1 INFO-VAX	Thu, 24 Jul 2003	Volume 2003 : Issue 405       Contents: Re: Alpha station boot problems  Re: Alpha station boot problems 5 Re: BACKUP/IGNORE=INTERLOCK (was: Re: OpenVms Backup) N Re: D.H. Brown: EV7 AlphaServers Deliver Enhanced RAS and Powerful PerformanceN Re: D.H. Brown: EV7 AlphaServers Deliver Enhanced RAS and Powerful PerformanceN Re: D.H. Brown: EV7 AlphaServers Deliver Enhanced RAS and Powerful PerformanceN Re: D.H. Brown: EV7 AlphaServers Deliver Enhanced RAS and Powerful PerformanceN Re: D.H. Brown: EV7 AlphaServers Deliver Enhanced RAS and Powerful PerformanceP Re: DEC-AXPVMS-VMS731_CDRECORD-V0100--4.PCSI v. DEC-AXPVMS-VMS731_PCSI-V0100--4.P Re: DEC-AXPVMS-VMS731_CDRECORD-V0100--4.PCSI v. DEC-AXPVMS-VMS731_PCSI-V0100--4.P Re: DEC-AXPVMS-VMS731_CDRECORD-V0100--4.PCSI v. DEC-AXPVMS-VMS731_PCSI-V0100--4.P Re: DEC-AXPVMS-VMS731_CDRECORD-V0100--4.PCSI v. DEC-AXPVMS-VMS731_PCSI-V0100--4.P Re: DEC-AXPVMS-VMS731_CDRECORD-V0100--4.PCSI v. DEC-AXPVMS-VMS731_PCSI-V0100--4.P Re: DEC-AXPVMS-VMS731_CDRECORD-V0100--4.PCSI v. DEC-AXPVMS-VMS731_PCSI-V0100--4. Re: Do ES45s run VMS 7.2-2F Re: Does RT-11 run on the PDP-11/70? (was Re: PDP-11 OS Release Dates) Re: EMC on VMS Re: EMC on VMS Re: EMC on VMS% Re: Finding installed images from DCL  Re: HP FUDBusting + Re: HP World: Why Alpha's Omega Makes Sense + Re: HP World: Why Alpha's Omega Makes Sense + Re: HP World: Why Alpha's Omega Makes Sense P Re: Intel's had Dirty Bombs Anthrax and everything! (Was: Re: HP World: Why Alph# Re: Mozilla... the new NULL process " RE: Packed decimal arithmetic in C" Re: Packed decimal arithmetic in C" Re: Packed decimal arithmetic in C" RE: Packed decimal arithmetic in C" RE: Packed decimal arithmetic in C" RE: Packed decimal arithmetic in C" Re: Packed decimal arithmetic in C" RE: Packed decimal arithmetic in C# Re: Sun Micro Profit, Revenues Fall  Re: Sybase client   F ----------------------------------------------------------------------  + Date: Wed, 23 Jul 2003 18:09:27 +0000 (UTC) 7 From: moroney@world.std.spaamtrap.com (Michael Moroney) ( Subject: Re: Alpha station boot problems( Message-ID: <bfmj0n$v79$1@pcls4.std.com>  H If control-C doesn't work, try control-P or the front panel halt button.   --   -Mike    ------------------------------  # Date: Wed, 23 Jul 2003 19:05:26 GMT # From: hoff@hp.nospam (Hoff Hoffman) ( Subject: Re: Alpha station boot problems0 Message-ID: <W%ATa.750$jf3.418@news.cpqcorp.net>  q In article <MDEGJFAOHGLNLHONIEBMGEDJCOAA.hvanderw@mansply.com>, "Hank Vander Waal" <hvanderw@mansply.com> writes: I :I have an Alpha station that was part of a cluster - doing a remote boot J :from the other system ...  I did not get that system :(  when this systemJ :starts up it tries to do a network boot and fails and keeps re-trying.  ID :can not control c or y out of it to change the boot to local drive.  8   Um, which particular AlphaStation model is involved?    I   Usually, you need to HALT the system using a BREAK on a serial console.   H   Depending on the particular AlphaStation and the front-panel settings,G   you might be able to halt the system using a front panel button, too.   G   (When operating with the graphics display, the front panel button is  F   the usual means I use to halt the system -- on some boxes, I have toE   switch a jumper to switch the button from "reset" to "halt", but as F   I don't know which AlphaStation is involved here...  Most boxes haveG   the ability of a serial console, too -- and input a BREAK key there.)   2   See the FAQ for some details on serial consoles.    N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com    ------------------------------  % Date: Wed, 23 Jul 2003 13:21:31 -0400 5 From: David Beatty <David.Beatty@qwertysasasdfgh.com> > Subject: Re: BACKUP/IGNORE=INTERLOCK (was: Re: OpenVms Backup)2 Message-ID: <xcIeP7yh4wXK5BYLtbRq7I=F9QuA@4ax.com>  F On Mon, 21 Jul 2003 21:37:20 GMT, hoff@hp.nospam (Hoff Hoffman) wrote:  b >In article <EeB0mFx4f4af@malvm7.mala.bc.ca.>, nothome@spammers.are.scum (Malcolm Dunnett) writes:2 >:In article <KyWSa.521$5X.420@news.cpqcorp.net>, ; >:    hammond@not@peek.ssr.hp.com (Charlie Hammond) writes:  >:   <snip>   > + >:What about the case of an Oracle database E >:where one wants to do a "hot backup". Of course one needs to follow G >:all Oracle's instructions about using ARCHIVELOG mode and setting the H >:tablespaces in BACKUP mode, but given all that it's still necessary to< >:use /IGN=INTERLOCK to actually copy the "hot" tablespaces. > F >  Please check with Oracle -- if Oracle is willing to assure you thatH >  there is nothing in cache and everything is out on disk on this, thenH >  the /IGNORE=INTERLOCK might well produce the desired results.  (WhileG >  I'd not generally expect a consistent file to be locked, there could H >  well be application-specific reason(s) to keep a lock on a file that & >  is consistent.  I don't know that.) > D >  Oracle Rdb certainly knows how to do this with the RMU tool, and D >  entirely avoids the requirements for /IGNORE=INTERLOCK.  Once RdbE >  tosses the data out, the resulting RMU/BACK archive files can then  >  be run through BACKUP.  >   F     In the case of Oracle, you still have to use the /IGNORE=INTERLOCK@ option on BACKUP because the file is still open.  When you set a< tablespace to START BACKUP, Oracle will not write changes to> that particular tablespace until you set the tablespace to END? BACKUP.  I can't remember the exact details, but on END BACKUP  D Oracle will either go back through the redo logs to write changes or? it saves changes to a rollback segment and writes them from the  rollback segment.    David R. Beatty    <snip>   ------------------------------  # Date: Wed, 23 Jul 2003 17:59:07 GMT 9 From: "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> W Subject: Re: D.H. Brown: EV7 AlphaServers Deliver Enhanced RAS and Powerful Performance 0 Message-ID: <L1ATa.739$6g3.285@news.cpqcorp.net>  K "Andrew Harrison SUNUK Consultancy" <Andrew_No.Harrison_No@nospamn.sun.com> ; wrote in message news:bfmatn$hlg$4@new-usenet.uk.sun.com...  > Fred Kleinsorge wrote:3 > > Buy a clue.  We'll all chip in a couple Euro's.  > > ' > > "Andrew Harrison SUNUK Consultancy" ' <Andrew_No.Harrison_No@nospamn.sun.com> ? > > wrote in message news:bfj0ir$cr6$1@new-usenet.uk.sun.com...  > >  > >  >  > 2 > Is that Fred the MBA or Fred the person who just > lost another argument. >   I Fred doesn't have an MBA.  Fred isn't a sales dweeb like Andrew the Troll  either.   3 > When will you learn Fred ? You role in this group 4 > is to make me look good, you seem to be doing this > for Bill as well keep it up. >   J Andrew, I learned long ago that it isn't worth arguing with you - you playF loose with facts and positions, and when cornered refuse to respond orJ continue to change the subject.  But I can hardly resist an occasional pot shot at you.  K The facts are that you haven't really a clue about the Itanium costs, sales C data, revenue or profit data, or how R&D is written off, or how P&L G statements are done.  For instance, you continue to quote "total sales" G numbers for Itanium that seem about the size of *specific sales* I have  seen.   L Tell you what.  How about you explain the financials for SPARC to us?  MaybeJ you have internal access to facts we don't have - just like you don't have1 access to facts that are internal to Intel or HP.    ------------------------------  % Date: Wed, 23 Jul 2003 13:18:19 -0700 3 From: Robert Klute <robert_klute_removethis@hp.com> W Subject: Re: D.H. Brown: EV7 AlphaServers Deliver Enhanced RAS and Powerful Performance 8 Message-ID: <onqthvsnc6jpvhrt1u91gpttd0dsuf50up@4ax.com>  E On Wed, 23 Jul 2003 15:59:36 +0100, Andrew Harrison SUNUK Consultancy . <Andrew_No.Harrison_No@nospamn.sun.com> wrote:      >What margins on Systems Sales ? > 6 >HP don't make a profit on systems sales currently and9 >historically they havn't either. Thats one of the things ) >that people were expecting Carly to fix.   1 Will you cut that out.  People in glass houses...   1 Sun, your favorite company, this last quarter had    Product Revenue		2,003 Product CoS		1,110 Product Gross Margin	  893  E Now, here is where you can argue with me, but let's assume 3/4 of R&D E was product R&D (I think it is a little low), as opposed to services, ; and that SGA was proportionl to revenue, so about 67%, then    Product R&D (est)	 361 Product SGA (est)	 554 Product Income		 (23)     E As for HPQ's current quarter, first, I don't know (and probably won't E till after you do); and, secone, even if I had some inkling, HP is in  it's quiet period.     Robert Klute Cupertino Solution Center  Hewlett-Packard Company  ----- ? The opinions are strictly those of the poster, not the company.    ------------------------------  % Date: Wed, 23 Jul 2003 18:07:34 -0400 * From: "Bill Todd" <billtodd@metrocast.net>W Subject: Re: D.H. Brown: EV7 AlphaServers Deliver Enhanced RAS and Powerful Performance 2 Message-ID: <o-2dnT7yP4ipmoKiXTWJig@metrocast.net>  3 "jlsue" <jefflsxxxz@sbcglobal.net> wrote in message 2 news:ta4thvcotv7hmi6i7ire682vqr39t5mt63@4ax.com...J > On Fri, 18 Jul 2003 18:01:15 -0400, "Bill Todd" <billtodd@metrocast.net> > wrote: >  > I > >While the idiots who killed EV8 don't seem to have had the wits to see  the K > >answer to that question before doing so, said answer is no longer in any E > >doubt.  The fall in combined VMS and Tru64 system profit after the 	 Alphacide H > >was announced was far greater than the stated cost of continued AlphaH > >development, so from a fiscal point of view EV8 would have been worth4 > >'pursuing' even if it had never actually shipped. > >  > L > While I agree that the fall in systems after the announcement hurt things,L > I don't know if that was planned-for or not - and I sincerely doubt you doC > either.  It's entirely possible that it was worse than originally 	 expected, ( > but that's water under the bridge now.  L The point, of course, was that a dramatic fall *should* have been expected -H by anyone other than the utterly incompetent (or those with other than aI strictly financial agenda who hoped they could obscure their real motives  with sufficient bluster).    > E > But there are no real numbers that you'd be able to find that would  "prove" * > this would have made EV8 worth pursuing.  $ Sure there are:  I've provided them.     There are many, many costsI > over-and-above what you may see in the Annual Report.  Groups like ESG, I > Tru64, OpenVMS, NSK, Customer Support, compilers, layered-products etc.  are I > just some examples.  And you've got to factor in the projected revenues  and / > margins, possible price-points, volumes, etc.   E No, I don't - because people like Rich Marcello have (at least if one K assumes minimal competence on *their* part) already done so.  Rich reported J that annual Alpha development costs ran $150 million (one might expect forJ both the EV7 and EV8 tracks combined, but even if you consider that figureI to have been purely EV8-related the conclusions don't change).  Rich also F reported that annual VMS system profit ran $800 million (in Y2K, on $4L billion in annual revenue, a figure which persisted through at least late Q1F of 2001).  Recently, Mark Gorham has been quoted as saying current VMSK annual system profit is $500 million on $2.5 - $3 billion in annual revenue G (and in December, 2001, a Compaq response to a Gartner report put VMS's J annual revenue at only $2 billion - so it, and likely profit as well, seemK to have rebounded a little from a low point shortly after the Alphacide but H to nowhere near the pre-Alphacide level).  And examination of the CompaqH quarterly report statements shows no significant decline in VMS revenues! before the date of the Alphacide.    > H > It is just far, far too complex for anyone on the outside to determine- > based on such little information available.   I That's why I've relied on the statements made by the insiders in the best . position to know what they were talking about.   - bill   ------------------------------    Date: 23 Jul 2003 17:26:11 -0500+ From: young_r@encompasserve.org (Rob Young) W Subject: Re: D.H. Brown: EV7 AlphaServers Deliver Enhanced RAS and Powerful Performance 3 Message-ID: <Vv7ebBqO+8Dc@eisner.encompasserve.org>   _ In article <o-2dnT7yP4ipmoKiXTWJig@metrocast.net>, "Bill Todd" <billtodd@metrocast.net> writes:  >    > G > No, I don't - because people like Rich Marcello have (at least if one M > assumes minimal competence on *their* part) already done so.  Rich reported L > that annual Alpha development costs ran $150 million (one might expect forL > both the EV7 and EV8 tracks combined, but even if you consider that figureK > to have been purely EV8-related the conclusions don't change).  Rich also H > reported that annual VMS system profit ran $800 million (in Y2K, on $4N > billion in annual revenue, a figure which persisted through at least late Q1H > of 2001).  Recently, Mark Gorham has been quoted as saying current VMSM > annual system profit is $500 million on $2.5 - $3 billion in annual revenue I > (and in December, 2001, a Compaq response to a Gartner report put VMS's L > annual revenue at only $2 billion - so it, and likely profit as well, seemM > to have rebounded a little from a low point shortly after the Alphacide but J > to nowhere near the pre-Alphacide level).  And examination of the CompaqJ > quarterly report statements shows no significant decline in VMS revenues# > before the date of the Alphacide.  >   E 	And imagine how much better their margins will be when all their R&D A 	for CPU is gone.  Sun did $482 million in R&D this last quarter. > 	They can't sustain that.  It isn't rocket science to conclude9 	a large portion of that $482 million is CPU related R&D:   i http://www.realworldtech.com/forums/index.cfm?action=detail&PostNum=1551&Thread=1&entryID=20161&roomID=11   J Sun squeaked out a $12m profit on $2.98B in its most recent quarter. TheirL revenue continues to fall, this is 13% less than the same period a year ago.I More significantly, services income rose 7% while hardware/software sales 	 fell 20%.   G A figure that catches the eye is the $482m Sun spent on R&D, up 4% from D a year ago. That is a whopping 16.2% of revenue compared to 5.7% forH IBM and 5.2% for HP. What kind of value is Sun getting for this money? AI significant fraction of it goes into MPU development but to what end? The J UltraSPARC family is far behind both proprietary processors like Alpha EV7F and POWER4x and more importantly, merchant processor families like IPFH and Opteron. How much more will it take before Sun shareholders start toG question whether a vertically integrated business model makes sense for 0 a company of Sun's dwindling size and influence.  - 	Go ahead.  Debate Paul about it.  Good luck.    				Rob    ------------------------------  % Date: Wed, 23 Jul 2003 18:52:22 -0400 * From: "Bill Todd" <billtodd@metrocast.net>W Subject: Re: D.H. Brown: EV7 AlphaServers Deliver Enhanced RAS and Powerful Performance 2 Message-ID: <wzmdnb5BQqApjIKiXTWJhw@metrocast.net>  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:Vv7ebBqO+8Dc@eisner.encompasserve.org... @ > In article <o-2dnT7yP4ipmoKiXTWJig@metrocast.net>, "Bill Todd"  <billtodd@metrocast.net> writes: > >  >  > > I > > No, I don't - because people like Rich Marcello have (at least if one F > > assumes minimal competence on *their* part) already done so.  Rich reportedJ > > that annual Alpha development costs ran $150 million (one might expect for G > > both the EV7 and EV8 tracks combined, but even if you consider that  figureH > > to have been purely EV8-related the conclusions don't change).  Rich alsoJ > > reported that annual VMS system profit ran $800 million (in Y2K, on $4H > > billion in annual revenue, a figure which persisted through at least late Q1 J > > of 2001).  Recently, Mark Gorham has been quoted as saying current VMSG > > annual system profit is $500 million on $2.5 - $3 billion in annual  revenue K > > (and in December, 2001, a Compaq response to a Gartner report put VMS's I > > annual revenue at only $2 billion - so it, and likely profit as well,  seemK > > to have rebounded a little from a low point shortly after the Alphacide  but L > > to nowhere near the pre-Alphacide level).  And examination of the CompaqL > > quarterly report statements shows no significant decline in VMS revenues% > > before the date of the Alphacide.  > >  > F > And imagine how much better their margins will be when all their R&D > for CPU is gone.  G One needn't imagine, Rob:  the numbers (or at least the upper bounds on E them) can easily be inferred from those I presented above - which, of H course, is exactly where the conclusion that it would have been far moreH profitable (even limiting one's view to VMS rather than Alpha sales as aG whole) to continue developing EV8 *even if it never shipped* came from.   F I've just snipped your irrelevant ramblings about Sun:  perhaps you're% confused about who Alpha's owner was.    - bill   ------------------------------  # Date: Wed, 23 Jul 2003 18:21:45 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond) Y Subject: Re: DEC-AXPVMS-VMS731_CDRECORD-V0100--4.PCSI v. DEC-AXPVMS-VMS731_PCSI-V0100--4. 0 Message-ID: <ZmATa.743$Bi3.534@news.cpqcorp.net>  : In article <bfmgjj$g7bth$1@ID-152801.news.uni-berlin.de>, & Michael Unger <unger@decus.de> writes:  9 >Is "/SAVE" the default behaviour [for PRODUCT INSTALL}?     No.    >If not, why not?     Compatibility with exiting code.   --  J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  # Date: Wed, 23 Jul 2003 19:03:28 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond) Y Subject: Re: DEC-AXPVMS-VMS731_CDRECORD-V0100--4.PCSI v. DEC-AXPVMS-VMS731_PCSI-V0100--4. 0 Message-ID: <4_ATa.749$Bi3.409@news.cpqcorp.net>  : In article <bfmgjj$g7bth$1@ID-152801.news.uni-berlin.de>, & Michael Unger <unger@decus.de> writes:  8 >Is "/SAVE" the default behaviour [for PRODUCT INSTALL]?   As previously stated, no.   D However, I understand that OpenVMS patch kits are being coded to useD an EXECUTE PREINSTALL procedure that will check and warn if /SAVE isC not specified.  (The details of how this is done are still a bit in / flux, but will be in the V7.3-2 documentation.)    --  J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  + Date: Wed, 23 Jul 2003 14:10:29 -0500 (CDT)  From: sms@antinode.orgY Subject: Re: DEC-AXPVMS-VMS731_CDRECORD-V0100--4.PCSI v. DEC-AXPVMS-VMS731_PCSI-V0100--4. ) Message-ID: <03072314102926@antinode.org>   3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond)   < > In article <bfmgjj$g7bth$1@ID-152801.news.uni-berlin.de>, ( > Michael Unger <unger@decus.de> writes: > ; > >Is "/SAVE" the default behaviour [for PRODUCT INSTALL}?   >  > No.  >  > >If not, why not?  > " > Compatibility with exiting code.  G    I can buy that, but it would sure help us morons if the question had * more answers than NO and YES, for example:      NO (Abort the installation.) *    YES (Continue, and lose recovery data.)0    SAVE (Continue, and save more recovery data.)  E Failing that, a suggestion to use /SAVE_RECOVERY_DATA would be better 
 than nothing.   %    Thanks to all for the useful info.   H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode,org     Saint Paul  MN  55105-2547    ------------------------------  # Date: Wed, 23 Jul 2003 19:52:52 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond) Y Subject: Re: DEC-AXPVMS-VMS731_CDRECORD-V0100--4.PCSI v. DEC-AXPVMS-VMS731_PCSI-V0100--4. 0 Message-ID: <oIBTa.756$8p3.594@news.cpqcorp.net>  * In article <03072314102926@antinode.org>,  sms@antinode.org writes:  G >                  ... it would sure help us morons if the question had + >more answers than NO and YES, for example:  >   >   NO (Abort the installation.)+ >   YES (Continue, and lose recovery data.) 1 >   SAVE (Continue, and save more recovery data.)  > F >Failing that, a suggestion to use /SAVE_RECOVERY_DATA would be better >than nothing.  @ I'm going to take this as a suggestion that a SAVE answer should9 be implemented.  This would be equivalent to having used  < /SAVE_RECOVERY_DATA on the original PRODUCT INSTALL command.  . I will pass this suggestion to the developers.   Thanks!    --  J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Wed, 23 Jul 2003 16:13:50 -0400  From: norm.raphael@metso.comY Subject: Re: DEC-AXPVMS-VMS731_CDRECORD-V0100--4.PCSI v. DEC-AXPVMS-VMS731_PCSI-V0100--4. ? Message-ID: <OF46564A8D.4FC355DB-ON85256D6C.006ED7AF@metso.com>   E While your suggesting, there should be logical names to  head-off and  substitute these answers forF command-procedure installs where an operator is not present similar to no_ask$backup.  ) From:  hammond@not on 07/23/2003 03:52 PM   ? Please respond to hammond@not@peek.ssr.hp.com (Charlie Hammond)    To:    Info-VAX@Mvb.Saic.Com cc:   ; Subject:    Re: DEC-AXPVMS-VMS731_CDRECORD-V0100--4.PCSI v. +        DEC-AXPVMS-VMS731_PCSI-V0100--4.PCSI     ) In article <03072314102926@antinode.org>,  sms@antinode.org writes:  G >                  ... it would sure help us morons if the question had + >more answers than NO and YES, for example:  >   >   NO (Abort the installation.)+ >   YES (Continue, and lose recovery data.) 1 >   SAVE (Continue, and save more recovery data.)  > F >Failing that, a suggestion to use /SAVE_RECOVERY_DATA would be better >than nothing.  @ I'm going to take this as a suggestion that a SAVE answer should8 be implemented.  This would be equivalent to having used< /SAVE_RECOVERY_DATA on the original PRODUCT INSTALL command.  . I will pass this suggestion to the developers.   Thanks!    --J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Wed, 23 Jul 2003 22:04:24 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> Y Subject: Re: DEC-AXPVMS-VMS731_CDRECORD-V0100--4.PCSI v. DEC-AXPVMS-VMS731_PCSI-V0100--4. ' Message-ID: <3F1F4CB8.424DF5DD@fsi.net>    Charlie Hammond wrote: > ; > In article <bfmgjj$g7bth$1@ID-152801.news.uni-berlin.de>, ( > Michael Unger <unger@decus.de> writes: > : > >Is "/SAVE" the default behaviour [for PRODUCT INSTALL}? >  > No.  >  > >If not, why not?  > " > Compatibility with exiting code.  H I was going to suggest extracting the the CLD for the PRODUCT verb using@ the freeware VERB utility and simply editing the .CLD to add theG "DEFAULT" attribute to the SAVE qualifier and then modifying DCLTABLES. 9 /SAVE does not appear to be in my V7.2-2 system, however.    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Wed, 23 Jul 2003 13:09:07 -0400 5 From: David Beatty <David.Beatty@qwertysasasdfgh.com> # Subject: Re: Do ES45s run VMS 7.2-2 2 Message-ID: <+8AePxy=ElxYr4rYq0bbqfcfRi8F@4ax.com>  @ On Sat, 19 Jul 2003 21:22:14 GMT, Beach Runner@nospam.com wrote:   >  >  >David Beatty wrote: > I >> On 15 Jul 2003 13:20:55 -0700, chessmaster1010@hotmail.com (JG) wrote:  >>J >> >Does anyone know if VMS 7.2-2 will run ok on an AlphaServer ES45?  TheG >> >HP VMS website says 7.3 is the minimum OS support, but I was hoping H >> >they may have backported support in 7.2-2, or in one if its patches. >> >I >> >I have a mixed VMS 6.2 and VMS 7.2 cluster and I can't upgrade all of I >> >the VMS 6.2 systems at the present time.  VMS 7.3 is not supported in ) >> >a mixed version cluster with VMS 6.2.  >>G >>     We've had this discussion on comp.os.vms before.  There is a CPU C >> routine specifically for the ES45 that is found in V7.3+ that is  >> absent in V7.2-2. >>D >>     I'm running a development cluster with the following versions >> of VMS booted:  >> >>     VAX V6.2  >>     Alpha V6.2  >>     Alpha V7.1  >>     Alpha V7.2-2  >>     VAX V7.3  >>     Alpha V7.3-1  > D >That doesn't mean you won't have problems.   Very often when you doB >something unsupported it may work for a while, or be intermittentC >and then strange things happen.  This will especially show up with @ >you stress the capabilities of the cluster or use new features. > B >And then what do you do, tell colorado spring, but even though itJ >wasn't supposed to work, I got it to work, make it work again. They won'tB >do it. YOu are doing something that a data structure viewpoint isI >unsupported. All you've been  is lucky and your luck could go out at any  >minute. > F >It is a bad practice to break all the rules and accept things to keep	 >working.  >  >Bob, Beach Runner.  >   C     True, but if you read the V7.3-1 release notes, it does support 1 mixed clusters, including versions V6.2 and V7.1.    David    >  >> >>? >> I've had no cluster related issues running with this type of 3 >> configuration; of course, your mileage may vary.  >>B >>     Reading the release notes for V7.3-2, *that* version of VMS= >> will only support a mixed version cluster with VAX V7.2 or  >> Alpha V7.2-2. >>: >>     With the release of VMS731_UPDATE, you only need to8 >> apply VMS731_PCSI and VMS731_UPDATE to bring a V7.3-1 >> system up to date.  >> >> David R. Beatty   ------------------------------  % Date: Thu, 24 Jul 2003 13:15:27 +1000 C From: "Christine Ricketts/Andrew Stewart" <u1276a@uxnxixtxe.com.au> O Subject: Re: Does RT-11 run on the PDP-11/70? (was Re: PDP-11 OS Release Dates) . Message-ID: <3f1f51c1$1_1@news.iprimus.com.au>  = "Zane H. Healy" <healyzh@shell1.aracnet.com> wrote in message & news:bfk7oq02blh@enews3.newsguy.com...3 > In comp.os.rsts John Santos <JOHN@egh.com> wrote: H > > I think you're thinking of the 85x0/8700/88x0's, which had Pros withK > > P/OS as consoles.  I can't remember if they were 350's, 380's, or both.  > H > That's what I was thinking of.  The consoles started as 350's and then whenI > they ran out of them switched to 380's.  I've got the console (380) and K > d|i|g|i|t|a|l badge off of one, that a coworkers Dad scrapped a few years  > ago.  H DEC PRO 350s as consoles?  The SECV got the first VAX 8500 in Australia.9 It had a DEC PRO 380.  Our later VAX 8800s used them too. = The last VAX 8820 had a MicroVAX II as the console processor.   G The Pro had an CTI option with Intel chips that gave one 8 bit parallel A output port, one 8 bit parallel input port and one 8 bit parallel G bidirectional port.  I'm fairly sure this was how the Console Processor  interfaced to the VAX.  F > > I think the 8600/8650 had a PDP-11 console with an RL02 instead of5 > > a floppy.  No idea what O/S (if any) was on them.   1 It was a Qbus PDP-11/23 with an RLV12 controller. * A Qbus extender fed into the VAX circuity.   Regards, Andy.   ------------------------------  % Date: Wed, 23 Jul 2003 21:35:16 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>  Subject: Re: EMC on VMS ' Message-ID: <3F1F45E4.40687350@fsi.net>    Rob Young wrote: > g > In article <a98cd882.0307230551.76245540@posting.google.com>, Bart.Zorn@xs4all.nl (Bart Zorn) writes:  > [snip]J > > Generally you use HBVS because you want the best RAS features. You getJ > > those using components which are fully supported by HP. If that is notE > > possible, for instance because the management of your company has I > > decided that you have to use EMC, than you would want to use at least F > > something that complies to the specifications that are required by0 > > HBVS. Unfortunately, EMC is not among those. > F >         Sure - an HBVS requirement.  Unless of course you will neverC >         use it, ignore the requirement for HBVS - and sleep well!   G ...until a system disk shadow-set member drops, and then eventually the F other, and you're left with an empty shadow-set at 03:45 (a.m.), a 40+E mile commute (circa. 1 hr) and who knows how many hours until you can E get EMC to configure new storage-sets for you so that you can restore ! your backup of the system disk...   G Total downtime: probably in excess of 15 hours for something that could " easily have been avoided entirely.   No, thank you.   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------    Date: 23 Jul 2003 21:51:26 -0500+ From: young_r@encompasserve.org (Rob Young)  Subject: Re: EMC on VMS 3 Message-ID: <oZeMT0OOwmd7@eisner.encompasserve.org>   [ In article <3F1F45E4.40687350@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:  > Rob Young wrote: >>  h >> In article <a98cd882.0307230551.76245540@posting.google.com>, Bart.Zorn@xs4all.nl (Bart Zorn) writes:	 >> [snip] K >> > Generally you use HBVS because you want the best RAS features. You get K >> > those using components which are fully supported by HP. If that is not F >> > possible, for instance because the management of your company hasJ >> > decided that you have to use EMC, than you would want to use at leastG >> > something that complies to the specifications that are required by 1 >> > HBVS. Unfortunately, EMC is not among those.  >>  G >>         Sure - an HBVS requirement.  Unless of course you will never D >>         use it, ignore the requirement for HBVS - and sleep well! > I > ...until a system disk shadow-set member drops, and then eventually the H > other, and you're left with an empty shadow-set at 03:45 (a.m.), a 40+G > mile commute (circa. 1 hr) and who knows how many hours until you can G > get EMC to configure new storage-sets for you so that you can restore # > your backup of the system disk...  >   F 	You are talking out of your hat.  EMC doesn't have to be involved at B 	all now in configuring Symmetrix storage.  Second, their storage  	doesn't go away.    				Rob    ------------------------------  % Date: Wed, 23 Jul 2003 22:15:12 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>  Subject: Re: EMC on VMS ' Message-ID: <3F1F4F40.ECF75B4F@fsi.net>    Rob Young wrote: > ] > In article <3F1F45E4.40687350@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:  > > Rob Young wrote: > >>j > >> In article <a98cd882.0307230551.76245540@posting.google.com>, Bart.Zorn@xs4all.nl (Bart Zorn) writes: > >> [snip] M > >> > Generally you use HBVS because you want the best RAS features. You get M > >> > those using components which are fully supported by HP. If that is not H > >> > possible, for instance because the management of your company hasL > >> > decided that you have to use EMC, than you would want to use at leastI > >> > something that complies to the specifications that are required by 3 > >> > HBVS. Unfortunately, EMC is not among those.  > >>I > >>         Sure - an HBVS requirement.  Unless of course you will never F > >>         use it, ignore the requirement for HBVS - and sleep well! > > K > > ...until a system disk shadow-set member drops, and then eventually the J > > other, and you're left with an empty shadow-set at 03:45 (a.m.), a 40+I > > mile commute (circa. 1 hr) and who knows how many hours until you can I > > get EMC to configure new storage-sets for you so that you can restore % > > your backup of the system disk...  > >  > N >         You are talking out of your hat.  EMC doesn't have to be involved atJ >         all now in configuring Symmetrix storage.  Second, their storage >         doesn't go away.  7 Remember that one of VMS's niche markets is healthcare.l  F To quote Capt. Mancuso in "The Hunt for Red October", "Are you willing( to bet your life on that?" I wouldn't...  E More importantly, will it pass a CAP inspection? Will the FDA certifyV it?h   -- s David J. DachteraA dba DJE Systemsr http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/o   ------------------------------  # Date: Wed, 23 Jul 2003 18:54:39 GMT.# From: hoff@hp.nospam (Hoff Hoffman)h. Subject: Re: Finding installed images from DCL0 Message-ID: <PRATa.747$jf3.570@news.cpqcorp.net>  k In article <7d1eda7ed227af56ffb462f203e6c6a7@free.teranews.com>, David Harrold <DHarrold@wi.rr.com> writes:V  K :I am trying to find a way to list all of the files that are installed (viaX5 :INSTALL) on a disk from DCL.  This is on VMS V7.3-1._ :IO :I am aware of F$FILE_ATTRIBUTES() and the KNOWN argument, however I would haveIN :to generate a list of all files on the disk and loop through them checking toJ :see if the file is installed.  Yes, this would work, but I was hoping for :something more streamlined.  F   If you wish to acquire the list under control of a DCL procedure andH   display of the output from INSTALL is not sufficient, you will want toE   code a trivial f$search loop for the entire target device, checkingh;   each file returned from the call using f$file_attributes.   H   This is slow(er) than parsing the output of INSTALL suggested in otherG   replies, but it is expected to continue to work if/when the format of.   INSTALL displays is changed.  I   Regarding other solutions posted, the parsing of the output of INSTALL cH   or of other utility output is not supported and not recommended -- the7   output displays of these utilities can and do change.e  F   One of the other suggested solutions included manual parsing of fileD   specifications and an assumption of angle brackets within the fileE   specification, and neither of these approaches/assumptions is good,2B   nor safe.  (Directly parsing a filename is a bad idea -- use theG   f$parse lexical function or an equivalent, and allow OpenVMS to parse-A   the specification for you.  There is unfortunately at least onecC   standard OpenVMS utility that has tried to self-parse a filespec,-A   and that has gotten it wrong -- please use the OpenVMS filespec    parsing routines.)    N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq.N  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comy   ------------------------------  # Date: Wed, 23 Jul 2003 17:54:04 GMTr& From: jlsue <jefflsxxxz@sbcglobal.net> Subject: Re: HP FUDBusting8 Message-ID: <77hthvsihqp4j1mn2ik34ihfeocvra3it6@4ax.com>  G On Mon, 21 Jul 2003 15:58:05 +0200, Stefaan A Eeckels <hoendech@ecc.lu>, wrote:     >L; >I did some work on VMS way back in 1986-1987, but I'm moren8 >of a Unix man by heart (I guess people who're primarly = >programmers prefer Unix, whereas administrators prefer VMS).l >t  J Actually, I always found programming on VMS a great experience.  IncludingC features like the following, my code could be much more concise and,  readable, and have fewer errors:  0 	1.	Use the help librarian to manage help topics3 	2.	Use the Command Definition Utility to create myM7 		command lines - don't have to create some error-prone * 		syntax checker for every program I write= 	3.	Use the message librarian to store my error codes - don't-< 		have to keep track of them by code #, can use the symbolic 		name instead8 	4.	Use LSE to help create code - include syntax for all7 		system calls, as well as access to help/documentation-, 		on the languages, RTL, SYS$ routines, etc.; 	5.	Actual callable interfaces for almost anything you need- 		to do on the system:: 		a. RTL and SYS$ routines for system/process info, string 		 manipulation, etc.s; 		b. Great signal passing mechanisms to signal those errorsO8 		with more meaningful info than "can't open file", etc.: 	6.	Very complete documentation sets that not only include: 		reference material, but "operations guides" that explain( 		how/why you do things a particular way6 	7.	Calling standard that makes programming modules in8 		multiple languages a snap (albeit "C" does break that, 		unfortunately) 	8.	Very exceptional compilers  . Oh, and the debugger was great to use as well.   ------------------------------  # Date: Wed, 23 Jul 2003 17:58:38 GMTe& From: Rick Jones <foo@bar.baz.invalid>4 Subject: Re: HP World: Why Alpha's Omega Makes Sense0 Message-ID: <i1ATa.738$Ke3.376@news.cpqcorp.net>  ) Bill Todd <billtodd@metrocast.net> wrote:sF > A 1.8 GHz Opteron competes at least as closely (as always, exceptingF > FP scores, which are interesting for HPC activity but rather less soD > for commercially-significant processing) with a 1.5 GHz Madison as  > it does with a 1 GHz McKinley.  0 > ... scores for Opteron, McKinley, and Madison:  2 > The respective 4-processor SPECweb99_SSL scores:  ( > 3498   2280   3702 (unofficial result)   A couple of things...   C The 4x1.5 GHz Itanium2 6M SPECweb99_SSL 3702 score completed review C and is now published at http://www.spec.org/ along with the updated $ 2-way numbers for rx2600 and rx5670.  > Also, WRT comparisons between Itanium2 and Opteron and IA32 onC SPECweb99_SSL the following was also just posted on SPEC's website:K  + http://www.spec.org/web99ssl/cd_notice.htmle  
 rick jones  E PS - for the sake of completeness, at HP's request, the _first_ threepE Madison SPECweb99_SSL scores have been marked "NC" because HP came tonA realize that OS used was not going to ship within the three monthh availability requirement.    --  G oxymoron n, Hummer H2 with California Save Our Coasts and Oceans platesfF 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...    ------------------------------  % Date: Wed, 23 Jul 2003 13:56:02 -0400w* From: JF Mezei <jfmezei.spamnot@istop.com>4 Subject: Re: HP World: Why Alpha's Omega Makes Sense) Message-ID: <3F1ECC31.106163E0@istop.com>K   jlsue wrote:I > it wouldn't be as nice as EV8, I'm confident that these people can (andnI > will) create a product that will have great capabilities and longevity.a  L People said that of IA64 5 years ago. People said that of Intel 2 years ago.; At one point, one can no longer keep talking in the future.n  M The impression I have is that EV7 was purposefully slowed down to quicken itseK death. Not a smart move when you continue to talk about IA64 in the future.t  I > Not sure what you mean by "dismiss", but there are just no server-class F > systems vendors - today - that I know of making plans for this chip.  L Please remember Winkler's famous "windows will eat Unix's underbelly" words.N With this in mind, consider that the wintel market is slowly eroding sales andL margins from enterprise class systems.  The 64 bit 8086 from AMD isn't goingI after IA64's throath, it is going after Intel's 8086 product line. If youhJ reduce Intel's 8086 profits, where will the money come from to continue to. subsidize the development of that IA64 thing ?  L > Why would you expect it to be profitable yet?  No business-critical serverK > market has been realized yet - it's just too early.  IIRC, the Alpha chipeE > was EV4 before most of us ever saw it in real business-use systems.a  M Did Digital spend 10 years spending megabucks before first commercial systemstM became available on Alpha ? It seems to me that Alpha's design costed Digital-= peanuts compared to what Intel spent on the IA64 chip design.0  K (Where DEC got bitten was the FAB which it insisted on reserving for itselfr> and refused extra fabbing business, except for the ARM chips).   ------------------------------   Date: 23 Jul 2003 20:43 CDT ' From: carl@gerg.tamu.edu (Carl Perkins)w4 Subject: Re: HP World: Why Alpha's Omega Makes Sense- Message-ID: <23JUL200320430897@gerg.tamu.edu>s  " jefflsxxxz@sbcglobal.net writes...K }Why would you expect it to be profitable yet?  No business-critical servernJ }market has been realized yet - it's just too early.  IIRC, the Alpha chipJ }was EV4 before most of us ever saw it in real business-use systems.  What, }is the current state of IA64?  EV2?  EV2.5?  : The EVx designation does not mean what you think it means.  K The "EV4" was really the "DECchip 21064". The "0" in that is the generationiL number. The "EV5" was the 21164, one generation later. The "EVx" designationI was originally just indicative of the process or scale it was built with,tG but has since then become synonymous with the various chips themselves.IH The "EVxy" version is now "the design that was originally expected to beK produced with the EVx process but that has been shrunk to the EVy process".nJ For example, there isn't actually an EV7 - the very first version is builtN at EV8 so the "EV7" is really the "EV78", which is why the first (and probably  last) shrink will be the "EV79".  G There was no "EV3" or earlier except as purely developmental stuff. TheEG "EV4" was not only the first version that was available in systems, I'mDH pretty sure it was the first version that was *supposed* to be available in systems.   E Thus, in some loose sense, the current IA64 is equivalent to the EV568H (21164A) - the shrunk version of the second generation of the processor.H (In both cases, ignoring all developmental test chips produced earlier.)G Of course it doesn't match exactly - the 21164 had pretty much the samesK core as the 21064, but the Itanium 2 has a core that has had a considerableiJ amount of redesign work done on it. You could say they skipped the EV5 andI went directly to the EV6 by doing extesnsive redesign of the core insteadwI of just doing a few tweaks and shrinking the thing - in that case the new 9 version would be roughly equivalent to the EV67 (21264A).e   --- Carl   ------------------------------  % Date: Wed, 23 Jul 2003 18:27:52 -0400r* From: "Bill Todd" <billtodd@metrocast.net>Y Subject: Re: Intel's had Dirty Bombs Anthrax and everything! (Was: Re: HP World: Why Alph 2 Message-ID: <IGWdnVDZQ_ZrloKiXTWJiQ@metrocast.net>  3 "jlsue" <jefflsxxxz@sbcglobal.net> wrote in messaget2 news:n6fthvkgjj86fl5iuac20dqqab9m1tk3he@4ax.com...J > On Fri, 18 Jul 2003 18:10:44 -0400, "Bill Todd" <billtodd@metrocast.net> > wrote: >n > >s6 > >"jlsue" <jefflsxxxz@sbcglobal.net> wrote in message5 > >news:h5aghvkavsukjcto77hkv3qsd1ciu3g7le@4ax.com...v > ry.E > >> > > >>; > >> Right.  Somehow you feel they owe you an apology.(???)a > >aI > >No, I feel they owe those who made decisions based upon their specifici andG1 > >unequivocal 'commitments to Alpha' an apology.e > 5 > You can only claim this "requirement" for yourself.d  B What part of "I feel that..." do you find difficult to understand?     You don't get toG > decide whether the way they did it was satisfactory for most of theirl > customers.  K But I *do* get to decide whether I think the behavior is worth making clearhI to people who might be interested in or affected by it - so that they cano decide for themselves.  @   Again, in the end, if you (or any other specific customer) areJ > not satisfied, you are fully able to vote with your feet.  At that pointK > you have no other reason to continue this ranting, though, because you'veu, > made your decision on how to deal with it.  K Au contraire:  that would be a very self-centered attitude to take (sort of G like driving by an accident because it didn't affect me personally).  IuK don't know about you, but when I see something dramatically wrong I'm often B motivated to try to correct it even if it doesn't really affect me@ personally - and this *is* my 'decision on how to deal with it'.   >l > >d > >  Cool.  No audacityh > >> there.d > >eJ > >I've never pretended to be a shrinking violet.  If you don't like that, > >tough shit. > H > It's the fact that you think you speak for everyone on how they should view > these actions that gets me.s  H Perhaps instead of putting words in my mouth you should examine your ownH more carefully.  While I make it clear exactly why *I* feel the way I doK about cHumPaq, I'm more than content after laying out the evidence to leave-) it to others to sort out how *they* feel.    >  > >_0 > >>  No real "treachery" either, just business. > >w > >Ah - the Mafia ethic. >nL > Yeah, right.  They decide that the direction the company should go to meetG > business goals, and they incur the risks to take it there.  Hey, it'sT theiriJ > responsibility to keep the company profitable, if they don't do it, it's7 > their job on the line.  Armchair refs need not apply.   J I really don't have any problem with that characterization, if it will getL people to stop complaining about my efforts to make sure that the risks theyH took by throwing away a business with immense potential profit, breakingD commitments in the process, and then lying about it have appropriateD consequences.  Treachery, perfidy, and/or incompetence *should* haveF consequences, after all:  that's the free-market philosophy in action.   >1 > >f > >>G > >> But do go on.  It's really good to dwell on the past indefinitely.s > > D > >Remember your Santayana:  "Those who do not remember the past are	 condemnedrK > >to repeat it."  Given the disturbing parallels between Compaq's handling  ofH > >the 'proprietary' Alpha platform and cHumPaq's continuing handling of VMS, > >that seems particularly apt.  > >l >cF > There's a world of difference between saying "I don't believe in 'X'L > because of past behavior", and ranting constantly, in excruciating detail,* > that your beliefs and opinions are fact.  K Ah, but they're mostly *not* opinions:  I document them carefully.  And you&J might note that they don't sprout up at random:  they're mostly offered inL rebuttal to some whack-a-mole who pops up thinking it may finally be safe toH start spewing the party-line lies again (or who starts reassuring people2 about cHumPaq's 'commitments' to VMS's future...).   - bill   ------------------------------  # Date: Wed, 23 Jul 2003 18:58:31 GMTc' From: Colin Blake <colin@theblakes.com>f, Subject: Re: Mozilla... the new NULL process< Message-ID: <rVATa.36576$8g6.632305@news1.news.adelphia.net>   Michael Unger wrote:  G > What about "(C)SWB", the officially supported HP browser for OpenVMS?u  G Polling. The code is identical. SWB = Mozilla + minor branding changes.%   ------------------------------  # Date: Wed, 23 Jul 2003 18:36:16 GMTe# From: hoff@hp.nospam (Hoff Hoffman) + Subject: RE: Packed decimal arithmetic in Cl0 Message-ID: <AAATa.745$jf3.360@news.cpqcorp.net>  _ In article <CIEJLCMNHNNDLLOOGNJIOEOFHJAA.tom@kednos.com>, "Tom Linden" <tom@kednos.com> writes:   C :>  If it were my code, I'd move it out of packed decimal entirely.u :iM :Sometimes you can't do that.  Try telling a bank or insurance company to usefI :float binary instead of scaled fixed decimal or float decimal. (and BTW,G> :that is why financial institutions stick with PL/I and Cobol)  L   Um, err, Tom, I do have a passing familiarity with floating point formats,J   and with the related application trade-offs -- and I even know where theI   hidden bit is hidden.  :-)  I've ported business applications over intouK   integer math, too -- and with a corresponding and substantial improvementcJ   in the application performance.  (One company was already using floatingH   point saw their performance _and_ their statement and billing accuracyE   improve with the conversion to integer, too -- so I most definitelyeH   understand your, um, point.  Sorry.  Pun Intended.  :-)  But your, um,1   floating point point is a good one, regardless.i  J   As for integer options, financial matters tend not to exceed a quadword,H   and Knuth and the OpenVMS RTLs have various algorithms and some simpleJ   extended-precision routines available for those that have exceeded that.I   Alpha can perform quadword math native, and various Alpha compilers canr7   typically generate quadword math operations directly.)  H   The trillion-dollar US Federal budget value and the US Debt value willK   each (presently :-) fit into a quadword -- after all, a quadword providesa#   room for well over a quintillion.s  H   (The US trillion and US quintillion cited here are equivalent to what H   much of the remainder of the planet refers to as billion and trillion,H   respectively, IIRC.  Units here are US.  Please localize accordingly.)  J   OpenVMS Engineering has been suggesting folks move out of packed decimalJ   for a while now, too -- there are much larger discussions of this datingI   from the early days of OpenVMS Alpha, as the VAX hardware-level supportlJ   of packed decimal was not propogated into Alpha.  The newsgroup archives7   should have various discussions dating from that era.i       N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faqiN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.coma   ------------------------------  # Date: Wed, 23 Jul 2003 19:15:49 GMT & From: John Reagan <john.reagan@hp.com>+ Subject: Re: Packed decimal arithmetic in C 0 Message-ID: <F9BTa.753$em3.357@news.cpqcorp.net>   Tom Linden wrote:A  N > Sometimes you can't do that.  Try telling a bank or insurance company to useJ > float binary instead of scaled fixed decimal or float decimal. (and BTW, > that: > is why financial institutions stick with PL/I and Cobol) >   H If they were only that smart...  I've seen dozens of examples of people H using binary floating to track currency (and each one is broken).  I've E even seen them in languages that provided scaled integers like BASIC.m  H I won't mention the customers involved to prevent a worldwide financial 
 meltdown. :-)m  E As Hoff mentioned in another post, an integer-only solution is often -I adequate.  If you have a language that provides scaled integers, that is     what I recommend.u     -- : John ReaganM' Compaq Pascal/{A|I}MACRO Project Leader- Hewlett-Packard Companye   ------------------------------  % Date: Wed, 23 Jul 2003 15:29:05 -04001* From: JF Mezei <jfmezei.spamnot@istop.com>+ Subject: Re: Packed decimal arithmetic in Ch) Message-ID: <3F1EE1FB.FE8CBFC4@istop.com>    John Reagan wrote:I > If they were only that smart...  I've seen dozens of examples of peopleuI > using binary floating to track currency (and each one is broken).  I'veoG > even seen them in languages that provided scaled integers like BASIC.   K Interesting since there are often standards on exactly how many decimals towF use when specifying currency and conversion factors, as in the case ofL international SWIFT transactions. With floating point, you can't specify howA many decimals to consider during arithmetic, only during display.-  I In the case of packed decimal n C, couldn't one simply write his own mathe/ routines in Cobol or PL1 and call them from C ?0   ------------------------------  % Date: Wed, 23 Jul 2003 14:12:00 -0700@# From: "Tom Linden" <tom@kednos.com>e+ Subject: RE: Packed decimal arithmetic in Cc9 Message-ID: <CIEJLCMNHNNDLLOOGNJIKEAAHKAA.tom@kednos.com>e   >-----Original Message-----:2 >From: JF Mezei [mailto:jfmezei.spamnot@istop.com]( >Sent: Wednesday, July 23, 2003 12:29 PM >To: Info-VAX@Mvb.Saic.Com, >Subject: Re: Packed decimal arithmetic in C >i >o >John Reagan wrote:nJ >> If they were only that smart...  I've seen dozens of examples of peopleJ >> using binary floating to track currency (and each one is broken).  I'veH >> even seen them in languages that provided scaled integers like BASIC. > L >Interesting since there are often standards on exactly how many decimals toG >use when specifying currency and conversion factors, as in the case of-A >international SWIFT transactions. With floating point, you can't5 >specify howB >many decimals to consider during arithmetic, only during display. >5J >In the case of packed decimal n C, couldn't one simply write his own math0 >routines in Cobol or PL1 and call them from C ?   Yes.   >  >---' >Incoming mail is certified Virus Free.4; >Checked by AVG anti-virus system (http://www.grisoft.com).aA >Version: 6.0.502 / Virus Database: 300 - Release Date: 7/18/2003d >r --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.502 / Virus Database: 300 - Release Date: 7/18/2003   ------------------------------   Date: 24 Jul 03 02:51:04 +0200) From: p_sture@elias.decus.ch (Paul Sture)t+ Subject: RE: Packed decimal arithmetic in C/) Message-ID: <$PdVxoHUiy4G@elias.decus.ch>o  _ In article <CIEJLCMNHNNDLLOOGNJIOEOFHJAA.tom@kednos.com>, "Tom Linden" <tom@kednos.com> writes:o >  >  >>-----Original Message-----, >>From: Hoff Hoffman [mailto:hoff@hp.nospam]& >>Sent: Tuesday, July 22, 2003 3:27 PM >>To: Info-VAX@Mvb.Saic.Comn- >>Subject: RE: Packed decimal arithmetic in CP >> >>3 >>From: James Gessling [mailto:jgessling@yahoo.com]nA >>:Anyone know of library functions (or anything) callable from Co5 >>:that can handle this [packed decimal format math]?' >>H >>  Packed decimal math is slow in general, and conversions are requiredK >>  for most math operations; OpenVMS performs math using native datatypes.-L >>  Convert to a native integer format, and do the math.  Then convert back.D >>  Use lib$cvt_dx_dx, or such, with the packed-decimal string type. >>E >>  You could also call a COBOL or PL/I routine, languages which have-* >>  integrated support for packed decimal. >>C >>  If it were my code, I'd move it out of packed decimal entirely.: > N > Sometimes you can't do that.  Try telling a bank or insurance company to useJ > float binary instead of scaled fixed decimal or float decimal. (and BTW,? > that is why financial institutions stick with PL/I and Cobol)  >   H You don't need to reach to the figures a bank or insurance company uses.L A signed 32-bit integer including pennies/cents cannot cope with a figure of 30 million.n  I IIRC, VAC COBOL had quadword integers by 1995 (maybe before, but that was,F the first time I came across them). Before that, currency calculationsA involving 5 significant figures plues large numbers coule only bew" done accurately in packed decimal.  H Accounts have to be accurate to the penny/cent, and this is where packed+ decimal has ruled the roost for many years.   J Floating point cannot get near, or should we be talking Enron and Worldcom here?e  L > Even though they aren't in HW on Alpha they are implemented for PL/I using > very+ > efficient algorithms, nines and excess 3.m >    ------------------------------  % Date: Wed, 23 Jul 2003 17:53:51 -0700p# From: "Tom Linden" <tom@kednos.com> + Subject: RE: Packed decimal arithmetic in C-9 Message-ID: <CIEJLCMNHNNDLLOOGNJIEEAGHKAA.tom@kednos.com>J   >-----Original Message-----n1 >From: Paul Sture [mailto:p_sture@elias.decus.ch]-' >Sent: Wednesday, July 23, 2003 5:51 PM4 >To: Info-VAX@Mvb.Saic.Com, >Subject: RE: Packed decimal arithmetic in C >  >a? >In article <CIEJLCMNHNNDLLOOGNJIOEOFHJAA.tom@kednos.com>, "Tom ! >Linden" <tom@kednos.com> writes:  >> >> >>>-----Original Message-----r- >>>From: Hoff Hoffman [mailto:hoff@hp.nospam]w' >>>Sent: Tuesday, July 22, 2003 3:27 PMg >>>To: Info-VAX@Mvb.Saic.Com. >>>Subject: RE: Packed decimal arithmetic in C >>>  >>>r4 >>>From: James Gessling [mailto:jgessling@yahoo.com]B >>>:Anyone know of library functions (or anything) callable from C6 >>>:that can handle this [packed decimal format math]? >>>hI >>>  Packed decimal math is slow in general, and conversions are requiredoL >>>  for most math operations; OpenVMS performs math using native datatypes.? >>>  Convert to a native integer format, and do the math.  Then: >convert back.E >>>  Use lib$cvt_dx_dx, or such, with the packed-decimal string type.S >>>uF >>>  You could also call a COBOL or PL/I routine, languages which have+ >>>  integrated support for packed decimal.s >>>oD >>>  If it were my code, I'd move it out of packed decimal entirely. >>@ >> Sometimes you can't do that.  Try telling a bank or insurance >company to use-K >> float binary instead of scaled fixed decimal or float decimal. (and BTW,e@ >> that is why financial institutions stick with PL/I and Cobol) >> >oI >You don't need to reach to the figures a bank or insurance company uses.dC >A signed 32-bit integer including pennies/cents cannot cope with a 
 >figure of >30 million. > J >IIRC, VAC COBOL had quadword integers by 1995 (maybe before, but that wasG >the first time I came across them). Before that, currency calculationsuB >involving 5 significant figures plues large numbers coule only be# >done accurately in packed decimal.i >aI >Accounts have to be accurate to the penny/cent, and this is where packed>, >decimal has ruled the roost for many years. >VK >Floating point cannot get near, or should we be talking Enron and Worldcom  >here?  I Well not Float Binary, but you can use Float Decimal.  A lot of financialdH application also use picture data type, available in both Cobol and PL/IL which for those not familiar, is character like string  version of a decimalI with embedded currency symbols, decimal points, db, cr and other goodies.p >gB >> Even though they aren't in HW on Alpha they are implemented for >PL/I usinga >> veryw, >> efficient algorithms, nines and excess 3. >> >u >---' >Incoming mail is certified Virus Free.e; >Checked by AVG anti-virus system (http://www.grisoft.com). A >Version: 6.0.502 / Virus Database: 300 - Release Date: 7/18/2003e >  ---i& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.502 / Virus Database: 300 - Release Date: 7/18/2003   ------------------------------  % Date: Wed, 23 Jul 2003 21:34:29 -0400y* From: JF Mezei <jfmezei.spamnot@istop.com>+ Subject: Re: Packed decimal arithmetic in Ca) Message-ID: <3F1F378B.63490175@istop.com>    Tom Linden wrote:aK > Well not Float Binary, but you can use Float Decimal.  A lot of financialhJ > application also use picture data type, available in both Cobol and PL/IN > which for those not familiar, is character like string  version of a decimalK > with embedded currency symbols, decimal points, db, cr and other goodies.   J Aren't those used for display purposes only and not for file access ? As IN recall, COBOL is unable to parse such data when it reads a record from a file.  I When I worked on IBM mainframes, I had the opportunity to examine program.N dumps to find out where/why they bombed (abended in IBM world) and I found outJ that COBOL generated IBM 370 assembler code that was very compact. A basicJ MOVE chocolate-bar to mouth. instruction might, if the fields matched sizeD etc, generate a single MVC instruction in assembler (as opposed to a+ subroutine call for strcpy or memcpy in C).   N When you read a record from a file, the data was simply moved to the structure0 without conversion. (same for writing to file).   H If the applicatioon has a file that contains packed decimal data, then aX single image should adapt to that. Rewriting the whole application may not be practical.  J Also, there is a big issue with floating point when dealing with financialK calculations. If the corporate rule stipulate that all calculations MUST berN with only 3 decimals points, rounded a certain way to 2 decimals at the end ofI the calculation, then using floating point variable does not allow you to  implement this rule.    N Packed decimal or integer are the only way to ensure that you maintain decimal limits during the calculations.a   ------------------------------   Date: 23 Jul 2003 21:59 CDTn' From: carl@gerg.tamu.edu (Carl Perkins) + Subject: RE: Packed decimal arithmetic in C2- Message-ID: <23JUL200321594635@gerg.tamu.edu>h  - p_sture@elias.decus.ch (Paul Sture) writes...oI }You don't need to reach to the figures a bank or insurance company uses. M }A signed 32-bit integer including pennies/cents cannot cope with a figure ofn }30 million.   So use 64 bit integers.e  H If your base unit is 1/1000000 of a dollar you can almost fit the entireL US GDP into a signed 64 bit integer (the US GDP is a smidge over 10 trillionJ in the most recent figures I've seen; using a millionth of a dollar as theJ unit you can fit a litle over 9.3 trillion into a signed 64 bit integer).   G If you switch the unit to 1/100000 of a dollar you can actually fit thenI entire Earth's combined GDP into a signed 64 bit integer (and then some -UI it's roughtly a factor of two under the maximum number that will fit). IfrI that isn't big enough for any conceivable financial transaction, there is H something seriously wrong with your software. If you are trying to trackJ the entire planet't GDP to an accuracy better than a thousandth of a cent,3 then you need to rethink what you are trying to do.s  = ("Trillion" as used in the US, of course: 1 000 000 000 000.)i   --- Carl   ------------------------------  % Date: Wed, 23 Jul 2003 13:46:01 -0400i* From: JF Mezei <jfmezei.spamnot@istop.com>, Subject: Re: Sun Micro Profit, Revenues Fall) Message-ID: <3F1EC9D9.5E475A8C@istop.com>A   Mark Buda wrote:M > Another article said "Its proprietary equipment based on the Unix operating K > system now must compete with rival machines that run less-expensive chips G > and software."   I thought SUN used industry standard hardware... :-)e  L Reminds me of media accusations against Digital in the late 1980s... made by SUN :-)n  J I wonder if Sun will be able to react to such "proprietary" accusations. IL think that in fairness, HP is in the same boat too, with its own proprietary expensive hardware.u  K Perhaps the whole Linux thing for IBM is just a strategy to avoid the mediao calling IBM "proprietary".  H > "Obviously, we're dealing with a pretty tough year," McNealy said on a  > conference call with analysts.  L Sun fell from much higher than others because it was the man benefactor fromN the .COM frenzy. The question becomes: how much longer will Sun continue to be) haunted by the leftovers of the .COM era?    ------------------------------  # Date: Thu, 24 Jul 2003 00:23:23 GMT  From: Rob.Buxton@wcc.govt.nz Subject: Re: Sybase client% Message-ID: <3f1f2657.248676578@news>o  < On Wed, 23 Jul 2003 12:03:45 +0200, dm <dimez@gmx.ch> wrote:  J >Has any body a running sybase client (license locally installed) for VMS H >7.1-2 (or higher) which can query with the IP stack (TCP-IP V5.1-15 or 
 >higher) ?4 >(Decnet is working, but on PC only TCP is working!) >t >thanks for any ideas or helps >t
 >Dieter Meiero >h > C Yes we're running Sybase Open Client 10.0 (something) under OpenVMSi1 7.2 (VAX) and 7.2-2 (Alpha) with TCPIP 5.3 ECO 2.y  / Is there a problem you're not telling us about?   # We only use TCPIP for connectivity.    ------------------------------   End of INFO-VAX 2003.405 ************************<3F1F4F40.ECF75B4F@fsi.net>    Rob Young wrote: > ] > In article <3F1F45E4.40687350@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:  > > Rob Young wrote: > >>j > >> In article <a98cd882.0307230551.76245540@posting.google.com>, Bart.Zorn@xs4all.nl (Bart Zorn) writes: > >> [snip] M > >> > Generally you use HBVS because you want the best RAS features. You get M > >> > those using components which are fully supported by HP. If that is g_6b_crop/jpeg-6b-crop/testimg.bmp (35606 bytes) started. ; >>> 226 Transfer completed.  35050 (8) bytes transferred.1
 <<< PASVA >>> 227 Entering passive mode; use PORT (198,151,12,104,18,119)i <<< RETR testimg.jpgt >>> 150 IMAGE retrieve of /disk$misc/decus/freewarev50/jpeg_6b_crop/jpeg-6b-crop/testimg.jpg (6532 bytes) started.: >>> 226 Transfer completed.  5756 (8) bytes transferred.
 <<< PASVA >>> 227 Entering passive mode; use PORT (198,151,12,104,18,120)d <<< RETR testimg.ppmv >>> 150 IMAGE retrieve of /disk$misc/decus/freewarev50/jpeg_6b_crop/jpeg-6b-crop/testimg.ppm (102292 bytes) started.< >>> 226 Transfer completed.  101484 (8) bytes transferred.
 <<< PASVA >>> 227 Entering passive mode; use PORT (198,151,12,104,18,121)d <<< RETR testimgp.jpg0u >>> 150 IMAGE retrieve of /disk$misc/decus/freewarev50/jpeg_6b_crop/jpeg-6b-crop/testimgp.jpg (6643 bytes) started..: >>> 226 Transfer completed.  5645 (8) bytes transferred.
 <<< PASVA >>> 227 Entering passive mode; use