1 INFO-VAX	Mon, 11 Sep 2006	Volume 2006 : Issue 498       Contents:# Re: All is not well at the HP board < Re: DISMOUNT foreign tape crashes with VMS83A_ADDENDUM-V0100< Re: DISMOUNT foreign tape crashes with VMS83A_ADDENDUM-V0100& Re: HP announces new Integrity servers& Re: HP announces new Integrity servers& Re: HP announces new Integrity servers4 Processor mode for UWSS image Init Routines on Alpha' Re: Pseudo-Device Driver example needed   F ----------------------------------------------------------------------  % Date: Mon, 11 Sep 2006 01:12:29 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> , Subject: Re: All is not well at the HP board, Message-ID: <4504F02A.60171D8A@teksavvy.com>   Update:   F On Sunday, the HP board met in an emergency meeting via teleconferenceA to discuss the issue. There was no announcement on Dunn's future.   E Dunn had told AP that she would resign if asked by the board. She has G also said that the problem isn't with her but with the board member who = leaked the info and the other who had an "angry resignation".   E AP reports another meeting to be held tomorrow late in the day. So no < formal news expected until well after close of business day.    E It was Perkin's laywer who asked the SEC to require HP to divulge the F details/reasons for Perkin's resignation in its next SEC filing, which: put the story in the public (and now legal/criminal) eyes.   ------------------------------  # Date: Mon, 11 Sep 2006 00:05:51 GMT 5 From: rdeininger@mindspringdot.com (Robert Deininger) E Subject: Re: DISMOUNT foreign tape crashes with VMS83A_ADDENDUM-V0100 [ Message-ID: <rdeininger-1009062005500001@dialup-4.233.167.188.dial1.manchester1.level3.net>   5 In article <4503A46D.490A30DD@teksavvy.com>, JF Mezei % <jfmezei.spamnot@teksavvy.com> wrote:    >Volker Halle wrote: >>  7 >> OpenVMS engineering has acknowledged this problem in F >> VMS83A_ADDENDUM-V0100 and is going to provide a new IO_ROUTINES.EXE >> image on monday   > I >That is very good.  This type of response where VMS engineering provides G >a patch within a couple of days of a problem being reported on the net 8 >should go into some of the marketing materials for VMS. > > >- you need to log a support call with your HP support center. > G >This is not something that needs to go into the marketing. Considering F >8.3 was released so recently, shouldn't the patch be freely available0 >without having to go through a support centre ?  I The support center can likely make an image file available almost as soon I as it is available.  Patch kits don't turn around that fast.  If the time 0 since V8.3 release is relevant, I don't see how.  J Do you think there is someone better able to field customer calls than theI support center?  Is there a constructive suggestion somewhere here that I  missed?    ------------------------------  % Date: Sun, 10 Sep 2006 21:40:31 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> E Subject: Re: DISMOUNT foreign tape crashes with VMS83A_ADDENDUM-V0100 , Message-ID: <4504BE8C.B2C1C15F@teksavvy.com>   Robert Deininger wrote:  >  If the time2 > since V8.3 release is relevant, I don't see how.  D Something about warrantee ? Aren't there  jurisdictions in the worldD where the vendor would be forced to provide major bug fixes for free@ within a certain amount of time of software becoming available ?    L > Do you think there is someone better able to field customer calls than the > support center?     @ This assumes everyone has access to the support centre. Are HP'sF business proactices such that anyone buying softrware or t5he right toB an upgrade gets free support (aka: access to support centre) for aH certain amount of time ? If not, then some customers may have purchangedG aa licence-only for 8.3 and have no access to the support centre to get  that critical patch.   ------------------------------  % Date: Sun, 10 Sep 2006 15:01:55 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> / Subject: Re: HP announces new Integrity servers , Message-ID: <45046122.C0B72505@teksavvy.com>   "Main, Kerry" wrote:H > Again, you are in the weeds. Customers move to platforms like Windows,G > or Linux or OpenVMS for many reasons other than the basic HW. That is  > not the issue.    E In the case of VMS, it is *an* issue. VMS' future is uncertain due to G the lack of HP truly committing to its growth (aka: lack of marketing). H And it makes matters worse that HP now expects customers to buy HW on anB even more uncertain platform, and refuses to commit to porting VMS beyond that IA64 thing.   G The issue of VMS' survival is very much in play here, whereas the issue ' of Windows or Linux' survivals are not.     B Asking a potential customer to commit to VMS on IA64 is asking theH customer to take a big risk that this platform won't be available in theE future. It may be OK to host a couple of short term applications, but Q you wouldn't want to build your whole IT infrastructure on such a risky platform.   D If HP were to port VMS to the 8086, it would remove any fears of theG platform being pulled from under VMS. And HP could then include the VMS @ word in every  8086 server/desktop advertisement it makes ("Runs@ Windows, Linux, VMS"). This alone would give a HUGE boost to VMS: compared to the token marketing it is allowed to have now.  D And with VMS on 8086, HP would not have any disincentive to sell VMSB because a VMS sale would be a 8086 sale and thus contribute to the- metrics by which HP is measured: 8086 sales.      J > Windows - that means .Net which is an OO approach to programming that inJ > most cases means a total re-write of your current 3GL based application.  D Or a simple unpacking of a colourful box, inserting a CD and runningE SETUP.EXE since Windows has so much available software that you don't A need to constantly re-invent the wheel like you have to do on VMS ; because there is so little modern software available on it.     H > Keep in mind re-testing and re-certifying your application will likely' > take longer than the actual porting,    C Yeah, perhaps now you understand why moving from Alpha to that IA64 E thing is far more than just a recompile for customers. And since even H the IDG study posted by HP on HP's website indicated a signifficant lossD of the installed base (roughly 1 in 3 customer), this forced more to IA64 is a VERY BAD IDEA.  G It isn't a comparison of how diffiocult it is to port your own code. It E is a comparison of just installing already available sopftware from a ? more popular platform and just ditching your own app. Moving to : shinkwrapped software has been a trend going on for years.   ------------------------------    Date: 10 Sep 2006 13:09:09 -0700  From: "Ian Miller" <ijm@uk2.net>/ Subject: Re: HP announces new Integrity servers C Message-ID: <1157918949.894670.103230@m79g2000cwm.googlegroups.com>    More fuel for the fire  @ http://www.techworld.com/opsys/features/index.cfm?featureID=2793  D ": Has OpenVMS gone away yet? Can it continue to exist in an ItaniumB world? A: People find that 16 and 32-way systems running Unix withG virtualised Windows and Linux partitions run much quicker. But we had a E conference recently that was full with people queuing outside and two C days of very interesting discussion. The problem is that many don't A know about the alternatives that are now available -- for example  Oracle 10G is here now.   G What's more, Windows and Linux have had tools such as capacity planning A for some time. OpenVMS people haven't had such goodies until now.   G Q: Does HP regret going into Itanium? A: I can't speak for HP but I was B an Alpha man. Can we give Alpha customers running OpenVMS a better6 place to go? Yes we can. They'll get four times betterB price-performance for OpenVMS on Alpha by moving to Integrity. AndD customers have proved that they see things in the same way since, in: the second quarter, Integrity servers outsold Superdomes.    "    ------------------------------  % Date: Sun, 10 Sep 2006 22:04:38 -0400 ( From: Bill Todd <billtodd@metrocast.net>/ Subject: Re: HP announces new Integrity servers G Message-ID: <uNydnWLxZqmqWZnYnZ2dnUVZ_sGdnZ2d@metrocastcablevision.com>    Ian Miller wrote:    ...   I > Q: Does HP regret going into Itanium? A: I can't speak for HP but I was  > an Alpha man.   H But perhaps more importantly, he is *now* "European product manager for B HP's Itanium-based products" (how's that for an unbiased source?).  6   Can we give Alpha customers running OpenVMS a better > place to go? Yes we can.  F No, you can't.  To have a *better* place to go, you have to have some H *other* option to be 'better' than.  As no such other option exists, by $ definition Itanic can't be 'better'.  G That's what happens when you burn your boats far out of sight of land:  F even clinging to a spar starts to look good (since it's 'better' than 
 drowning).     They'll get four times better @ > price-performance for OpenVMS on Alpha by moving to Integrity.  D Perhaps English is not Ken's native language - unless Boob is right F after all and there really *is* an Alpha hiding inside that Integrity  system.   H If what Ken meant was that Integrity offers 4x better price/performance G *than* VMS on Alpha, well, that's what tends to happen when one vendor  H owns both products and can decide to use one as a cash cow while making @ the other as competitive as possible:  reality, perhaps, but no D reflection whatsoever on the actual relative *potential* of the two - platforms had their owner chosen differently.      And F > customers have proved that they see things in the same way since, in; > the second quarter, Integrity servers outsold Superdomes.   I Now, what do you suppose he meant by that?  After all, Integrity servers  I *include* Superdomes, so managing to out-sell them shouldn't be all that  
 difficult.  < But wait - they don't include *all* Superdomes - the HP9000 E PA-RISC-based ones, for example.  So perhaps he meant that Integrity  F servers (including Integrity Superdomes) out-sold HP9000 Superdomes - D again, hardly a surprise, given the high-end-niche character of the 9 latter compared with the far broader scope of the former.   I And, of course, the fact that Integrity servers get sold not just to run  B HP-UX but Linux, Windows, and even (once in a while) VMS as well, C whereas HP9000 Superdomes aren't likely to see much action outside  D HP-UX.  Gee, it's beginning to sound kind of like stating that more I passengers fly from NYC to LA than from NYC to Oshkosh, WI:  most likely  / true, but also most likely utterly meaningless.   > Would it even mean much if Itanic Superdomes out-sold PA-RISC I Superdomes, given the last-gasp status of the latter?  Sounds a lot more  F like customers seeing the writing on the wall than 'seeing things the F same way', and as for constituting any form of 'proof' - well, that's 
 spin for you.   H While we're here (since you brought it up), let's examine a bit more of H Ken's "HP Wisdom" (a lame ad campaign if ever I saw one, but perhaps it E plays better to people unfamiliar with the company):  how's this for  G trying to suggest that Itanics out-perform and out-price/perform Xeons  I (the context in which he offers up the statement below) without directly   trashing a sibling HP product:  I "We are delivering up to four times better price performance -- but it's  H really about virtualisation, ease of use, and performance/watt. We've a " 2x performance per watt advantage"  G Now, it's really difficult to characterize the above as anything other  F than outright, bald-faced lies.  Leaving aside SPECint (where the new B Woodcrest Xeons demolish Itanic by a factor of nearly 2:1 - while E drawing 20% *less* power in the process), moving into the commercial  H realm (where Itanic should shine if it shines anywhere), if you use the I figures in HP's own 4-core TPC-C submissions for Woodcrest and Montecito  G you'll find that the bare Woodcrest platform costs a small fraction of  F what the bare Montecito platform costs (and thus offers significantly F *better* price/performance than Montecito:  don't get taken in by the C $/tpmC figures - HP discounts Itanics a lot more aggressively than  H Woodcrests, at least for TPC-C configurations...) while offering almost D 92% as much performance/Watt (rather than only half as much, as Ken H claims - and if Woodcrest could support as much RAM as Montecito can it K would likely produce *more* TPC-C performance per Watt than Montecito can).   I If one were charitable, one might guess that what Ken really *meant* was  I that Montecito enjoys a 2x performance per Watt advantage over *Madison*  D - and that's likely true, since Madison's performance per Watt was, E shall we say, unimpressive in today's competitive context.  But when  E you're talking about competition, and using the word 'advantage', it  F kind of (perhaps by design) leaves the impression that you're talking > about advantage *over that competition* rather than over some   now-obsolete ancestral platform.  I Ken concluded that sentence with "... and have taken a lead over [IBM's]  F Power":  could he have meant for the 2x performance per Watt to apply G there?  Probably not:  Montecito's best commercial performance against  I POWER is in 4-core TPC-C, where it's almost equal to POWER (it beats the  H best POWER5 submission but IBM hasn't bothered to submit POWER5+, which F clocks 22% faster in peak trim, at the 4-core system size), and it is B unlikely that POWER5+ consumes anything like twice the power that I Montecito does.  In large TPC-C systems, of course, POWER5+ beats Itanic  H by factors of well over 2:1 (that's extrapolated to Montecito Superdome G results that for some reason HP hasn't been all that anxious to submit  I yet:  POWER5+ usually beats Madison by over 3:1 core-for-core, though by  A only a bit over a projected 2.5:1 for a recent Fujitsu 32-socket    top-of-the-line Madison system).  G The immortal phrase "Would you buy a used car from this man?" comes to  B mind - but, realistically, Ken probably isn't much worse than the I average less-than-completely-honest agenda-driven flack.  I guess that's  I what people have come to expect these days, but it's still kind of a sad  H commentary on civilization as we know it and I don't think that another ( Ken familiar to many here would approve.   - bill   ------------------------------  % Date: Mon, 11 Sep 2006 07:15:24 +0800 3 From: "Richard Maher" <maher_rj@hotspamnotmail.com> = Subject: Processor mode for UWSS image Init Routines on Alpha 1 Message-ID: <ee264f$ooo$1@news-02.connect.com.au>    Hi,   I John if you read this could you please just look at the following snippet L and tell me what's wrong with it? My theory is, if Auth_Init is executing inE Exec Mode then the process would terminate with ss$_abort. And if the L process terminates with ss$_abort then Auth_Init was executing in Exec mode.7 Any thoughts? (If we agree it's in Exec Mode then why?)    .call_entry label=auth_init   	 movpsl r0 , extzv #psl$v_curmod,#psl$s_curmod,r0,cur_mod cmpl #psl$c_exec,cur_mod bneq 20$   $exit_s code=#44  F That code branches to 20$ on VAX. (There was a also post about "I love  Stuart Davidson" some time agao)   For more se: -L http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=1058548  ? I guess I'll just put Auth_Init in  a UWSS and call it from the 4 Initialization routine of a 2nd User-Mode shareable.   Cheers Richard Maher   ------------------------------    Date: 10 Sep 2006 12:55:39 -0700  From: "Ian Miller" <ijm@uk2.net>0 Subject: Re: Pseudo-Device Driver example neededB Message-ID: <1157918139.128972.33450@m79g2000cwm.googlegroups.com>   It appears to be here   3 http://vmsone.com/~decuslib/vaxsig/vax88b2/addbook/    ------------------------------   End of INFO-VAX 2006.498 ************************                                                                                                                                                                                                                                                                                                                                                                                                                                      