when it came to building hardware.- > >- > >Such as?e > " > - Not building large enough SMPs5 > - Not focusing on memory performance (esp. latency)lA > - Aversion to obsoleting their own systems (esp. the 8200/8400)    ------------------------------  # Date: Sun, 01 Jul 2001 06:37:30 GMTl From: newsuser@news.com (dondo)o  Subject: Re: IA64 Rocks My World8 Message-ID: <Xns90D11A7CBB6C6newsusernewscom@24.24.0.22>  - Alexis Cousein <al@brussels.sgi.com> wrote in % <3B3C85E8.7040700@brussels.sgi.com>:     >Brannon Batson wrote: >h? >> bengtl.net@telia.nospam.com (Bengt Larsson) wrote in messagen1 >> news:<3b3b4181.30126558@enews.newsguy.com>... a >> nE >>>In comp.arch, Eric Smith <eric-no-spam-for-me@brouhaha.com> wrote:n >>>  >>>eB >>>>Casper.Dik@Holland.Sun.Com (Casper H.S. Dik - Network Security >>>>Engineer) writes:  >>>>C >>>>>Engineers shouldn't blame everything on marketing; Digital hasM= >>>>>made serious mistakes when it came to building hardware.a >>>>>i  ; ... And every DEC computer - even I believe workstations -  = have at least had ecc on the cache since 1982.  Someone couldt learn from that... y   ------------------------------  % Date: Sun, 01 Jul 2001 10:49:12 -0400p2 From: rdeininger@mindspring.com (Robert Deininger)2 Subject: Re: One more dreadful thought to considerL Message-ID: <rdeininger-0107011049120001@user-2ive66e.dialup.mindspring.com>  8 In article <9hkr8b$psm$1@odo.ecs.umass.edu>, Joe Heimann" <heimann@nog.ecs.umass.edu> wrote:  . > Marty Kuhrt <kuhrt@encompasserve.org> wrote:  H > > And the rumor I heard was that Quantum only wanted the disk divison H > > and didn't want anything to do with the tape drive stuff.  Digital, L > > in typical "baby and the bath water" style, insisted that they take the H > > tape division, too.  Quantum may not be making money with disks, butF > > you have to think that they are cleaning up on tape drive margins.H > > Granted, this may be an unsubstantiated rumor, but is outside of the > > realm of believability?  > I > Quantum is no longer making any money on their disk division, they soldrG > it entirely to Maxtor.  Now they are just making DLT tape drives and sI > what they call "storage solutions" as well as integrated backup/archiveiG > systems.  Funny that the tape drives from Digital are what they endedtH > up making their money on out of the deal.  Of course they had some funE > in the last 6-9 months separating out the DLT and disk firmware and H > other developement groups.  Some of it was still being done at the oldG > DEC group in eastern Mass., forget which plant that was.  Quantum hadoF > to make sure they transferred the full build environment for the DLTG > firmware before they trabsferred the disk operation to Maxtor.  HeardmJ > some of this direct from a Quantum person handling part of the transfer.  I The DECie who invented the TK50 / DLT drives sure earned his or her pay. nJ Amazing longevity.  95 MB to whatever the current high end is, some of the parts haven't changed at all.    -- e Robert Deininger rdeininger@mindspring.comi   ------------------------------  # Date: Sun, 01 Jul 2001 15:26:45 GMTi4 From: "Terry C. Shannon" <terryshannon@mediaone.net>? Subject: Re: Question to Charlie Matco - reply to Terry Shannona; Message-ID: <ViH%6.828$9r6.1504902@typhoon.ne.mediaone.net>i  ? "Robert Deininger" <rdeininger@mindspring.com> wrote in message & news:rdeininger-0107011035320001@user- >AI > A clever VP of operations would be figuring out how to squeeze a betterrJ > deal out of Compaq on the next sale, using this week's news as leverage.> > You should be finding ammo for him to use, not relaying FUD.  K Well, here's an interesting piece of ordnance that can be used by GS-SeriesaL customers. The last US price list I saw (which was dated 25 May IIRC) showedL 731MHz and 1GHz GS-Series systems at the same price. Given the upgrade costsF associated with QBB and Firebox replacement, anyone who's got a 731MHzH system on order now should tell CPQ to deliver a brand spanking new 1GHz machine at the same price.   ------------------------------  % Date: Sun, 01 Jul 2001 10:35:32 -0400 2 From: rdeininger@mindspring.com (Robert Deininger)? Subject: Re: Question to Charlie Matco - reply to Terry ShannontL Message-ID: <rdeininger-0107011035320001@user-2ive66e.dialup.mindspring.com>  = In article <d56d1c2d.0106300924.120a311b@posting.google.com>,s# cstranslations@msn.com (Joe) wrote:A    I > I didn't personalyy say anything to the VP of operations until ThursdayhI > afternoon. His response was centered around "would you spend $500,000 -rC > $1,000,000 on something for which will be unsupport in 2 years?" -  C Well, you did a lousy job of informing this VP if that's what he isdF thinking.  Compaq has already said they will continue to support alphaC systems for much more than 2 years.  In 2 years, they will still be@F developing them (though maybe near the end of development). If you buyC $1,000,000 worth of alphaservers now, normal Compaq support will bebE available for the normal service life of the system.  They won't even D start trying to shove you onto an intel-based system until those are production-quality.d  G A clever VP of operations would be figuring out how to squeeze a betteroI deal out of Compaq on the next sale, using this week's news as leverage. i< You should be finding ammo for him to use, not relaying FUD.   -- - Robert Deininger rdeininger@mindspring.com0   ------------------------------  $ Date: Sun, 1 Jul 2001 21:29:04 +0800" From: "Kenneth" <best@hotmail.com> Subject: RMS file performance40 Message-ID: <9hn8j7$eh61@imsp212.netvigator.com>  H Beside file fragmentation, what statistic should I check to for the poor' performance of the index type RMS file?    ------------------------------   Date: 01 Jul 2001 04:45:34 GMT! From: jlahman1@aol.com (Jlahman1)a0 Subject: Re: Thanks Compaq for the new business!: Message-ID: <20010701004534.18080.00000687@ng-mm1.aol.com>   This is not a troll!  @ Maybe you should wake up and see what is really going on here.    N I'm sorry that I can't divulge our customers, but needless to say they are all9 very large automobile companies (domestic and foreign).  x  C I was at one of our customer's site on Monday (a very large tractorrO manufacturer).  They have many, many VMS (VAX and Alpha) systems.  For example, + their material flow system runs on a VAX.  e  J It was very interesting to see their reaction to the announcement: "VMS is
 dead, too".  M  J They wanted to convert the material handling system to Alpha/VMS.  Now, inN their minds, this is no longer an option (even with the promise to convert VMSO to IA64).  They see that their only solution is to convert their VMS systems totH windows NT/2K.  And, their preferred PC vendor is Gateway, not Compaq.    M I received a call from an old friend of mine at a steel company.  He was in a.I panic.  What is he supposed to do now?  The decisions he made in the past K concerning the supervisory automation systems were all based on the promise M that Compaq will enhance and support VMS.  This may will be true, but the wayxM Compaq has suddenly dropped Alpha made him feel very uncomfortable.  Being in L charge of the automation systems. Compaq has mad he job much more difficult.  N Our customers need answers and assurances for they do not like to operate in aJ world of uncertainty.  And, this is what Compaq has provided them with the  announcement of killing Alpha.    O It appears to me that the management at Compaq do not have their fingers on thetO pulse of their customer base.  This seems to be strictly a decision made by thesK beancounters without regard to keeping their own customer base satisfied.      ------------------------------  % Date: Sat, 30 Jun 2001 21:03:52 -0400y' From: "Bill Todd" <billtodd@foo.mv.com>h Subject: The Alpha/IA64 Hybrid( Message-ID: <9hlsnc$om2$1@pyrite.mv.net>  I In an attempt to climb out of the hole it jumped into last Monday, CompaqrE has apparently been energetically leaking information (from employeesfD purporting to speak on their behalf) about Intel's having internallyJ perceived the error of its ways and now wishing to fix the problems in theK IA64 architecture by enlisting much or all of the Alpha architecture in the K service of the IA64 instruction set.  Particularly explicit quotes include:g  J From a Compaq employee in comp.os.linux.alpha (6/26 4:48 A.M. EDT and 6/28 4:03 A.M. EDT):M  I "Alpha is not dead, its technologie will be implemented in the Itanium v3sC processor of intel which will be released end of 2003, early 2004."o  L "All what was planned to be in the EV8 chip will be implemented in the intelI itanium v3.  The engineers who are developing the alpha are transfered too intel."e  K And in comp.os.vms from a non-employee reporting what he was told by CompaqeJ engineers (and Mark Gorham, who may be the new head VMS honcho - I've lost track):a   6/28 9:28 P.M. EDT:i  K "With Alpha engineers working on Itanium-II (read: fixing what's wrong withoL Itanium-I with Alpha technology) ... Compaq is helping Intel design the nextL generation of IA64 which will be based on Alpha technology and optimized for Compaq operating systems."   6/29 9:11 A.M. EDT:e  L "Simply put, Intel has finally admitted, at least privately, they don't knowI how to design a 64 bit processor.  They need decent engineers, and that'saI where the xfer of Alpha engineers to Intel fits ... That's the EV8 team's.J task.  Build a better Itanium than anything Intel or HP could, and call it Itanium-II."  J See also 6/28 8:28 P.M. EDT for more near-quotes-from-Compaq-types leadingF up to the above.  There were other earlier similar (but less explicit)J allegations by others, but I didn't have the patience to go back and trackK them down.  The Inquirer also took note of such activity in an article lastr	 Thursday:h  ' http://www.theinquirer.net/28060117.htms  G "Today's spin from Compaq employees has ranged from assertions that thetA Alpha line was incapable of further performance increases, to the4E reassurance that the next generation of IA64 chips will be re-brandedp Alphas."  D Leaving aside the fact that none of this will necessarily help avoidE increased doubts about Compaq's commitment to anything non-Wintel andmF short-and-medium-term negative impact of the coming platform-switch onJ Alpha, VMS, and Tru64 sales, there's still the interesting question of howL Intel and HP feel about such allegations about the inadequacy of their shinyL new IA64 architecture.  Even if the allegations are true, reports that IntelL is now going to start over from an Alpha base won't exactly be good news forI Itanium/McKinley acceptance over the next couple of years, so what's goodeK for Compaq's ability to rationalize its divestiture of Alpha may be bad fortJ Intel/HP business, and vice versa.  And if the allegations are false, then> one might expect Intel and HP to be more than a little miffed.  L I took a quick look today at comp.sys.hp.hardware and comp.sys.hp.hpux - notL a peep there about anything Alpha-related.  Comp.sys.intel discussion seemedE to center on how Compaq needed Alpha to maintain credibility in theiriL high-end business - i.e., surprisingly positive sentiments about Alpha (fromL a group with their name), even before the announcement was official and they& knew they could afford to be generous.  I So I clearly haven't done any exhaustive research (and have no idea whereOH such discussion might hang out), but so far haven't stumbled upon a peepK from Intel, HP, or anyone purporting to speak on their behalf on the mattersH of Compaq's (alleged) allegations of IA64 inadequacies and its rescue byD Alpha technology.  If anyone else has, could they provide a pointer?   Thanks,I   - bill  L [I don't usually cross-post this widely, but I did take care to include onlyL groups where active discussion of Alpha's demise both seemed appropriate and had already occurred.]   ------------------------------  # Date: Sun, 01 Jul 2001 01:31:37 GMT 4 From: "Terry C. Shannon" <terryshannon@mediaone.net>" Subject: Re: The Alpha/IA64 Hybrid; Message-ID: <Z3v%6.748$9r6.1258838@typhoon.ne.mediaone.net>   2 "Bill Todd" <billtodd@foo.mv.com> wrote in message" news:9hlsnc$om2$1@pyrite.mv.net...K > In an attempt to climb out of the hole it jumped into last Monday, CompaqiG > has apparently been energetically leaking information (from employeesnF > purporting to speak on their behalf) about Intel's having internallyL > perceived the error of its ways and now wishing to fix the problems in theI > IA64 architecture by enlisting much or all of the Alpha architecture ine the D > service of the IA64 instruction set.  Particularly explicit quotes include: >-L > From a Compaq employee in comp.os.linux.alpha (6/26 4:48 A.M. EDT and 6/28 > 4:03 A.M. EDT):l >tK > "Alpha is not dead, its technologie will be implemented in the Itanium v3gE > processor of intel which will be released end of 2003, early 2004."$  C Now that's inconsistent with the roadmap showing convergence in the  POST-Madison era.m   >eH > "All what was planned to be in the EV8 chip will be implemented in the inteliK > itanium v3.  The engineers who are developing the alpha are transfered to 	 > intel."i  H Now that's a tall order. EV8 is (well, was) by no means a done deal, andL would not have been a done deal until 2004 at the earliest. Since IIRC IntelH has shown McKinley first silicon, Intel's already working on Itanium V3, right?   >=F > And in comp.os.vms from a non-employee reporting what he was told by CompaqL > engineers (and Mark Gorham, who may be the new head VMS honcho - I've lost	 > track):g >s > 6/28 9:28 P.M. EDT:  >jH > "With Alpha engineers working on Itanium-II (read: fixing what's wrong withI > Itanium-I with Alpha technology) ... Compaq is helping Intel design theh nextJ > generation of IA64 which will be based on Alpha technology and optimized fori > Compaq operating systems." >a > 6/29 9:11 A.M. EDT:v >mI > "Simply put, Intel has finally admitted, at least privately, they don't0 knowK > how to design a 64 bit processor.  They need decent engineers, and that'seK > where the xfer of Alpha engineers to Intel fits ... That's the EV8 team'soL > task.  Build a better Itanium than anything Intel or HP could, and call it > Itanium-II."   Rest of breast-beating omitted.t   >dF > Leaving aside the fact that none of this will necessarily help avoidG > increased doubts about Compaq's commitment to anything non-Wintel andlH > short-and-medium-term negative impact of the coming platform-switch onL > Alpha, VMS, and Tru64 sales, there's still the interesting question of howH > Intel and HP feel about such allegations about the inadequacy of their shiny H > new IA64 architecture.  Even if the allegations are true, reports that IntelaJ > is now going to start over from an Alpha base won't exactly be good news foroK > Itanium/McKinley acceptance over the next couple of years, so what's goodaI > for Compaq's ability to rationalize its divestiture of Alpha may be bad  foroL > Intel/HP business, and vice versa.  And if the allegations are false, then@ > one might expect Intel and HP to be more than a little miffed. >l  B Yes, one might. If the comments from CPQ employees are deliberate,L sanctioned leaks, they certainly are not well-orchestrated nor well-crafted!   ------------------------------  # Date: Sun, 01 Jul 2001 02:27:10 GMTh< From: apkenned@removethis.telalink.net (Alan P. Kennedy, Sr)" Subject: Re: The Alpha/IA64 Hybrid% Message-ID: <m3ae2pxu2w.fsf@spam.net>o  E Really nice try, but FUD does not you no where on Usenet. However, asoD a minor league player which I am, therefore I will have no effect onE compaq's bottom line. Compaq does not see what it has done. Customersn5 want and need continuity even in the computer market.m  E Please no flames intended but if I really wanted, I could pull out mynB old OS/2 disk and boot it up and run it on i386 hardware. The realC shocker is that it would have support for the current hardware. Not @ exactly the same thing as hardware but the point is folks wantedF alphas and not IA-64's, no matter if they are IA-64->alpha++. A little1 redirection there, and in the future so it seems.e  F Whatever as I have learned long time ago once the ship leaves port youF better get on board or get another ship. Because the decision has beenC made and there is no turning back. Well this is more like a rocket.l= You can't recall a launch. Hay, sounds like a good reason for4C anti-ballistic missile to some, more like a jobs program to others.7  D When customers that pay with real money expect continued support andF don't get it they are pissed no matter what the reason. Bottom line or' whatever, and folks have a long memory.   A The question is weather compaq is looking out for compaq or their D customers? Compaq should be looking out for both. Who knows it could be both. Time will tell. s  F However, it was expected that some of these risc processors would biteC the dust sooner or later. Dec failed to market the alpha and compaqaB failed to market the alpha. So I guess the problem was the alpha.   E Hay, it really does not matter in the long run. Folks will compute ondC anything :-)). I sure will even if it is a 8086. Yes, I have one of D those also no hard disk two floppies! I did like my alpha but that'sD life. This is not about liking anything. Compaq better better make aA profit or else!!! Now the question that I cannot answer was alpha9
 making money?m   Later,   Alan  7 >>>>> "Bill" == Bill Todd <billtodd@foo.mv.com> writes:l  A  Bill> In an attempt to climb out of the hole it jumped into lasti?  Bill> Monday, Compaq has apparently been energetically leakingu?  Bill> information (from employees purporting to speak on theiraE  Bill> behalf) about Intel's having internally perceived the error of ?  Bill> its ways and now wishing to fix the problems in the IA64nF  Bill> architecture by enlisting much or all of the Alpha architecture?  Bill> in the service of the IA64 instruction set. Particularly   Bill> explicit quotes include:n <snip the propaganda>>   ------------------------------  % Date: Sat, 30 Jun 2001 22:44:21 -0400a' From: "Bill Todd" <billtodd@foo.mv.com>s" Subject: Re: The Alpha/IA64 Hybrid( Message-ID: <9hm2jc$rvr$1@pyrite.mv.net>  I "Alan P. Kennedy, Sr" <apkenned@removethis.telalink.net> wrote in messagee news:m3ae2pxu2w.fsf@spam.net...j   ...i  C >  Bill> In an attempt to climb out of the hole it jumped into lastaA >  Bill> Monday, Compaq has apparently been energetically leakingsA >  Bill> information (from employees purporting to speak on theiroG >  Bill> behalf) about Intel's having internally perceived the error oftA >  Bill> its ways and now wishing to fix the problems in the IA64.H >  Bill> architecture by enlisting much or all of the Alpha architectureA >  Bill> in the service of the IA64 instruction set. Particularlyr! >  Bill> explicit quotes include:t > <snip the propaganda>>  K While some of the rest of your post had at least some merit, I suspect mostrJ people tend to be able to differentiate better between documented evidenceK and propaganda.  Some of the quotes came from Compaq employees, others were.E attributed to Compaq employees by the poster - and all were carefully  attributed.l  K The main reason I included them was because I suspected that people who had-K not seen them would simply assume they hadn't occurred or had been far less<H explicit (I might have myself:  they're pretty far-out).  In combinationL with other similar material that has appeared directly from (alleged) CompaqH employees and from people who allege to have talked directly with CompaqI employees, a real pattern does seem to be present, as The Inquirer noted.   K At any rate, since the issue has been raised (repeatedly and explicitly), I,J was curious whether any response had occurred that I hadn't seen.  I don'tK question that there may well be FUD involved:  I'd simply like to be bettert% able to determine which side it's on.T   - bill   ------------------------------  % Date: Sun, 01 Jul 2001 15:44:44 +0200S, From: Toon Moene <toon@moene.indiv.nluug.nl>" Subject: Re: The Alpha/IA64 Hybrid4 Message-ID: <3B3F294C.90F71085@moene.indiv.nluug.nl>   Bill Todd wrote:  I > "Today's spin from Compaq employees has ranged from assertions that the C > Alpha line was incapable of further performance increases, to the0G > reassurance that the next generation of IA64 chips will be re-brandedt
 > Alphas."  G Well, this would be quite nice.  Has anyone have any success in suing a1D company for explicitly incorrect statements covered by an NDA (or is that what NDA's are for ;-)>  , In that case, my employer could get rich ...   -- tG Toon Moene - mailto:toon@moene.indiv.nluug.nl - phoneto: +31 346 214290n6 Saturnushof 14, 3738 XG  Maartensdijk, The NetherlandsG Maintainer, GNU Fortran 77: http://gcc.gnu.org/onlinedocs/g77_news.htmlaE Join GNU Fortran 95: http://g95.sourceforge.net/ (under construction)    ------------------------------   Date: 1 Jul 2001 08:54:55 GMTp( From: nmm1@cus.cam.ac.uk (Nick Maclaren)" Subject: Re: The Alpha/IA64 Hybrid0 Message-ID: <9hmogv$5np$1@pegasus.csx.cam.ac.uk>  ; In article <Z3v%6.748$9r6.1258838@typhoon.ne.mediaone.net>, 3 Terry C. Shannon <terryshannon@mediaone.net> wrote:m >sM >> From a Compaq employee in comp.os.linux.alpha (6/26 4:48 A.M. EDT and 6/28  >> 4:03 A.M. EDT): >>L >> "Alpha is not dead, its technologie will be implemented in the Itanium v3F >> processor of intel which will be released end of 2003, early 2004." >tD >Now that's inconsistent with the roadmap showing convergence in the >POST-Madison era.  C Or, if it is true, we can expect that project to slip in a way that-$ will make the Merced look prompt :-)  I >> "All what was planned to be in the EV8 chip will be implemented in thee >intelL >> itanium v3.  The engineers who are developing the alpha are transfered to
 >> intel." >sI >Now that's a tall order. EV8 is (well, was) by no means a done deal, and M >would not have been a done deal until 2004 at the earliest. Since IIRC IntelhI >has shown McKinley first silicon, Intel's already working on Itanium V3,  >right?D  D If not, Intel's board should start shooting the management, startingA from the CEO and working down until they have eliminated the dead @ wood.  Unless they are being COMPLETELY incompetent, Madison andD Deerfield are already being simulated and their successors are being mapped out.a  D The thing that puzzles me is the lack of all mention of HP.  ReportsB are that the original plans were that Intel and HP would alternateC with being lead designer, which would give the following 'roadmap':n  0 Merced    Intel designed, just in production :-)A McKinley  HP designed, taped out and being tested in real systemsd7 Madison   Intel (?) designed, currently being simulatedn4 Deerfield HP (?) designed, currently being simulatedC [ The last two were announced as having the same schedule, with oneV# for desktops and one for servers. ]tI Followons Intel/HP designed, being mapped out, but perhaps to be scrappedmC Alpha     Intel designed, either to follow on from the previous, ort1           to replace them (probably with a delay)n  C >Yes, one might. If the comments from CPQ employees are deliberate,dM >sanctioned leaks, they certainly are not well-orchestrated nor well-crafted!r  E They are almost certainly sanctioned, but my guess is that the CompaqcB management panicked, and insisted that some positive slants be putC out ASAP, without actually assigning anyone competent into plannings a suitable story.   * Unless they are EXTREMELY subtle sabotage!     Regards, Nick Maclaren,* University of Cambridge Computing Service,> New Museums Site, Pembroke Street, Cambridge CB2 3QG, England. Email:  nmm1@cam.ac.uk/ Tel.:  +44 1223 334761    Fax:  +44 1223 334679e   ------------------------------  $ Date: Sun, 1 Jul 2001 11:44:41 -0500) From: "T. S. Murphy" <murphts@swbell.net> : Subject: Re: The death of VMS has been greatly exaggerated- Message-ID: <_rI%6.8$Y36.17784@nnrp3.sbc.net>@  0 "John Vottero" <John@mvpsi.com> wrote in message) news:tjs68m3p89kn64@news.supernews.com...   J > I'm wrong?  Care to explain how a Xeon accesses memory beyond 4GB with a 32 > bit pointer?  K When you set the PAE (Physical Address Extension) bit in CR4, the PTE's are)J 64 bits instead of 32 bits. With 32 bit PTE's, there are 20 bits availableD for the physical page address + 12 bits for the page offset, so yourJ translation is 32 bits. The PAE adds an additional 32 bits to the PTE, theH bottom 4 of which are used to extend the physical base address (32 + 4 =D 36). It would be trivial to extend this to up to 64 bits of physicalH addresss space since the top 28 bits are reserved. You are limited to 32K bits of virtual address space per page table (which basically boils down towG a 32 bits address space per process). The PAE is useful when you have atL bunch of process which together take up more than 4GB, but not useful if you+ have one process which needs more than 4GB.e  E You can see the IA-32 developer manual volume 3, section 3.8 for morea detailed information.o   ------------------------------  $ Date: Sun, 1 Jul 2001 12:58:14 -0500) From: "T. S. Murphy" <murphts@swbell.net>s- Subject: Using SCSI controller on QBUS system.- Message-ID: <ZwJ%6.9$pk4.32808@nnrp2.sbc.net>e  L I have a SCSI QBUS controller which I am trying to switch from one system toJ another. Originally it was in a KA-655 system in a H9642 chassis, and I am( trying to switch it to a BA-123 chassis.  @ The controller is a CMD-440/TM board, and it is MSCP compatible.  K I moved the KA-655 board (along with RAM) and the controller along with oneEJ disk to the BA-123 system (those are the only three boards in the system).L From there, I can run the on-board software on the controller, and I can seeI the disk on the SCSI chain, as well as format it and test it. It shows merK the mapping between SCSI devices and DUx drives, and tells me that the disko2 on SCSI ID0 is DU0. Everything appears to be fine.  K However, I cannot see the controller/disk from the console. When I do "showrH dev" it returns right away with nothing. When I do "show uqssp" it againJ doesn't find anything. I do something like "set host/uq/man <csraddr>" andJ it recognizes the controller as a UQSSP device but finds no disks on it. IB try to do things like "b dua0", etc. but it doesn't find the disk.  L This same configuration (same CPU, controller, and disk) worked fine when itI was in the other system. I was able to see the disk from the console whenoL doing "show dev", and boot from it. In general what sort of configuration isK required to set up an MSCP controller? Should I just be able to see it fromtL the console? I don't understand what changed when I moved it from one system
 to the other.n  F I have two of th CMD-440/TM controllers, and the both exhibit the same& behavior in the new system. Any clues?   ------------------------------  + Date: Sun, 01 Jul 2001 08:52:55 -0600 (MDT)t" From: John Nebel <nebel@csdco.com> Subject: VMS 7.3 experiences? G Message-ID: <Pine.OSF.4.21.0107010849270.12112-100000@athena.csdco.com>b  I What experiences have people had with V7.3?  Is it too early to use it inwF a mixed-architecture and wide area production cluster, or can one just forge ahead?   Thanks.s  
 John Nebel   ------------------------------  % Date: Sun, 01 Jul 2001 08:56:21 -0700e! From: Tom Linden <tom@kednos.com>s! Subject: RE: VMS 7.3 experiences? 9 Message-ID: <CIEJLCMNHNNDLLOOGNJIKEFMCOAA.tom@kednos.com>o  I Well this may not be germane to your question, but I just put up 7.3 on an PWS in a smallL cluster (6.2 & 7.1 on AXP, 7.1 on VAX).  Two of the systems run phase IV and two Phase V'B (IV on 7.3, because it was easier o install)  I have a development environment.  I In light of the recent announcments I think that Compaq will have to do adE much better job of simplifying the installation.  For a knowledgeabledJ sysadmin I'm sure it is easy, but I had problems with decnet, tcpip5.1 and  integrating it into the cluster.  I It seems to me that it is entirely possible to automate the install as MSe has done with W2K.I For example, it might ask the question, is this being added to a cluster,3K and if you answer yes it would figure what it needed to do, what files weree needed where, and so on.  F Configuring and enabling ftp did not create the anonymous account, for example.  TCPIP J needs to be configured after cluster is setup, sysuaf imported, but it may be necessary8 to have, telnet running before doing so, as was my case.   > -----Original Message-----+ > From: John Nebel [mailto:nebel@csdco.com]a% > Sent: Sunday, July 01, 2001 7:53 AM  > To: Info-VAX@Mvb.Saic.Comr > Subject: VMS 7.3 experiences?o >x >e >oK > What experiences have people had with V7.3?  Is it too early to use it intH > a mixed-architecture and wide area production cluster, or can one just > forge ahead? >d	 > Thanks.r >a > John Nebel >i >o   ------------------------------   Date: 1 Jul 2001 16:07:54 GMT 1 From: JONESD@er6.eng.ohio-state.edu (David Jones)m! Subject: Re: VMS 7.3 experiences?c9 Message-ID: <9hnhsq$35$1@charm.magnus.acs.ohio-state.edu>0  F In message <Pine.OSF.4.21.0107010849270.12112-100000@athena.csdco.com>&   John Nebel <nebel@csdco.com> writes:J >What experiences have people had with V7.3?  Is it too early to use it inG >a mixed-architecture and wide area production cluster, or can one just 
 >forge ahead?o  J I've been playing with it on a standalone system from time to time, mostlyD so far to check out the extended file cache (XFC).  It seems to helpF performance, probably due in equal measure to its dynamic sizing (i.e.E more memory in cache) and it holding more closed files than VIOC.  ItaD still doesn't look like they were ready to support writeback howeverC and I can't get the volumes to not be set to VIOC compability mode.T  D For a production system, my biggest concern would be all the changesF they made to the C runtime.  They ought sell a separate source listing$ CD with just the C runtime listings.    < David L. Jones               |      Phone:    (614) 292-6929- Ohio State University        |      Internet:1L 140 W. 19th St. Rm. 231a     |               jonesd@er6s1.eng.ohio-state.edu: Columbus, OH 43210           |               vman+@osu.edu  1 Disclaimer: I'm looking for marbles all day long.f   ------------------------------  % Date: Sun, 01 Jul 2001 10:52:06 -0400i2 From: rdeininger@mindspring.com (Robert Deininger) Subject: Re: VMS V7.3 SPD ErroreL Message-ID: <rdeininger-0107011052060001@user-2ive66e.dialup.mindspring.com>  A In article <Pfa%6.154244$Mq.4633135@e420r-sjo3.usenetserver.com>,m dittman@dittman.net wrote:  B > The VMS V7.3 SPD lists the MicroVAX II in both the supported and/ > unsupported hardware list.  Which is correct?s  F I think we'd heard rumors about that system going off support.  Is theH minimum memory listed?  I think it's more than the maximum a MicroVAX III can have, using DEC-standard parts.  That would be enough to desupport, It think.  G It's also possible a 3rd party came up with a way to put more memory incG the system, and then VMS would work.  But it probably still wouldn't bei
 supported.   --   Robert Deininger rdeininger@mindspring.com    ------------------------------  # Date: Sun, 01 Jul 2001 17:34:50 GMTn From: dittman@dittman.neta Subject: Re: VMS V7.3 SPD ErrorfA Message-ID: <_aJ%6.185273$Mq.5477342@e420r-sjo3.usenetserver.com>d  3 Robert Deininger <rdeininger@mindspring.com> wrote:lC : In article <Pfa%6.154244$Mq.4633135@e420r-sjo3.usenetserver.com>,  : dittman@dittman.net wrote:  C :> The VMS V7.3 SPD lists the MicroVAX II in both the supported andu0 :> unsupported hardware list.  Which is correct?  H : I think we'd heard rumors about that system going off support.  Is theJ : minimum memory listed?  I think it's more than the maximum a MicroVAX IIK : can have, using DEC-standard parts.  That would be enough to desupport, Iu : think.  : The MicroVAX II CPU can have 16MB with DEC-standard parts.  I : It's also possible a 3rd party came up with a way to put more memory indI : the system, and then VMS would work.  But it probably still wouldn't bet : supported.  B The MicroVAX II can't have more than 16MB, even with a third-party	 solution.p -- i Eric Dittman dittman@dittman.net4   ------------------------------   Date: 1 Jul 2001 16:26:09 GMTr- From: djweath@attglobal.net (Dave Weatherall)aE Subject: Re: Wailing and moaning... (was: Question to Charlie Matco.)t5 Message-ID: <DTiotGxQ0bj6-pn2-sbLdYdUx45jd@localhost>   F On Thu, 28 Jun 2001 05:07:48, "Bill Todd" <billtodd@foo.mv.com> wrote:  
  <Snip...>  K > I suspect the industry already knows that IA64 archtecture member by someoE > other name, however, so it might be better to use that one to avoidhL > confusion (see below).  Unless, of course, the member is so distant in theF > future that no industry-known name yet exists for it - in which caseL > speaking of it with such specificity is a bit misleading (as is suggestingM > any particular time-frame in which it might actually appear so that VMS canm
 > run on it).   F These 2 posts of mine are about 3 days behind the thread now but this D event has generated so much traffic and I've still got 338 to go in  c.o.v alone...   My point :-I   What frightens me :t  F Compaq make a corporate decision to drop Alpha as a platform and adoptD IA64 for all their OS's but both Fred and Hoff imply that they have F only just _started_ to assess what is required for VMS. (Tru64 and NSK seem to be further on).o  B The NG is already debating mythical/possible changes to IA64 that ! might be required to support VMS.s  F However , there is a statement that it will take 18 months for iVMS toF be up-and-running and the whole thing ready for production in 3 years.A Who assessed the timescale, on what is it based? Who's doing the tF planning? What were the criteria and, crucially, what are the caveats  and contingiencies?i    -- >  , Cheers - Dave. (tapping on AMD running OS/2)  , Ok that was negative but it is On-Topic. :-)   ------------------------------  # Date: Sun, 01 Jul 2001 16:37:46 GMT 4 From: "Terry C. Shannon" <terryshannon@mediaone.net>E Subject: Re: Wailing and moaning... (was: Question to Charlie Matco.)c; Message-ID: <ulI%6.844$9r6.1535251@typhoon.ne.mediaone.net>-  : "Dave Weatherall" <djweath@attglobal.net> wrote in message/ news:DTiotGxQ0bj6-pn2-sbLdYdUx45jd@localhost...IH > On Thu, 28 Jun 2001 05:07:48, "Bill Todd" <billtodd@foo.mv.com> wrote: >  >  <Snip...> >hH > > I suspect the industry already knows that IA64 archtecture member by someG > > other name, however, so it might be better to use that one to avoidBJ > > confusion (see below).  Unless, of course, the member is so distant in theiH > > future that no industry-known name yet exists for it - in which caseC > > speaking of it with such specificity is a bit misleading (as isn
 suggestingK > > any particular time-frame in which it might actually appear so that VMSr cano > > run on it).t >uG > These 2 posts of mine are about 3 days behind the thread now but this E > event has generated so much traffic and I've still got 338 to go ine > c.o.v alone... >p
 > My point :-t >m > What frightens me :r >eH > Compaq make a corporate decision to drop Alpha as a platform and adoptE > IA64 for all their OS's but both Fred and Hoff imply that they havewH > only just _started_ to assess what is required for VMS. (Tru64 and NSK > seem to be further on).n >aC > The NG is already debating mythical/possible changes to IA64 that # > might be required to support VMS.d > H > However , there is a statement that it will take 18 months for iVMS toH > be up-and-running and the whole thing ready for production in 3 years.B > Who assessed the timescale, on what is it based? Who's doing theG > planning? What were the criteria and, crucially, what are the caveatss > and contingiencies?h  H I suspect that a lot of the plotting and scheming is being done by SteveH Jenkins and his merry band in the newly-reorganized AlphaServer Platform Group.   ------------------------------  # Date: Sun, 01 Jul 2001 05:10:31 GMTl. From: "mulp" <michaelpettengill@earthlink.net> Subject: Re: What is lockstep ?-D Message-ID: <bhy%6.7814$ck5.790190@newsread2.prod.itd.earthlink.net>  < "Chris Townley" <news@townleyc.demon.co.uk> wrote in message; news:993948882.11027.0.nnrp-12.d4e45fa5@news.demon.co.uk...gD > Loads of posts talk about Tandem using lockstep - but what is it ? >gB > Sorry if I am being a bit naive, but never come across Tandem...  K Tandem is another company swallowed by Compaq, one that has survived so farcG with its name intact and even most of its products.  (A number of otherHL companies have been effectively flushed in name and product and technology.)  J Lockstep refers to running two or more circuits together with some sort ofI check circuit verifying that they are producting the same results.  FaultvG tolerant systems, such as Tandem's generally combine two CPUs which arejL tightly coupled running in lock step.  Hence the name Tandem.  If the resultJ the two differ, the pair is shutdown.  Using various other techniques, the pair is bypassed or replaced.-  K The circuits have to be very deterministic, or at least appear that way, toM operate in lockstep.   ------------------------------  # Date: Sun, 01 Jul 2001 02:27:32 GMT - From: "Richard L. Dyson" <rickdyson@home.com>m* Subject: Re: where should SYSUAF.DAT live?' Message-ID: <3B3E8A8E.378A3A0@home.com>o   Tom Linden wrote:. > I > I have copies in both SYSMGR and SYSEXE on 7.3 AXP.  By trial and error 7 > I found that AUTHORIZE picks it up out of the former.t  G As others have pointed out, the default location should be Sys$System. 0 However,J it can be moved around.  Believe me, you don't want more than one of these files . laying around or you will regret it someday...  M It is recommended that even when you have it in the default location (as wellt asK about a dozen other similar files like rights, proxy, mail, stuff etc.) youmI should also make logical name definitions to point to the files. (See thelL Sys$Manager:SyLogicals.Template file since at least v7.x for a list and more
 suggestions!)t  K This makes it possible to run AUTHORIZE and other programs without worryingn aboute= your current default being in Sys$System.  Very convenient!!!t  M This also allows you to offload these files to a commonly available disk whent youOM want/need to maintain only one set of UAF-related data for a cluster of nodesb thatJ all have their own system disk.  If you don't do this, I believe your only	 option iscJ to either copy around the data files to the other nodes or perform all the+ same operations on each node one at a time.i  M Imagine this in a multi-thousand user setup and you tell your users that they M have to change their password on EACH OF THE VMS NODES every time they changec it... :)   Regards, Rick   ------------------------------   Date: 1 Jul 2001 14:45:19 +0200  From: gazso@ludens.elte.hu4 Subject: Re: Why Compaq won't succeed on IA64 either! Message-ID: <KdSlQXGTLMXI@ludens>n  \ In article <3B3BEC34.E29F082B@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:N > Face the fact: Compaq's actions (or lack thereof) have demonstrated that VMSN > is not a key product to its strategy. It may be a cash cow right now but not > in the future.  3 In this case they might going to drop the price? :)y    Gaspar    ------------------------------  % Date: Sun, 01 Jul 2001 10:39:06 -0400e2 From: rdeininger@mindspring.com (Robert Deininger)4 Subject: Re: Why Compaq won't succeed on IA64 eitherL Message-ID: <rdeininger-0107011039060001@user-2ive66e.dialup.mindspring.com>  5 In article <3B3CCDA3.19299E9D@videotron.ca>, JF Mezeiy% <jfmezei.spamnot@videotron.ca> wrote:s  E > Look at how hard Marcello must have fought to get a trivially small  > advertising budget last year.o  G Yes, and look how much profit he brought to the bottom line.  It's damn1I clear that, in a period where Compaq was doing poorly overall, the profitaH from VMS-related (and Tru64-related) business was noticed by the CEO andC BOD.  Wiser folks would have noticed sooner, and would have noticeds6 alpha's contribution as well.  Better late than never.   -- o Robert Deininger rdeininger@mindspring.com8   ------------------------------  $ Date: Sun, 1 Jul 2001 14:18:19 +10000 From: "Dave Gaukroger" <bangalla@ozemail.com.au>8 Subject: Re: Yet Another Reason Why Windoze .ne. OpenVMS/ Message-ID: <nyx%6.131$GL5.7038@ozemail.com.au>i  0 "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote:  K > It will be most interesting to see if Microsoft will tolerate that Compaqo not3J > load Windows on every IA64 machine. Perhaps another reason VMS won't run onL > commodity hardware since that way, Compaq will have a clear defiition of aL > windows Pc and a enterprise machine and won't be forced to load windows on > enterprise machines.  J I don't see this as a significant issue. Proliant boxes don't ship with anJ OS and they support MS, Netware, Linux and SCO. Drivers also exist to loadF win95 on a Proliant (yes, the logic escapes me too) so the commodity /' enterprise issue gets a little blurred.i  K That said, one of the few positive things about x86 based operating systemscK is that I can load any of them onto a PC if I want to evaluate a particular K app or scenario. If Compaq is smart (which becomes more questionable by thenJ minute) they will ensure that VMS can run on any IA-64 that they sell (notL necassarily any IA-64 if they want to hang on to customers). This will allowK smaller operations to move to a real OS on their servers and encourage ISVsa to take VMS seriously.  L This now follows the argument about low cost alpha, people will be much moreA willing to port and develop for VMS if they can put it on a cheapiI development box. I'd like to be able to install VMS on a IA-64 Deskpro toeL evaluate an app the way I do with NT or 2K now on a PII. The company where IH work doesn't have alphas, I have a VAXstation I'm trying to learn VMS on. because we support some people who do use VMS.  I I guess my line of thought mirrors many others in the group - The port ofnK VMS to IA-64 is good, the death of Alpha is unnecassary and a technical and & PR disaster no matter how you spin it.   Dave   ------------------------------  % Date: Sun, 01 Jul 2001 07:15:55 -0700n! From: Tom Linden <tom@kednos.com>h8 Subject: RE: Yet Another Reason Why Windoze .ne. OpenVMS9 Message-ID: <CIEJLCMNHNNDLLOOGNJIIEFJCOAA.tom@kednos.com>   J The past is often prologue.  There was a version of the 21064 (A, I think)J that was fused in such a way that it could only run the ARC console, hence NT.o   > -----Original Message-----7 > From: Dave Gaukroger [mailto:bangalla@ozemail.com.au]t' > Sent: Saturday, June 30, 2001 9:18 PMr > To: Info-VAX@Mvb.Saic.Coma: > Subject: Re: Yet Another Reason Why Windoze .ne. OpenVMS >o >l >I2 > "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote: > A > > It will be most interesting to see if Microsoft will toleratea
 > that Compaqi > not L > > load Windows on every IA64 machine. Perhaps another reason VMS won't run > on? > > commodity hardware since that way, Compaq will have a cleare > defiition of aC > > windows Pc and a enterprise machine and won't be forced to load1 > windows on > > enterprise machines. >rL > I don't see this as a significant issue. Proliant boxes don't ship with anL > OS and they support MS, Netware, Linux and SCO. Drivers also exist to loadH > win95 on a Proliant (yes, the logic escapes me too) so the commodity /) > enterprise issue gets a little blurred.o >e; > That said, one of the few positive things about x86 basedC > operating systemsFB > is that I can load any of them onto a PC if I want to evaluate a > particular9 > app or scenario. If Compaq is smart (which becomes moreX > questionable by the L > minute) they will ensure that VMS can run on any IA-64 that they sell (notC > necassarily any IA-64 if they want to hang on to customers). Thise > will allow> > smaller operations to move to a real OS on their servers and > encourage ISVs > to take VMS seriously. >dA > This now follows the argument about low cost alpha, people willi > be much moreC > willing to port and develop for VMS if they can put it on a cheapiK > development box. I'd like to be able to install VMS on a IA-64 Deskpro to > > evaluate an app the way I do with NT or 2K now on a PII. The > company where IdJ > work doesn't have alphas, I have a VAXstation I'm trying to learn VMS on0 > because we support some people who do use VMS. >mK > I guess my line of thought mirrors many others in the group - The port ofe? > VMS to IA-64 is good, the death of Alpha is unnecassary and au > technical andg( > PR disaster no matter how you spin it. >g > Dave >r >  >n   ------------------------------  % Date: Sun, 01 Jul 2001 18:18:54 +0100 + From: "antonio.carlini" <arcarlini@iee.org>38 Subject: Re: Yet Another Reason Why Windoze .ne. OpenVMS' Message-ID: <3B3F5B7E.266EBF65@iee.org>t   Robert Deininger wrote:e > F > In article <CIEJLCMNHNNDLLOOGNJIIEFJCOAA.tom@kednos.com>, Tom Linden > <tom@kednos.com> wrote:i > N > > The past is often prologue.  There was a version of the 21064 (A, I think)N > > that was fused in such a way that it could only run the ARC console, hence > > NT.  > 3 > There was a 21064PC, which was probably NT-only. a  . I remember a 21*1*64PC and it was not NT only.  . There was a fuse that could have been blown to/ only allow NT-style memory management (at least * that's my recollection of what it did). I 0 don't believe any systems ever shipped with that0 fuse blown. Systems were shipped that would only+ run NT out of the box but that was achieved * by mangling the console software. With the, right tricks at the console you can unmangle. such systems and persuade them to run OpenVMS.   Antonio      -- r   ---------------h- Antonio Carlini             arcarlini@iee.org    ------------------------------   End of INFO-VAX 2001.362 ************************