0 INFO-VAX	Thu, 15 Jan 2004	Volume 2004 : Issue 29      Contents: Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 500.000 AMD64's shipped... Re: 7.3-1 to 7.3-2 Re: 7.3-1 to 7.3-2? Re: ANN: FreeeTDS 0.62 Sybase & MS SQL client software released  DECLASER 5100 output paper jam Re: Detaching processes on VMS Re: Detaching processes on VMS Re: Detaching processes on VMS  Re: DHCP/BOOTP/PXE under OPENVMSH Re: Faced with flagging PC sales,  HPand others are pushing aggressivelyH Re: Faced with flagging PC sales,  HPand others are pushing aggressively Re: file and record formats  Re: file and record formats  Re: file and record formats D Re: HP Integrity Superdome server sets world record for 10-terabyte,P Re: HP Integrity Superdome server sets world record for 10-terabyte, single-syst2 Re: HPQ stock to be dual-listed on NYSE and NASDAQ- Humongus bug in DECTERM (postscript printing) 8 IBM announces DB2 for HP Integrity Itanium-based servers< Re: IBM announces DB2 for HP Integrity Itanium-based serversP Re: Intel to chip away at Itanium prices <- or ... I want my cheap workst. to ruP Re: Intel to chip away at Itanium prices <- or ... I want my cheap workst. to ruP Re: Intel to chip away at Itanium prices <- or ... I want my cheap workst. to ruP Re: Intel to chip away at Itanium prices <- or ... I want my cheap workst. to ru3 Re: Modify the logical SYS$STARTUP, why or why not? 3 Re: Modify the logical SYS$STARTUP, why or why not? 3 Re: Non-WS Process Quotas not in Performance Manual F Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions perF Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions perF RE: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions perF Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions perF RE: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions perF RE: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions perF Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions perF Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions perF Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions perF Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions perP Re: OT - English (UK) humor - (was Re: The Register: OpenVMS among most-secure oP Re: OT - English (UK) humor - (was Re: The Register: OpenVMS among most-secure o OT: AA pilot arrested in Brazil ! OT: Good luck to Glenn C Everhart 8 Re: Secure Web Server Version 2.0 available for download Re: ssh on Ovms 7.3-2 0 Re: SYS$QIOW, INSTAT and OFFSET - what are they?@ Re: The Register: OpenVMS among most-secure of operating systems@ Re: The Register: OpenVMS among most-secure of operating systems@ Re: The Register: OpenVMS among most-secure of operating systems Re: VAX architecture and Charon  Re: VAX architecture and Charon  Re: VAX architecture and Charon  Re: VAX architecture and Charon  Re: VAX architecture and Charon ! Re: VMS runs well on HP Superdome ! Re: VMS runs well on HP Superdome  Will HP *ever* publicize OVMS?  F ----------------------------------------------------------------------  + Date: Thu, 15 Jan 2004 07:07:41 +0000 (UTC) % From: bdc@world.std.com (Brian Chase) ' Subject: Re: 500.000 AMD64's shipped... ( Message-ID: <bu5e7t$71r$1@pcls4.std.com>  3 In article <$blvA10Xq9lE@eisner.encompasserve.org>, : Bob Koehler <koehler@eisner.aspm.encompasserve.org> wrote:: > In article <4005A927.7986EBF5@yahoo.com>, Tim MacEachern* > <Tim.MacEachern@ns.sympatico.ca> writes:  K > > Nonsense, as said by a previous poster, who in the mainstream market is 5 > > going to want more than 40 physical address bits.  > H >    The same guy who wanted more than 512KB RAM in his PC 18 years ago.  F At current market prices, 1TB (40-bits) of cheap PC RAM will run aboutE $400,000 US. Assuming Moore's Law holds the course, it'll be sometime G around the years 2015-2017 before power users would find that amount of  memory reasonably affordable.   I I think AMD have adequately sized their current implementation of x86-64. H There's nothing in the architecture which would present them from addingD more physical and virtual address bits, up to 64, as they need them.G Putting them in now would just be... well a significant waste of effort B and resources with no practical benefit from their target markets.  E The notion of Itanium being somehow /better/ because it currently has F 50-bit physical addressing is nothing but a meaningless sales gimmick.J It's going to be quite a while before anyone can afford 1024TB of high-endA server ECC RAM.  What 50-bits means to me is that current Itanium I processor offerings are probably a good bit more expensive than they need  to be.   -brian.  --  F --- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ -----5                    Do not fold, mutilate, or spindle.    ------------------------------  % Date: Thu, 15 Jan 2004 02:32:31 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>' Subject: Re: 500.000 AMD64's shipped... ) Message-ID: <400641ED.5A51BAC0@istop.com>    Fred Kleinsorge wrote:K > You are turning the world on it's head.  The R&D investments are done and M > paid for.  I've heard no proof that Intel is losing money on IPF.  Intel is N > *highly* profitable, and still maintains the vast majority of the uProcessor	 > market.   ; I do not think I agree with the "R&D is done and paid for".   N For on thing, even the 8086 is still getting tons of R&D. And since IA64 has aL long way to go to get respectable power/heat/performance, and has to competeH against Power at one end, and the 8086 at the other end,  if intel stopsG spending R&D money on IA64, IA64 will fall by the wayside very quickly.   N Secondly, I doubt very much that it has been "paid for".  Sure, Intel may haveK subsidized IA64 with 8086 profits, but that woudl really mean that intel is M dumping IA64 on the market with no expectation that it will ever pay back the L initial R&D. (and is there any indication that IA64 is generating sufficient? Intel revenus to justified continued development of the chip ?)   N Thirdly, I just saw one of the senior Intel accountants interviwed on the BBC.N He mentioned wireless and higer volumes of commodity chips as the generator of, the big profit increase, no mention of IA64.   ------------------------------  % Date: Thu, 15 Jan 2004 02:46:41 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>' Subject: Re: 500.000 AMD64's shipped... ) Message-ID: <4006453E.51861944@istop.com>    Keith Parris wrote: G > And Intel is telling us they plan to drive Itanium prices down to the @ > level of Xeons, so they can phase out Xeons.  That's commodity
 > pricing.  5 In 2007.  By then, the market may have changed a lot.   ? > plus a couple other big ones you happened to miss: IBM, Dell.   L Neither have bet their business on IA64. They are only producing token IA64sN to please Intel in exchange for a better deal on other Intel hardware which si key to their business.  G > A lot of the potential of Alpha didn't pan out, like Windows support, D > and its use in multiple vendors' systems.  DEC wanted it to become$ > industry-standard.  Didn't happen.  G It isn't because it was impossible to do, it was simply because Digital N prevented itself from doing it. (think of cyclist sticking his own pump in his own wheel).   M Compaq however was in a position to rescue Alpha and succeed with it. But its ? relationashiop with Intel was more important than the company's  profits/survival.     F > We're in a better position with Itanium now than we were with Alpha.F > Windows, Linux, HP-UX, and OpenVMS on the same hardware, at the same > hardware prices.  K No you are in exactly the same position as Alpha. Replace HP-0UX with Tru64 K and the above line applied to Alpha, until Compaq started to pull the plug.   N Secondly, IA64 is in a much weaker position with regartds to Alpha because VMSM will have far fewer software available on it (especially initially). And that M will hurt VMS substantially because it will end up reducing the portfolion of  availbale software on VMS.  C > of scale that VMS never before enjoyed.  Looks like Itanium-based B > systems are going to price out at only 40-70% of what equivalent > Alphas cost.  J But remember that under Digital, Alphas were prevented from competing withD wintel boxes and were artificially priced higher. It isn't a case of! "couldn't" but rather "wouldn't".   H > And it's an Intel chip.  The #1 microprocessor vendor.  Huge bankroll.H >  Top-notch fabrication technology.  And the ability to actually market > its products.   K But if intel subsidizes IA64, it can then be accused of dumping the chip on I the market. Intel will eventually have to show real profits on that chip.    ------------------------------  % Date: Thu, 15 Jan 2004 08:35:19 +0100  From: Dirk Munk <munk@home.nl>' Subject: Re: 500.000 AMD64's shipped... 2 Message-ID: <bu5gh4$k0i$1@news4.tilbu1.nb.home.nl>   Keith Parris wrote: ^ > David Froble <davef@tsoft-inc.com> wrote in message news:<40048CD6.7050707@tsoft-inc.com>... > H >>This is HP's problem, if and only if its Intel that gives up on IA-64. >  > ...  > N >>The subject line says it all.  If AMD starts eating into Intel's profitable $ >>products, Intel will have to react >  > G > We've already seen Intel's reaction.  They're not planning to give up D > on IA64; instead, they plan to make Itanium chips as cheap as Xeon2 > chips, so they can phase out Xeon by about 2006:; > http://news.com.com/2100-1006_3-5140486.html?tag=nefd_top   Q That is nothing new, and was to be expected! The Itanium is a server cpu, and so   is the Xeon.  Q However AMD has *one* 32/64 bit architecture for the laptop, the desktop and the  O server environment. Intel will have two very distinctive architectures, 64 bit  @ for the servers and a rather different 32 bit for anything else.   ------------------------------    Date: 15 Jan 2004 02:25:43 -0600+ From: young_r@encompasserve.org (Rob Young) ' Subject: Re: 500.000 AMD64's shipped... 3 Message-ID: <sB5lG0CN3lHt@eisner.encompasserve.org>   V In article <4006453E.51861944@istop.com>, JF Mezei <jfmezei.spamnot@istop.com> writes: > Keith Parris wrote: H >> And Intel is telling us they plan to drive Itanium prices down to theA >> level of Xeons, so they can phase out Xeons.  That's commodity  >> pricing.  > 7 > In 2007.  By then, the market may have changed a lot.  >   = 	But you aren't considering how cheap Xeon's will be in 2007. 1 	They certainly won't be more expensive.  SGI has = 	a very good story to tell and this hints as to where Itanium  	pricing is headed.     . http://www.sgi.com/features/2004/jan/altix350/  N Starting at just $12,199 USLP, with an average configuration USLP of $5400 perK processor, Altix 350 offers the demonstrated scalability and reliability of # Altix at a very competitive price.    ? 	See that?  Additional Itanium CPUs are $5400.  Sun and IBM are @ 	quaking in their boots as this is a glimpse of things to come - 	much cheaper Itanium prices.    > I > It isn't because it was impossible to do, it was simply because Digital P > prevented itself from doing it. (think of cyclist sticking his own pump in his
 > own wheel).  > O > Compaq however was in a position to rescue Alpha and succeed with it. But its A > relationashiop with Intel was more important than the company's  > profits/survival.  >   A 	Nah, it's impossible.  In fact, I'll guess Shoemaker, Zander and D 	Joy bailed out of Sun because of Itanium scuttlebutt.  The industry= 	is very leaky at a high level you know.  Fister gives little D 	glimpses of where it is headed to squash traderag mongering.  Think< 	about it.  High-end CPU prices are screaming down - even as( 	we speak.  Sun (Scotty) doesn't get it.   > P > Secondly, IA64 is in a much weaker position with regartds to Alpha because VMSO > will have far fewer software available on it (especially initially). And that O > will hurt VMS substantially because it will end up reducing the portfolion of  > availbale software on VMS. >   @ 	Much stronger position.  Itanium server prices are going to be 0 	incredible a few years out and are sweet today.  6 	(Hint:  scroll back up and read the SGI link - okay?)  M > But if intel subsidizes IA64, it can then be accused of dumping the chip on K > the market. Intel will eventually have to show real profits on that chip.   @ 	Ha.  Microsoft Office subsidizes Microsoft.  You won't even getC 	anybody poking around Intel.  $2.2 billion in profit last quarter, > 	a 107% increase compared to last year's quarter (annualized).   				Rob    ------------------------------  + Date: Thu, 15 Jan 2004 09:33:53 +0000 (UTC) % From: bdc@world.std.com (Brian Chase) ' Subject: Re: 500.000 AMD64's shipped... ( Message-ID: <bu5mq1$rko$1@pcls4.std.com>  ) In article <4006453E.51861944@istop.com>, , JF Mezei  <jfmezei.spamnot@istop.com> wrote: > Keith Parris wrote:   I > > And Intel is telling us they plan to drive Itanium prices down to the B > > level of Xeons, so they can phase out Xeons.  That's commodity > > pricing. > 7 > In 2007.  By then, the market may have changed a lot.  > A > > plus a couple other big ones you happened to miss: IBM, Dell.  > H > Neither have bet their business on IA64. They are only producing tokenD > IA64s to please Intel in exchange for a better deal on other Intel* > hardware which si key to their business.  I Besides, it just makes good business sense for both IBM and Dell to offer G Itanium systems--on a number of levels.  Since HP have bet their server C business on Itanium; there's simply no turning back for HP.  From a J strategic standpoint, even if Dell and IBM never sell one of their ItaniumG systems, their existence puts competitive pressure on HP's pricing (and  profit margins).  G There's also a good chance that Itanium is here to stay; it wouldn't be - sensible for either IBM or Dell to ignore it.    -brian.  --  F --- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ -----5                    Do not fold, mutilate, or spindle.    ------------------------------  % Date: Thu, 15 Jan 2004 04:53:35 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>' Subject: Re: 500.000 AMD64's shipped... ) Message-ID: <400662F2.69A1A060@istop.com>    Brian Chase wrote:I > Itanium systems--on a number of levels.  Since HP have bet their server E > business on Itanium; there's simply no turning back for HP.  From a L > strategic standpoint, even if Dell and IBM never sell one of their ItaniumI > systems, their existence puts competitive pressure on HP's pricing (and  > profit margins).  L Correct. IBM and Dell would be pricing those IA64 systems at a loss to forceN HP to lower its IA64 proices and lose money on the systems. IBM abnd Dell knowF that they won't be selling many such systems so they can afford to tag unrealistically low prices.   J Nobody will buy a Dell IA64 system to run HP-UX or VMS since they won't be supported by HP.   ------------------------------  % Date: Thu, 15 Jan 2004 13:43:58 +0000 O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> ' Subject: Re: 500.000 AMD64's shipped... 0 Message-ID: <bu65g3$kig$1@new-usenet.uk.sun.com>   Rick Jones wrote: R > Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote: > C >>40 bit addressing will allow the Opteron to address more physical @ >>memory than is currently available in any microprocessor based	 >>system.  >  > D > I'll probably get the math wrong, but 40 bits of addressing is 1TBG > yes? (unless some of those physical bits get used for other nefarious H > porpoises I suppose).  I suspect you were very careful to use the wordH > "currently" as the Madison Superdome will shortly be supporting 1TB of > RAM in a single OS instance. >   ? No you got the maths right and currently the SuperDome supports 9 512 GB half what the current Opteron currently addresses.   ; I would also wager that you have a tiny tiny tiny number of : SuperDomes currently installed at customers with 1/2 TB of RAM.   Regards  Andrew Harrison  > rick jones   ------------------------------  % Date: Thu, 15 Jan 2004 13:50:57 +0000 O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> ' Subject: Re: 500.000 AMD64's shipped... 0 Message-ID: <bu65t7$kss$1@new-usenet.uk.sun.com>   Keith Parris wrote:  > Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<bu3gu1$mm2$1@new-usenet.uk.sun.com>...  > ; >>But Opteron Integer performance is stubbornly faster than 7 >>Itanium despite being saddled with 1/8 the registers.  >> >>1477 SPECint vs 1322 >  > H > At the moment, AMD's x86-64 is competing mostly against Intel's 32-bit- > chips, which soundly trounce it in SPECint:  >   Pentium 4EE (3.2 Ghz)  1620  >   AMD Opteron (2.2 Ghz)  1477  >   Itanium 2   (1.5 Ghz)  1322  >   Alpha EV68  (1.25 Ghz)  928  >   SPARC Fujitsu(1.35 Ghz) 905  > E > Right now Itanium is artificially constrained in clock speed due to B > heat generation -- 130 watts is the power budget today's systemsF > allow.  Intel just announced that after 5-6 years of hard work, theyH > now have the keys to producing low-k dielectric transistors, with 100XE > lower leakage current.  This can reduce the power consumption of an C > overall chip by 50%, allowing the clock rate to be cranked up and $ > still fit within the power budget. >   ; Keith you have just displayed a total lack of understanding 1 about the design constraints for Microprocessors.   9 Itanium is not artifically constrained in clock speed due 4 to heat generation. Intel designed it to produce the3 ammount of heat it produces for the target process.   " There is no artificial constraint.  8 And the Intel process you refer to is scheduled to start8 production at best in in the last quarter of this decade< far to late to influence the sucess or otherwise of Itanium.     > ? >>>Floating-point registers 8 for x86-64 (it's saddled with the A >>>now-ancient x87 floating-point architecture); 128 for Itanium.  >>8 >>Ahh finally you have found a feature that might matter >  > ...  > 9 >>The trouble is that this one feature thats positive for 6 >>Itanium doesn't actually matter to the vast majority& >>of people who will buy your systems. >  > ? > You really think floating-point performance doesn't matter to D > customers? But on second thought, we should have guessed you mightG > have that impression of customers' needs, since for the Sun customers D > you deal with, given Sun has been so far behind in the performanceE > race for so long, performance probably matters very little, or they $ > wouldn't be Sun customers anymore.  ; No read my post again, FP does matter to some customers but 6 the market is tiny compared with integer applications.  ; The HPC market as a whole is tiny compared with the general 7 server market which is dominated by web, app, DBMS type 9 applications the vast majority of which use no FP at all.    Regards  Andrew Harrison    ------------------------------  % Date: Thu, 15 Jan 2004 13:57:21 +0000 O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> ' Subject: Re: 500.000 AMD64's shipped... 0 Message-ID: <bu6696$kui$1@new-usenet.uk.sun.com>   Fred Kleinsorge wrote:D > "Tim MacEachern" <Tim.MacEachern@ns.sympatico.ca> wrote in message% > news:4005A927.7986EBF5@yahoo.com...  > I >>Nonsense, as said by a previous poster, who in the mainstream market is  >  > going to want  > % >>more than 40 physical address bits.  >  > L > Who in the "mainstream market" is going to want more than 32 virtual bits?N > And at what cost?  Will all the 32-bit non-privleged code run when the OS isJ > in 64-bit mode?  What about drivers and devices?  What about things thatL > will run *slower* which is often the case when VA requirements are not the > big-ticket problem?  >  >  >   " Who was refering to 32bit anything    You seem to have gone off topic.  0 Back on topic 40 bits or 1 TB of physical memory. is probably enough for most of the market that0 Opteron and Itanium for that matter is trying to address.  . SuperDome currently supports 1/2 TB of RAM and, will this year or next year support 1 TB for* which the market will be vanishingly tiny.  * Trying to establish that Itanium is better* than Opteron because it supports 50 rather. than 40 bit addressing when there is currently- no roadmap from HP that includes Systems that + will require more than 40 bit addressing is  a really pointless argument.  ( Its just as well that you arn't actually# supporting the argument (are you ?)    Regards  Andrew Harrison    ------------------------------    Date: 15 Jan 2004 06:40:44 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) ' Subject: Re: 500.000 AMD64's shipped... < Message-ID: <cf15391e.0401150640.88743d5@posting.google.com>  [ JF Mezei <jfmezei.spamnot@istop.com> wrote in message news:<4005CD94.4CAD34A3@istop.com>... M > But the same technology will also allow the 8086, Power , Sparc etc to also  > boost their clock rates.  F x86 and SPARC (don't know about Power) chips don't dissipate 130 watts@ today, so they aren't constrained by heat generation -- they are constrained by other factors.    ------------------------------    Date: 15 Jan 2004 06:52:42 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) ' Subject: Re: 500.000 AMD64's shipped... = Message-ID: <cf15391e.0401150652.672d26a3@posting.google.com>   [ JF Mezei <jfmezei.spamnot@istop.com> wrote in message news:<4006453E.51861944@istop.com>... N > Neither have bet their business on IA64. They are only producing token IA64sP > to please Intel in exchange for a better deal on other Intel hardware which is > key to their business.  F The reason you give (better IA32 pricing) is pure speculation.  But itD is clear that Dell and IBM think there's enough potential in ItaniumA that they don't want to miss out.  Dell even sat things out for a  while, and dove back in.  I > It isn't because it was impossible to do, it was simply because Digital P > prevented itself from doing it. (think of cyclist sticking his own pump in his
 > own wheel).    Was, was, was.  It's history.    H > > We're in a better position with Itanium now than we were with Alpha.H > > Windows, Linux, HP-UX, and OpenVMS on the same hardware, at the same > > hardware prices. > 3 > No you are in exactly the same position as Alpha.   A No, Alpha lost Windows support, and never had HP-UX support. (And B HP-UX is about a $20B annual business, much bigger than Tru64 ever was.)   P > Secondly, IA64 is in a much weaker position with regartds to Alpha because VMSO > will have far fewer software available on it (especially initially). And that N > will hurt VMS substantially because it will end up reducing the portfolio of > available software on VMS.  A Initially Alpha didn't have as much software as VAX.  We got past E that.  In the interim, if a given piece of software is only available 4 on Alpha, folks can continue to run on their Alphas.  L > But remember that under Digital, Alphas were prevented from competing withF > wintel boxes and were artificially priced higher. It isn't a case of# > "couldn't" but rather "wouldn't".   C HP isn't making that same mistake.  Same HW price regardless of OS.   M > But if intel subsidizes IA64, it can then be accused of dumping the chip on K > the market. Intel will eventually have to show real profits on that chip.   F All new products are subsidized when under developement and when firstB introduced.  Otherwise we'd never have new products.  Profits come- later, and they subsidize other new products.   @ Intel has told us they plan to drive Itanium prices down to XeonB levels, and move the market from 32 bits to 64 bits using Itanium.   ------------------------------    Date: 15 Jan 2004 07:00:03 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) ' Subject: Re: 500.000 AMD64's shipped... = Message-ID: <cf15391e.0401150700.77623e22@posting.google.com>   [ JF Mezei <jfmezei.spamnot@istop.com> wrote in message news:<400662F2.69A1A060@istop.com>... N > Correct. IBM and Dell would be pricing those IA64 systems at a loss to forceP > HP to lower its IA64 proices and lose money on the systems. IBM abnd Dell knowH > that they won't be selling many such systems so they can afford to tag > unrealistically low prices.   D If IBM and Dell were to price IA64 systems below cost, it would makeF them so attractive that it would shift business away from their other,D priced-for-profit systems and cause them to lose money.  They're not
 that dumb.  F No, they're building IA64 systems so they won't be left in the dust ifF Itanium does take off in the marketplace.  IA64 systems will be priced5 to make a small profit like the rest of their lineup.    ------------------------------    Date: 15 Jan 2004 07:25:29 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) ' Subject: Re: 500.000 AMD64's shipped... = Message-ID: <cf15391e.0401150725.7cbb7fed@posting.google.com>   X Dirk Munk <munk@home.nl> wrote in message news:<bu5gh4$k0i$1@news4.tilbu1.nb.home.nl>...S > However AMD has *one* 32/64 bit architecture for the laptop, the desktop and the  Q > server environment. Intel will have two very distinctive architectures, 64 bit  B > for the servers and a rather different 32 bit for anything else.  > x86-64 may do well on the desktop, but the design is seriously< compromised for the server space.  Maximum of 8-way (currentF Superdomes are already at 64-way), 40 bits or 1 TB of physical addressD space (current Superdomes already support half that address space atC 512 GB), maximum of 48-bit virtual address space instead of 64-bit,  and only 16 integer registers.   ------------------------------    Date: 15 Jan 2004 07:36:15 -0800' From: icerq4a@spray.se (David Svensson) ' Subject: Re: 500.000 AMD64's shipped... = Message-ID: <734da31c.0401150736.65c94cd6@posting.google.com>    Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<bu3rct$qce$2@new-usenet.uk.sun.com>...  > David Svensson wrote:  > > Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<bu1eh4$1ct$1@new-usenet.uk.sun.com>...  > >  >   I > > It is currently expensive and hotter, but in my real world experience J > > with testing customer database setups, nothing beats Itanium today andJ > > that matters. (this includes testing of all current CPU architectures) >  > I > So you have tried an unspecified DBMS on a Power4+ P690, a USIII 1.2Ghz H > SF Sun and a GS1280. You must have access to a great deal of expensive > kit, congratulations.  >    Thanks, F but they are 4-way boxes, so there is no need to have P690 and GS1280.   ------------------------------    Date: 15 Jan 2004 07:48:02 -0800 From: dennis@etinc.com (Dennis) ' Subject: Re: 500.000 AMD64's shipped... = Message-ID: <3bcf70cd.0401150748.443b1806@posting.google.com>   F I'll take a 3.06Ghz Xeon over either. The stability and performance ofF the peripheral chipsets, particularly the bus, are more important, andD intel has a wide lead in that category. All of the AMD chipsets haveB flaws. Additionally, most high end machines are I/O bound, not cpuE bound. So for the vast majority of applications your little arguments * and benchmarks are just plain meaningless.  D Bravo to AMD however, for doing good marketing. Thank goodness thereD is always a benchmark to make your case, no matter what it might be.   DB     Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<bu3gu1$mm2$1@new-usenet.uk.sun.com>...  > Keith Parris wrote: _ > > JF Mezei <jfmezei.spamnot@istop.com> wrote in message news:<4003A02E.23A91A6F@istop.com>...  > >  > >>the inferior iA64 thing. > >  > > I > > Do you realize the AMD x86-64 architecture isn't even a true 64 bits? J > > Yes, it has 64-bit registers, but its virtual address space is only 48J > > bits. That is actually smaller than Madison's _physical_ address spaceI > > of 50 bits. (Opteron has a 40-bit physical address space.) IA64 has a C > > full 64-bit virtual address space. While x86-64 contains 64-bit J > > extensions to x86-32, Itanium was designed for 64 bits from the ground > > up.  > . > Does any of this make any difference at all. > : > 40 bit addressing will allow the Opteron to address moreC > physical memory than is currently available in any microprocessor  > based system.  > = > The fact that Madison has 50 bit addressing and Opteron has 9 > 40 will make absolutely no difference at all to current 
 > systems. > < > You need to find a feature that matters Keith 40 bit vs 50" > bit physical addressing doesn't. > ? > You have however found a feature that does matter, Opteron is < > based on x86 Itanium isn't, Opteron has all of the current9 > IA32 SW catalogue available to it Itanium has ~500 apps ; > for each OS it supports. Thats a feature that does matter # > but not in the way you hoped for.  >  > > 6 > > Integer registers: 16 for x86-64; 128 for Itanium. > >  > ; > But Opteron Integer performance is stubbornly faster than 7 > Itanium despite being saddled with 1/8 the registers.  >  > 1477 SPECint vs 1322 > 8 > So you have found another feature that doesn't matter. > ; > You should also have know that the measure was wortheless . > because Itanium needs more registers anyway. > @ > > Floating-point registers 8 for x86-64 (it's saddled with theB > > now-ancient x87 floating-point architecture); 128 for Itanium. > >  > ? > Ahh finally you have found a feature that might matter though . > it again may not be the number of registers. > 9 > The trouble is that this one feature thats positive for 6 > Itanium doesn't actually matter to the vast majority& > of people who will buy your systems. > ; > As ever a misplaced effort on your part has been rewarded  > with a poke in the eye.  > 	 > regards  > Andrew Harrison I > > If you look at the SPECfp2000 results at http://spec.org, all the top C > > results, from HP, SGI, Bull, Dell, Supermicro, and ION Computer D > > Systems, are with the Itanium 2.  The #1 top result is for an HPE > > system with Itanium 2 (1.5 Ghz) at 2119. Power4+ (1.7 Ghz) is way G > > behind at 1699, Pentium 4EE (3.2 Ghz) at 1516, Opteron (2.2 Ghz) at I > > 1514, EV7 (1.15 Ghz in GS1280) at 1482, and SPARC (1.32 Ghz) at 1350.    ------------------------------  % Date: Thu, 15 Jan 2004 13:24:17 -0500 ( From: David Froble <davef@tsoft-inc.com>' Subject: Re: 500.000 AMD64's shipped... , Message-ID: <4006DAD1.8040006@tsoft-inc.com>   Keith Parris wrote:   Z > Dirk Munk <munk@home.nl> wrote in message news:<bu5gh4$k0i$1@news4.tilbu1.nb.home.nl>... > S >>However AMD has *one* 32/64 bit architecture for the laptop, the desktop and the  Q >>server environment. Intel will have two very distinctive architectures, 64 bit  B >>for the servers and a rather different 32 bit for anything else. >> > @ > x86-64 may do well on the desktop, but the design is seriously> > compromised for the server space.  Maximum of 8-way (current$ > Superdomes are already at 64-way),    N Lets see, we're comparing a chip to a system?  Credibility takes another dip. K My understanding is that Opteron has on-chip glue for up to 8-way systems.  O IA-64 doesn't have any such, it's all done off-chip.  Can you explain why such  1 couldn't also be done in an Opteron based system?   % > 40 bits or 1 TB of physical address F > space (current Superdomes already support half that address space atE > 512 GB), maximum of 48-bit virtual address space instead of 64-bit,   > and only 16 integer registers. >    Dave   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------  # Date: Thu, 15 Jan 2004 18:25:10 GMT & From: jlsue <jefflsxxxz@sbcglobal.net>' Subject: Re: 500.000 AMD64's shipped...V8 Message-ID: <82md00ppv9b62lha0ll4trv7hrdpihe1fj@4ax.com>  H On Wed, 14 Jan 2004 17:59:59 -0500, JF Mezei <jfmezei.spamnot@istop.com> wrote:  
 >jlsue wrote: K >> So you're saying that it's not valid for HP to reduce 4 server platforms'J >> (PA-RISC, IA32/Proliant, Alpha, MIPS) down to two (and eventually one)? >eG >No, I am saying it is not right for HP to make such a move under false N >pretense.  IA64/epic was a pet project that should have never been allowed to" >go beyond the pet project status.  H Lots of words, all devoid of meaning.  There is no such thing as a falseE pretense here.  Almost any chip can succeed given the right amount of H dedication, commitment, and investment.  It appears that, perhaps, IntelK had a much better combination of those characteristics for IA64 than Compaqe had for Alpha.   > L >When Carly started to date Curly, it was already known that IA64 was a dog.O >Carly could have told Curly that she wanted Alpha real bad, and told Compaq touO >continue work on Alpha at full speed so that EV7 could be unleashed on May 7thgI >when the merger was consumed. HP could have publicly ditched IA64 before M >merced came out, relieving Intel of its responsability to lose more money ono >that losing chip.  G Lots of question begging there.  HP had ALREADY invested in IA64 as thetK replacement for PA-RISC.  HP-UX migration was already in the works, and for J Compaq's side, Tandem had already done some work that way as well (or so I thought).     K What makes you so sure that HP wouldrather have Alpha over IA64?  Alpha was G more middle-aged, and IA64 was just beginning it's lifecycle.  Assuming G that Intel is fully dedicated, comitted, and investing appropriately in H IA64, it's a chance to get in on the gound floor, and get out from underJ their own need to invest people in chip biz (and, to some extent, compiler biz).   F But you're ignoring my response to your initial premise:  "It makes noI sense to drop Alpha and go to IA64 if the IA64 c hip will just be anotherMD niche chip, like Alpha was."  My question shows that, yes, there canI actually be very valid business reasons for making that move anyway.  AndsB it's not clear yet that IA64 will be only a niche chip yet anyway.   >pI >Then, HP could have outsourced Alpha to Intel and give the engineers the 0 >budgets they neeed to progress at a rapid pace.  H Alpha has been outsourced already - it's Intel and IA64, though, instead now.   >sK >HP would have been able to compete head to head with Power with a superior J >archicteture, Carly could have better justified the purchase of Compaq by/ >stating that Alpha was a big asset she needed.   I Architecture smarchitecture.  Only a few people in this world (relativelyaI speaking) care about architecture.  How much business processing can flowoK can they run through the systems for some $$$ amount?  That's what businessS care about.e   >oN >Or better yet, Carly could have focused her efforts on her company and fix itP >up instead of wasting monety to buy Compaq and just ditch Compaq's assets ASAP.  I HP is doing pretty well, tyvm.  From where I sit (consulting), I see sometJ pretty good increase in business.  IA64-based servers don't impact much ofI my job yet, but hen 4 years ago storage didn't either (and now storage isgJ probably 70-80% of my business).  I think HP got some pretty good businessJ out fo the Compaq deal (enterprise customers, consulting, storage, to nameA a few).  They seem to know how to leverage the strengths of those  businesses too.H   --- jls 0 The preceding message was personal opinion only.6 I do not speak in any authorized capacity for anyone,  and certainly not my employer.- (get rid of the xxxz in my address to e-mail)n   ------------------------------  % Date: Thu, 15 Jan 2004 13:20:42 -0500$( From: David Froble <davef@tsoft-inc.com>' Subject: Re: 500.000 AMD64's shipped...e, Message-ID: <4006D9FA.6030806@tsoft-inc.com>   > Keith Parris wrote:7I >> At the moment, AMD's x86-64 is competing mostly against Intel's 32-biti. >> chips, which soundly trounce it in SPECint:  >>   Pentium 4EE (3.2 Ghz)  1620  >>   AMD Opteron (2.2 Ghz)  1477    D The Pentium 4EE was announced to attempt to minimalize the AMD chip Q announcements.  Can you actually find one of these animals?  Do they exist?  How   many have been sold?  P At least Andy boy sticks mainly to reality.  Not that he wants to, but he knows I we'll question any BS he posts.  The above doesn't help your credibility.l   Dave   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Road  Vanderbilt, PA  15486t   ------------------------------  # Date: Thu, 15 Jan 2004 17:50:05 GMTf From: LBohan@mail.esignal.comc Subject: Re: 7.3-1 to 7.3-2e8 Message-ID: <dnjd00t420v8en4qu2i6rod5cc7pqv3dqr@4ax.com>  E On 9 Jan 2004 00:45:18 -0800, bhushann@hotmail.com (Bhushan Narkhede)e wrote:  	 >Hi Guys,c > C >  I am looking at the efforts that might be involved in moving ourh1 >system applications from OpenVMS 7.3-1 to 7.3-2.t > . >can anyone guide on how to proceed on this???G >What all changes have gone in from 7.3-1 to 7.3-2 ??? or point to some-! >kind of documentation on it ???? F >I am trying to get the release notes for 7.3-2, but looks like the HPG >site is undergoing some kind of work or due to some reason not able to0 >get any thing from their site.i >f >Rgds, >Bhushan   if the following is true,  D# 1.  your 7.3-1 systems disk is ODS5E9 2.  you plan to make a stand-alone 7.3-1 on a spare disk.      < SYS$SYSTEM:AXPVMS$PCSI_INSTALL_MIN.COM,  may hang (forever?)9 unless you rename the following files to all uppercase.     ? [SYS0.SYSCOMMON]DEC-AXPVMS-OpenVMS-V0703-1-5.PCSI$DESCRIPTION;1v7 [SYS0.SYSCOMMON]DEC-AXPVMS-OpenVMS-V0703-1-5.PCSI$TLB;1i  5 I believe I found this soln, only after searching on n; HP's Technical Knowledge Base (login/registration required)    ------------------------------  # Date: Thu, 15 Jan 2004 18:20:14 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond)6 Subject: Re: 7.3-1 to 7.3-2s3 Message-ID: <yRANb.12443$Ic5.6421@news.cpqcorp.net>h  9 In article <dnjd00t420v8en4qu2i6rod5cc7pqv3dqr@4ax.com>, i LBohan@mail.esignal.com writes:e   >if the following is true,  $ >1.  your 7.3-1 systems disk is ODS5: >2.  you plan to make a stand-alone 7.3-1 on a spare disk. >    t= >SYS$SYSTEM:AXPVMS$PCSI_INSTALL_MIN.COM,  may hang (forever?)i: >unless you rename the following files to all uppercase.   >j@ >[SYS0.SYSCOMMON]DEC-AXPVMS-OpenVMS-V0703-1-5.PCSI$DESCRIPTION;18 >[SYS0.SYSCOMMON]DEC-AXPVMS-OpenVMS-V0703-1-5.PCSI$TLB;1   This is correct.6 However, this has been fixed for OpenVMS Alpha V7.3-2. (Also for OpenVMS I64.)i  4 So this should not be an issue for V7.3.1 -> V7.3-2.& Please let me know if you think it is.   --  J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Thu, 15 Jan 2004 08:16:02 -0600r6 From: "Craig A. Berry" <craigberry@mac.com.spamfooler>H Subject: Re: ANN: FreeeTDS 0.62 Sybase & MS SQL client software releasedT Message-ID: <craigberry-5E16C8.08160215012004@dsl081-159-101.chi1.dsl.speakeasy.net>  * In article <40060F1B.9040704@pacbell.net>,E  "bgInc. - You'll go where we want you to go" <ereiamjh@pacbell.net> e  wrote:l   > Craig A. Berry wrote:iI > > FreeTDS is a collection of free, open source libraries and utilities eL > > for communicating with databases that use the TDS (Tabular Data Stream)  > > protocol.  o  I > This built OK for me on 7.3-1, but tsql fails in iconv.c. It's looking yH > for (from memory) UTF-8_OSI8859-1.ICONV and others (that's the one it @ > appears to fail on). Standard (?) VMS only has a few files in G > sys$i18n_iconv, and not the ones it's looking for. I also rebuilt it lI > using the 'minimal' iconv supplied, but this also fails (I haven't yet h" > had time to look much into why).  < Hmm.  Both options worked in my testing, but you may have a @ configuration that fell through the cracks of what gets handled ? properly.  The configuration process should detect whether the e= necessary character set translations are present and use the t? replacement iconv if not.  What I test for are the presence of aC translations between UCS-2 and IS0-8859-1, but perhaps I should be uH testing for UTF-8 rather than (or in addtion to) UCS-2; if you have the H latter but not the former (which seems unlikely) then the configuration ' script might set things up incorrectly.i  : In order to force use of the replacement iconv, you can do   $ define SYS$I18N_ICONV NL:A  E before configuring.  If you have a previous build, you should also dol   $ mms cleanr  C before configuring and rebuilding.  If that doesn't get you going, a define the logicals    $ define TDSDUMP tdsdump.log( $ define TDSDUMPCONFIG tdsdumpconfig.log  G and see if anything shows up in the logs that reveals what the problem b is.e  7 > Is this working for anybody? Do I need to install an   > 'internationalization kit'?a  E Installing the internationalization kit would almost certainly solve  G the problem you are seeing.  I believe it's on the second CD ("layered  H products") that comes with the system CD; unfortunately that means it's G unavailable to hobbyists dependent on the Montagar distribution.  It's nH a bit backward that in this day and age character set translations have : to be separately installed and aren't part of the base OS.   ------------------------------  % Date: Thu, 15 Jan 2004 05:00:15 -0500t* From: JF Mezei <jfmezei.spamnot@istop.com>' Subject: DECLASER 5100 output paper jaml) Message-ID: <40066481.11CD1BBF@istop.com>l  M Some time ago, someone mentioned they were having problems with paper gettinga; jammed at the very last rollers on top of the 5100 printer.   I I started getting this problem too. I lived with it for some time by just N pressing on the roller whenever a page would come out to make sure it woudln'tK get stuck. But I solved it, it seems, by simply cleaning the rubber rollersa
 with alchool.r   ------------------------------  + Date: Thu, 15 Jan 2004 09:13:36 +0000 (UTC)h% From: bdc@world.std.com (Brian Chase) ' Subject: Re: Detaching processes on VMS-( Message-ID: <bu5lk0$7s7$1@pcls4.std.com>  < In article <LRmNb.8025$Vr7.7892@newssvr22.news.prodigy.com>,@ Michael Austin  <maustin@no-more-spam.firstdbasource.com> wrote: > Steve Young wrote:  P > >   I was wondering if anyone could suggest a tool that would enable me to runQ > > detached processes (not in the spawn/detach sense), much like one can do withcP > > screen under UNIX.  i.e., I start up a process, detach and log out and leaveO > > the process running.  When I log back in I can re-attach to the process and % > > pick up exactly where I left off.  > >  > >   Any ideas?   > H > Using ConsoleWorks allows me to do this on the console, however, only K > one person can use it at a time -- actually, multiple people can connect iK > to it, it is still a single session so you end up with dualing keyboards.e > G > A better description of the problem you are trying to solve might be r > warranted...  I I'm not sure it's so much a problem as it is a desired feature.  The tooli@ "screen" that Steve is describing is an almost infinitely usefulF productivity enhancer and desktop clutter eliminator.  I have it on myH short list of indispensable bits of non-standard software that I like toI have on any Unix system I admin.  And it's a pretty short list-- the onlyi other item on it is "wget".   G On a Unix system with screen installed, and end-user could do somethingd like the following:   H * Pull up a Unix shell window on your desktop workstation at the office.B * Run the screen program, which presents the user with an new but :   otherwise identical shell prompt within the same window.J * Type Ctrl-A-C, which creates a second shell session, placing the user at   this new shell.iG * Type Ctrl-A-C again, which creates a third shell session, placing thei   user at this new shell. > * Run pine (an text based e-mail client) in the current shell.= * Type Ctrl-A-P to go to the previous (second) shell session.g8 * Run a "tail -f" command to monitor the web server log.1 * Type Ctrl-A-P to go to the first shell session. 2 * Run nethack or tetris for terminals until bored.D * Type Ctrl-A-2 to take you to the third shell session where you get=   caught up on your mail (the session numbering starts at 0).iI * Type Ctrl-A-N to cycle around to your first session again to play a bitn   more nethack.nH * Quickly type Ctrl-A-N to cycle to the second session, with the web log*   being monitored, when the boss walks by.C * After the danger has passed, type Ctrl-A-P to get back to playingf
   nethack.I * At the end of the work day, type Ctrl-A-D to detach screen and leave it B   running in the background (where it still holds the nethack, log"   monitoring, and pine processes).C * Drive a miserable hour long commute from the office to your home.30 * Get on your home computer and dial up to work.? * SSH into your desktop workstation, getting to a shell prompt.58 * Type in "screen -r" to reattach your screen processes. * Play nethack a bit more.> * Type Ctrl-A-2 and check your e-mail under your pine session.. * Type Ctrl-A-D to detach your screen process.	 * Logout.n * Sleep.C * Drive a miserable hour long commute to the office from your home. E * Login to your workstation, pull up a shell window, and reattach to     the screen process.  * lather, rinse, repeat, etc...s  J And that's really just the basics of what screen offers.  Each instance ofC screen you run can hold up to 10 sessions, and you can run multiple J instances of screen on a given system, attaching and dettaching to them asI necessary.  You can also take snapshots of a given session's text, savingmG it to file.  You can log a session to file.  You can cut-and-paste texttG between a screen's sessions.  You can lock your screen session when youi walk away from your terminal.s  A Another way to describe screen is that it's a software version ofrD something like a multi-session VT420 terminal, with the exception ofJ screen being a magnitude or two more useful and flexible than its physicalF counterpart.  That ability to detach from the screen sessions and thenH later login and reattach to them from some completely different locationA is immensely useful, even without screen's multi-session support.   G So, as an example, let us say that I've just run the edit command on my>J VMS system.  Is there some way that I can either "background" (in the UnixH sense) this edit or barring that, is there a way that I can logout whileJ still, somehow, leaving my edit session running on the VMS system?  And isH it then possible to reattach to that session at some later time, to pick up where I left off?  @ TOPS-10 supported this behavior with the 'detach', 'attach', andE 'reattach' commands, so the concept has been around for rather a long L while.  Certainly the original designers of VMS would've been aware of it.  D The VMS attach command looks promising, but it's not clear how or if1 there's a way to explicitly detach a VMS process.g   -brian.i --  F --- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ -----5                    Do not fold, mutilate, or spindle.-   ------------------------------  % Date: Thu, 15 Jan 2004 03:42:49 -0600i% From: "Mike Naime" <mnaime@kc.rr.com>o' Subject: Re: Detaching processes on VMS09 Message-ID: <7ktNb.40791$VV4.16805@twister.rdc-kc.rr.com>t  0 Brian Chase <bdc@world.std.com> wrote in message" news:bu5lk0$7s7$1@pcls4.std.com...> > In article <LRmNb.8025$Vr7.7892@newssvr22.news.prodigy.com>,B > Michael Austin  <maustin@no-more-spam.firstdbasource.com> wrote: > > Steve Young wrote: >eK > > >   I was wondering if anyone could suggest a tool that would enable met to runK > > > detached processes (not in the spawn/detach sense), much like one can  do with L > > > screen under UNIX.  i.e., I start up a process, detach and log out and leavetE > > > the process running.  When I log back in I can re-attach to then process andi' > > > pick up exactly where I left off.t > > >r > > >   Any ideas? > >oI > > Using ConsoleWorks allows me to do this on the console, however, only L > > one person can use it at a time -- actually, multiple people can connectB > > to it, it is still a single session so you end up with dualing
 keyboards. > > H > > A better description of the problem you are trying to solve might be > > warranted... >oK > I'm not sure it's so much a problem as it is a desired feature.  The tool0B > "screen" that Steve is describing is an almost infinitely usefulH > productivity enhancer and desktop clutter eliminator.  I have it on myJ > short list of indispensable bits of non-standard software that I like toK > have on any Unix system I admin.  And it's a pretty short list-- the onlyt > other item on it is "wget".n >aI > On a Unix system with screen installed, and end-user could do somethinge > like the following:s >mJ > * Pull up a Unix shell window on your desktop workstation at the office.C > * Run the screen program, which presents the user with an new but < >   otherwise identical shell prompt within the same window.L > * Type Ctrl-A-C, which creates a second shell session, placing the user at >   this new shell. I > * Type Ctrl-A-C again, which creates a third shell session, placing thet >   user at this new shell.e@ > * Run pine (an text based e-mail client) in the current shell.? > * Type Ctrl-A-P to go to the previous (second) shell session. : > * Run a "tail -f" command to monitor the web server log.3 > * Type Ctrl-A-P to go to the first shell session. 4 > * Run nethack or tetris for terminals until bored.F > * Type Ctrl-A-2 to take you to the third shell session where you get? >   caught up on your mail (the session numbering starts at 0).nK > * Type Ctrl-A-N to cycle around to your first session again to play a bitn >   more nethack.pJ > * Quickly type Ctrl-A-N to cycle to the second session, with the web log, >   being monitored, when the boss walks by.E > * After the danger has passed, type Ctrl-A-P to get back to playingl >   nethack.K > * At the end of the work day, type Ctrl-A-D to detach screen and leave it D >   running in the background (where it still holds the nethack, log$ >   monitoring, and pine processes).E > * Drive a miserable hour long commute from the office to your home.u2 > * Get on your home computer and dial up to work.A > * SSH into your desktop workstation, getting to a shell prompt.u: > * Type in "screen -r" to reattach your screen processes. > * Play nethack a bit more.@ > * Type Ctrl-A-2 and check your e-mail under your pine session.0 > * Type Ctrl-A-D to detach your screen process. > * Logout. 
 > * Sleep.E > * Drive a miserable hour long commute to the office from your home.wF > * Login to your workstation, pull up a shell window, and reattach to >   the screen process.y! > * lather, rinse, repeat, etc...o >nL > And that's really just the basics of what screen offers.  Each instance ofE > screen you run can hold up to 10 sessions, and you can run multiple L > instances of screen on a given system, attaching and dettaching to them asK > necessary.  You can also take snapshots of a given session's text, saving I > it to file.  You can log a session to file.  You can cut-and-paste text-I > between a screen's sessions.  You can lock your screen session when youo > walk away from your terminal.r >oC > Another way to describe screen is that it's a software version ofsF > something like a multi-session VT420 terminal, with the exception ofL > screen being a magnitude or two more useful and flexible than its physicalH > counterpart.  That ability to detach from the screen sessions and thenJ > later login and reattach to them from some completely different locationC > is immensely useful, even without screen's multi-session support.i >oI > So, as an example, let us say that I've just run the edit command on my2L > VMS system.  Is there some way that I can either "background" (in the UnixJ > sense) this edit or barring that, is there a way that I can logout whileL > still, somehow, leaving my edit session running on the VMS system?  And isJ > it then possible to reattach to that session at some later time, to pick > up where I left off?  J This sounds exactly like LAT sessions on the old decservers.  You had a VTH with up to 8 sessions on the DECserver port that you where plugged into.L You were able to switch between the active session by (programmable) controlG codes.   I used CTRL-F for Forward, and CTRL-B for Backwards.  The onlyIL problem was that the terminal servers had to be on the same network, or on aK network that allowed you to pass LAT traffic from one LAN to another LAN if  you didi not want to SET HOST.  H I remember our old Mainframe used to be setup with the Virtual TerminalsK that you could re-attach too if you lost connection.  But I'm not real sureoL how they set those up.  We did not use them on the 3100's that I set/used at my end.r  B > TOPS-10 supported this behavior with the 'detach', 'attach', andG > 'reattach' commands, so the concept has been around for rather a long L > while.  Certainly the original designers of VMS would've been aware of it.F > The VMS attach command looks promising, but it's not clear how or if3 > there's a way to explicitly detach a VMS process.  > 	 > -brian.d > --H > --- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ -----7 >                    Do not fold, mutilate, or spindle.o   ------------------------------  # Date: Thu, 15 Jan 2004 13:50:39 GMT5( From: g_hakansson@kafsv3.CTH.CPQCORP.NET' Subject: Re: Detaching processes on VMS03 Message-ID: <PUwNb.12408$MK4.9232@news.cpqcorp.net>2  B     I seem to recall a program called BOSS that would do something?     similar to the description that Brian has provided. Also, IeB     recall something called SWIM that would do something like that     as well.  +     You can get BOSS.ZIP and SWIM.ZIP from:I  1     http://www.tmk.com/ftp/vms-freeware/fileserv/h  5     I haven't tried either so if you do try them out,o#     let us know in this news group.M  
     Have fun,o  
     Graham     P In article <bu5lk0$7s7$1@pcls4.std.com>, bdc@world.std.com (Brian Chase) writes:= >In article <LRmNb.8025$Vr7.7892@newssvr22.news.prodigy.com>, A >Michael Austin  <maustin@no-more-spam.firstdbasource.com> wrote:t >> Steve Young wrote:g >sQ >> >   I was wondering if anyone could suggest a tool that would enable me to runnR >> > detached processes (not in the spawn/detach sense), much like one can do withQ >> > screen under UNIX.  i.e., I start up a process, detach and log out and leaveyP >> > the process running.  When I log back in I can re-attach to the process and& >> > pick up exactly where I left off. >> > o >> >   Any ideas?  n >> mI >> Using ConsoleWorks allows me to do this on the console, however, only pL >> one person can use it at a time -- actually, multiple people can connect L >> to it, it is still a single session so you end up with dualing keyboards. >> sH >> A better description of the problem you are trying to solve might be  >> warranted...  >nJ >I'm not sure it's so much a problem as it is a desired feature.  The toolA >"screen" that Steve is describing is an almost infinitely usefultG >productivity enhancer and desktop clutter eliminator.  I have it on my I >short list of indispensable bits of non-standard software that I like toiJ >have on any Unix system I admin.  And it's a pretty short list-- the only >other item on it is "wget". >aH >On a Unix system with screen installed, and end-user could do something >like the following: > I >* Pull up a Unix shell window on your desktop workstation at the office. C >* Run the screen program, which presents the user with an new but m; >  otherwise identical shell prompt within the same window.wK >* Type Ctrl-A-C, which creates a second shell session, placing the user atn >  this new shell.H >* Type Ctrl-A-C again, which creates a third shell session, placing the >  user at this new shell.? >* Run pine (an text based e-mail client) in the current shell.7> >* Type Ctrl-A-P to go to the previous (second) shell session.9 >* Run a "tail -f" command to monitor the web server log.12 >* Type Ctrl-A-P to go to the first shell session.3 >* Run nethack or tetris for terminals until bored.eE >* Type Ctrl-A-2 to take you to the third shell session where you getp> >  caught up on your mail (the session numbering starts at 0).J >* Type Ctrl-A-N to cycle around to your first session again to play a bit >  more nethack.I >* Quickly type Ctrl-A-N to cycle to the second session, with the web log + >  being monitored, when the boss walks by.bD >* After the danger has passed, type Ctrl-A-P to get back to playing >  nethack. J >* At the end of the work day, type Ctrl-A-D to detach screen and leave itC >  running in the background (where it still holds the nethack, log # >  monitoring, and pine processes).tD >* Drive a miserable hour long commute from the office to your home.1 >* Get on your home computer and dial up to work.-@ >* SSH into your desktop workstation, getting to a shell prompt.9 >* Type in "screen -r" to reattach your screen processes.: >* Play nethack a bit more. ? >* Type Ctrl-A-2 and check your e-mail under your pine session. / >* Type Ctrl-A-D to detach your screen process.u
 >* Logout.	 >* Sleep..D >* Drive a miserable hour long commute to the office from your home.F >* Login to your workstation, pull up a shell window, and reattach to  >  the screen process.  >* lather, rinse, repeat, etc... > K >And that's really just the basics of what screen offers.  Each instance of D >screen you run can hold up to 10 sessions, and you can run multipleK >instances of screen on a given system, attaching and dettaching to them as J >necessary.  You can also take snapshots of a given session's text, savingH >it to file.  You can log a session to file.  You can cut-and-paste textH >between a screen's sessions.  You can lock your screen session when you >walk away from your terminal. >-B >Another way to describe screen is that it's a software version ofE >something like a multi-session VT420 terminal, with the exception of.K >screen being a magnitude or two more useful and flexible than its physicalaG >counterpart.  That ability to detach from the screen sessions and then I >later login and reattach to them from some completely different locationoB >is immensely useful, even without screen's multi-session support. >sH >So, as an example, let us say that I've just run the edit command on myK >VMS system.  Is there some way that I can either "background" (in the Unix,I >sense) this edit or barring that, is there a way that I can logout whileoK >still, somehow, leaving my edit session running on the VMS system?  And isDI >it then possible to reattach to that session at some later time, to picke >up where I left off?  >eA >TOPS-10 supported this behavior with the 'detach', 'attach', andwF >'reattach' commands, so the concept has been around for rather a longM >while.  Certainly the original designers of VMS would've been aware of it.  eE >The VMS attach command looks promising, but it's not clear how or if.2 >there's a way to explicitly detach a VMS process. >  >-brian. >-- G >--- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ -----i6 >                   Do not fold, mutilate, or spindle.   ------------------------------  % Date: Thu, 15 Jan 2004 01:40:33 -0500,* From: JF Mezei <jfmezei.spamnot@istop.com>) Subject: Re: DHCP/BOOTP/PXE under OPENVMSi) Message-ID: <400635C2.72DC5E55@istop.com>r   Harald Pollak wrote:I > PXE is an enhancement for the DHCP for remote booting, i'm not sure bute > ithink its part of PC99. > * > PXE means Preboot eXecution Environment.9 > http://www.pix.net/software/pxeboot/archive/pxespec.pdfo  7 Ok, it looks like it is the intel IA64 specific stuff. e  J THE VMS DHCP server predated IA64 by many years so it have no knowledge of those additional PXE options.   J However, it *MIGHT* be possible to manually add those options (if you knowK what numbers they use) in the text configuration files (but not through thei GUI DHCP configuration program.s   ------------------------------    Date: 15 Jan 2004 07:21:42 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)5Q Subject: Re: Faced with flagging PC sales,  HPand others are pushing aggressively 3 Message-ID: <GXgEVPdTDKtJ@eisner.encompasserve.org>   b In article <100bhi3tik2po18@corp.supernews.com>, Greg Cagle <news@removethisgregcagle.com> writes: > I > I've worked with and for HP since the late 80s, and have never heard ofr? > "HP-PARC". A Google search brings up a few hits, but they are B > all obvious typos or errors. An internal search of HP's intranet > reveals no hits.  K    That could be like looking for PDP-8 at DEC/Compaq/HP sites.  It's been  ,    a long time since I saw the HP-PARC name.   ------------------------------   Date: 15 Jan 2004 14:27:58 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)Q Subject: Re: Faced with flagging PC sales,  HPand others are pushing aggressivelyc9 Message-ID: <bu681d$e2ssh$1@ID-135708.news.uni-berlin.de>0  0 In article <100bhi3tik2po18@corp.supernews.com>,2 	Greg Cagle <news@removethisgregcagle.com> writes:6 > Bob Koehler said the following on 1/14/2004 1:48 PM: > e >> In article <100b2n1lp6kdua9@corp.supernews.com>, Greg Cagle <news@removethisgregcagle.com> writes:n >> d. >>>I assume you're talking about PA-RISC here. >> 3 >>  K >>    HP used to talk about HP-PARC and PA-RISC, but always made them soundaE >>    the same.  I always assumed they only had one RISC architecture C >>    back then, but two different names for two slightly different  >>    aspects. >> eI >>    Does anyone know what the real meaning/difference is of HP-PARC and  >>    PA-RISC? > I > I've worked with and for HP since the late 80s, and have never heard of ? > "HP-PARC". A Google search brings up a few hits, but they aretB > all obvious typos or errors. An internal search of HP's intranet > reveals no hits.  G I always thought that HP-PARC stood for tht Hewlett Packard - Palo AltoaA Research Center.  In a report on Venture Capital I find the line:rC     "Few anchor companies or organizations (eg. HP-PARC, Stanford)"c    But I could be wrong.......  :-)   bill   -- ,J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   .   ------------------------------  % Date: Thu, 15 Jan 2004 02:15:02 -0500)* From: JF Mezei <jfmezei.spamnot@istop.com>$ Subject: Re: file and record formats) Message-ID: <40063DD5.B276A715@istop.com>o   Kaleb Pederson wrote:AB > Actually, I was saying that it *did* work.  To be more specific: > 
 > these work:  >  > char tmp[1024];o > ifstream in("filename"); > in.getline(tmp,1024);u  I No, they doN,t work under VMS since for you, they bypass the standard RMSlL processing whcih would feed your application Unix style data from the file.   N I have looked at some of the documentation and I have seen no reference in theE iostream documents to any RMS extensions which are available in the CtH language. The iostream documentation does specify that it provides a rawM stream of data from/to a file and there is also mention that C++ doesn't havenH any "binary" flag when you open a file (which implies that all access is binary, aka: raw data).p  M if you can use the standard C IO routines, then you'll find your program willcG behave properly no matter what the type of file you are reading, excepteJ certain types without any "record" concepts where a single read may exceedK your process's internal buffer size (to get to those files, you need to uset RMS routines).   ------------------------------  % Date: Thu, 15 Jan 2004 07:18:47 -0800 - From: Kaleb Pederson <kpederson@mail.ewu.edu>2$ Subject: Re: file and record formats2 Message-ID: <200401150718.47803.kpederson@ewu.edu>  6 On Wednesday 14 January 2004 11:15 pm, JF Mezei wrote: > Kaleb Pederson wrote: D > > Actually, I was saying that it *did* work.  To be more specific: > >r > > these work:x > >b > > char tmp[1024];: > > ifstream in("filename"); > > in.getline(tmp,1024);b >hK > No, they doN,t work under VMS since for you, they bypass the standard RMS<M > processing whcih would feed your application Unix style data from the file.i  E I must not have made myself clear -- it *did* work as they correctly lK recognized the Variable Length records (the ones preceeded by the two byte  # length that contained no CR or LF).v  K It was for those where I had specified my delimiter where it failed.  Even bK though I had not specifically done anything to avoid using the RMS system, iI when I specified that newline was my delimiter it bypassed the specifics.y  L > I have looked at some of the documentation and I have seen no reference inK > the iostream documents to any RMS extensions which are available in the CeJ > language. The iostream documentation does specify that it provides a rawJ > stream of data from/to a file and there is also mention that C++ doesn'tL > have any "binary" flag when you open a file (which implies that all access > is binary, aka: raw data).  C Actually, there is a binary flag.  The default is to use text mode.e  M In ios_base, which is inherited by the different stream classes, there is an  5 'openmode' which can be specified after the filename:i   // for reads) ifstream in("filename.txt",ios_base::in);  or // for write or append8 ifstream in("filename.txt",ios_base::out|ios_base::app); or // for binary mode* ifstream in("filename.txt",ios_base::bin);  I You could find this in "The C++ Programming Language," third edition, by d" Bjarne Stroustrup, section 21.5.1.  I He also states in that section "the actual values of openmodes and their  M meanings are implementation-defined. Please consult your systems and library 4I manual for details - and experiment.  The comments [describing the above  C flags] should give some idea of the intended meaning of the modes."   J > if you can use the standard C IO routines, then you'll find your programG > will behave properly no matter what the type of file you are reading,tL > except certain types without any "record" concepts where a single read mayM > exceed your process's internal buffer size (to get to those files, you needn > to use RMS routines).l  L I wasn't aware that fgets read in an entire line, but I'm sure that you are , right and that it will do the correct thing.   I hope that makes more sense.d   --Kalebl   ------------------------------  % Date: Thu, 15 Jan 2004 17:15:00 +0100.- From: "Winfried Bergmann" <dummy@empuron.com>d$ Subject: Re: file and record formats9 Message-ID: <bu6e6f$e3aab$1@ID-170759.news.uni-berlin.de>s  @ "Kaleb Pederson" <kpederson@mail.ewu.edu> schrieb im Newsbeitrag, news:200401150718.47803.kpederson@ewu.edu...8 > On Wednesday 14 January 2004 11:15 pm, JF Mezei wrote: > > Kaleb Pederson wrote:rF > > > Actually, I was saying that it *did* work.  To be more specific: > > >' > > > these work:a > > >t > > > char tmp[1024];t > > > ifstream in("filename"); > > > in.getline(tmp,1024);  > >uI > > No, they doN,t work under VMS since for you, they bypass the standardA RMS I > > processing whcih would feed your application Unix style data from the  file.g > F > I must not have made myself clear -- it *did* work as they correctlyL > recognized the Variable Length records (the ones preceeded by the two byte% > length that contained no CR or LF).s > L > It was for those where I had specified my delimiter where it failed.  EvenL > though I had not specifically done anything to avoid using the RMS system,K > when I specified that newline was my delimiter it bypassed the specifics.i >nK > > I have looked at some of the documentation and I have seen no reference  inK > > the iostream documents to any RMS extensions which are available in the- C-L > > language. The iostream documentation does specify that it provides a rawL > > stream of data from/to a file and there is also mention that C++ doesn'tG > > have any "binary" flag when you open a file (which implies that all  access > > is binary, aka: raw data). >tE > Actually, there is a binary flag.  The default is to use text mode.i >4K > In ios_base, which is inherited by the different stream classes, there is  an7 > 'openmode' which can be specified after the filename:t > 
 > // for ready+ > ifstream in("filename.txt",ios_base::in);H > or > // for write or append: > ifstream in("filename.txt",ios_base::out|ios_base::app); > or > // for binary mode, > ifstream in("filename.txt",ios_base::bin); >oJ > You could find this in "The C++ Programming Language," third edition, by$ > Bjarne Stroustrup, section 21.5.1. > J > He also states in that section "the actual values of openmodes and theirF > meanings are implementation-defined. Please consult your systems and library-J > manual for details - and experiment.  The comments [describing the aboveE > flags] should give some idea of the intended meaning of the modes."c >qL > > if you can use the standard C IO routines, then you'll find your programI > > will behave properly no matter what the type of file you are reading,iJ > > except certain types without any "record" concepts where a single read maytJ > > exceed your process's internal buffer size (to get to those files, you need > > to use RMS routines).o >oI > I wasn't aware that fgets read in an entire line, but I'm sure that youe are . > right and that it will do the correct thing.  L Maybe you got confused with scanf, which stops after the first space (e.g.),J using it this way: fscanf (stdin, "%s", string). But there are ways around this limitation.   ------------------------------  % Date: Thu, 15 Jan 2004 11:53:01 -0600 ( From: Wayne Sewell <wayne@tachysoft.com>M Subject: Re: HP Integrity Superdome server sets world record for 10-terabyte,a/ Message-ID: <00A2BEE6.974D097E.5@tachysoft.com>3  2 >From: keithparris_NOSPAM@yahoo.com (Keith Parris) >X-Newsgroups: comp.os.vmsN >Subject: Re: HP Integrity Superdome server sets world record for 10-terabyte,' >         single-system TPC-H benchmark ! >Date: 15 Jan 2004 07:59:37 -0800    > G >The HP Corporate advertising folks are interested only in building theeE >overall HP brand image. They tell me the only way a specific product6D >name would be mentioned in such an ad is if the customer insists onE >it. (So now, loyal OpenVMS customers, if you're to be featured in ans. >HP corporate ad, you know what your duty is.)    K That seems like a strange policy, expecting the customers to do the productXL promotion.  Even if the corporate people have this attitude, the parts of hpN that *do* care about the particular product should at least *suggest* that theL customer do this, if that's the only way that it can be done.  Assuming that3 the vms proponents know about the ad ahead of time.r  H The customer may not expect that it would be necessary to coerce hp intoH promoting their own product, especially one available *only* from hp, asL opposed to a billy/eunuchs "solution" that any dickweed vendor could supply.   Wayne O ===============================================================================sN Wayne Sewell, Tachyon Software Consulting  (281)812-0738   wayne@tachysoft.com; http://www.tachysoft.com/www/tachyon.html and wayne.html   CO ===============================================================================yB Jed Clampett, checking into hotel: "This place got a cement pond?", 	Ellie May: "And do yuh let critters in it?"   ------------------------------    Date: 15 Jan 2004 07:59:37 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris).Y Subject: Re: HP Integrity Superdome server sets world record for 10-terabyte, single-systr= Message-ID: <cf15391e.0401150759.63ab0513@posting.google.com>s  { "bgInc. - You'll go where we want you to go" <ereiamjh@pacbell.net> wrote in message news:<400614C4.6070903@pacbell.net>...s > Great, for HP-UX...   F Also shows the performance of the Superdome platform that VMS now also  has been demonstrated to run on.  3 > now why don't they mention VMS in the usps story i  F The HP Corporate advertising folks are interested only in building theD overall HP brand image. They tell me the only way a specific productC name would be mentioned in such an ad is if the customer insists on D it. (So now, loyal OpenVMS customers, if you're to be featured in an- HP corporate ad, you know what your duty is.)w   ------------------------------  % Date: Thu, 15 Jan 2004 02:49:14 -0500t* From: JF Mezei <jfmezei.spamnot@istop.com>; Subject: Re: HPQ stock to be dual-listed on NYSE and NASDAQ5) Message-ID: <400645D7.14FC0A5A@istop.com>    Keith Parris wrote:nH > HP (NYSE:HPQ) today announced that it has filed an application to dual3 > list its common stock on the Nasdaq Stock Market.t   read:p  + HP wants to keep NASDAQ as a good customer.r   ------------------------------  % Date: Thu, 15 Jan 2004 04:58:23 -0500i* From: JF Mezei <jfmezei.spamnot@istop.com>6 Subject: Humongus bug in DECTERM (postscript printing)) Message-ID: <40066411.15900218@istop.com>,  : Digital/Compaq/HP are guilty of destroying a whole forest.  L On DECTERM (VAX VMS 7.2), you can select text and ask to print it. The printM dialogue asks you what format you want to print, and this include Postscript.e  L So, figuring I was printing to a postscript printer, I chose "postcript" and3 queued the print job to the postscript print queue.b  M Low and behold, paper started spewing out of the printer indicating all sorts  of errors.... :-)e  N Seems that even if you specify postscript output, DECTERM still generates just ASCII text.b  K perhaps DECTERM should override this and add a /DATA=ANSI flag to the printoS job. (or remove the postscript option if it is incapable of generating postscript).-   ------------------------------    Date: 15 Jan 2004 10:20:00 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) A Subject: IBM announces DB2 for HP Integrity Itanium-based servers = Message-ID: <cf15391e.0401151020.313247ec@posting.google.com>t  7 Lest anyone think IBM doesn't take Itanium seriously...   0 This quote is taken directly from IBM's website:0 http://www-306.ibm.com/software/data/hp/ega.html  C "Itanium-based solutions from HP have come of age, providing servernE and workstation users with the applications, powerful performance and B return on investment (ROI) benefits that are enabling customers to- quickly respond to their business challenges.d  @ Itanium-based solutions are here today, performing the real-lifeF information management jobs that require "heavy lifting". HP Integrity7 servers, based on the Intel Itanium processors, are themA standards-based, innovative building blocks you can use today, to:@ realize price and costs savings while gaining greater stability,5 efficiency and agility within your IT infrastructure.s  D IBM Data Management solutions featuring IBM DB2 UDB can be installedE on HP Integrity servers running Linux, Windows Server 2003, and HP-UXs v2."  ( Quite an endorsement from IBM there. :-)  < It will be interesting to see, as the UNIX/Linux Portability? Initiative makes such tasks easier, if IBM eventually announces & support for DB2 on OpenVMS on Itanium.   ------------------------------   Date: 15 Jan 2004 18:29:50 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)E Subject: Re: IBM announces DB2 for HP Integrity Itanium-based servers 9 Message-ID: <bu6m6u$edro0$1@ID-135708.news.uni-berlin.de>   = In article <cf15391e.0401151020.313247ec@posting.google.com>, 4 	keithparris_NOSPAM@yahoo.com (Keith Parris) writes: > F > IBM Data Management solutions featuring IBM DB2 UDB can be installedG > on HP Integrity servers running Linux, Windows Server 2003, and HP-UXwG                                   ^^^^^  ^^^^^^^^^^^^^^^^^^^      ^^^^^h > v2." >   I And, as the faithful are quick to point out, this is comp.os.VMS, so whatl6 exactly does this OT posting have to do with it?   ;-)   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   s   ------------------------------  + Date: Thu, 15 Jan 2004 10:19:44 +0000 (UTC)s% From: bdc@world.std.com (Brian Chase) Y Subject: Re: Intel to chip away at Itanium prices <- or ... I want my cheap workst. to ru ( Message-ID: <bu5pg0$q2e$1@pcls4.std.com>  3 In article <ClB0LAsGLVxt@eisner.encompasserve.org>,r, Rob Young <young_r@encompasserve.org> wrote:  G > http://www.tpc.org/results/individual_results/HP/hp_tpcc_sd1.5_es.pdfe > - > 	You see a 16 CPU cell board costs $275000.s > > > 	If you trip out to Dell, you see additional 2.8 GHz 2 MByte> > 	Xeon CPUs cost about $4500.  That appears to where Intel isJ > 	headed with Itanium pricing.  Makes you wonder how much market pressureA > 	Sun and IBM can sustain with UltraSparc and Power.  AMD 64-bite. > 	at the low-end and Itanium at the high-end.  H I'd mostly agree with you on the state of Sun, though I never personallyH never hit any issues with their CPU performance.  I've found I/O to be a) more relevant concern in my applications.,  G I don't think IBM have any cause for worry with the Power architecture,mA they already know how to do it cheaply.  The Apple G5 PowerPC 970 J processor, made by IBM, is a variant of the Power4, and they're selling atH $1500/proc in Apple's Xserve.  The Power5 derived PowerPC 980 is rumoredJ to be out by the end of this year.  I'd expect it to be roughly comparable in price to the 970.   -brian.  -- sF --- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ -----5                    Do not fold, mutilate, or spindle.k   ------------------------------  % Date: Thu, 15 Jan 2004 11:05:03 +0100 * From: Paul Sture <nospam@sture.homeip.net>Y Subject: Re: Intel to chip away at Itanium prices <- or ... I want my cheap workst. to ru 0 Message-ID: <400673DF.7851C1B6@sture.homeip.net>   John Smith wrote:g > * > Andrew Harrison SUNUK Consultancy wrote: > > John Smith wrote:l > >> Fabio Cardoso wrote:n > >> > >>> Clickv > >>>t? > >>> http://news.com.com/2100-1006_3-5140486.html?tag=nefd_topt > >>>e > >>>  > >>>sC > >>> Intel wants to remove price as a barrier to the acceptance ofmF > >>> Itanium servers, a goal that could allow the processor to becomeI > >>> the company's primary server chip in the second half of the decade.e > >>>fH > >>> The Santa Clara, Calif.-based chipmaker is working on chipsets andE > >>> other products and technologies that will make an Itanium-basedhI > >>> server no more expensive than a similar machine powered by its XeonnI > >>> chip by 2007, said Mike Fister, senior vice president of the server  > >>> products group at Intel. > >>>tA > >>> Because Itanium can provide more performance than Xeon, the F > >>> elimination of the current, often substantial, price discrepancyB > >>> could then permit Itanium to become Intel's principal serverI > >>> offering. If successful, the strategy could allow Intel to begin toh  > >>> phase out Xeon after 2006. > >> > >> > >>I > >> Looks like EV79 will be needed after all....possibly even EV8 by thehD > >> time Intel meets its aggressive 1999 (oooops) , er...um... 2007* > >> launch date for lower cost IA64 chips > >> > >>C > > IDC have just downgraded their estimates for the total value ofg! > > the Itanium market yet again.r > > @ > > Early on in the project/experiment/dissaster (you chose) IDC? > > were projecting a market of ~28 billion dollars for Itanium  > > based systems by 2007. > >uG > > Subsequent events led them to reduce this in a series of reductionsaD > > to $8.7 billion by 2007. IDC has now reduced this further to 7.5B > > billion for 2007 because of increased competition from AMD and > > other vendors. > >tA > > This suggests that even if choristers on this newsgroup thinks@ > > that AMD64 doesn't compete with Itanium that this view isn't  > > shared by industry analysts. > >m> > > Even worse HP's EBU currently does ~$5.7 billion a year in@ > > server revnues which is getting perilously close to the size> > > that people are expecting the entire Itanium market to be. > >s; > > Or put another way Intel end up fabbing Itaniums for HPn > > and no one else. > J > Not to worry....soon Epson or Canon will develop an jet ink printer thatK > uses photodegradable conductive inks that can be applied in layers, so HPlL > can have one printer sitting on a desktop fabbing all the IA64 chips it'll > ever need.  ;-)i > F > Who knows...maybe the printer will be able to use HP printer paper!!  E But remember, when the ink cartridge runs out, it might be cheaper too( buy a new printer than a new cartridge.    --  
 Paul Sture   ------------------------------    Date: 15 Jan 2004 02:35:39 -0800. From: fabiopenvms@yahoo.com.br (Fabio Cardoso)Y Subject: Re: Intel to chip away at Itanium prices <- or ... I want my cheap workst. to rui= Message-ID: <f30679fb.0401150235.5b74ca83@posting.google.com>l   Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<bu3m9e$oka$1@new-usenet.uk.sun.com>...  > John Smith wrote:r > > Fabio Cardoso wrote: > > 	 > >>Clickw > >>= > >>http://news.com.com/2100-1006_3-5140486.html?tag=nefd_topu > >> > >> > >>I > >>Intel wants to remove price as a barrier to the acceptance of Itanium J > >>servers, a goal that could allow the processor to become the company's9 > >>primary server chip in the second half of the decade.w > >>F > >>The Santa Clara, Calif.-based chipmaker is working on chipsets andJ > >>other products and technologies that will make an Itanium-based serverH > >>no more expensive than a similar machine powered by its Xeon chip byH > >>2007, said Mike Fister, senior vice president of the server products > >>group at Intel.t > >>? > >>Because Itanium can provide more performance than Xeon, thenJ > >>elimination of the current, often substantial, price discrepancy couldG > >>then permit Itanium to become Intel's principal server offering. If-I > >>successful, the strategy could allow Intel to begin to phase out Xeonu > >>after 2006.o > >  > >  > > M > > Looks like EV79 will be needed after all....possibly even EV8 by the time N > > Intel meets its aggressive 1999 (oooops) , er...um... 2007 launch date for > > lower cost IA64 chips  > >  > > A > IDC have just downgraded their estimates for the total value of  > the Itanium market yet again.  > > > Early on in the project/experiment/dissaster (you chose) IDC= > were projecting a market of ~28 billion dollars for Itanium  > based systems by 2007. > E > Subsequent events led them to reduce this in a series of reductionsoB > to $8.7 billion by 2007. IDC has now reduced this further to 7.5@ > billion for 2007 because of increased competition from AMD and > other vendors. > ? > This suggests that even if choristers on this newsgroup think > > that AMD64 doesn't compete with Itanium that this view isn't > shared by industry analysts. > < > Even worse HP's EBU currently does ~$5.7 billion a year in> > server revnues which is getting perilously close to the size< > that people are expecting the entire Itanium market to be. > 9 > Or put another way Intel end up fabbing Itaniums for HPi > and no one else. > 	 > Regardst > Andrew Harrisonb    C As HP began the Processor Consolidation (Itanium = PA-RISC + Alpha) < I believe other companies will follow the same way ! May be < something like SPARC + PPC or one of these processors going G to AMD hands ! So, would be good to the market to have less processors.nE Less processors means less operetaing systems and applications ports.eE So the customers can worry aboit their business not how many OSes andnL processors they have in their Data Center ! May be a merge of AIX / Solaris.C May be less Linux versions ! I think about 3 or 4 processors in thes@ market is good enough ! I feel the lack of the Open Processor ! D May be SPARC could do that in AMDs hands ! AMD looks more flexible  than Intel.    Regardsf   FC   ------------------------------  % Date: Thu, 15 Jan 2004 07:16:21 -0500m* From: JF Mezei <jfmezei.spamnot@istop.com>Y Subject: Re: Intel to chip away at Itanium prices <- or ... I want my cheap workst. to ruh) Message-ID: <40068491.C76BF1F2@istop.com>    Fabio Cardoso wrote:= > I believe other companies will follow the same way ! May bei= > something like SPARC + PPC or one of these processors going I > to AMD hands ! So, would be good to the market to have less processors.sG > Less processors means less operetaing systems and applications ports.i  
 Not quite.  M Look at Apple. They have, since the Macintosh era, had very specific hardware / requirements to ensure stability of the system.p  M VMS engineers have stated that at least for now, VMS will only run on systemsiM equipped with certain IO devices permitting certain types of drivers. Not allr1 IA64 system will have that type of io capability.a  N In fact, whnen you compare stable enterprise systems you can bet your businessF on and those that are cheaper but less reliable, the difference is the. coordination between OS and specific hardware.   ------------------------------    Date: 15 Jan 2004 07:27:38 -0800. From: spamsink2001@yahoo.com (Alan E. Feldman)< Subject: Re: Modify the logical SYS$STARTUP, why or why not?= Message-ID: <b096a4ee.0401150727.4951f915@posting.google.com>e  a brandon@dalsemi.com (John Brandon) wrote in message news:<04011217214316@dscis6-0.dalsemi.com>... K > I just came across some changes to the logicals SYS$STARTUP in one of our D > remote sites.  I would tend to think that this is not a good idea. > * > Anyone else doing the same type changes? >  > ANy other comments?  >  >  > " > $ show logical /full sys$startup: >    "SYS$STARTUP" [exec] = "MYSTARTUP" (LNM$SYSTEM_TABLE)' >         = "SYS$SYSROOT:[SYS$STARTUP]"- >         = "SYS$MANAGER",@ > 1  "MYSTARTUP" [exec] = "DISK2:[SYSCOMMON]" (LNM$SYSTEM_TABLE)E > 1  "SYS$MANAGER" [exec] = "SYS$SYSROOT:[SYSMGR]" (LNM$SYSTEM_TABLE)M  9  Q 0: Careful. Enough changes like this and you might end up with SYS$SPAGHETTI! :-)m  B 1: Ah, but then you'd have a place to put your spaghetti code. :-)   0: Oh, right! :-)i   Disclaimer: JMHO Alan E. Feldmang   ------------------------------  % Date: Thu, 15 Jan 2004 15:46:03 +0000 0 From: Chris Sharman <chris.sharman@sorry.nospam>< Subject: Re: Modify the logical SYS$STARTUP, why or why not?4 Message-ID: <bu6cjr$naf$1$8300dec7@news.demon.co.uk>   John Brandon wrote:a  K > I just came across some changes to the logicals SYS$STARTUP in one of ourID > remote sites.  I would tend to think that this is not a good idea. > * > Anyone else doing the same type changes?   Done this for years:  G $ def/sys/exec sys$sysroot 'f$trnlnm("sys$sysroot")'/tran=(conc,term),-w(                  sys$common:,clu_common:  + (in SYCONFIG, because it seems to be first)e. It's worked (on VAX & Alpha) up to 7.3 so far.D It may not be ideal, but it saves having to maintain cluster common  stuff on several disks.k> Keep lots of stuff in clu_common, including most of the other K [sysmgr]sy*.com files (and yes, we do have to fix them up at each upgrade).o   Chrisd   ------------------------------  % Date: Thu, 15 Jan 2004 10:33:02 +0100-* From: Paul Sture <nospam@sture.homeip.net>< Subject: Re: Non-WS Process Quotas not in Performance Manual0 Message-ID: <40066C5E.32112551@sture.homeip.net>   Paul Repacholi wrote:o > 7 > "Richard B. Gilbert" <rgilbert88@comcast.net> writes:  > G > > It was a long time ago but I believe the VMS V3.7 supplied a lot of G > > the underpinnings for CI clusters like support for HSCs and the CI.s > > This was in 1984 > G > 3.7 was to a good degree a cluster of one. Used $ENQ/$DEQ etc. 3.4 ona? > had MSCP disk support AIR, forget when the PA driver apeared.  > H > > VMS V4.0, if I recall correctly, was the first release that actuallyC > > allowed you to build a cluster.  Not sure when VMS V4.0 hit the C > > streets but I think 1985 would be a close guess.  I didn't haveaI > > software support at the time and the release after V3.6  was, for me,t	 > > V4.4.t > J > I installed 4.0 on 750s in Jan '85 in NL, so I think it was well out andJ > about by then. I first saw a cluster at the fall 83 DECUS in the chicken& > cooker. Next to the CFS-20 system :) >   F From page 60 (it's page 64 in Acrobat) of "OpenVMS at 20 Nothing Stops it" book (pdf 2.5MB) at   2 http://h71000.www7.hp.com/openvms/20th/vmsbook.pdf  " VMS V3 April 1982 ~10,000 licenses o Support for new processors$   VAX-11/730, VAX 11/725, VAX-11/782# o Asymmetric multiprocessing (ASMP)    for VAX-11/782+ o Support for new architectures, protocols,    busses# o System communication architecture    (SCS)h& o Mass storage control protocol (MSCP)! o Lock management system services. o MONITOR utility fory   performance monitoring o BACKUP$ o Command definition utility for DCL& o Terminal autobaud detection, CTRL/T,   and hangup on logout o SPAWN and ATTACH  ' VAX V4 September 1984 ~ 40,000 licensesr& o Support for new processor  VAX 8600   MicroVAX I/II (v4.1)   VAXstation I/II (v4.1)
 o VAXclusters, o Connection manager o Distributed lock manager# o Distributed file system (F11BXQP)i o Security enhancements  o Command line editing and   command recall o Local area terminal server" o Access control lists implemented o Cluster wide operator controla o Variable prompt strings     G And IIRC, we actually got it in the UK in early 1985, upgrading our 750t in something like March.   -- :
 Paul Sture   ------------------------------  % Date: Thu, 15 Jan 2004 00:25:21 -0700s( From: Lorin Ricker <lorin@locktrack.com>O Subject: Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions pert8 Message-ID: <jhec00htujrs7jqarfau5ppk9phijfofdo@4ax.com>  F On Sat, 10 Jan 2004 18:46:58 -0600, Wayne Sewell <wayne@tachysoft.com> wrote:O >However, all of this depends on applications following the convention of using J >the time system services.  Rogue applications such as the oracle databaseP >product and rdb as polluted by oracle corporation (the original rdb, which usedM >the system services, *did* work with datesim) bypass the time conversion anda# >the real system time is displayed.s > O >Which of course is not what the customer wants, but there's not a damn thing I ( >can do about it.  Bingo, lost customer. > J >We had a customer a while back who really wanted to do this, and he triedO >really hard to convey this issue to someone at oracle in some oracle-sponsored.N >forum (don't remember which of the two database products it was), but nothingN >ever came of it.  I wasn't directly involved, but I gathered the response wasK >something along the lines of: "we're oracle and you're just a customer; ite >doesn't matter what you want".'  C IIRC, *I* was the customer (representative) who'd contacted Wayne'sjD organization and happily discovered his DateSim solution, only to beE disappointed when it didn't work because of Oracle/Rdb's rogue methodaF of accessing system time -- I may have actually coined the term "rogue? application" in this particular context, too.  And yes, Wayne'slC portrayal of Oracle's response to us when we communicated with them * about this problem is reasonably accurate.  & >>What is the actual business problem?L >Loss of sales for a product that some people *might* have bought if not for' >Oracle's superior attitude about this.SK >Like I said, some people *do* want this capability.  I don't how how many,iO >because we have so many customers shut down by this problem that it never gets " >far enough for me to be involved. >Wayne  E In my experience, *lots* of customers & developers would want this...f@ all you have to do is deploy a generally useful application on aF server or cluster located in one timezone, with Internet/WAN-accessingB customers located in multiple timezones.  Users generally think ofF their application in terms of their own local time, rarely or never inE terms of "system time" whatever/wherever that might be.  The Internetp? now makes real DEC's own original vision of *truely distributedt< computing*... we shouldn't have to care, from an application4 perspective, where-in-the-heck is the server system!  D Our own domain is a law enforcement/jail/inmate-records application,B and you can bet that the *only* time that matters to a correctionsB officer is *local* time... when did we book him (the bad guy), how@ long has he been here, when, exactly, do we release him.  Little> things like Congress moving the sun twice yearly (standard vs.E daylight) exasperate them a bit... "system time" in another time zonevD flabbergasts them.  This user perspective (requirement) will *never*' change, regardless of what Oracle says.g  B Our own solutions to this, once we determined that Wayne's DateSim? couldn't help us with Rdb, was to create a set of callbacks and.E calculated fields which adjust Rdb's current_timestamp returned-value E based on the username (we actually store a bit of timezone-id info in,E the "user defined" space in the SYSUAF user-auth record, and fetch it'' at application start-up for each user).   F Brilliant tho' this approach may be, it takes a lot of work to set up,@ and it's code to maintain (for numerous tables & fields)... muchC better would be for both VMS and products like Rdb to *wake up* andiA recognize the world-wide, far-flung nature of users connecting totE server systems, and that users perceive only *local time*, not systeme? time.  And what the heck happens when we colonize the moon &/orc Mars?!!e  F In the best of worlds, provisions for local-time adjustments, per user? &/or process, would be provided by VMS as a native part of timecF reporting, where any/all applications could use it.  But then, it'd beF even more critical for *rogue applications* like Oracle/Rdb (et al) toE *abandon* their mistakenly cherished direct-access to the system time C quadword in favor of appropriate system service calls which provide  this enhanced functionality.  C There's a common, real-world need out there... it's probably rarelyoC been vocalized very well, but here's one thread that's trying to do>7 so.  Hope I've helped.  Oracle, HP, do you hear us yet?    Lorin Ricker   ------------------------------    Date: 15 Jan 2004 02:47:13 -0600+ From: young_r@encompasserve.org (Rob Young).O Subject: Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions per 3 Message-ID: <utof1efWsEYx@eisner.encompasserve.org>e  c In article <jhec00htujrs7jqarfau5ppk9phijfofdo@4ax.com>, Lorin Ricker <lorin@locktrack.com> writes:t   > H > Brilliant tho' this approach may be, it takes a lot of work to set up,B > and it's code to maintain (for numerous tables & fields)... muchE > better would be for both VMS and products like Rdb to *wake up* andrC > recognize the world-wide, far-flung nature of users connecting to G > server systems, and that users perceive only *local time*, not systemaA > time.  And what the heck happens when we colonize the moon &/oro	 > Mars?!!> >    	[snip]f   > E > There's a common, real-world need out there... it's probably rarelyeE > been vocalized very well, but here's one thread that's trying to doe9 > so.  Hope I've helped.  Oracle, HP, do you hear us yet?f >   B 	Nice post, really.  Before you think I'm blasting you... I'm not.E 	It may be as simple as you are nickel/dime compared to big customerswB 	that want the other behavior - always.  For instance, JSTARS.  I D 	imagines JSTARS flys through many timezones many times a day.  The D 	only thing that matters to them is system time (or UTC).  Clusters F 	spanning timezones that aren't working with UTC time are surely in a " 	tiny minority.  My guess/opinion.   				Rob    ------------------------------  % Date: Thu, 15 Jan 2004 05:42:15 -0500i' From: "Main, Kerry" <kerry.main@hp.com>aO Subject: RE: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions pernR Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB237C91@tayexc19.americas.cpqcorp.net>   [snip]  F > UTC).  Clusters spanning timezones that aren't working with UTC time1 are surely in a tiny minority.  My guess/opinion.t >=20	 > 				Robe >=20  
 G'day Rob,  G While that may certainly have been true in the past, one of the popular C solutions to address high availability requirements associated withfG large server consolidation efforts these days (white hot btw) is to use>D cluster technologies. In large consolidation efforts, it is actuallyF fairly common to be collapsing servers from different timezones into a/ centralized (and sometimes multi-site) cluster.h   Regards]  
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660y Fax: 613-591-4477M Email: kerryDOTmainAThpDOTcome. (remove the DOT's and AT for email address)=20   ------------------------------  # Date: Thu, 15 Jan 2004 11:05:59 GMTsL From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr")O Subject: Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions per)6 Message-ID: <00A2BE9C.F6E12433@SSRL.SLAC.STANFORD.EDU>  a In article <utof1efWsEYx@eisner.encompasserve.org>, young_r@encompasserve.org (Rob Young) writes:,d >In article <jhec00htujrs7jqarfau5ppk9phijfofdo@4ax.com>, Lorin Ricker <lorin@locktrack.com> writes: >  >> eI >> Brilliant tho' this approach may be, it takes a lot of work to set up, C >> and it's code to maintain (for numerous tables & fields)... muchtF >> better would be for both VMS and products like Rdb to *wake up* andD >> recognize the world-wide, far-flung nature of users connecting toH >> server systems, and that users perceive only *local time*, not systemB >> time.  And what the heck happens when we colonize the moon &/or
 >> Mars?!! >> x >o >	[snip] >s >> ,F >> There's a common, real-world need out there... it's probably rarelyF >> been vocalized very well, but here's one thread that's trying to do: >> so.  Hope I've helped.  Oracle, HP, do you hear us yet? >> d >fC >	Nice post, really.  Before you think I'm blasting you... I'm not.nF >	It may be as simple as you are nickel/dime compared to big customersC >	that want the other behavior - always.  For instance, JSTARS.  I rE >	imagines JSTARS flys through many timezones many times a day.  The vE >	only thing that matters to them is system time (or UTC).  Clusters eG >	spanning timezones that aren't working with UTC time are surely in a n# >	tiny minority.  My guess/opinion.e    L Think this through.  If Rdb used the system services to get time, that wouldH make zero impact on JSTARS (even if JSTARS used Rdb).  It's to nobody's K benefit to make DATESIM work erratically, not even to the benefit of peopletG who don't use it.  (It wouldn't have any impact on people who want the x: system time to be the same for all users, in other words.)   -- Alane --  O ==============================================================================='0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056-M  Paper mail to: SSRL -- SLAC BIN 99, 2575 Sand Hill Rd, Menlo Park CA   94025gO ===============================================================================6   ------------------------------  # Date: Thu, 15 Jan 2004 11:16:56 GMT@4 From: brad@.gateway.2wire.net (Bradford J. Hamilton)O Subject: RE: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions perr- Message-ID: <IEuNb.71345$8H.111998@attbi_s03>y  | In article <FD827B33AB0D9C4E92EACEEFEE2BA2FB237C91@tayexc19.americas.cpqcorp.net>, "Main, Kerry" <kerry.main@hp.com> writes: !m ![snip]n !"G !> UTC).  Clusters spanning timezones that aren't working with UTC time 2 !are surely in a tiny minority.  My guess/opinion. !>=20i
 !> 				Rob !>=20  !r !G'day Rob,U ! H !While that may certainly have been true in the past, one of the popularD !solutions to address high availability requirements associated withH !large server consolidation efforts these days (white hot btw) is to useE !cluster technologies. In large consolidation efforts, it is actuallytG !fairly common to be collapsing servers from different timezones into ai0 !centralized (and sometimes multi-site) cluster. !a  G And not only clusters - I've been looking at DateSim to consolidate ourfN stand-alone machines to run our application in multiple timezones.  Of course,O once the consolidation was complete, the next logical step would be to create amJ cluster from the "consolidated" machine, and the other "unused" machines.	 :-)f   !Regards !s !Kerry Main  !Senior Consultant !HP Services Canadal !Voice: 613-592-4660 !Fax: 613-591-4477 !Email: kerryDOTmainAThpDOTcom/ !(remove the DOT's and AT for email address)=20   J __________________________________________________________________________A Bradford J. Hamilton                    "All opinions are my own"hK bMradAhamiPltSon-at-coMmcAast.nPeSt     "Lose the MAPS, and replace '-at-' l0                                          with @"   ------------------------------    Date: 15 Jan 2004 06:10:38 -0600+ From: young_r@encompasserve.org (Rob Young)lO Subject: RE: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions perg3 Message-ID: <xd$myrMWisBP@eisner.encompasserve.org>n  d In article <IEuNb.71345$8H.111998@attbi_s03>, brad@.gateway.2wire.net (Bradford J. Hamilton) writes:~ > In article <FD827B33AB0D9C4E92EACEEFEE2BA2FB237C91@tayexc19.americas.cpqcorp.net>, "Main, Kerry" <kerry.main@hp.com> writes: > !t	 > ![snip]  > !]I > !> UTC).  Clusters spanning timezones that aren't working with UTC timey4 > !are surely in a tiny minority.  My guess/opinion. > !>=20o > !> 				Rob > !>=20r > !y
 > !G'day Rob,  > !yJ > !While that may certainly have been true in the past, one of the popularF > !solutions to address high availability requirements associated withJ > !large server consolidation efforts these days (white hot btw) is to useG > !cluster technologies. In large consolidation efforts, it is actuallyoI > !fairly common to be collapsing servers from different timezones into ao2 > !centralized (and sometimes multi-site) cluster. > !. > I > And not only clusters - I've been looking at DateSim to consolidate our P > stand-alone machines to run our application in multiple timezones.  Of course,Q > once the consolidation was complete, the next logical step would be to create aeL > cluster from the "consolidated" machine, and the other "unused" machines.	 > :-)c >   B 	Interesting comments.  I'm missing the point - been up all night, 	so pardon me while I ruminate.e   				Robi   ------------------------------  % Date: Thu, 15 Jan 2004 07:44:16 -0600s( From: Wayne Sewell <wayne@tachysoft.com>O Subject: Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions perb0 Message-ID: <00A2BEC3.D76B688E.15@tachysoft.com>  ) >From: Lorin Ricker <lorin@locktrack.com>l >X-Newsgroups: comp.os.vmsP >Subject: Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions per     >aG >In the best of worlds, provisions for local-time adjustments, per usere@ >&/or process, would be provided by VMS as a native part of time6 >reporting, where any/all applications could use it.     True.t  H I hope they don't, though.  At least not until we've made a few sales of
 datesim.  :-)t    O =============================================================================== N Wayne Sewell, Tachyon Software Consulting  (281)812-0738   wayne@tachysoft.com; http://www.tachysoft.com/www/tachyon.html and wayne.html    O ===============================================================================cB Jed Clampett, checking into hotel: "This place got a cement pond?", 	Ellie May: "And do yuh let critters in it?"   ------------------------------  % Date: Thu, 15 Jan 2004 07:53:45 -0600(( From: Wayne Sewell <wayne@tachysoft.com>O Subject: Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions per=0 Message-ID: <00A2BEC5.2A8DB886.19@tachysoft.com>  G >From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin CmptgE >      Mgr") >X-Newsgroups: comp.os.vmsP >Subject: Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions per$ >Date: Thu, 15 Jan 2004 11:05:59 GMT     >jM >Think this through.  If Rdb used the system services to get time, that wouldeI >make zero impact on JSTARS (even if JSTARS used Rdb).  It's to nobody's bL >benefit to make DATESIM work erratically, not even to the benefit of peopleH >who don't use it.  (It wouldn't have any impact on people who want the ; >system time to be the same for all users, in other words.)o >l    O And before anybody brings up performance, we aren't talking about *all* uses ofsH system time by oracle/rdb, just those that are displayed to the user.  IG understand that there are a lot of internal time stamps during databaseuL operations.  We don't care about those; they can continue to access the timeL quadword.  We just want the current_timestamp field (and perhaps transactionN start time, which can apparently be interrogated by the user) to go though the system services.  7 This is ALL IT WOULD TAKE to make timezone mode work.  u   Wayne:O ===============================================================================iN Wayne Sewell, Tachyon Software Consulting  (281)812-0738   wayne@tachysoft.com; http://www.tachysoft.com/www/tachyon.html and wayne.html   NO =============================================================================== B Jed Clampett, checking into hotel: "This place got a cement pond?", 	Ellie May: "And do yuh let critters in it?"   ------------------------------  % Date: Thu, 15 Jan 2004 08:02:33 -06000( From: Wayne Sewell <wayne@tachysoft.com>O Subject: Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions pera/ Message-ID: <00A2BEC6.64F1DF1A.3@tachysoft.com>l  , >From: young_r@encompasserve.org (Rob Young) >X-Newsgroups: comp.os.vmsP >Subject: Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions per! >Date: 15 Jan 2004 02:47:13 -0600t  C >	Nice post, really.  Before you think I'm blasting you... I'm not.tF >	It may be as simple as you are nickel/dime compared to big customersC >	that want the other behavior - always.  For instance, JSTARS.  I  E >	imagines JSTARS flys through many timezones many times a day.  The  E >	only thing that matters to them is system time (or UTC).  Clusters eG >	spanning timezones that aren't working with UTC time are surely in a 0# >	tiny minority.  My guess/opinion.O >M  ? Another thing to keep in mind is the particular users involved.r  L Technical people would probably have no difficulty in understanding utc or aJ different timezone for the server.  Lorin's correction officers might be a different story.  K Bubba:  "Huh?  I got to use East Coast time when I book these muhfugs?  But  we're in San Francisco!"   WaynehO ===============================================================================pN Wayne Sewell, Tachyon Software Consulting  (281)812-0738   wayne@tachysoft.com; http://www.tachysoft.com/www/tachyon.html and wayne.html    O ===============================================================================aB Jed Clampett, checking into hotel: "This place got a cement pond?", 	Ellie May: "And do yuh let critters in it?"   ------------------------------  % Date: Thu, 15 Jan 2004 08:34:50 -0700z4 From: Norman Lastovica <norman.lastovica@oracle.com>O Subject: Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions pern* Message-ID: <4006B31A.78DF0A35@oracle.com>   Lorin Ricker wrote:  > H > On Sat, 10 Jan 2004 18:46:58 -0600, Wayne Sewell <wayne@tachysoft.com> > wrote:  E > There's a common, real-world need out there... it's probably rarelyaE > been vocalized very well, but here's one thread that's trying to do 9 > so.  Hope I've helped.  Oracle, HP, do you hear us yet?-  3 	while I'm sure that it is personally satisfying to 7 vent in public forums, if you want stuff to happen, youu6 really ought to get in touch with the vendor directly.= And if you don't get the answer that you need, then escalate.c   ------------------------------  # Date: Thu, 15 Jan 2004 11:24:33 GMTA" From:   VAXman-  @SendSpamHere.ORGY Subject: Re: OT - English (UK) humor - (was Re: The Register: OpenVMS among most-secure ot0 Message-ID: <00A2BEB8.C11AC30D@SendSpamHere.ORG>  T In article <bu5i0k$j56$1@lore.csc.com>, Nic Clews <sendspamhere@[127.0.0.1]> writes: >Fred Kleinsorge wrote:f >> e >> kN >> Yes, but for a small investment, you can purchase the complete set of DVDs.N >> A very worthwhile investment... and longterm cheaper than Digital Cable ;-) >mB >But NOT in the UK, it is only possible to purchase a compilation. >o? >I might as well have my head nailed to the floor (eh, VAXman?)    You'll have to insist on that. s  2 Such violence and sarcasm... when will it all end? --B http://www.legacy-2000.com  for the *best* OpenVMS system securityC                             solutions that others only claim to be.- -- .K VAXman- A Bored Certified VMS Kernel Mode Hacker   VAXman(at)TMESIS(dot)COM             l5   "Well my son, life is like a beanstalk, isn't it?" g   ------------------------------  % Date: Thu, 15 Jan 2004 08:12:03 +0000u* From: Nic Clews <sendspamhere@[127.0.0.1]>Y Subject: Re: OT - English (UK) humor - (was Re: The Register: OpenVMS among most-secure o8' Message-ID: <bu5i0k$j56$1@lore.csc.com>0   Fred Kleinsorge wrote: >  > M > Yes, but for a small investment, you can purchase the complete set of DVDs. M > A very worthwhile investment... and longterm cheaper than Digital Cable ;-)-  A But NOT in the UK, it is only possible to purchase a compilation.s  > I might as well have my head nailed to the floor (eh, VAXman?)   -- P? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciencesm nclews at csc dot como   ------------------------------    Date: 15 Jan 2004 07:44:44 -0800. From: fabiopenvms@yahoo.com.br (Fabio Cardoso)( Subject: OT: AA pilot arrested in Brazil= Message-ID: <f30679fb.0401150744.5b477602@posting.google.com>c  M As americans are too puritane to show it over  the Net I am putting the link:e  A The photo of the AA pilot ! He paid about US $ 12.000,00 to leave  the prision !   3 http://odia.ig.com.br/arte/04/01/capa/15_piloto.jpg/   CNN news ! ~  K http://www.cnn.com/2004/WORLD/americas/01/14/finger.gesture.reut/index.html-  # By the way... no ... no comments ! a     Regards0   FC   ------------------------------  % Date: Thu, 15 Jan 2004 03:17:07 -0500a* From: JF Mezei <jfmezei.spamnot@istop.com>* Subject: OT: Good luck to Glenn C Everhart) Message-ID: <40064C5E.6064E29A@istop.com>e  A JP Morgan Chase is buying Bank One promising billions in savings.o  U Hopefully Mr Everhart will come out on the winning side or tbe inevitable downsizing.1   ------------------------------   Date: 15 Jan 2004 14:33:04 GMT1 From: JONESD@er6.eng.ohio-state.edu (David Jones)eA Subject: Re: Secure Web Server Version 2.0 available for downloade: Message-ID: <bu68b0$271$1@charm.magnus.acs.ohio-state.edu>  ; In message <KHmNb.8024$Ym7.4552@newssvr22.news.prodigy.com>CB   Michael Austin <maustin@no-more-spam.firstdbasource.com> writes:H >It seemes to me that since you cannot determine the content-length of aI >generated page, what's the difference in transmitting a file that has an G >  indeterminate file length??  Seems like someone may need to read ando, >interpret the RFC a little more closely....  L The original Apache 1.3 that CSWS 1.x was based on had a bug in its chunkingO code and would fail if the chunks were too large (>4 digits IIRC).  This wasn'teN a problem on Unix because the chunks were always generated from data read fromJ pipes, which limited the chunk size.  The broken code was fixed in CSWS to@ allow var format files to be read and sent as very large chunks.  J Perhaps there was a regression when they went to the Apache 2.0 code base.    < David L. Jones               |      Phone:    (614) 292-6929- Ohio State University        |      Internet:lL 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.    ------------------------------   Date: 15 JAN 2004 15:14:36 GMT4 From: karcher@thuria.waisman.wisc.edu (Carl Karcher) Subject: Re: ssh on Ovms 7.3-26 Message-ID: <15JAN04.15143671@thuria.waisman.wisc.edu>  M In a previous article, "Irving Fewkes Snurd III" <something@lycos.com> wrote:f  H ->We just upgraded to -2, and now ssh2 works, but tunneling for Xwindows" ->doesn't. Anyone have a solution? ->   ->    Bill Manwaring, Indiana Un  C I have a vague memory of asking about this in NH last November. TheuH response, if I recall correctly, was that X forwarding was not there yet: but was planned. Have you checked the TCPIP release notes?   --G -- Carl Karcher, Waisman Computing Services, Waisman Center, UW-Madison 8 --                 karcher.nomorespxm@waisman.wisc.edu     ------------------------------    Date: 15 Jan 2004 06:17:06 -0800$ From: rbrooke@intlsteel.com (Roscoe)9 Subject: Re: SYS$QIOW, INSTAT and OFFSET - what are they? = Message-ID: <2c811a5d.0401150617.17117919@posting.google.com>   7 Exactly what I needed!  Thanks for the great info guys!k   Ross A. Brooke ISG Steel Companyp Hennepin Works   ------------------------------  % Date: Thu, 15 Jan 2004 02:23:45 -0500-* From: JF Mezei <jfmezei.spamnot@istop.com>I Subject: Re: The Register: OpenVMS among most-secure of operating systemsh) Message-ID: <40063FE0.DA64A846@istop.com>m  G When they discovered the flaw in the POP server that would allow one tooM overwrite any file on the system because the impage was installed with SYSPRVsN and could be called interactively with a log file specification, did that make it to CERT ?  J I suspect that there are many issues specific to VMS that don't make it toJ CERT.  CERT is prebably much more concerned about widely used systems, not* niche systems with very vew installations.   ------------------------------    Date: 15 Jan 2004 07:41:40 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) I Subject: Re: The Register: OpenVMS among most-secure of operating systemso3 Message-ID: <EuOzEZ7kqyfe@eisner.encompasserve.org>h  Z In article <40055a04$1@cpns1.saic.com>, Mark Berryman <Mark.Berryman@Mvb.Saic.Com> writes: > K > I've never looked for Multinet issues in CERT so I'll have to table this  * > until I have a little time for research.  F    I have.  I recall seeing one or two that Multinet would ACCVIO and F    restart the service (no impact on security or availability).  IIRC D    the CERT wording made things look worse, it just claimed Multinet>    was affected.  So yes, I agree with Andrew that CERT is notF    completely reliable; they'd have had you rushing out to get a patch    you didn't need to rush for.V  D    I also recall seeing that Multinet SSH1 had the same limitations G    that all SSH1 had due to SSH1 design.  Guess which vendor no longer eI    ships an SSH1 client in its standard distribution and leaves the SSH1 k    server off by default?    ------------------------------  # Date: Thu, 15 Jan 2004 17:07:46 GMT & From: jlsue <jefflsxxxz@sbcglobal.net>I Subject: Re: The Register: OpenVMS among most-secure of operating systemst8 Message-ID: <p7id00hpnmtvdt5jn92dniniq90ghjl110@4ax.com>  H On Thu, 15 Jan 2004 02:23:45 -0500, JF Mezei <jfmezei.spamnot@istop.com> wrote:  H >When they discovered the flaw in the POP server that would allow one toN >overwrite any file on the system because the impage was installed with SYSPRVO >and could be called interactively with a log file specification, did that maker
 >it to CERT ?7  J Whose POP server?  Do I run that on my systems?  If not, then now could it2 affect me and why would it actually be a VMS CERT?   >eK >I suspect that there are many issues specific to VMS that don't make it tonK >CERT.  CERT is prebably much more concerned about widely used systems, nots+ >niche systems with very vew installations.   K Define many.  By one defiintion of that term I could infer that you believeSG there are many security-related fixes in VMS that are not reported.  Is  that what you're implying?   --- jlso0 The preceding message was personal opinion only.6 I do not speak in any authorized capacity for anyone,  and certainly not my employer.- (get rid of the xxxz in my address to e-mail)C   ------------------------------  # Date: Thu, 15 Jan 2004 07:54:46 GMTt2 From: "Robert Boers" <Robert.boers@softresint.com>( Subject: Re: VAX architecture and Charon* Message-ID: <40064743$1@news.deckpoint.ch>  < "Chris Scheers" <chris@applied-synergy.com> wrote in message, news:4005BE2A.B0BD083@applied-synergy.com... > Bob Koehler wrote: > >0K > > In article <DTiotGxQ0bj6-pn2-Hs2bHgj6HB1d@localhost>, "Dave Weatherall"a" <djw-nothere@nospam.nohow> writes: > > >h	 > > > Bob L > > >           when researching the possibility of using Charon-VAX for oneK > > > of our applications, I noticed  a reference to a DRQ3B equivalent PCIeI > > > card, on SRI's page. I even proposed it for last years's budget butb, > > > have been too busy to take it forward. > > >sJ > > > The card itself was manufactured a compamy in California. I'm sure I@ > > > saved the page somewhere but can't find it a t the moment. > >cJ > >    I've seen references to the hardware components.  One company makesJ > >    a Qbus PCI adapter and Qbus DRV11-W and DRQ3B look-alikes.  AnotherH > >    company makes a single PCI card that will emulate DR11-B, DR11-W, > >    DRV11-W, or DRQ3B.t > > / > >    It's the timing I'm not willing to risk.s >l > It's not just timing.c >bH > As I understand it, these PCI cards are designed to plug into an AlphaC > and come with Alpha drivers which are QIO compatible with the VAX-
 > drivers. >-F > There is no reason to think that these cards are register compatibleF > with the Q-bus cards. (They really can't be, they are on a different > bus.)o > I > Unless Charon-VAX does something really strange, I'ld be very surprisedaH > if it was able to do anything with these cards without some additional	 > coding.. > I > ----------------------------------------------------------------------- & > Chris Scheers, Applied Synergy, Inc. >aD > Voice: 817-237-3360            Internet: chris@applied-synergy.com >   Fax: 817-237-3074o  J The Logical Company designed the PCI parallel and IEQ-11 replacement cardsL specifically to be register compatible with the original Qbus cards, so thatI we could integrate them in CHARON-VAX in a transparent manner. CHARON-VAX]J maps their registers in the virtual VAX environment in such a way that theD standard binary VAX drivers work (not just VMS, also VAXEln etc). No additional coding is required.  G For devices for which no compatible PCI replacement cards have yet been L developed, the BCI-2104 PCI-Qbus adapter can be used to support the original Qbus interfaces.   Robert Boers  Software Resources International   ------------------------------    Date: 15 Jan 2004 07:29:47 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)0( Subject: Re: VAX architecture and Charon3 Message-ID: <NYtDMv0QCfgI@eisner.encompasserve.org>c  _ In article <40064743$1@news.deckpoint.ch>, "Robert Boers" <Robert.boers@softresint.com> writes:.  L > The Logical Company designed the PCI parallel and IEQ-11 replacement cardsN > specifically to be register compatible with the original Qbus cards, so thatK > we could integrate them in CHARON-VAX in a transparent manner. CHARON-VAXeL > maps their registers in the virtual VAX environment in such a way that theF > standard binary VAX drivers work (not just VMS, also VAXEln etc). No  > additional coding is required.  B    Can I map them to non-standard addresses so VMS won't load it'sF    drivers?  Sometimes during boot we can't prevent VMS from trying toF    load the DRV-11W drivers, so we put the boards at DR11-B addresses.C    There's no driver shipped for DR11-B so VMS fails to load it anda    we can load ours instead.  F    And since we have several DR11-WA on each Qbus we need to map them     all to different addresses.   ------------------------------    Date: 15 Jan 2004 07:26:41 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)i( Subject: Re: VAX architecture and Charon3 Message-ID: <JASNSndn17gc@eisner.encompasserve.org>r  d In article <4005BE2A.B0BD083@applied-synergy.com>, Chris Scheers <chris@applied-synergy.com> writes:  H > As I understand it, these PCI cards are designed to plug into an AlphaC > and come with Alpha drivers which are QIO compatible with the VAX 
 > drivers. > F > There is no reason to think that these cards are register compatibleF > with the Q-bus cards. (They really can't be, they are on a different > bus.)n  A    And since we do our own drivers, we'd have to port them again, B    which increases the cost of moving to Charon-VAX.  But we couldE    do that (I did the UNIBUS DR-11W to Qbus DRV11-W port).  We we're s'    not ready to do is upset the timing.t   ------------------------------  # Date: Thu, 15 Jan 2004 14:14:32 GMT/2 From: "Robert Boers" <Robert.boers@softresint.com>( Subject: Re: VAX architecture and Charon* Message-ID: <4006a045$1@news.deckpoint.ch>  H "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in message- news:JASNSndn17gc@eisner.encompasserve.org...cB > In article <4005BE2A.B0BD083@applied-synergy.com>, Chris Scheers# <chris@applied-synergy.com> writes:O >tJ > > As I understand it, these PCI cards are designed to plug into an AlphaE > > and come with Alpha drivers which are QIO compatible with the VAXt > > drivers. > >HH > > There is no reason to think that these cards are register compatibleH > > with the Q-bus cards. (They really can't be, they are on a different	 > > bus.)c >nC >    And since we do our own drivers, we'd have to port them again,XD >    which increases the cost of moving to Charon-VAX.  But we couldF >    do that (I did the UNIBUS DR-11W to Qbus DRV11-W port).  We we're) >    not ready to do is upset the timing.b >p  H There is no need to port or modify custom VMS drivers. They see the sameI device addresses and the same Qbus in the virtual VAX environment as in a-I hardware VAX when using these PCI cards. The CHARON-VAX enviroment is the I same as a hardware VAX including any I/O device, you run the unmodified -e2 existing - binary VAX driver code for that device.   Robert   ------------------------------  # Date: Thu, 15 Jan 2004 14:34:39 GMTn2 From: "Robert Boers" <Robert.boers@softresint.com>( Subject: Re: VAX architecture and Charon* Message-ID: <4006a4fd$1@news.deckpoint.ch>  H "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in message- news:NYtDMv0QCfgI@eisner.encompasserve.org...l; > In article <40064743$1@news.deckpoint.ch>, "Robert Boers"i% <Robert.boers@softresint.com> writes:e >aH > > The Logical Company designed the PCI parallel and IEQ-11 replacement cards,K > > specifically to be register compatible with the original Qbus cards, so  thatB > > we could integrate them in CHARON-VAX in a transparent manner.
 CHARON-VAXJ > > maps their registers in the virtual VAX environment in such a way that the(H > > standard binary VAX drivers work (not just VMS, also VAXEln etc). No" > > additional coding is required. >lD >    Can I map them to non-standard addresses so VMS won't load it'sH >    drivers?  Sometimes during boot we can't prevent VMS from trying toH >    load the DRV-11W drivers, so we put the boards at DR11-B addresses.E >    There's no driver shipped for DR11-B so VMS fails to load it ande >    we can load ours instead. >bG >    And since we have several DR11-WA on each Qbus we need to map thema  >    all to different addresses. >pL If a custom Qbus device is connected via a PCI Qbus adapter, its Qbus deviceH addresses and vectors - as set on the board - re-appear identical in theD virtual VAX environment. As always, you can and use the existing VAX drivers.  L If a Qbus device is replaced by a PCI clone (e.g. the IEQ11 by the DCI3100),B place an entry in the CHARON-VAX configuration file to specify itsK device/vector address in the emulated environment. Example configuration of / a DCI3100 on the secundary IEQ11 Qbus location:o   [...]e LOAD IEQ11 IXB Set IXB interface="DCI3100 1"l& SET IXB address=0x03FE900 vector=0x0C0 [...].  K and you have an IEQ11 showing up in your virtual VAX environment that works 1 as usual with its existing VMS or VAXELN drivers.a  D There is a description of these options in chapter 3 of this manual:/ http://www.softresint.com/pub/doc/30-16-015.pdfa   Regards, Roberte   ------------------------------    Date: 15 Jan 2004 07:36:24 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris)o* Subject: Re: VMS runs well on HP Superdome= Message-ID: <cf15391e.0401150736.2ff5eb64@posting.google.com>-  Z JF Mezei <jfmezei.spamnot@istop.com> wrote in message news:<4005C7BF.67190AD@istop.com>...) > She may have performed the merger well,i   Nice to see you admit this.s  H > but has the money they invested in buying/destroying Compaq paid off ?  ( Where do you get this "destroying" idea?  E ProLiants live on and thrive, as do iPaqs. I still see Compaq PCs fordF sale in all the retail outlets. StorageWorks and the EVA have paid offD handsomely, and the SmartArray continues and has spawned the popular= MSA1000. NonStop continues and has booted on Itanium. OpenVMSiC continues and now runs on Itanium. TruClusters and AdvFS from Tru64s will live on in HP-UX.   The synergy has been excellent.d   ------------------------------  % Date: Thu, 15 Jan 2004 08:15:56 -0800u/ From: Greg Cagle <news@removethisgregcagle.com>n* Subject: Re: VMS runs well on HP Superdome0 Message-ID: <100df605p1gj104@corp.supernews.com>  5 Keith Parris said the following on 1/15/2004 7:36 AM:h  \ > JF Mezei <jfmezei.spamnot@istop.com> wrote in message news:<4005C7BF.67190AD@istop.com>... >>H >>but has the money they invested in buying/destroying Compaq paid off ? >  > * > Where do you get this "destroying" idea? > G > ProLiants live on and thrive, as do iPaqs. I still see Compaq PCs foreH > sale in all the retail outlets. StorageWorks and the EVA have paid offF > handsomely, and the SmartArray continues and has spawned the popular? > MSA1000. NonStop continues and has booted on Itanium. OpenVMSaE > continues and now runs on Itanium. TruClusters and AdvFS from Tru64  > will live on in HP-UX. > ! > The synergy has been excellent.i  B If you are a member of the hp_alumni list (as I am) you would find@ that a lot of people think she destroyed HP with the merger, not Compaq.    -- h
 Greg Cagle gregc at gregcagle dot com   ------------------------------  # Date: Thu, 15 Jan 2004 16:25:23 GMTo# From: "John Smith" <a@nonymous.com>s' Subject: Will HP *ever* publicize OVMS?pK Message-ID: <T9zNb.164079$AAe1.151059@news01.bloor.is.net.cable.rogers.com>-  L http://story.news.yahoo.com/news?tmpl=story&cid=1738&ncid=738&e=9&u=/zd/2004 0115/tc_zd/116601g    > Hewlett-Packard Co. on Thursday will announce that it earned aF record-breaking $2.5 billion Linux (news - web sites)-based revenue inH fiscal 2003, with its Linux services and solutions business posting a 40 percent rise over fiscal 2002.  I While the revenue was derived from the sale of Linux-related products and I services, the Palo Alto, Calif., company did not specify exactly what wasr, included and counted as Linux-based revenue.    ( Where's the HP press release about OVMS?/ carly(tm) and the Bored - deaf, dumb, and blindr   ------------------------------   End of INFO-VAX 2004.029 ************************ck happens when we colonize the moon &/oro	 > Mars?!!> >    	[snip]f   > E > There's a common, real-world need out there... 5.185 at Host 24.84.96.254 accepted.n
 <<< LIST >>> 150 List started./ >>> 226 Transfer completed. 
 <<< noop >>> 200 No-operation OK.7 <<< CWD /disk$misc/decus/freewarev40/finger/iufinger/dD >>> 250 Connected to /disk$misc/decus/freewarev40/finger/iufinger. <<< TYPE A >>> 200 Type A ok. <<< PORT 24,84,96,254,5,186c3 >>> 200 Port 5.186 at Host 24.84.96.254 accepted.m
 <<< LIST >>> 150 List started.  >>> 226 Transfer completed.
 <<< noop >>> 200 No-operation OK.8 <<< CWD /disk$misc/decus/freewarev40/finger/iufingerd/E >>> 250 Connected to /disk$misc/decus/freewarev40/finger/iufingerd.> <<< TYPE A >>> 200 Type A ok. <<< PORT 24,84,96,254,5,187e3 >>> 200 Port 5.187 at Host 24.84.96.254 accepted./
 <<< LIST >>> 150 List started.Y >>> 226 Transfer completed.<
 <<< noop >>> 200 No-operation OK.8 <<< CWD /disk$misc/decus/freewarev40/finger/mg_finger/E >>> 250 Connected to /disk$misc/decus/freewarev40/finger/mg_finger.  <<< TYPE A >>> 200 Type A ok. <<< PORT 24,84,96,254,5,188/3 >>> 200 Port 5.188 at Host 24.84.96.254 accepted.r
 <<< LIST >>> 150 List started.Y >>> 226 Transfer completed.<
 <<< noop >>> 200 No-operation OK.7 <<< CWD /disk$misc/decus/freewarev40/finger/psfinger/>D >>> 250 Connected to /disk$misc/decus/freewarev40/finger/psfinger. <<< TYPE A >>> 200 Type A ok. <<< PORT 24,84,96,254,5,189e3 >>> 200 Port 5.189 at Host 24.84.96.254 accepted./
 <<< LIST >>> 150 List started.  >>> 226 Tran