0 INFO-VAX	Sat, 19 Feb 2005	Volume 2005 : Issue 99      Contents:# Re: HP financials: Alpha sales down # Re: HP financials: Alpha sales down # Re: HP financials: Alpha sales down # Re: HP financials: Alpha sales down # Re: HP financials: Alpha sales down # RE: HP financials: Alpha sales down # Re: HP financials: Alpha sales down # RE: HP financials: Alpha sales down # Re: HP financials: Alpha sales down P Itanium Montecito Wins Microprocessor Report Analysts' Choice Award for Best SerP Re: Itanium Montecito Wins Microprocessor Report Analysts' Choice Award for BestP Re: Itanium Montecito Wins Microprocessor Report Analysts' Choice Award for BestP Re: Itanium Montecito Wins Microprocessor Report Analysts' Choice Award for BestP Re: Itanium Montecito Wins Microprocessor Report Analysts' Choice Awardfor Best - Re: linux kernel has major security flaws ...  Re: looking for a new job 1 Re: Looking for suggestions for bootcamp sessions = Re: Micro$oft warns of undetectable spyware security risk ... ! Re: Mounting disks during STARTUP ! Re: Mounting disks during STARTUP ! Re: Mounting disks during STARTUP ! Re: Mounting disks during STARTUP 8 MY patching procedures (was Re: 7.32 patching questions) Re: Need performance help ! RE: OpenVMS 8.2 docs and freeware ! Re: OpenVMS 8.2 docs and freeware 4 Re: OpenVMS Cluster to Replace News.Individual.NET ?" Re: OpenVMS Small Partner Strategy" Re: OpenVMS Small Partner Strategy" Re: OpenVMS Small Partner Strategy" Re: OpenVMS Small Partner Strategy" Re: OpenVMS Small Partner Strategy" Re: OpenVMS Small Partner Strategy" Re: OpenVMS Small Partner Strategy& Re: Port Print facility Update for I64& Re: Port Print facility Update for I64& Re: Port Print facility Update for I64/ Re: read monitor (>>>) variables from OpenVMS ?  Re: Shark Tank mentions VMS  Re: Shark Tank mentions VMS  The results of advertising... ! Re: The results of advertising... ! Re: The results of advertising... ! Re: The results of advertising...   Re: VMS 8.2 Itanium Distribution  Re: VMS 8.2 Itanium Distribution  Re: VMS 8.2 Itanium Distribution) [OT]: SHA-1 broken by Chinese researchers - Re: [OT]: SHA-1 broken by Chinese researchers   F ----------------------------------------------------------------------  # Date: Fri, 18 Feb 2005 19:18:55 GMT 1 From: Keith Parris <keithparris_NOSPAM@yahoo.com> , Subject: Re: HP financials: Alpha sales down/ Message-ID: <zcrRd.317$LO7.89@news.cpqcorp.net>    Dave Froble wrote:8 > Don't forget that T64 was sold on more Alphas than VMS  E Actually, although the Tru64 Alpha ship rates eventually did grow to  @ exceed those of VMS by about a 60/40 ratio for a few years, the I cumulative number of Alphas sold with VMS ended up being about about 1.5  9 times the number for Tru64, based on the last data I saw.    ------------------------------  # Date: Fri, 18 Feb 2005 19:21:56 GMT 1 From: Keith Parris <keithparris_NOSPAM@yahoo.com> , Subject: Re: HP financials: Alpha sales down/ Message-ID: <ofrRd.318$qE.226@news.cpqcorp.net>    JF Mezei wrote: H > VMS is on its way there too, with only 660 customer left that matter.   I I don't know where you pulled that number from, but it's low by multiple   orders of magnitude.   ------------------------------  # Date: Fri, 18 Feb 2005 19:47:49 GMT 1 From: Keith Parris <keithparris_NOSPAM@yahoo.com> , Subject: Re: HP financials: Alpha sales down/ Message-ID: <FDrRd.326$2H.321@news.cpqcorp.net>    JF Mezei wrote: 8 > IA64 sales aren't really a factor in the past quarter.  F They're fast becoming a factor, with Integrity sales now representing < 18% of BCS revenues, having doubled year over year for 1Q05.  5 > Also, VMS runs only on a small subset of IA64 boxes   * Better take a closer look at the Roadmap: > http://h71000.www7.hp.com/openvms/roadmap/openvms_roadmaps.htm  G The vast majority of IA64 volumes will be in the boxes VMS 8.2 already  F supports: rx1600/1620, rx2600/2620, and rx4640. VMS basically runs on F all the current IA64 models except the Superdome and the rx7620/8620, E and the rx8620 and Superdome are slated for support in the new 8.2-1  H release that was added to the roadmap to get support for >8 CPU systems C out to customers faster, while the rx7620 will be supported in 8.3.    ------------------------------  % Date: Fri, 18 Feb 2005 14:28:02 -0500 # From: "John Smith" <a@nonymous.com> , Subject: Re: HP financials: Alpha sales down, Message-ID: <6bSdnZXZ595Y3IvfRVn-jg@igs.net>   Keith Parris wrote:  > JF Mezei wrote: H >> VMS is on its way there too, with only 660 customer left that matter. > A > I don't know where you pulled that number from, but it's low by  > multiple orders of magnitude.     L The 660 came from Gorham/Marcello in a published interview - as key accounts that HP talks to.   E So by your reckoning there are a total of about 660 x 100= 66,000 VMS  accounts out there.   F Divide the mythical 411,000 by that and we get an average of about 6.3L systems per account. But we know that there are very few 'average' accounts," so here's how it might break down:   Average  Systems + Per Key    Total     Total          Average , Acct.         Key      Other           Other) ----------------------------------------- *   10          6,600    404,400        6.19)   50        33,000    378,000        5.79 ( 100        66,000    345,000        5.28' 200      132,000    279,000        4.27 ' 300      198,000    213,000        3.26 ' 311      205,260    205,740        3.15   K Don't forget that this would include organizations with a single VAXstation ; too, so the averages have to be taken with a grain of salt.   K To put it all in perspective, wasn't Sun selling >300,000 units per year in = the late 90's. I'll be a lot of them are still out there too.        --- OpenVMS - The classics never go out of style.    ------------------------------  % Date: Sat, 19 Feb 2005 04:20:59 +0800  From: prep@prep.synonet.com , Subject: Re: HP financials: Alpha sales down- Message-ID: <87mzu1livo.fsf@prep.synonet.com>   % "John Smith" <a@nonymous.com> writes:    > Tom Linden wrote:   & >> Which computer systems did IBM EOL?  D > 709 - sometime long ago, but I bet if you called and asked IBM forC > help with one you still had around, they'd find a way to help you  > out.  F No, you can probably still run your 709 stuff on the 7090 emulation on8 the 360 emulation on a virtual partition on a zSeries...  % Note that I say nothing about COST...   ) The system 3, 36 and 9000s have died off.    --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Fri, 18 Feb 2005 17:45:35 -0500 ' From: "Main, Kerry" <kerry.main@hp.com> , Subject: RE: HP financials: Alpha sales downR Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB53F60A@tayexc19.americas.cpqcorp.net>   > -----Original Message-----4 > From: Bill Todd [mailto:billtodd@metrocast.net]=20! > Sent: February 18, 2005 1:45 PM  > To: Info-VAX@Mvb.Saic.Com . > Subject: Re: HP financials: Alpha sales down >=20 > Main, Kerry wrote: > >>-----Original Message-----6 > >>From: Bill Todd [mailto:billtodd@metrocast.net]=20# > >>Sent: February 17, 2005 8:41 PM  > >>To: Info-VAX@Mvb.Saic.Com 0 > >>Subject: Re: HP financials: Alpha sales down > >> > >>Keith Parris wrote:  >=20 > ...  >=20 > >>  "HP Services (HPS) > >>F > >>>revenue grew 20% year-over-year to a record $3.8 billion. On a=20@ > >>>year-over-year basis, Managed Services revenue grew 44%,=20 > >> > >>Consulting and=20  > >>> > >>>Integration grew 20% and Technology Services grew 14%.=20 > >> > >>Operating profit=20  > >>@ > >>>was $281 million, or 7.4% of revenue, compared with $261=20 > >> > >>million in the=20  > >>? > >>>prior-year period." Based on past experience, one could=20  > >> > >>estimate that=20 > >>? > >>>roughly 15% of the revenues and 25% of the profits here=20  > >> > >>were from VMS. > >>F > >>The 'past experience' you are basing this on would appear to be=20< > >>pre-Alphacide experience (very much as that perpetual=20 > >>'411,000 installed=20 > > >>systems' seems to be), unless you're asserting that the=20 > Alphacide had=20I > >>very little effect in this area:  I could have believed that a few=20 @ > >>months or even a year or more after the event (due to the=20 > >>time-lag noted=20 J > >>above), but it seems rather less believable coming up on its fourth=20 > >>anniversary. > >>
 > >>- bill > >> > >=20 > >=20	 > > Bill,  > >=20, > > The services revenue is very significant >=20@ > You are, as usual, side-stepping the specific issue raised,=20 > Kerry.  VMS=20A > services revenue could qualify as 'very significant' even if=20  > were only=20 > 1/4 what it was 4 years ago. >=20? > If you're not willing to go on record with actual numbers,=20  > you're just=20G > (again as usual) spewing spin.  By the way, the value of that $784=20 ? > million service contract you cited was not only spread out=20  > over a full=20> > decade (i.e, less than $80 million annually) but, as John=20 > noted, merely=20= > reflects the extension of *existing* VMS service revenue=20  > rather than new=20 > business.  >=20 > - bill >=20  F The point that needs to always be re-enforced (unfortunately) is that,D as has been mentioned a number of times, a significant amount of theB success and growth of the HP services business has been because ofF OpenVMS services (consulting, CS maint, OS software, outsourcing etc).  > To many people get caught up in HW numbers only. Yes, they are6 important, but there are also other considerations.=20  F As I mentioned before, vendors are looking to make revenue off Linux -F primarily not by HW, but rather from services (support, consulting, sw
 license etc).   F The VA pointer was simply a recent example of OpenVMS services revenue being huge.=20  C Are the services numbers as big as it used to be? Probably not. =20   G Course, the same could be said for Solaris, HP-UX, AIX, MVS etc etc ...   D Have the overall numbers of OpenVMS servers been shrinking? I do notD have access to the numbers (and I would *not* provide them even if IG did), but the reality is that overall, ALL server platform numbers from @ all vendors in most med-large companies are shrinking because of consolidation initiatives.=20   " That is not spin. That is reality.  @ The only question might be which ones are shrinking the fastest?  G The absolute biggest target for server consolidation right now (by far) 9 in almost all med-large businesses are Wintel servers.=20   F No, that does not mean they are getting rid of Wintel. Rather it meansB they are making better use of fewer, larger servers and VMware for partitioning.=20  F However, it does mean the HW maint revenue for Windows servers is also going way, way down as well.  D Just because the numbers of servers in a company is going down, thatF does not necessarily mean that company is dropping that platform. ThisD is not unlike when many small VAXes or Alpha's are moved into larger" more centralized OpenVMS clusters.  " That is not spin. That is reality.  5 Now, insert coulda, shoulda marketing responses next.   
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  $ "OpenVMS has always had integrity .. Now, Integrity has OpenVMS .."   ------------------------------  % Date: Fri, 18 Feb 2005 18:55:37 -0500 ( From: Bill Todd <billtodd@metrocast.net>, Subject: Re: HP financials: Alpha sales down= Message-ID: <vLmdnVGp4KzmHYvfRVn-2g@metrocastcablevision.com>    Main, Kerry wrote:   ...   H > The point that needs to always be re-enforced (unfortunately) is that,F > as has been mentioned a number of times, a significant amount of theD > success and growth of the HP services business has been because ofH > OpenVMS services (consulting, CS maint, OS software, outsourcing etc).  H That would certainly be a point worth reinforcing with HP's management, C but hardly needs reinforcing here.  Furthermore, it has absolutely  G nothing to do with the point that was actually under discussion at the  B time you chimed in, which was not the relative magnitude of VMS's G contribution to HP's service but rather the direction it's been moving  L in and whether any *recent* figures were consistent with Keith's suggestion.   ...   C > Are the services numbers as big as it used to be? Probably not.     E Ah - finally at least one small step in at least a slightly relevant  
 direction.   > I > Course, the same could be said for Solaris, HP-UX, AIX, MVS etc etc ...   G Which even if true (as you point out later) simply shifts the question  E to what the *relative* rates of decline are.  Of course, even *that*  E question isn't relevant to the issue I raised with Keith's assertion.    > F > Have the overall numbers of OpenVMS servers been shrinking? I do notF > have access to the numbers (and I would *not* provide them even if II > did), but the reality is that overall, ALL server platform numbers from B > all vendors in most med-large companies are shrinking because of > consolidation initiatives.    A Neither IDC not Gartner would seem to share that view - see, for  	 example,  X http://www.computerworld.com/hardwaretopics/hardware/server/story/0,10801,95476,00.html  and X http://www.tekrati.com/T2/Analyst_Research/ResearchAnnouncementsDetails.asp?Newsid=4077 I (the latter notes that this marks the 6th consecutive quarter of overall  G server revenue growth, incidentally, and both point out that growth in  J Unix servers, while lower than that in Wintel servers, has been positive).   > $ > That is not spin. That is reality.  I As just noted it is indeed not only spin but demonstrably false, and so,  5 obviously, is your pious assertion immediately above.    - bill   ------------------------------  % Date: Fri, 18 Feb 2005 22:23:31 -0500 ' From: "Main, Kerry" <kerry.main@hp.com> , Subject: RE: HP financials: Alpha sales downR Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB53F62A@tayexc19.americas.cpqcorp.net>   > -----Original Message-----4 > From: Bill Todd [mailto:billtodd@metrocast.net]=20! > Sent: February 18, 2005 6:56 PM  > To: Info-VAX@Mvb.Saic.Com . > Subject: Re: HP financials: Alpha sales down >=20  	 [snip ..]    =20 E > Neither IDC not Gartner would seem to share that view - see, for=20 
 > example,=20 @ > http://www.computerworld.com/hardwaretopics/hardware/server/st > ory/0,10801,95476,00.html=20 > and=20@ > http://www.tekrati.com/T2/Analyst_Research/ResearchAnnouncemen  > tsDetails.asp?Newsid=3D4077=20B > (the latter notes that this marks the 6th consecutive quarter=20 > of overall=20 A > server revenue growth, incidentally, and both point out that=20  > growth in=20? > Unix servers, while lower than that in Wintel servers, has=20  > been positive).  >=20 > >=20& > > That is not spin. That is reality. >=20> > As just noted it is indeed not only spin but demonstrably=20 > false, and so,=20 7 > obviously, is your pious assertion immediately above.  >=20 > - bill >=20  H These analysts are only looking at new servers - not the total number of servers in a company.   F How do the analysts explain the exponential number of requests for newH services dealing with server consolidation happening all over the place?  G Not just with HP, but Sun, IBM and every med-large SI company that does H IT related work. As a test - go to www.ibm.com and www.sun.com and enterG "server consolidation" into their main search engine and see what comes  up.   E One Customer I personally worked with is planning to drop from 600 to  approx 250 Wintel servers.=20   ? Yes, they are buying blades and bigger Wintel servers to do the H consolidation, but the total number of Wintel servers the company has isH dropping drastically - and this hits the CS maint line of all vendors as% the new servers include warranty etc.   F I might even suggest that the reason for the UNIX/Wintel servers doingD relatively well right now is exactly because of server consolidationF i.e. you have to make an upfront investment now in new servers (bladesF make for good Windows/Linux servers) to save operations and reduce TCO! later after the consolidation.=20   D Course, these Customers are not planning to buy much new stuff afterB they get their consolidation done (they are trying to avoid serverH sprawl remember, so are building in growth head room now), so it will beD interesting to see what the Wintel/UNIX server growth numbers are in 6-12 months.  C Next time you talk to these analysts, ask them if they think server G consolidation is hot. If they say no, then they have absolutely no idea ' of what Customers are up to these days.    Regards   
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  $ "OpenVMS has always had integrity .. Now, Integrity has OpenVMS .."   ------------------------------  % Date: Sat, 19 Feb 2005 01:21:10 -0500 ( From: Bill Todd <billtodd@metrocast.net>, Subject: Re: HP financials: Alpha sales down= Message-ID: <aoSdnYZwuIhKR4vfRVn-uA@metrocastcablevision.com>    Main, Kerry wrote: >>-----Original Message-----2 >>From: Bill Todd [mailto:billtodd@metrocast.net] ! >>Sent: February 18, 2005 6:56 PM  >>To: Info-VAX@Mvb.Saic.Com . >>Subject: Re: HP financials: Alpha sales down >> >  >  > [snip ..]  >  >    > C >>Neither IDC not Gartner would seem to share that view - see, for   >>example,  @ >>http://www.computerworld.com/hardwaretopics/hardware/server/st >>ory/0,10801,95476,00.html  >>and @ >>http://www.tekrati.com/T2/Analyst_Research/ResearchAnnouncemen >>tsDetails.asp?Newsid=4077 @ >>(the latter notes that this marks the 6th consecutive quarter 
 >>of overall  ? >>server revenue growth, incidentally, and both point out that   >>growth in = >>Unix servers, while lower than that in Wintel servers, has   >>been positive).  >> >>% >>>That is not spin. That is reality.  >>< >>As just noted it is indeed not only spin but demonstrably  >>false, and so,  7 >>obviously, is your pious assertion immediately above.  >> >>- bill >> >  > J > These analysts are only looking at new servers - not the total number of > servers in a company.   G So your thesis seems to be that sales are up (temporarily) but service  F revenues are down, because larger numbers of older machines are being G replaced with smaller numbers of newer machines of larger capacity but  I lower total service costs.  This was not the way I first understood your  H earlier statements about shrinkage in 'the overall numbers of VMS [and, F as you then extended it, all] servers', which is why I responded with = the Gartner/IDC data, but in rereading them I can see it now.   G While I'll leave it to the business folk here to comment upon just how  F realistic a scenario this is, it seems a bit strange to me.  I'd much F more expect businesses to perform rolling upgrades, both to avoid the F magnitude of the catastrophe should something go awry in the move and G because their systems vary in age and the newer ones are worth a great  I deal more to them left up and running in place than they would be on the  F used market (and for that matter, if they decided to dump them on the G used market anyway, many would then wind up generating service revenue  D again in their new homes).  I'm also inclined to suspect that while G consolidation may be all the rage in the U.S., most of the rest of the  F world is still ramping up its primary server use and hasn't gotten to  the consolidation stage yet.  B So I still don't find that explanation entirely convincing in the I description of its nature or of its effect on service revenue, but, as I  C said, that's more for the kinds of people who actually manage such   transitions to comment upon.   > H > How do the analysts explain the exponential number of requests for newJ > services dealing with server consolidation happening all over the place?  I Now, *that* would suggest that service revenues should be *up*, at least   temporarily.  E None of which, of course, gets back to the basic question of whether  I Keith's estimates were based on pre-Alphacide numbers (which they seemed  0 suspiciously close to) or something more recent.   - bill   ------------------------------  # Date: Fri, 18 Feb 2005 21:12:45 GMT 1 From: Keith Parris <keithparris_NOSPAM@yahoo.com> Y Subject: Itanium Montecito Wins Microprocessor Report Analysts' Choice Award for Best Ser . Message-ID: <hTsRd.337$WJ.29@news.cpqcorp.net>  Q http://h21007.www2.hp.com/dspp/files/unprotected/Itanium/MPRonline-Itanium105.pdf   D "Intels Itanium processors continue to improve year after year. In I 2004, Intel introduced another version of the 130nm Madison die with 9MB  H of L3 cache. ... But this is only a warm-up for the next version of the < Itanium, code-named Montecito, which began sampling in 2004.  I By far the most ambitious server-processor design in terms of transistor  I count is Intels Montecito, using 1.72 billion transistors (Figure 5) in  C 90nm to create a dual-core processor with more than 26MB of on-die  I cache. Intels manufacturing prowess continued to amaze with its ability  I to build gigantic chips with these enormous caches. The chip promises to  H produce improvements in performance, due to faster clock speeds, larger & caches, and coarse multithreading. ...  G Even with all this logic integration, Intel managed to lower the power   envelope as well. ...   @ Montecito will likely be the highest-performing mainframe-class C processor in 2005 and the highest-scoring processor on SPEC2000_fp.   B Despite all the criticisms ... Intel continues to show unwavering ! support for the architecture. ...   G By late 2005, it will ... have become the standard to judge against in   bigiron systems.   And the Winner Is...  C This year the decision came down to AMDs dual-core Opteron, IBMs  < Power5, Intels Montecito, and Suns Niagara processors. ...  G The Intel Montecito processor is the winner of our MPR AnalystsChoice  ( Award for Best Server Processor of 2004.   ------------------------------  % Date: Fri, 18 Feb 2005 16:32:37 -0500 % From: "vax, 9000" <vax9000@gmail.com> Y Subject: Re: Itanium Montecito Wins Microprocessor Report Analysts' Choice Award for Best : Message-ID: <cv5ml3$4gn$2@charm.magnus.acs.ohio-state.edu>   Keith Parris wrote:   A > Montecito will likely be the highest-performing mainframe-class E > processor in 2005 and the highest-scoring processor on SPEC2000_fp.   % Who scores highest SPEC2000-int? AMD.   	 vax, 9000    ------------------------------  # Date: Fri, 18 Feb 2005 22:26:27 GMT 1 From: Keith Parris <keithparris_NOSPAM@yahoo.com> Y Subject: Re: Itanium Montecito Wins Microprocessor Report Analysts' Choice Award for Best / Message-ID: <nYtRd.343$7M.103@news.cpqcorp.net>    vax, 9000 wrote: > Keith Parris wrote: A >>Montecito will likely be the highest-performing mainframe-class E >>processor in 2005 and the highest-scoring processor on SPEC2000_fp.  > ' > Who scores highest SPEC2000-int? AMD.   H But there are no Montecito results yet at spec.org. Madison 9M is there H at 1590 at 1.6 Ghz. MPR estimates 1700 for Montecito at 1.7 Ghz. AMD is B presently at 1854. Interesting how MPR deftly excludes AMD in its 6 statement above by using the phrase "mainframe-class".   ------------------------------  % Date: Fri, 18 Feb 2005 18:20:46 -0500 ( From: Bill Todd <billtodd@metrocast.net>Y Subject: Re: Itanium Montecito Wins Microprocessor Report Analysts' Choice Award for Best = Message-ID: <d_CdnUoFg-zS5YvfRVn-3Q@metrocastcablevision.com>    Keith Parris wrote: T > http://h21007.www2.hp.com/dspp/files/unprotected/Itanium/MPRonline-Itanium105.pdf   E What an utter farce.  I dealt with it a week ago when it appeared at  6 RealWorldTech, so can just recycle that response here:    G Wow - I suppose one could call this 'audacious', but on the face of it  6 it looks a great deal more like outright incompetence.  I I mean, this *is* a comparison of *server* processors, right? So SPECint  H and SPECfp aren't quite the be-all and end-all here, even if some might 4 consider them of great importance in other settings.  ? I particularly enjoyed his single-sentence dismissal of POWER5:   I "The POWER5 is a dual-core, dual-threaded processor that is near the top   in many system benchmarks."   G Well, if by 'near the top' he meant 'so far out in front that the next  A competitor might as well not have showed up' and by 'many system  A benchmarks' he meant 'most common commercial server benchmarks',  H perhaps. Let's actually look at a few, though, rather than spew dubious  generalities as Krewell did:  G TPC-C: POWER5 leads its closest competitor by a factor of almost 3.2:1  H at the 64-processor node, by a factor of over 2.6:1 at the 16-processor C node, by a factor of over 1.73:1 at the 8-processor node, and by a  D factor of 1.2:1 at the 4-processor node (in the last two cases, the B closest competitor was a system based on the new, top-of-the-line F Madison 9M; in the first two, the Madison 6Ms which were their top of H the line a few months ago). So it's a clear winner even in small system 6 sizes, and wins running away as system sizes increase.  D TPC-H: Since IBM has only submitted a couple of 16-processor POWER5 H systems at the 1000 scale factor level, the only direct observation one I can make is that POWER5 beats its nearest competitor (16 1.3 GHz Itanics  > in a Legend system) by a factor of a bit over 2.6:1 (and on a D per-processor basis beats an HP 2.2 GHz Opteron system by nearly as D much). Other comparisons are difficult due both to use of different B scale factors and to potential differences due to the presence or  absence of clustering.  I SAP SD 2-tier: HP has always been kind of shy about submitting Superdome  I Itanic results for this benchmark (perhaps reflecting Superdome's rather  G mediocre scaling characteristics in commercial use), preferring to let  D its aging Alpha system carry the flag above the 16-processor system G level. But a bit under a year ago NEC's 32-processor Madison 6M system  G finally managed to beat the 32-processor score of 4500 users which EV7  F established over 2 years ago, and just yesterday bumped its score all D the way up to 5210 using the new Madison 9Ms (though a 32-processor I Fujitsu SPARC64 notched up a score of 5200 6 months ago). Unfortunately,  C POWER5's score of 5056 nearly matches this using only half as many  A processors, while POWER5's 64-processor score of 20,000 close to  F quadruples it using only twice as many processors, suggesting another A advantage of close to 2:1 over the second-place contenders (plus  E excellent scaling characteristics to boot). The situation isn't much  G better with HP's smaller Integrity line: POWER5's 8 processor score of  D 2600 is only 10% lower than Integrity's 16-processor score of 2880, A while 73% higher than Integrity's 8-processor score of 1500. And  G POWER5's 4-processor score of 1313 is over 49% higher than Integrity's  ? 880 (though Opteron is in second place here with a score of 914 C [late-breaking update:  now 1017]). So it's not *just* a matter of  I scaling: POWER5 is a *lot* faster, period, across a wide range of server   sizes in this benchmark.  F SAP SD 3-tier: Lots fewer comparisons possible here than with 2-tier, C but (just to give the flavor of things) the brand-new 64-processor  E Madison 9M Superdome manages a score of 100,000 users, which is only  E about 4.6 times the 4-processor POWER5 score of 21,712. An HP system  D with eight 1.1 GHz 4 MB Itanics using their mx2 card managed to hit I 17,184, while a 4-processor Madison 6M system only hit 7,800 (go figure:  8 perhaps they learned a lot in the intervening 6 months).  E JBB2000: Now here's a benchmark that a lot of vendors appear to take  F seriously, given the number of submissions and credible contenders. A F while ago Fujitsu edged out Madison 6Ms at the 32-processor node size H and smaller and more recently started clobbering them fairly decisively G across the board, and similarly AMD first edged out and then clobbered  F Madison 6Ms at the 4-processor node size - while at 32 cores HP's own I new PA8800 runs neck-and-neck with Madison 6M systems (both at 574K plus  H change). But once again POWER5 puts them all to shame, with scores more H than 50% above those of its closest competitor (usually Fujitsu) across I the size range save at 4 processors and below, where it's only about 50%  E higher than Madison 6M and only about 15% ahead of Opteron (I didn't  3 find a recent 4-processor submission from Fujitsu).   F SPECweb99_SSL: This isn't nearly as generally-popular a benchmark and H was a favorite of HP's to quote for Itanic early-on, when it led there. I More recently, Opteron took away that lead, but the new Madison 9Ms have  I edged just a smidge ahead of it again. However, POWER5 is slightly ahead  G of them both - using a crypto accelerator card, to be sure, but that's   legal.  I SPECweb99: A much more popular benchmark than its sibling above, and one  H that HP was very shy about submitting Itanic to until *very* recently - G quite possibly because it would have had difficulty beating the scores  E posted by Opteron and POWER4+. I guess it thought the new Madison 9M  F might be able to cut the mustard, and indeed it surpasses those two - # but still can't quite catch POWER5.     D Now, that's not all the commercial server benchmarks out there, and F possibly not even all the generally respected ones (I did check OASB, I but there weren't any reasonably apples-to-apples comparisons there). So  F by all means remind me of any significant ones that I may have missed.  D Of course, Montecito will presumably improve on the current Madison H figures, but well before Montecito ships POWER5+ will be with us, which H should also be good for a noticeable boost. Given that the scores above D suggest that the *existing* POWER5 platforms may not have very much I difficulty staying ahead of Montecito in most commercial benchmarks, the  D main remaining question would seem to be whether Montecito can beat H POWER5+ in any non-commercial benchmarks that might possibly qualify as G significant to 'servers' (the category in which MPR was presenting its   accolade, in case you forgot).  G Perhaps Krewell was just so awestruck (or would 'dumbstruck' be a more  E apt characterization?) at how many transistors Intel managed to cram  E into Montecito that the analytical areas in his brain just shut down  E completely. But it remains a mystery why others at MPR didn't gently  E take him aside and suggest that he take another look: are they *all*   that incompetent?    - bill   ------------------------------  % Date: Fri, 18 Feb 2005 18:21:29 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> Y Subject: Re: Itanium Montecito Wins Microprocessor Report Analysts' Choice Awardfor Best  B Message-ID: <1108768174.85763da66c574107941ab1f4ee56ead8@teranews>   Keith Parris wrote: D > 90nm to create a dual-core processor with more than 26MB of on-die	 > cache.    C So, Intel is increasing effective cache from 9 M to 13m. (since the D cache is not shared between teh 2 cores, each core only gets 13meg).    C > Intels manufacturing prowess continued to amaze with its ability 5 > to build gigantic chips with these enormous caches.   E How fast would the 8086 go if it were given such an oversized cache ?   C > Despite all the criticisms ... Intel continues to show unwavering # > support for the architecture. ...   D No, they are just executing the plan of record. Last year, they took< many actions to narrow IA64's market and potential. And more; importantly, will get the 8086 to compete against the IA64.   H > By late 2005, it will ... have become the standard to judge against in > bigiron systems.  ) Note the frequent use of the word "will".   H > The Intel Montecito processor is the winner of our MPR AnalystsChoice* > Award for Best Server Processor of 2004.  D Sorry, but that sentence makes the whole article non-credible. SinceA this chip is not available commercially on systems yet, and it is A already 2005, how can such a virtual chip get an award for 2004 ?   E It may get some speed wins when it becomes availale commercially, but B until that happens, it is wrong to talk of Montecito as if it wereP "current". Remember  the frequent use of the word "will" earlier in the article.   ------------------------------  % Date: Sat, 19 Feb 2005 04:04:32 +0800  From: prep@prep.synonet.comw6 Subject: Re: linux kernel has major security flaws ...- Message-ID: <87u0o9ljn3.fsf@prep.synonet.com>a  / Kilgallen@SpamCop.net (Larry Kilgallen) writes:   F > The question I would ask (also not as a Linux expert) is how easy itC > is for the unprivileged user to get read access to the swap area.oC > If this is the installation default or similarly easy to achieve,A# > then I think the hole is genuine.t  ? As I understand it, it in effect allows you to map swap :( Thiso is, as they say, not good.    ? > As you point out, physical access beats all, even if it meansuA > leaving a hardware device to intercept the key which is used to $ > encrypt the page and swap file :-)  > You are on the decrypted side of the swap, so it only protects> some one doing a scavange on the drive afterwards. MSCP drives< could do this, but data security is not industry standard so
 it had to go.r   -- R< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.e@                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Fri, 18 Feb 2005 20:48:23 +0100tB From: Michiel Erens <I.dont.want.spam@this.mailaddress.is.invalid>" Subject: Re: looking for a new job# Message-ID: <42164676$1@news.nb.nu>c   Heinz wrote: > Hi everybody > L > After the company I use to work for closed the doors I was left without a  > job.N > At this time it is very hard to find a job as vax/vms system manager in the  > netherlands.O > so if there is somebody out there in the world who can use a vax/vms - NT -  uE > system manager please let me know and I be happy to send you my cv.n  @ http://www.cdc.nl/dto/werken_bij/204804_senior_ictbeheerder.html   -- n ME Posted by news://news.nb.nu-   ------------------------------  # Date: Fri, 18 Feb 2005 20:24:09 GMT # From: hoff@hp.nospam (Hoff Hoffman)a: Subject: Re: Looking for suggestions for bootcamp sessions/ Message-ID: <J9sRd.333$fH.269@news.cpqcorp.net>7  \ In article <avfb119bmu4vl9pt0umdvc0ejub86ii8cj@4ax.com>, Nigel Barker <nigel@hp.com> writes:B :On 16 Feb 2005 19:16:02 -0800, kenneth.randell@verizon.net wrote: :gH :>On an Alpha, I simply type B DKA0 and I'm all set -- this is very easyF :>for my users to remember.  My experience on Itanium is that the bootH :>information has some kind of disk-partition/other-magic-code stored inF :>the NVRAM, so even if the disk is in the same slot and bootable, theA :>saved boot information is different so the system won't boot.  c  F   We have ways to get at the boot alias information out in the consoleD   from the running operating system environment, FWIW.  OpenVMS V8.2B   uses these to slam the boot options, to avoid the skewing of theA   GUIDs.  (Sounds rather weird, "skewing of the GUIDs", eh? :-)  -  J :>My remote users are not the kind to use BCFG commands in order to changeG :>things.  This part can likely be done via a script, but it seems moree :>hackerly to write a program. :2K :It's true that boot will fail if you swap out one system disk for another  L :set up to autoboot from a disk with a particular Globally Unique Identifier :(GUID).  E   Looks like it's time for an update session; we've been busy in this D   area.  :-)  There are two or three of us working within this area.  O :However it is still possible to do the equivalent of >>> b dka0 it's just thatd/ :you have to type in a few more characters e.g.l : $ :shell>  fs0:\efi\vms\vms_loader.efi  @   That's the fixed-disk path to the OpenVMS EFI bootstrap script   in the first FAT partition.t  O :Creating a new entry in the boot menu is just a question of selecting a coupleu :of menu items.8     Ayup.     N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq-N  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com6   ------------------------------  % Date: Fri, 18 Feb 2005 14:05:22 -0700 . From: "Michael D. Ober" <mdo.@.wakeassoc..com>F Subject: Re: Micro$oft warns of undetectable spyware security risk .... Message-ID: <mMsRd.22$jx4.733@news.uswest.net>  H This is why you need to be sure you know where your patches (for any OS,L including VMS) are coming from.  One of VMS's predecessors on the PDP-11 wasH hacked by a "official" upgrade that contained a backdoor.  This hack wasD done during a security evaluation and was packaged to appear to have originated at DEC.  L If this becomes a common problem, I suspect it will spell the end of free OS downloads (Linux).  
 Mike Ober.  & <bob@instantwhip.com> wrote in message< news:1108733310.343931.80970@g14g2000cwa.googlegroups.com...< > so you think everything can be patched in windoze land ... > + > http://www.theinquirer.net/?article=21326  >  >6   ------------------------------  % Date: Sat, 19 Feb 2005 04:33:34 +0800  From: prep@prep.synonet.com * Subject: Re: Mounting disks during STARTUP- Message-ID: <87ekfdliap.fsf@prep.synonet.com>g  4 David J Dachtera <djesys.nospam@comcast.net> writes:   > Bart Zorn wrote:  aI >> Not really. If you find a disk which appears to be a shadowset member,0D >> you can mount it with the /INCLUDE qualifier and it will find its  >> fellow members automatically.  % > Has anyone seen that actually work?   4 Yes, I was doing this on my home system for a while.   -- 5< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.s@                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Fri, 18 Feb 2005 21:08:03 GMTn# From: Beach Runner <bob@nospam.com> * Subject: Re: Mounting disks during STARTUP; Message-ID: <TOsRd.91691$qB6.89052@tornado.tampabay.rr.com>   I The best reason to use standard procedure is that someday you may not be  G there, or you will be on vacation and your system needs help.  Keeping  . things in standard locations facilitates that.   John Brandon wrote:u   > VMS V7.2-1h1 Alpha> > I mount various disks drives in the following command files: >  >   SYS$MANAGER:SYLOGICALS.COM! >   SYS$MANAGER:SYPAGSWPFILES.COM, >   SYS$MANAGER:SYSECURITY.COM! >   SYS$MANAGER:SYSTARTUP_VMS.COM  >  >   (basic order of startup) > K > In all cases, the disks being mounted are always online as they are beingeB > served by HSG80 controllers - so there is not spin-up wait time. > Q > I started thinking why not mount all the disk drives in one command file ratheraO > than the multiple locations as is now.  I was thinking that either one of theu, > following would be ideal for this purpose: >  >   SYS$MANAGER:SYCONFIG.COM >   SYS$MANAGER:SYLOGICALS.COM > 3 > I am aware that if I use SYS$MANAGER:SYCONFIG.COM ( > I would need to execute the following: > ( >   $ MCR SYSMAN SYSMAN IO AUTOCONFIGURE > E > The reason to use SYCONFIG is to keep the startup strategy in tact.  >  > By Startup Strategy I mean:r/ >   - in SYLOGICALS perform logical definitionsi7 >   - in SYPAGESWPFILES perform page/swap file installse
 >   - etc. > P > I tend to believe that there should be a separate startup process for mountingK > disks - since most everything you do at startup requires not only CPU and P > Memory but Disk drives as well.  Pretty much a rudimentary requirement to have7 > your disk drives mounted prior to completing startup.o >  > 6 > Any thoughts on using either SYCONFIG or SYLOGICALS? >  > B > Anyone doing something similar they would care to share with us? >  >  > TIAp >  >  > John "REBOOT" Brandoni > VMS Systems Administrator , > firstname.lastname.spam.me.not@dalsemi.com   ------------------------------  # Date: Fri, 18 Feb 2005 21:08:48 GMTs# From: Beach Runner <bob@nospam.com>h* Subject: Re: Mounting disks during STARTUP; Message-ID: <APsRd.91692$qB6.45523@tornado.tampabay.rr.com>v   John Brandon wrote:   ( > helbig@astro.multiCLOTHESvax.de wrote: > F >>>The reason to use SYCONFIG is to keep the startup strategy in tact.   It can break autogen.      >>+ >>Can you elabourate on this last sentence?= >  > J > As I tried to explain in the first thread - each startup routine has itsP > purpose.  I guess I am expanding on the intent of SYCONFIG by making it a diskP > startup procedure - right or wrong - I just got it in my head that SYCONFIG is) > the right procedure to mount the disks.  >  > ( > helbig@astro.multiCLOTHESvax.de wrote: > J >>I have a procedure called MOUNT.COM which resides on a disk which is notE >>a system disk for any node in the cluster (actually, it is the useraF >>disk).  This also contains SYSUAF and other such files.  The idea isI >>that these should be common across the cluster, so there is no point invE >>having them on a) just one system disk or b) maintained in multiplehD >>copies on several system disks.  I mount all disks but 2 here (see
 >>below).  >  > Q > I use a similiar concept - I called mine DISKS$MOUNT_STARTUP.COM.  However I doiQ > not hard-code my devices.  The procedure reads a text file (DISKS_DATABASE.DAT)rJ > that contains pertinent information for each device (device name, label,P > logical, and shadowset.)  The procedure will perorm checks on the device priorC > to mount and reports back any problems prior and after the mount.  > H > I created a CLUSTER$DISK that contains all my common startup routines.O > The SYLOGICALS and SYSTARTUP_VMS reference this disk for the startup routinesu > and executes them. > M > So we have similiar concepts in startup routines.  Just different names andb? > variants on hard-coding (I hate it when I have to hard-code).a >  > / > Ken Fairfield [my.full.name@intel.com] wrote:a > ? >>Is there a reason you don't want to wait until SYSTARTUP_VMS?i >  > O > I feel that one routine should be responsible for mounting all your disks.  IsM > view disks no differently than CPU or Memory.  Therefore my disks should be , > mounted very early in the startup process. > Q > Mounting them in (say) SYCONFIG I avoid having to put the same mount mechanismse5 > in SYLOGICALS, SYPAGSWPFILES, SYSECURITY, SYSTARTUPs* > Why not just have them in ONE procedure? > P > It is easier to manage.  And if I happen to shuffle my startup disks around itL > will be transparent to the startup - as long as I do not change the volume  > label then I am in good shape. >  >  >  >  > John "REBOOT" Brandono > VMS Systems Administrator., > firstname.lastname.spam.me.not@dalsemi.com   ------------------------------  % Date: Fri, 18 Feb 2005 20:59:30 -0600i2 From: David J Dachtera <djesys.nospam@comcast.net>* Subject: Re: Mounting disks during STARTUP+ Message-ID: <4216AB92.B95547F1@comcast.net>o   Bart Zorn wrote: > H > There is one example which demonstrates that it works: when you have aH > shadowed system disk with more than one member, the second (and third)@ > member get automatically added to the shadow set at boot time.  $ That's how it SUPPOSED to work, yes.  H When my system boots, I see messages about shadow-set inconsistency, and> then the boot disk is reMOUNTed as a single-member shadow-set.  H First guess time would be because the primary swap/page files are on the> system disk, and it can be DISMOUNTed cleanly during OPCCRASH.  ? > But yes, I have seen it work and we use it here in a separatee6 > MOUNT_DISKS.COM which is called frol SYLOGICALS.COM.  G I'll try messing with it on my little Alpha at home. I still think I'lleG need to finish up SSCKB to get what I want, though (a central knowledgewF repository for shadow-set constituency information, other than between the SysAdmin's ears).t   --   David J Dachtera dba DJE Systemst http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:c" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/n  " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/e   Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------    Date: 18 Feb 2005 12:13:14 -0800 From: tadamsmar@yahoo.com-A Subject: MY patching procedures (was Re: 7.32 patching questions) B Message-ID: <1108757593.967434.76060@f14g2000cwb.googlegroups.com>   Steven M. Schweda wrote: > From: tadamsmar@yahoo.comi >o& > > > Peter 'EPLAN' LANGSTOEGER wrote: > > > > TCPIP_ECO		V5.4-154o > > >n) > > > First, I don't see that patch here:i > > >lC > > > ftp://ftp.itrc.hp.com/openvms_patches/layered_products/alpha/e	 > > [...]m8 > > Opps! I was running PROD SHOW PROD on the wrong box! > >  > > I'm at TCPIP V5.4-15 >n >    That makes more sense.  >)9 > > but the ECO is still not on the ftp site, so where doe > > I find it? >dG >    I had the same problem.  The key is not to be distracted by these:  >t/ >       ALP_TCPIP_ECO-V0501-155.PCSI-DCX_AXPEXEt# >       ALP_TCPIP_ECO-V0501-155.txtl/ >       ALP_TCPIP_ECO-V0503-182.PCSI-DCX_AXPEXE/# >       ALP_TCPIP_ECO-V0503-182.txt0 >m  > but to continue down to these: >n8 >       DEC-AXPVMS-TCPIP_ECO-V0504-151-4.PCSI-DCX_AXPEXE, >       DEC-AXPVMS-TCPIP_ECO-V0504-151-4.txt8 >       DEC-AXPVMS-TCPIP_ECO-V0504-152-4.PCSI-DCX_AXPEXE, >       DEC-AXPVMS-TCPIP_ECO-V0504-152-4.txt? >       DEC-AXPVMS-TCPIP_ECO-V0504-154-4.PCSI-DCX_AXPEXE   <---e? >       DEC-AXPVMS-TCPIP_ECO-V0504-154-4.txt               <---s >o% >    It's a clever naming convention.n >a > H ------------------------------------------------------------------------ > 6 >    Steven M. Schweda               (+1) 651-699-98185 >    382 South Warwick Street        sms@antinode-orgi >    Saint Paul  MN  55105-2547c  " Here are the command files I used.# No doubt some can improve on these.@   GET.COM:   !Usage: @GET "carly@hp.com"iA $ftp ftp.itrc.hp.com/username="anonymous"/p="''p1'"/input=get.ftpt   GET.FTPt    cd /openvms_patches/alpha/V7.3-2 get *VMS732_UPDATE-V0300*.txtc get *VMS732_SYSINI-V0100*.txte get *VMS732_MQ-V0100*.txt  get *VMS732_BACKUP-V0300*.txte get *VMS732_CPU270F-V0100*.txt get *VMS732_RMS-V0200*.txt get *VMS732_PTHREAD-V0200*.txt! get *VMS732_FIBRE_SCSI-V0400*.txtt get *VMS732_SYS-V0600*.txt get *VMS732_TRACE-V0200*.txt get *VMS732_PCSI-V0100*.txth get *VMS732_DRIVER-V0100*.txtl* cd /openvms_patches/layered_products/alpha get *TCPIP_ECO-V0504-154-4*.txtv set type image get *VMS732_UPDATE-V0300*.PCSI*g get *VMS732_SYSINI-V0100*.PCSI*n get *VMS732_MQ-V0100*.PCSI*0 get *VMS732_BACKUP-V0300*.PCSI*n  get *VMS732_CPU270F-V0100*.PCSI* get *VMS732_RMS-V0200*.PCSI*  get *VMS732_PTHREAD-V0200*.PCSI*# get *VMS732_FIBRE_SCSI-V0400*.PCSI*  get *VMS732_SYS-V0600*.PCSI* get *VMS732_TRACE-V0200*.PCSI* get *VMS732_PCSI-V0100*.PCSI*s get *VMS732_DRIVER-V0100*.PCSI* * cd /openvms_patches/layered_products/alpha! get *TCPIP_ECO-V0504-154-4*.PCSI*l   RUNEXE.COM:v  - (assumes that the files are in [ADAMS.PATCH]) 0 (note that this may not extract all files, since/ some of the downloads may already be extracted)l   $ LOOP:l= $       REP_SPEC = F$SEARCH("[ADAMS.PATCH]*.PCSI-DCX_AXPEXE")o $       IF REP_SPEC .NES. "" $       THEN $           RUN 'REP_SPEC' $           GOTO LOOPr
 $       ENDIFo  9 to test the files I used the following command procedure, 5 but you will want to run this in a scratch directory,o since is makes a big mess.   EXTRACT.COM:  ' !Define PCSI$SOURCE before you run thisr, $PROD EXTRACT FILE/OPTI=NOCONF VMS732_UPDATE  , $PROD EXTRACT FILE/OPTI=NOCONF VMS732_SYSINI( $PROD EXTRACT FILE/OPTI=NOCONF VMS732_MQ, $PROD EXTRACT FILE/OPTI=NOCONF VMS732_BACKUP- $PROD EXTRACT FILE/OPTI=NOCONF VMS732_CPU270Fn) $PROD EXTRACT FILE/OPTI=NOCONF VMS732_RMSl- $PROD EXTRACT FILE/OPTI=NOCONF VMS732_PTHREADa0 $PROD EXTRACT FILE/OPTI=NOCONF VMS732_FIBRE_SCSI) $PROD EXTRACT FILE/OPTI=NOCONF VMS732_SYSs+ $PROD EXTRACT FILE/OPTI=NOCONF VMS732_TRACEe* $PROD EXTRACT FILE/OPTI=NOCONF VMS732_PCSI, $PROD EXTRACT FILE/OPTI=NOCONF VMS732_DRIVER( $PROD EXTRACT FILE/OPTI=NOCONF TCPIP_ECO  / This is the install, it asks lots of questions.   / Caution, I am not sure that it is OK to installs3 all the patches without rebooting between them, bute that is what this does:   0 A less trusting soul may want to add /SAVE, some# kits assume no save, and don't ask.   5 Note that the kit from the mandatory security CD thats3 came with the 8.2 distribution is not included heree  , I believe I ordered these as required by the
 dependencies.r  ' !Define PCSI$SOURCE before you run thisa% $PROD INSTALL/OPTI=NOCONF VMS732_PCSI4  $SET COMMAND SYS$UPDATE:PCSI.CLD' $PROD INSTALL/OPTI=NOCONF VMS732_UPDATEa$ $PROD INSTALL/OPTI=NOCONF VMS732_SYS' $PROD INSTALL/OPTI=NOCONF VMS732_SYSINI-# $PROD INSTALL/OPTI=NOCONF VMS732_MQ:' $PROD INSTALL/OPTI=NOCONF VMS732_BACKUPe( $PROD INSTALL/OPTI=NOCONF VMS732_CPU270F$ $PROD INSTALL/OPTI=NOCONF VMS732_RMS( $PROD INSTALL/OPTI=NOCONF VMS732_PTHREAD+ $PROD INSTALL/OPTI=NOCONF VMS732_FIBRE_SCSIg& $PROD INSTALL/OPTI=NOCONF VMS732_TRACE' $PROD INSTALL/OPTI=NOCONF VMS732_DRIVER2# $PROD INSTALL/OPTI=NOCONF TCPIP_ECO.   ------------------------------  % Date: Fri, 18 Feb 2005 21:07:58 -0600.2 From: David J Dachtera <djesys.nospam@comcast.net>" Subject: Re: Need performance help+ Message-ID: <4216AD8D.67EA9B26@comcast.net>h   "Main, Kerry" wrote: >  > > -----Original Message-----= > > From: David J Dachtera [mailto:djesys.nospam@comcast.net]c# > > Sent: February 17, 2005 8:48 PM4 > > To: Info-VAX@Mvb.Saic.Comt& > > Subject: Re: Need performance help > >o > > Kenneth Farmer wrote:g > > >i> > > > Please email this guy directly if you're able to assist. > > >h& > > > Rajat.Gupta(at)gb.vodafone.co.uk > > >>< > > > I've also posted in OpenVMS Tech forum on OpenVMS.org: > > > > > > > http://www.openvms.org/phorum/read.php?f=1&i=1533&t=1533 > > >a
 > > > Hello ,i > > >i@ > > > Thank you very much for the reponse. I want to know of any > > profiling tool/ > > > avalaible for openVMS like gprof on unix.n > > >fA > > > Actually I have a process that gives a performance of 20000  > > transactions perA > > > sec on linux machine but gives only 3000 transactions on anc > > OpenVMS platform= > > > with maximum CPU utilization. I want to get hold of theb > > limiting issues. > >oI > > Look into using Extended File Cache, and make sure your disk units ono= > > your storage arrays are set with writeback cache enabled.  > >d > 1 > Also, have not used this myself, but check out:h2 > http://h71000.www7.hp.com/openvms/products/dcpi/  G Looked at that today. I was hoping it could be a replacement for SPM oruD PSPA, but it seems much more engineering-oriented, and assumes a farB more intimate knowledge of the software than the average CTO, dataG center manager, etc. is likely to possess when seeking performance datat from the system.   Still hoping...a   -- g David J Dachtera dba DJE Systemse http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:b" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/d  " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Fri, 18 Feb 2005 15:10:32 -0500r' From: "Main, Kerry" <kerry.main@hp.com>'* Subject: RE: OpenVMS 8.2 docs and freewareR Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB53F5E8@tayexc19.americas.cpqcorp.net>   > -----Original Message----- > From: Sander, Warren=20 ! > Sent: February 18, 2005 1:09 PMf > To: Info-VAX@Mvb.Saic.Comu, > Subject: Re: OpenVMS 8.2 docs and freeware >=20B > the 7.3-2 version of the guide to security is the most recent=20 > copy. it was > not updated for 8.2n" > according to the doc group folks >=20< > "Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message/ > news:uDyJyfLtgR3Z@eisner.encompasserve.org...e@ > > In article <42152be4$1@usenet01.boi.hp.com>, "warren sander"  > <warren.sander@hp.com> writes:4 > > > The OpenVMS 8.2 docs are new up and available.# > > > http://h71000.www7.hp.com/doc- > >-> > > From here in Massachusetts, the 8.2 link gets me a page=20 > that says 8.2pB > > at the top but has only a column for 7.3-2.  The one manual=20 > I checkedn? > > (Guide to System Security, of course) is the 7.3-2 version.n >=20 >=20  C Minor note, but the security Volume 2 (SSL) and Volume 3 (Kerberos)i guides were updated for V8.2.I   Regardse  
 Kerry Main Senior Consultantt HP Services Canada Voice: 613-592-4660h Fax: 613-591-44773 kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  $ "OpenVMS has always had integrity .. Now, Integrity has OpenVMS .."   ------------------------------  % Date: Sat, 19 Feb 2005 04:31:45 +0800o From: prep@prep.synonet.comn* Subject: Re: OpenVMS 8.2 docs and freeware- Message-ID: <87is4plidq.fsf@prep.synonet.com>t   Thank you Warren.p     -- w< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.g@                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------   Date: 18 Feb 05 18:07:54 EST) From: cook@wvnvms.wvnet.edu (George Cook)l= Subject: Re: OpenVMS Cluster to Replace News.Individual.NET ?E! Message-ID: <QghMrTaJGPFJ@wvnvms>c  U In article <cv57mm$j4s$1@naig.caltech.edu>, David Mathog <mathog@caltech.edu> writes:t > George Cook wrote:F >  > I may be misunderstanding your data, but the greatest part of the8 >  > slowness appears related to file create/open/close? > = > In some cases that was true, but not for the examples givens > in the benchmark.. > ' > I've occasionally seen Unix code thatn8 > opens a file for append, writes to it, then closes it,= > then does it all over again.  This causes only a very minor,= > performance hit on Unix/linux (at least the ones I've used)c? > but is horrifically slow on VMS.  You note that DNEWS is muchnC > faster than ANUNEWS.  I'm not familiar with either one.  However,s6 > since you state that the latter keeps each post in a4 > separate file by definition it must be doing a lot= > of open/close, and it's almost axiomatic from that point onS/ > that it must be very slow.  I'm also going toA? > guess that it started as a port from some Unix program, where,D > all of those and opens and closes wouldn't have been as much of a @ > problem.  One can pretty much guess that DNEWS stores posts in@ > a more efficient manner, in some sort of container or database? > file, so that it doesn't need to open/close zillions of files>9 > all the time. Perhaps it uses RMS indexed files, and iflE > not, reimplements something fairly similar at the disk block level._ > F > Bypass RMS and then VMS can write data onto disk at a sustained rateJ > as fast as the hardware will go.  On the other hand, use RMS, especially: > through C, and typical smallish read/write programs are,8 > or at least were, much slower on VMS than on identicalG > hardware running linux.  This was a real problem for me since roughly-. > 100% of the software we use is written in C!  B Both ANUNEWS and DNEWS are written in C.  I am not sure of DNEWS'sE origin (I think UNIX), but it wasn't VMS.  ANUNEWS I believe was only D developed on VMS.  So we have a UNIX C program which runs as well asD I can ask on VMS, and a VMS C program which couldn't run much worse.  A It all has to do with the design and implementation.  Another wayiC of looking at it is that poorly designed/written C programs tend tocF have better I/O performance on UNIX.  From your analysis above I thinkC you would agree that ANUNEWS would run a great deal better on UNIX.g  B DNEWS has an even more efficient I/O scheme which I never botheredF to convert to because everything idles along just fine using the older scheme.e   > prep@prep.synonet.com wrote:, >> David Mathog <mathog@caltech.edu> writes: >> rH >> David, if you take a broken IO design and do a brain-dead port of it,F >> what do you expect... If you do it RIGHT, vms will pound unix IO toJ >> pulp. That is why unix is now getting such new ideas as MMIO and Direct	 >> IO. ;|t > H > Well, take a look at the benchmarks and tell me in which way they are E > broken.  Most of them do nothing more than open an input file, openLA > an output file (or files) and copy from one to the other.  TheyeK > represent a pretty typical IO load for the sorts of programs I encounter.a >  >> lG >> A News server is a notorious system killer. google for `i-node hack'1F >> for an intro. The normal news implementation uses a simplistic fileE >> design to store articles, and hammers file creation to a pulp. Thes! >> 50 GB/day is also problematic.s >>  C >> There is already a reasonable News server for VMS, ANU News, and C >> it has had several updates and overhauls by <some one who's nameiB >> I should be able to remember> in it's comercial supported form. > 8 > George said that ANUNEWS "was a total I/O dog" whereasA > DNEWS isn't.  You guys must be grabbing different parts of thata, > elephant.  Is ANUNEWS an I/O dog or not???  F ANUNEWS was written by Geoff Huston at Australian National University.I Version 1 is dated 17-Jul-1986.  I don't know of any commercial versions.sB It is a total I/O pig.  It requires multiple large RMS index filesG and large directory files (all the articles for one group are stored asrD individual files in one directory).  It required multiple batch jobsE running concurrently to add, delete and forward articles plus a greatiC deal of manual intervention.  In other words it would have required  effort to make it any worse.  E FWIW, I keep a small ANUNEWS setup (small enough that it doesn't neednD any attention) running for less than ten groups just so I can accessG them quickly with the ANUNEWS reader.  That is the one actual advantage B of ANUNEWS; the article database can be accessed directly and very quickly without using NNTP.n     George Cookb WVNETt   ------------------------------    Date: 18 Feb 2005 13:58:58 -0800 From: jordan@ccs4vms.com+ Subject: Re: OpenVMS Small Partner Strategy C Message-ID: <1108763938.863609.111130@f14g2000cwb.googlegroups.com>l  F > Not that you don't have better things to do with your time, but have you  > written to Ann Livermore?y    : Nope.  I used to write to Ms. Fiorina on occasion.  Got anB auto-responder once, thats it.  When HP dropped the bomb on us, weE wrote, called, yelled, and pulled strings with every contact we stillsD had from the DEC days, to no avail.  The decision had been made at a> very high level in HP by non-DEC people and was not subject to= complaints from expendable resources like us small resellers.O   ------------------------------  % Date: Fri, 18 Feb 2005 17:18:50 -0500 # From: "John Smith" <a@nonymous.com>r+ Subject: Re: OpenVMS Small Partner Strategy:, Message-ID: <YpOdnXc_D4hQ9IvfRVn-jA@igs.net>   jordan@ccs4vms.com wrote:fG >> Not that you don't have better things to do with your time, but have   >> you written to Ann Livermore? >n >o< > Nope.  I used to write to Ms. Fiorina on occasion.  Got anD > auto-responder once, thats it.  When HP dropped the bomb on us, weG > wrote, called, yelled, and pulled strings with every contact we still F > had from the DEC days, to no avail.  The decision had been made at a@ > very high level in HP by non-DEC people and was not subject to? > complaints from expendable resources like us small resellers.n    K During the Jan. 18th web chat session Ann said that she wanted to hear fromlJ people on how to increase VMS sales. Write her with your take on the smallH reseller situation and ask for what you'd really like to have occur as a$ solution and see what she says/does.   Let us know how it goes.   --- OpenVMS - The classics never go out of style.n   ------------------------------    Date: 18 Feb 2005 14:21:46 -0800 From: jordan@ccs4vms.com+ Subject: Re: OpenVMS Small Partner Strategy C Message-ID: <1108765306.093631.135690@g14g2000cwa.googlegroups.com>B  G > they are not cheaper ... there is the big lie ... total tco is highero  % > with any other non vms solution ...l  E You are correct.  And that matters not at all to the vast majority ofiD people buying computers who only look at the up front cost no matterF what kind of additional info you try to provide to them.  Thats why anE inexpensive (PC competitive pricing) complete VMS starter system withrC user licenses and apps/developer tools would, even now, be the bestl8 thing HP could do for the OS in addition to advertising.  B In any case, we were still HP resellers for (less than enterprise)F equipment, meaning wintel peecees.  When we tried to sell them, we gotD shut out on up front pricing.  Dell was selling stuff off the net as@ cheap or cheaper than the discount we got from HP for equivalentC hardware (yes, as a weenie little reseller we didn't get much breakoG from the distributors), and even the big retailers were cheaper than webE could sell HP boxes for  So we sold them Dells and installation/setupe services instead.s  C There is also a perception out there in smaller corporate that Dell ? small boxes, servers, etc. are 'better' for the price, and moresD reliable, than Compaq or HP boxes.  I haven't seen any proof of thatC but its a very real perception.  Maybe its from all those folks who C bought presarios and consumer HP systems in the late '90s and livedn through the support nightmares.c   ------------------------------    Date: 18 Feb 2005 14:22:42 -0800 From: jordan@ccs4vms.com+ Subject: Re: OpenVMS Small Partner StrategyhB Message-ID: <1108765362.522536.21180@z14g2000cwz.googlegroups.com>  E I'll work on it over the weekend and try to send something on Monday.    ------------------------------  % Date: Fri, 18 Feb 2005 21:11:34 -0600 2 From: David J Dachtera <djesys.nospam@comcast.net>+ Subject: Re: OpenVMS Small Partner Strategya+ Message-ID: <4216AE66.84373CD5@comcast.net>    Dave Froble wrote: >  > David J Dachtera wrote:aJ > > O.k. I've got the kernel of an idea. I can usually count on this groupH > > to pick such things apart down to the sub-atomic level, so let's see > > what anyone thinks of this:n > >tJ > > hp/VMS has this unrealistic thing about $1-megabuck in volume per yearK > > for "qualified resellers". Great for super-size VARs and resellers, bad  > > for us "little guys".t > >e# > > So, here's what I was thinking:  > >i. > > 1. Identify existing "qualified resellers"( > > 2. Establish relationships with themL > > 3. Enlist their support to negotiate with hp for support of (or at leastG > > non-interference with) a multi-tiered distribution network with theeG > > "qualified resellers" near the top, and us "bottom feeders" further 	 > > down.eC > > 4. We order VMS, Alphas and I64 machines through the "qualifiedd# > > resellers" as our sales happen.s > >f@ > > Seems to me, this can be sold (to hp) because all sides win. > > + > > Comments? Commendations? Condemnations?  > >e > > > Why not just form a combine of all the small VARs and such?   @ Some may recall me posting recent using a .SIG billing myself asF "Unoffical Co-ordinator, OpenVMS Consortium". Sound like what you have in mind?   > If theC > total of sales through the combine exceeds the $1M then it should. > qualify as a reseller.  B Depends what other roadblocks hp/VMS have erected to keep VMS from? reaching the stockholders' expectations for revenue generation.h  + > Ok, who's going to do all the paperwork? t  E I guess I'll have to, until we start making enough money to hire somer staff.   > Another idea shot in the ass.e  , Geez, Dave, can you be any more pessimistic?   -- e David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:e" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/t   Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Fri, 18 Feb 2005 21:17:31 -0600 2 From: David J Dachtera <djesys.nospam@comcast.net>+ Subject: Re: OpenVMS Small Partner Strategy-+ Message-ID: <4216AFCB.44C6356E@comcast.net>0   JF Mezei wrote:  >  > David J Dachtera wrote:uJ > > hp/VMS has this unrealistic thing about $1-megabuck in volume per yearK > > for "qualified resellers". Great for super-size VARs and resellers, bad  > > for us "little guys".k > F > The argument in favour of this policy is ensuring quality control ofJ > vendors and making sure that anyone making a pitch to a customer has had. > full training by Digital/Compaq/HP/whatever.  1 The intent and the result seem to be incongruous.f  I > The problem is that ever since the Palmer years, they have used this asuE > a barrier to selling VMS instead of as a facilitator to doing greata- > sales and recruiting more VMS salespersons.d  B Presumably, at least part of the fodder that feeds the "conspiracy theory" engine.,  . > > 1. Identify existing "qualified resellers"( > > 2. Establish relationships with them? > > 3. Enlist their support to negotiate with hp for support of  > C > One would have to know if the contracts between the vendor and HP,J > require that only "employees" with full HP training/accreditation handle > VMS sales.  D Would tend to be counter-productive, and possibly violate anti-trust; laws, since that would preclude hiring consultants or other $ third-parties to handle the process.  J > If there is no such restriction, then you could negotaite with Island orE > whoever some deal where you get a cut fo sales you generate. Nut ifiP > there are contract restrictions, the vendors may not be able to do such deals.  , "There's more than one way to skin any cat."   > (But yes, it is a good idea).  > B > The problem is that tiny vendors such as you and I cannot make aG > convincing sales pitch without the support of HP and by that, I mean:. > MARKETING AND ADVERTISING.  G ...which, if you recall recent posts, was one of the primary thrusts ofs  the ersatz "OpenVMS Consortium".  F Stop waiting for HP/VMS to pick-up the ball and run with it - it ain't happ'nin'. Get over it.t  J > If I go to a bank and pitch a VMS solution, I will be laughed at. But ifC > HP advertises VMS and banks become curious at VMS, then they willR > welcome my presentations.r  2 Depends how you present yourself. Think: teamwork.   -- p David J Dachtera dba DJE Systems9 http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:O" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/e   Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Fri, 18 Feb 2005 21:23:28 -0600 2 From: David J Dachtera <djesys.nospam@comcast.net>+ Subject: Re: OpenVMS Small Partner Strategyo+ Message-ID: <4216B130.9E730689@comcast.net>b   jordan@ccs4vms.com wrote:t > [snip]I > So in a way we can do what you described now, though with major caveatssH > and very little margin (and considering you're competing against otherG > end resellers who get a better deal than you on hardware and softwaresD > cost, you can't count on winning many sales... at least on price).  F Pricce alone is almost always a losing proposition. You'll end up with% an unsatisfied / insatiable customer.t  H > I don't see HP even bothering to pay attention to any such suggestion.  E They don't need to pay attention to it. In fact, better, if they justDH ignore it - let it be, and not interfere with it. We're putting money in, their pocket (whether they like it or not!).  D > They are just flat NOT INTERESTED in selling VMS into small/medium > business,>  
 Hence, us.  = > and so have no interest in vendors that specialized in that  > area.   F I wouldn't say "specialized", but perhaps "most experienced" would not not at all inaccurate.  A > Thats why we may well have sold our very last VMS system into auH > site last year, the end of a ~30 year run that started with PDP-8s andH > -11s and ended with Dells (because they are cheaper and better wintelsH > and we get a better deal retail from them than we can get from HP as aH > reseller of their wintel products, and Dell will sell _ANYTHING_ to us; > without imposing a million dollar minimum requirement...)E  # Well, do what ya gotta to survive. -  G Don't sell this short before it ever gets off the ground. If it crashescA and burns, so be it - at least we'll have tried. A mentor of mineoA teaches, "I'd rather see a crooked furrow than a field unplowed"..   -- s David J Dachtera dba DJE Systemsd http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:a" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/k  " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/d   Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Fri, 18 Feb 2005 21:29:57 -0600L2 From: David J Dachtera <djesys.nospam@comcast.net>/ Subject: Re: Port Print facility Update for I64c+ Message-ID: <4216B2B4.CE381156@comcast.net>h   John Reagan wrote: >  > David J Dachtera wrote:w > J > > Thanx to the maintainers of the test-drive cluster site for the use of > > the IA64 machine.0 > >3 > E > Just curious...  How long did it take you to port?  Any interestingS > problems you ran into?  G Its two fairly small pieces of BASIC code. I compiled, I built .OLBs, I>A edited the LINK procedure, I built it, I ran it (not exhaustivelys	 tested). e  E The source code is in the kit; so, I've always assumed that if anyonel/ downloads it, they adapt it to their own needs.o  G I could use some help trying AEST WRQ's ALPHALK2.EXE (the Alpha-versioniB of VAXLINK2). I got an image, but it ACCVIOs when I try to run it.  G WRQ also provides (unsupported, of course) a pair of X/Y/ZMODEM images, E for VAX only. I managed to VEST those a hile back, but no longer havetE the Alpha images (gotta VEST them again, I guess). I'd like to see ifW$ the VESTed images will AEST as well.   -- h David J Dachtera dba DJE Systemsb http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:1" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/g  " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/o   Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Fri, 18 Feb 2005 23:00:26 -0500g- From: "John E. Malmberg" <wb8tyw@qsl.network>a/ Subject: Re: Port Print facility Update for I64a1 Message-ID: <9uednQlZcJFBJIvfRVn-gA@adelphia.com>o   David J Dachtera wrote:A  I > I could use some help trying AEST WRQ's ALPHALK2.EXE (the Alpha-versionED > of VAXLINK2). I got an image, but it ACCVIOs when I try to run it.  D Why not simply compile it?  WRQ does not provide it in binary, they E provide it in source code, which they download to the host, and then p they compile it.  C It does that if it can not find the program, so if actions are not nC taken, you can end up with a copy of the transfer program in every i* directory that users have write access to.  I To prevent having copies of those files and the source to them scattered EA all over the systems I use to manage, I put a single copy of the II executables in a directory XFER_TOOLS:, and then I set both a symbol and AC a logical name to point to them so that no matter how the terminal t, emulator was trying find the image it would.  I > WRQ also provides (unsupported, of course) a pair of X/Y/ZMODEM images,PG > for VAX only. I managed to VEST those a hile back, but no longer have G > the Alpha images (gotta VEST them again, I guess). I'd like to see ifi& > the VESTed images will AEST as well.  I Look on the WRQ CD-ROM, you may find that they are all there in MACRO-32 u+ source code, but possibly with funny names.d   -Johns wb8tyw@qsl.network Personal Opinion Only    ------------------------------  % Date: Fri, 18 Feb 2005 22:43:38 -0600e2 From: David J Dachtera <djesys.nospam@comcast.net>/ Subject: Re: Port Print facility Update for I64r+ Message-ID: <4216C3FA.FC7BF877@comcast.net>n   "John E. Malmberg" wrote:e >  > David J Dachtera wrote:- > K > > I could use some help trying AEST WRQ's ALPHALK2.EXE (the Alpha-versioncF > > of VAXLINK2). I got an image, but it ACCVIOs when I try to run it. > E > Why not simply compile it?  WRQ does not provide it in binary, theyaF > provide it in source code, which they download to the host, and then > they compile it.  A If you know where to find the source code, please provide a link.e  A (Be careful not to confuse VAXLINK2 and ALPHALK2 (WRQ's host-sidet> protocol programs for VMS on VAX and Alpha, respectively) withC WRQUPLOA.MAR which is a "mini-Kermit" supplied in Macro/32 with therD output file name hard-coded. WRQUPLOA.MAR is uploaded as ASCII text,F commands are then sent to assemble/compile and link it, then it is runE to receive the VAXLINK.EXE program as binary over a presumably serialeE link. See WRQUPLOA.RCL (mostly on older R/2 and R/4 distro.'s) and/ore WRQUPLOA.RBS.)   > [snip]K > > WRQ also provides (unsupported, of course) a pair of X/Y/ZMODEM images,hI > > for VAX only. I managed to VEST those a hile back, but no longer haveaI > > the Alpha images (gotta VEST them again, I guess). I'd like to see ifs( > > the VESTed images will AEST as well. > J > Look on the WRQ CD-ROM, you may find that they are all there in MACRO-32- > source code, but possibly with funny names.o   Will look for them...T   -- e David J Dachtera dba DJE SystemsA http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:c" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/r  " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Sat, 19 Feb 2005 01:20:49 -0000r6 From: "Alex Daniels" <AlexNoSpamDaniels@themail.co.uk>8 Subject: Re: read monitor (>>>) variables from OpenVMS ?6 Message-ID: <42169474$0$32621$db0fefd9@news.zen.co.uk>  ? "David J Dachtera" <djesys.nospam@comcast.net> wrote in messagea% news:42140DA6.BD0CF220@comcast.net...tJ > ...or, type those commands into Notepad, and copy them to the clipboard.C > THen, connect to your console, hit CTRL+P, and press SHIFT-InsertnI > (faster than ALT E P or mousing). Your terminal program should send they. > data to the computer as if you had typed it. >a
 > Caveats: >v* > o Do not attempt on a production machine0 > o VMS may bugcheck after the exx0_mode change. >U    J It is also worth noting doing a CTRL/P when you are intending to continue,G is inherently more risky on SMP systems. A CTRL/P only halts CPU 0, then; other CPU's are still processing, with obvious limitations.a  K Having said that, I do use CTRL/P on test/development SMP machines and theny	 continue.i   Alex   ------------------------------  % Date: Fri, 18 Feb 2005 13:21:48 -0600-( From: brandon@dalsemi.com (John Brandon)$ Subject: Re: Shark Tank mentions VMS1 Message-ID: <05021813214843@dscis6-0.dalsemi.com>i  4 As long as we are talking about stupid human tricks.  A Though this has nothing to do with a command or keyboard entry...h    J One night as I was leaving the data center for dinner I informed the night8 cleaning person how to lockup the room before they left.N The door utilized a three digit push button lock - which you could lock in theL open position - which the cleaning person did so she could get in/out of the room.k  N Well I came back from dinner and found all the power off in the computer room.  / Guess what button the cleaning person pushed...,     John "REBOOT" Brandonl VMS Systems Administrator * firstname.lastname.spam.me.not@dalsemi.com   ------------------------------  % Date: Fri, 18 Feb 2005 14:53:54 -0800 % From: DeanW <dean.woodward@gmail.com> $ Subject: Re: Shark Tank mentions VMS6 Message-ID: <3f119ada0502181453c783722@mail.gmail.com>  M On Fri, 18 Feb 2005 13:21:48 -0600, John Brandon <brandon@dalsemi.com> wrote:t6 > As long as we are talking about stupid human tricks. > C > Though this has nothing to do with a command or keyboard entry...u >rL > One night as I was leaving the data center for dinner I informed the night: > cleaning person how to lockup the room before they left.P > The door utilized a three digit push button lock - which you could lock in theN > open position - which the cleaning person did so she could get in/out of the > room.1 >4P > Well I came back from dinner and found all the power off in the computer room. >c1 > Guess what button the cleaning person pushed...0  F Stupid Human Tricks, Part the second:  I used to work for a fair sizedE (650+ retail branch) bank. The data center was required to do monthlyCC tests of backup generators and fire alarm systems. The alarms had aiD bypass switch just inside the door to the platform. Procedure calledC for a guard to hold that button in while the test was conducted, sorF the fire department wouldn't get called out. They sent a new guy back,D he said he was holding the button in... Yup. chopped the power feed,7 initiated a Halon dump, and the fire department rolled.s  C Stupid Human Tricks, Chapter Three:  One of my customers (a largishCD correctional facility) also does monthly generator testing. One day,? they toggled the "test" switch, the generator fired up... and a.? radiator hose burst, spraying the mains relays with coolant andSF shorting them. Result: All the non-critical equipment stayed ON, whileF the mission-critical generator-protected stuff powered DOWN. Took themD six hours to get another relay box onsite and wired in before things! like doors and such worked again.<   ------------------------------  % Date: Fri, 18 Feb 2005 18:53:24 -0500 # From: "John Smith" <a@nonymous.com>n& Subject: The results of advertising..., Message-ID: <qpydnRbBMbBq4ovfRVn-uA@igs.net>  6 ....and the implications for VMS ......think about it.   http://www.bobparsons.com/  I From Bob's blog....Bob owns GoDaddy.com....one of the largest domain namem registrars on the planet.t   Thursday, February 17. 2005 > Early Super Bowl Ad results. Go Daddy settles with Fox Sports.J Since we had our first Super Bowl ad air, and our second airing of that adJ censored, I've been asked time and again to disclose what the results have been for GoDaddy.com.1  / Here's how to see our Super Bowl and other ads.3K By the way, if you haven't seen our Super Bowl ad or our new ads, and wouldtK like to, please click on the following link, or copy and paste it into your3 browser:  5 http://www.godaddy.com/gdshop/superbowl05/landing.aspR  , Our Super Bowl ad exceeded all expectations.J What I do know for certain is that the results from the ad exceeded all ofL my wildest expectations. You see, I didn't expect much more from the ad thanG GoDaddy.com benefiting from increased brand recognition, and maybe someeL additional business. All along we only intended the ad to be a launch of our brand into mainstream media.  A After the Super Bowl our ad was watched online millions of times.ID One of the first things to consider is the amount of post Super BowlL interest the ad (and our special 2 minute extended version) received. A goodH gauge for this is the number of views the commercial had on our Web siteD (and on other Web sites that we know about) as well as the number of
 downloads.  K A few of the Web sites on which our ads were available for view or downloadRL collapsed during the day after the Super Bowl. During the three months priorF to the Super Bowl we worked hard on our web site, and invested severalG million dollars more in our system infrastructure, just to make sure we F could handle anything that would come our way, our systems handled the traffic just fine.  L From our Web site alone, the two commercials together were viewed a total ofJ 2.6 million times. In addition to our Web site, these two commercials wereK also available on many other Web sites such as iFilm.com, aol.com, msn.com, L foxsports.com and many others. We estimate that the commercial was viewed atL least another million or so times on these sites (and perhaps much more thanL this). So our message was viewed at least 3.6 million additional times after the Super Bowl.S  ? The GoDaddy.com ad actually won the FoxSports.com on-line poll.CH Foxsports.com and aol.com both conducted a live voting poll on their WebJ sites. I am very pleased to report that as of this writing the GoDaddy.comJ Super Bowl ad leads the voting on the FoxSports.com Web site garnering 21%L of over 300,000 votes cast. In the aol.com poll, the GoDaddy.com ad finished  4th with 7% of the overall vote.  # Visits to our Web site skyrocketed..J Visits to our Web site also increased tremendously. During the 7 days thatK began with Super Bowl Sunday, visits to our Web sites increased by 367% (toNJ get this number we used a baseline period where our sites received what weK consider an average number of visitors). Business also increased (using theN same base period) by over 70%.  D Transfers to GoDaddy.com from other registrars are up significantly.H We also saw a large surge of customers from other domain name registrarsH transferring their domain names to GoDaddy.com. Transfers to GoDaddy.comL from other domain name registrars, during Super Bowl week, were up 36%. ManyI of the customers transferring domain names to us, told us that until they-I watched the commercial and checked out GoDaddy.com, that they had no ideamJ they could pay only $8.95 a year for their domain name and even get thingsK like domain forwarding (with or without masking) for free. Others said they J simply could not believe that our beginning Web site hosting plan providesH 500MB of storage with 25GB of bandwidth, and many other extras and costsH only $3.95 a month with no setup fee or commitment. (Note: Sorry for theC shameless self promotion, but this is what people were telling us).o  K There has been no discernable backlash from people upset by our commercial. H I am also very happy to report that in spite of all the controversy thatF resulted from our Super Bowl ad that there has been no backlash in ourE business. I've always been very proud that domain transfers away fromkK GoDaddy.com tend to be very low. This trend not only continued during SuperxH Bowl week but was at an all time low with domain names being transferredE away from us down 5%, from our base period.. This is in line with the0F on-line survey we asked viewers of our commercial to complete. In thatI survey more than 96% of males and more than 89% of females said they wereiK not offended by our commercial. Our low transfer away numbers indicate thatoK even the "offended" viewers (assuming they were in fact doing business witht@ us) decided not to move their domain names to another registrar.  ( Overall business continues to be strong.I Business itself continues to be strong and has yet to abate from its postcK Super Bowl peaks. I think that this is the case for two reasons: First, oureJ Super Bowl ad worked well. It brought a large amount of traffic to our webK site, and many people got to see just how good the deals are that we offer.-H Second, as I mentioned in an earlier article, we also kicked off our newH "There's a name for people like you" advertising campaign on television, radio and print.  I There's no doubt that our decision to advertise on the Super Bowl had theSJ effect we were looking for. It introduced GoDaddy.com to those individuals> who use the Internet but who don't spend a lot of time online.  5 I would rather Fox not have censored the second spot.KF Many people believe that it was the decision made by Fox to censor theH second airing that made the ad work so well. I personally am not so sureL that this is the case. I happen to know at many venues where when our ad wasJ shown, people actually stood up and cheered. I think this happened becauseK the commercial appealed to them in a slap stick sort of way and also struckyC a chord with them given some of the ridiculous and unevenly appliedoL censorship we are now seeing. Many people tell me that they're simply fed upK with it. So I believe that if the ad was in fact shown a second time (as itaK was planned to be shown) we would have had as much if not more traffic come.I to our Web site and would have been just as well know. Of course, we willtB never really know if Fox's decision actually helped us or hurt us.  . A few self-proclaimed experts were dead wrong.K I do know is that our ad worked because it was edgy and hit home with folkseK on both sides of the censorship issue. Since the ad aired there have been a=L number of self proclaimed pundits step up and say that our ad was completelyJ worthless. Two of these "experts" that I became particularly aware of wereI Jerry Della Femina, who appeared on the MSNBC show "Countdown;" and MosesTH Anshell CEO Jos Anshell, who joked in print and on a local Phoenix radioL station: "I would say that [Go Daddy's] commercial was a bust." Knowing thatL both of these two fellows make their living in the advertising business, andL that their public evaluation of our Super Bowl ad was dead wrong, gives me a9 rather satisfying smile. Better luck next time gentlemen!   - Other experts knew a winner when they saw it.-K In contrast to Jerry Della Femina and Jos Anshell, I look back with respect L to Donny Deutch, host of CNBC's "The Big Idea," Matt Lauer who co-hosts NBC'K s "The Today Show" and Dylan Ratigan, host of CNBC's BullsEye, who all cameuJ out early and said that our Super Bowl ad had potential. Donny Deutch even9 predicted that it would be the best ad on the Super Bowl.G  . GoDaddy.com has reached a settlement with Fox.A Finally, I announced when I appeared on The O'Reilly Factor, thatnJ GoDaddy.com and Fox Sports had reached an agreement in principle to settleH our differences resulting from Fox deciding not to air our ad the second time.    I continue to be a happy guy.sL Today we finalized and signed the agreement. Like most contracts that settleI disputes, this one has a clause that requires all parties to be silent onoH its specific terms. The one thing I can tell you is that I am very happy with the settlement.  D Share of voice numbers reaching out to Feb 12 are not yet available.L We are still waiting to learn what the "experts" are going to say concerningJ what our post Super Bowl "SOV" ("SOV" is an industry acronym for "Share ofL Voice" or mentions in subsequently published media such as television, radioK and print) numbers actually have been. I believe that when this informationaK becomes available, our censored commercial will prove to be one of the most'K widely talked about commercials ever. But then again, I'm looking at thingshK from the Foxhole (pun intended) where all the action is taking place; so my K slant at this moment, regarding "SOV" for our ad, might be somewhat off. I'cJ ll have these numbers early next week, and when I get them, I'll post them here on the Blog.s     --- OpenVMS - The classics never go out of style.M   ------------------------------  % Date: Fri, 18 Feb 2005 17:56:43 -0700h+ From: "Barry Treahy, Jr." <Treahy@MMaz.com>e* Subject: Re: The results of advertising...' Message-ID: <42168ECB.2070703@MMaz.com>.   John Smith wrote:u  6 >...and the implications for VMS ......think about it. >i >http://www.bobparsons.com/s >eK >>From Bob's blog....Bob owns GoDaddy.com....one of the largest domain name  >registrars on the planet. >  l > D No argument that his ad attracted a lot of attention and served the I purpose he desired (ie. increased interest in their company), but I have iF to believe that HP could be successful at promoting VMS with a little F more 'taste' but I also have to agree that at this point, any type of ( VMS advertising would be an improvement.   Barrys   -- w  < Barry Treahy, Jr                     E-mail: Treahy@MMaz.com< Midwest Microwave, Inc.                  Phone: 480/314-1320< Vice President & CIO                      FAX:  480/661-7028  I                        ... but it's a DRY HEAT!                          N   ------------------------------  % Date: Fri, 18 Feb 2005 21:13:06 -0500t- From: JF Mezei <jfmezei.spamnot@teksavvy.com>i* Subject: Re: The results of advertising...B Message-ID: <1108778463.b82de75f7abe103cfe7a7dadc29fe762@teranews>   John Smith wrote: 7 > http://www.godaddy.com/gdshop/superbowl05/landing.asp   ? For those who have compliant browsers but whcih GoDaddy doesn'tpC recognize and redirects to an "update your browser" page instead oft? giving them the content, you can few one of the commercials at:m  k > http://a1848.g.akamai.net/7/1848/13927/v001/godaddysof1.download.akamai.com/13755/1FULLCOVERAGE_SMALL.swfi    H So there you go GODADDY. Design your web site to attract viewers instead of reject viewers.  D I was able to get to the page on my good old MOSAIC on a vaxstation,+ view source and extract the flash file URL.     E If companies want business, why do they put restrictions on their webd1 sites to actively turn away potential customers ?t  G It is like REUTERS telling me I don't have a standard compliant browserdB when their web page fails miserably in the validator.w3.org test !  G Whatever happend to "just feed the content and let the client's browsero" figure out how to display it" ????  F In fairness, the ad is superb.  Definitely puts the name "godaddy" outE there. I had never heard of them before. (It is a shame they web sitep% permanently turned me off from them).    ------------------------------  % Date: Fri, 18 Feb 2005 21:22:17 -0500e- From: JF Mezei <jfmezei.spamnot@teksavvy.com>a* Subject: Re: The results of advertising...B Message-ID: <1108779021.9a55e7be9680b7c5734388d85616402e@teranews>   "Barry Treahy, Jr." wrote:E > No argument that his ad attracted a lot of attention and served therJ > purpose he desired (ie. increased interest in their company), but I haveG > to believe that HP could be successful at promoting VMS with a little G > more 'taste' but I also have to agree that at this point, any type ofE* > VMS advertising would be an improvement.   "Digital has it now"  E Back in the very early 1980s (or was it late 1970s), DEC's booth at aNF Montreal computer show was huge and placed very prominently. You could? see "Digital" balloons in many places downtown. DEC's booth was @ extremely popular. They had very "healthy" females in thin white- t-shirts handing out the Digital balloons :-)d   ------------------------------  # Date: Fri, 18 Feb 2005 22:07:32 GMT & From: John Reagan <john.reagan@hp.com>) Subject: Re: VMS 8.2 Itanium Distributionv/ Message-ID: <EGtRd.342$9L.189@news.cpqcorp.net>o   JF Mezei wrote:n > Nigel Barker wrote:  >  e > R >>It's on DVD for the very good reason that it doesn't fit on a CD. In fact AFAICR  >>it wouldn't even fit on 2 CDs. >  > C > I can understand executables bloating from VAX to Alpha since alleF > pointers/adresses double in size from 32 to 64 bits, and if you haveE > instructions containing 2 adddresses, that makes things even worse.d  D Not all pointers.  Most pointers on OpenVMS Alpha are still 32-bits  unless otherwise requested.1   > I > But from Alpha to IA64, since both are 64 bits, how come there would beP? > so much difference in size between 600meg and over 1200 meg ?  > B > Is the bloat due to truly bigger .exe files ? Or are there extra2 > components bundled into the OS installtion CDs ?  H In general images run 2-3 times larger.  A combination of the ELF/DWARF I formats not being as tightly compact and a larger number of instructions  H (many of which are 'nop's to fill bundles).  In general, you can ZIP an 7 .EXE file more than you can the source that created it.e  B I also think there are other components as well, but I'm not sure.   -- o John Reagans/ HP Pascal/{A|I}MACRO for OpenVMS Project Leaders Hewlett-Packard Company8   ------------------------------   Date: 18 Feb 2005 23:47:29 GMT From: healyzh@aracnet.com.) Subject: Re: VMS 8.2 Itanium Distributione, Message-ID: <cv5uqh01u9k@enews3.newsguy.com>  ' John Reagan <john.reagan@hp.com> wrote:aJ > In general images run 2-3 times larger.  A combination of the ELF/DWARF K > formats not being as tightly compact and a larger number of instructions  J > (many of which are 'nop's to fill bundles).  In general, you can ZIP an 9 > .EXE file more than you can the source that created it.w  D > I also think there are other components as well, but I'm not sure.  G How does this effect the Memory footprint?  Does this mean you need 2-3r4 times as much RAM on Itanium as you did on an Alpha?   		Zane   ------------------------------  % Date: Fri, 18 Feb 2005 21:04:27 -0600t2 From: David J Dachtera <djesys.nospam@comcast.net>) Subject: Re: VMS 8.2 Itanium Distribution * Message-ID: <4216ACBB.56A896E@comcast.net>   Nigel Barker wrote:d > H > On Thu, 17 Feb 2005 03:47:34 GMT, rdeininger@mindspringdot.com (Robert > Deininger) wrote:n > A > >(Yes, the VMS installation kit for Itanium is on DVD, not CD.)) > R > It's on DVD for the very good reason that it doesn't fit on a CD. In fact AFAICR  > it wouldn't even fit on 2 CDs.  F The laptop they gave me at work has a DVD drive. If I ever get hold of- an I64 kit, I'll take a good long look at it.   G Anyone know if the hobbyist folks have recovered and have done any workg on a hobbyist kit for I64?   -- o David J Dachtera dba DJE Systemsh http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/d  " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Fri, 18 Feb 2005 14:06:41 -05000# From: "John Smith" <a@nonymous.com>h2 Subject: [OT]: SHA-1 broken by Chinese researchers, Message-ID: <6JudncJRC5xeoYvfRVn-sQ@igs.net>  7 http://www.pcworld.com/news/article/0,aid,119726,00.aspg  2 Another Form of Encryption Goes Down for the Count  E Standard that is key to virtually all secure online communications iso broken.   
 Andrew Brandt- PC World   Thursday, February 17, 2005:  C News that a nine-year-old encryption method--one that underlies thecI protection of virtually all secure online communications--appears to haveaJ been cracked by a team of three Chinese researchers has spurred encryption3 experts around the world to issue a call to action.h  I The standard, known as SHA-1, "is used in pretty much every cryptographicsL protocol out there," says encryption expert Bruce Schneier. "[SHA-1 is] usedC in SSH, in SSL, in S/MIME, in PGP. It's used in IPSec. VPNs use it.e Everybody uses it."s  J The scope of the problem is enormous. Virtually all application and serverL software that incorporates SHA-1 into its functions--including Web browsers,K e-mail clients, instant messaging programs, secure shell clients, and file-tC and disk-encryption software--will need to be replaced or upgraded.a  I "We all sort of knew this could happen, but we didn't expect it this bad,t@ this soon," says Schneier, who also blogs about security topics.     It's Academic, So FarhJ "This is a critical break in SHA that is just at the edge of feasibility,"G Schneier says. But even though SHA-1 has been broken by academics, that D doesn't mean the government or criminals will be able to spy on your% encrypted communications immediately.u  ? For regular computer users, the breaking of SHA-1 has no suddenlK repercussions. Secure online communications have not been thrown wide open. K A tougher standard that hasn't been broken, called SHA-256, already exists.eJ Encryption experts are urging software companies to integrate SHA-256 into& applications that currently use SHA-1.  K Coincidentally, the news about SHA-1 has come out during one of the largest G conferences about computer security and encryption, the annual RSA Data @ Security Conference, which runs through Friday in San Francisco.    # Not a 'Run for the Exits' SituationeJ "We've all been discussing what we're going to do for some time," says JonL D. Callas, chief technology officer for PGP, a company that makes encryptionH products for individual and business computer users, as well as high-endG mail encryption gateways for enterprises. "The next release of PGP willmJ incorporate SHA-256 into the software," Callas says. "PGP 9 will likely go into beta in a few weeks."  G "At PGP, we've been working on this for a long time, but we're a little@@ quicker about this kind of stuff than most people," Callas adds.  L "This is not a 'Run for the exits, the place is on fire' kind of situation,"J Callas says. "It's 'The fire alarm is on, this is not a drill, please move to the exits.'"o     Hashing Takes a BeatingtG Schneier posted a brief item about SHA-1 on his blog Tuesday, crediting H three Shandong University researchers--Xiaoyun Wang, Yiqun Lisa Yin, and  Hongbo Yu--with the achievement.  L "They are respected cryptographers, their work is phenomenally good. This isG not a fly-by-night group, and there's no reason not to believe this [isb real]," he says.  H He describes SHA-1, invented by the National Security Agency in 1995, asI "the most common cryptographic primitive" on the Internet. (Cryptographic D primitive is an academic term describing a mathematical formula that9 cryptographers can use to scramble and unscramble codes.)g  F In the arcane language of encryption, SHA-1 is known as a one-way hashH function. Cryptographers use these tools to calculate a hash value for aH secret message. Hash values help guarantee that a secret message has notA been tampered with in transit, and they can't be used by spies toa reconstruct the message.  K "We know less about hashing than anything else in crypto--and we thought welL knew more," Callas says. "It will probably take us another two to five yearsH until we really understand hashing algorithms, and in the meantime there/ will be more dramatic things that will happen."u     Immense Computing Power UsedA Breaking encryption takes immense amounts of computing power. The J researchers who cracked SHA-1 didn't have banks of supercomputers at theirF disposal, so instead they used a distributed computing program--CallasJ describes it as "basically something like SETI@Home"-- to harness the idleJ computing power of thousands of PCs around the world to complete the task.  E "The best attack anyone has ever done [on current encryption] was thetI distributed attack on MD5-RC64, which took 300,000 computers--and it tooknH them five years," Callas says. "[Breaking SHA-1] is 16 times harder than? that; it'd take those same 300,000 computers roughly 74 years."o  H But faster home computers, and the power of distributed computing (whichJ shares portions of a monumental task among many thousands of users), seemsK to have shortened the time scale. "Cryptographic attacks always get better,pI sometimes by a factor of two or four, but they never get worse," Schneier  says.w  G In an essay he wrote for last August's Computerworld magazine, SchneierdI hinted that researchers at the time were perhaps close to breaking SHA-1.eF The essay urged cryptographers to start work on the next generation ofL one-way hash functions, before the current generation became so broken as to be unusable.       --  - OpenVMS - The classics never go out of style.    ------------------------------  % Date: Fri, 18 Feb 2005 19:21:17 -0800r# From: "Tom Linden" <tom@kednos.com> 6 Subject: Re: [OT]: SHA-1 broken by Chinese researchers( Message-ID: <opsmfcxrb0zgicya@hyrrokkin>  F On Fri, 18 Feb 2005 14:06:41 -0500, John Smith <a@nonymous.com> wrote:  8 > ttp://www.pcworld.com/news/article/0,aid,119726,00.asp4 > Another Form of Encryption Goes Down for the CountG > Standard that is key to virtually all secure online communications ish	 > broken.y > Andrew Brandt 
 > PC World > Thursday, February 17, 2005aE > News that a nine-year-old encryption method--one that underlies theaK > protection of virtually all secure online communications--appears to havelC > been cracked by a team of three Chinese researchers has spurred  S > encryption5 > experts around the world to issue a call to action.t  ? I think it is time to go to elliptic functions, ala Weierstrasss   -- cC Using Opera's revolutionary e-mail client: http://www.opera.com/m2/r   ------------------------------   End of INFO-VAX 2005.099 ************************