1 INFO-VAX	Tue, 08 Jun 2004	Volume 2004 : Issue 317       Contents:0 Re: ANN: HG FILESERV updates: Zip, DBS utilities ANN: MMK for IA64 0 Re: ANN: UnZip V5.51 available, now for IA64 too Re: AST level I/O  Re: AST level I/O  Re: AST level I/O  Re: AST level I/O  Re: AST level I/O  Re: DEC PWS RAM  Re: DEC PWS RAM ! Re: DECnet-Plus DECdts Management ! Re: DECnet-Plus DECdts Management  Exit and Exception Handlers  Re: Exit and Exception Handlers  Re: Exit and Exception Handlers  Re: Invaild SMTP in My Name  Re: Invaild SMTP in My Name 	 Re: Merci 	 Re: Merci + Re: Merrill Lynch to HP: Time for a breakup 5 Re: More on Processors from May issues of "PROCESSOR" 5 Re: More on Processors from May issues of "PROCESSOR"  Re: Not adding up? Re: Not adding up? Re: Not adding up? Re: Not adding up? Re: Not adding up? RE: Not adding up? Re: Not adding up? Re: Not adding up?& Re: OpenVMS Primer for System Managers Start Ingres with VAX time back # Re: Start Ingres with VAX time back # Re: Start Ingres with VAX time back # Re: Start Ingres with VAX time back  Re: Storing system files in CMS  Re: Sun, SAP and MicrosoftJ Re: The Loon UpChucketh (Was:  Re: Merrill Lynch to HP: Time for a breakupJ Re: The Loon UpChucketh (Was:  Re: Merrill Lynch to HP: Time for a breakupP Re: The Loon UpChucketh (Was:  Re: Merrill Lynch to HP: Time for a breakup a breP Re: The Loon UpChucketh (Was:  Re: Merrill Lynch to HP: Time for a breakup a breP Re: The Loon UpChucketh (Was:  Re: Merrill Lynch to HP: Time for a breakup a bre Re: Xwindows: SAVE-AS question# Re: [OpenVMS V7.3-2] Mount question # Re: [OpenVMS V7.3-2] Mount question   F ----------------------------------------------------------------------  % Date: Tue, 08 Jun 2004 01:22:29 -0500 - From: Hunter Goatley <goathunter@goatley.com> 9 Subject: Re: ANN: HG FILESERV updates: Zip, DBS utilities * Message-ID: <40C55B25.6050501@goatley.com>   Hunter Goatley wrote:   = > The following packages have been updated in my VMS freeware 
 > archive: >  >    - Zip  (Updated)  > ? >      Zip V2.3 is Info-ZIP's portable zip compression utility. ? >      This is not a new release, but binaries for OpenVMS IA64 = >      have been added to the kit (in addition to the VAX and , >      Alpha binaries previously available). >   >    - DBS-BUILD_LIBRARIES (New)  >    - DBS-ETHERWATCH  (Updated) >    - DBS-SCANLOCKS  (New)  >    - DBS-PATCH  (Updated)  >    - DBS-SYSRTL  (Updated) > C >      Thanks to Dave Sneddon, the packages above have been updated @ >      and now include binaries for OpenVMS IA64, in addition to >      VAX and Alpha.  > B I've also just added DBS-NULLSYMBIONT (a print queue symbiont that< does nothing) and an update to DBS-TAPEUTILS.  Thanks, Dave!  G > The packages can be downloaded via anonymous FTP from FTP.PROCESS.COM D > in [.VMS-FREEWARE.FILESERV], or you can visit the Process Software > OpenVMS Resource Center: > ! > http://www.process.com/openvms/  > B > The packages will show up on the mirror sites within the next 24 > hours. >    Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/ ; <goathunter@GOATLEY.COM>     http://www.goatley.com/hunter/    ------------------------------  % Date: Tue, 08 Jun 2004 01:38:49 -0500 - From: Hunter Goatley <goathunter@goatley.com>  Subject: ANN: MMK for IA64* Message-ID: <40C55EF9.4030309@goatley.com>  7 I never got around to posting this.  MMK (MadGoat MaKe, = which is compatible with MMS but has additional features) now < runs on OpenVMS IA64, in addition to OpenVMS VAX and OpenVMS Alpha.   http://www.process.com/openvms/   8 http://vms.process.com/ftp/vms-freeware/fileserv/mmk.zip3 ftp://ftp.process.com/vms-freeware/fileserv/mmk.zip   ( It can also be found on ftp.madgoat.com.   Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/ ; <goathunter@GOATLEY.COM>     http://www.goatley.com/hunter/    ------------------------------  % Date: Tue, 08 Jun 2004 01:21:00 -0500 - From: Hunter Goatley <goathunter@goatley.com> 9 Subject: Re: ANN: UnZip V5.51 available, now for IA64 too : Message-ID: <yOcxc.17108$Iu6.13327@bignews5.bellsouth.net>   JF Mezei wrote:  > Hunter Goatley wrote:  > : >>A week or so ago, Info-ZIP formally released UnZip V5.51< >>to correct a couple of directory-traversal vulnerabilities7 >>in the code.  I've updated my repackaging of UnZip to  >>provide this new release.  >  > K > Does VMS come with some unzip utility now ?  I have some unzip program in = > SYS$SYSTEM. (and I normally never deposit shareware there).  > I > If there is some unzip utility that comes with VMS, what is the correct N > procedure to properly disable the VMS one and enable the shareware one ? (isM > it worth removing the verb and defining a symbol in sylogin ? Or should one A > just replace the executable with the better shareware version ?   C HP distributes a version with GNV.  It's not based on V5.51, AFAIK.   9 Also, the Info-ZIP utilities are freeware, not shareware.    Hunter   ------------------------------  # Date: Tue, 08 Jun 2004 06:48:14 GMT 6 From: "Craig A. Berry" <craigberry@mac.com.spamfooler> Subject: Re: AST level I/O@ Message-ID: <4fbc241def1fb5ac3931c8b52d14113f@news.teranews.com>  3 In article <PO9V$ERLxnVf@eisner.encompasserve.org>, /  Kilgallen@SpamCop.net (Larry Kilgallen) wrote:   L > In article <a8df5820334ebf42467216d96d0b61b1@news.teranews.com>, JF Mezei ( > <jfmezei.spamnot@teksavvy.com> writes:? > > I have seen, while debugging, a C language printf in an AST J > > interrupting a printf in mainline (inserting text in the middle of the > > mainline printf).   > There are four reentrancy models for the CRTL, and unless you B explicitly choose AST reentrancy when doing C standard I/O at AST E level, then you deserve what you get, and it does sound like you got   it, JF.  See  H http://h71000.www7.hp.com/doc/732FINAL/5763/5763pro_007.html#index_x_212  0 > Priorities for CRTL seem to be Unix emulation,- > so one should not expect VMS level quality.   > It's not a question of quality, Larry, it's a question of the @ reentrancy model in use and/or the buffering model in use.  For G reentrancy see above.  As far as buffering goes, if you do a $QIO to a  F device, then when the $QIO completes (synchronously or via AST), then E it's a fair assumption the device driver has done its all to get the  D I/O as far as it can go (if the device has its own buffers that the 6 driver doesn't know about, that's a separate issue).    F But if you do a printf(), all you've done is tell a virtual construct E known as a filehandle, "Gosh, here's a buffer of stuff, and whenever  H it's convenient to you (and I do totally trust your judgement on this), D please send this stuff one step closer to its ultimate destination."  C If you don't trust the judgement of printf(), then you should call  > fflush(), which forces printf() to let go of the stuff in its A possession and pass it along to the next construct, which may be  E virtual or physical, depending on the device.  If the next construct  H happens to be RMS (or any other file system or pseudo-device layer with H its own buffering scheme), you may also want to call fsync(), which, in G the words of POSIX, causes the stuff in question to be "transferred to  F the storage device associated with the file," or, in the words of the A CRTL manual, "flushes data all the way to the disk, while fflush  8 flushes data only as far as the underlying RMS buffers."   ------------------------------  # Date: Tue, 08 Jun 2004 09:09:46 GMT - From: JF Mezei <jfmezei.spamnot@teksavvy.com>  Subject: Re: AST level I/O@ Message-ID: <d030064312723b0dcb69e40ad41c2074@news.teranews.com>   "Craig A. Berry" wrote: G > But if you do a printf(), all you've done is tell a virtual construct F > known as a filehandle, "Gosh, here's a buffer of stuff, and wheneverI > it's convenient to you (and I do totally trust your judgement on this), F > please send this stuff one step closer to its ultimate destination."  I But deep down, doesn't printf result in a $QIO ?  Are there other ways of 1 performing an IO operation without using $QIO ???   I So the question is really whether a single printf will result in multiple  $QIOs being called or not.  L (and in my case, it was simply debugging statements, not program design that had printfs inside an AST).    ------------------------------   Date: 8 Jun 2004 07:10:33 -0600 - From: Kilgallen@SpamCop.net (Larry Kilgallen)  Subject: Re: AST level I/O3 Message-ID: <NyDZm2ywpdFC@eisner.encompasserve.org>   y In article <4fbc241def1fb5ac3931c8b52d14113f@news.teranews.com>, "Craig A. Berry" <craigberry@mac.com.spamfooler> writes: 5 > In article <PO9V$ERLxnVf@eisner.encompasserve.org>, 1 >  Kilgallen@SpamCop.net (Larry Kilgallen) wrote:   1 >> Priorities for CRTL seem to be Unix emulation, . >> so one should not expect VMS level quality. > @ > It's not a question of quality, Larry, it's a question of the B > reentrancy model in use and/or the buffering model in use.  For   C Sorry, output is output.  Buffering is a technique that can be used A behind the scenes for performance purposes in certain situations. C If I can detect that it is going on from an ordinary program, it is = insufficiently behind the scenes, otherwise known as "wrong".   H > But if you do a printf(), all you've done is tell a virtual construct G > known as a filehandle, "Gosh, here's a buffer of stuff, and whenever  J > it's convenient to you (and I do totally trust your judgement on this), F > please send this stuff one step closer to its ultimate destination." > E > If you don't trust the judgement of printf(), then you should call  @ > fflush(), which forces printf() to let go of the stuff in its C > possession and pass it along to the next construct, which may be  G > virtual or physical, depending on the device.  If the next construct  J > happens to be RMS (or any other file system or pseudo-device layer with J > its own buffering scheme), you may also want to call fsync(), which, in I > the words of POSIX, causes the stuff in question to be "transferred to  H > the storage device associated with the file," or, in the words of the C > CRTL manual, "flushes data all the way to the disk, while fflush  : > flushes data only as far as the underlying RMS buffers."  ( ...and some people think TECO is arcane.   ------------------------------  % Date: Tue, 08 Jun 2004 13:57:02 +0000 - From: David B Sneddon <dbsneddon@bigpond.com>  Subject: Re: AST level I/O* Message-ID: <40C5C5AE.2070607@bigpond.com>  % Larry Kilgallen was overheard to say:  > * > ...and some people think TECO is arcane. >    I don't :-)    Regards, Dave.  --  I David B Sneddon (dbs)    VMS Systems Programmer     dbsneddon@bigpond.com I Sneddo's quick guide ...          http://www.users.bigpond.com/dbsneddon/ I DBS freeware at ...   http://www.users.bigpond.com/dbsneddon/software.htm I "Life is what happens to you while you're busy making other plans" Lennon    ------------------------------   Date: 8 Jun 2004 09:58:38 -0700 " From: cstranslations@msn.com (Joe) Subject: Re: AST level I/O= Message-ID: <d56d1c2d.0406080858.36599ecc@posting.google.com>   ] hoff@hp.nospam (Hoff Hoffman) wrote in message news:<qH0xc.3076$r%6.1697@news.cpqcorp.net>... d > In article <d56d1c2d.0406061125.13954a15@posting.google.com>, cstranslations@msn.com (Joe) writes: > E >   I usually use a hardware queue or a mailbox or such to pass along I >   such tasks to a server thread or to a loop in the mainline, as having E >   an AST doing comparatively slow tasks such as terminal I/O is not . >   something I prefer to implement in an AST. > F >   The AST drops work packets onto a queue processed by the mainline,C >   and the mainline chugs aware until there is no more work in the D >   queue.  The mainline then hibernates.  (I tend to also implementC >   a scheduled wakeup call, in addition to the $wake issued by the C >   AST.  This means the code can recover should a $wake get lost.) C >   When the code awakens, it starts pulling entries off the queue, F >   of course, and (usually) releasing packets back onto a free queue. > I >   I also prefer to avoid blocking the AST activities via the big-hammer J >   $setast operation, of course.  (There are a few cases where I will useL >   this as an expedience, but I tend to prefer to use queues or safe-update >   sequences where I can.)    >   ? Six of one half a dozen of the other and depends on what you're  doing...  D More or less it's a series of 80 or so batch jobs that generate dataB which is then collectively fed into a process that "finalizes" theD data from all the batch jobs. Run in series it takes about 2 days. IB figure that with 4 CPUs in the box and a few lines of code wrappedE around $SNDJBC/SJC$_SYNCHRONIZE_JOB keeping 4 or maybe 5 running at a B time I can cut 2 days down to about 6 hours tops so basically theB sole purpose in life for the program in question it to hang around< waiting (for the next job to finish) to submit the next one.  D I suppose I could have the completion AST plug something into a workC queue and have the main loop pull it out but it seems like overkill : since all I'm really doing to begin with is "hanging out."  D (my concern was with what happens when a less experienced programmerD gets hold of it, adds a bunch of print statements to the main loop E then ends up with I/O failures referencing FAB/RABs in a stack dump).    ------------------------------  % Date: Tue, 08 Jun 2004 08:22:50 +0200 * From: Paul Sture <nospam@sture.homeip.net> Subject: Re: DEC PWS RAM* Message-ID: <2il49rFnjbd3U1@uni-berlin.de>   healyzh@aracnet.com wrote:   > K > Thankfully on Saturday I decided to reseat everything and give the system K > one more try.  That didn't solve the problem, but I think I finally found H > the culprit.  It looks like my problem is one of the sets of RAM I wasN > trying to use.  The RAM was PC133 instead of PC100 (IIRC, RAM for the PWS isK > supposed to be PC66), and I now have the system running with 1GB of PC100 N > RAM.  It is working just fine, and doesn't crash when I try to run java or a= > transfer a large amount of data to the system via ethernet.  > J > For some reason the PC133 RAM was preventing me from using the other twoJ > banks (and from seeing RAM in Bank 1).  All three Banks are working just > fine now that it is gone.  >   G This reminds me of a conversation we had at work a few years ago, when  I RAM prices dropped and everyone was looking at adding more to their home   PCs.  H The generally accepted opinion was that it was OK to buy RAM rated at a D higher speed than the PC, but you should not mix RAM with different  speed ratings.   ------------------------------   Date: 8 Jun 2004 05:45:30 -0700 2 From: williamwebb@openvms-rocks.com (William Webb) Subject: Re: DEC PWS RAM= Message-ID: <bf98c417.0406080445.5b05a191@posting.google.com>   L healyzh@aracnet.com wrote in message news:<ca2rsm02pe@enews1.newsguy.com>...N > Recently I got what should have been enough RAM to upgrade my PWS 433au fromN > 832MB to 1.5GB and what resulted can serve as a cautionary tail to others... > K > While trying to get the three banks of 512MB installed, I started getting M > indications that the middle RAM bank was dead.  Further troubleshooting led J > me to believe that at least "Bank 1" if not "Bank 2" were dead, but thatI > "Bank 0" was more or less functional.  This wasn't to surprising as the L > system has had a dead PCI slot for about a year (and it was flakey for far > longer than that). > J > Finally I downgraded the system to an AlphaStation 200 4/233 and startedF > looking for a replacement system (the AS200 being way to slow to run > everything). > K > Thankfully on Saturday I decided to reseat everything and give the system K > one more try.  That didn't solve the problem, but I think I finally found H > the culprit.  It looks like my problem is one of the sets of RAM I wasN > trying to use.  The RAM was PC133 instead of PC100 (IIRC, RAM for the PWS isK > supposed to be PC66), and I now have the system running with 1GB of PC100 N > RAM.  It is working just fine, and doesn't crash when I try to run java or a= > transfer a large amount of data to the system via ethernet.  > J > For some reason the PC133 RAM was preventing me from using the other twoJ > banks (and from seeing RAM in Bank 1).  All three Banks are working just > fine now that it is gone.  > N > Of course I'm still planning to get it up to 1.5GB one of these days, I justE > got some brand new Kingston kits that should work in the system :^)  >  > 		Zane   Check your specs, my friend.    . Don't be fooled by what you can cram in there.  T That last half a gig won't do you any good unless you're running Tru64 or (Ack!) NT-  ' PW433au + OpenVMS = 1GB maximum memory.    Regards,   WWWebb   ------------------------------  $ Date: Tue, 8 Jun 2004 10:54:52 -0400< From: "Peter Weaver" <WeaverConsultingServices@sympatico.ca>* Subject: Re: DECnet-Plus DECdts Management* Message-ID: <2im2a0Fo8e6aU1@uni-berlin.de>   Michael Austin wrote:  > Peter Weaver wrote:  >> Peter Weaver wrote: >>> >>> Thanks for the e-mails I received. The document is also atH >>> http://vaxus.free.fr/doc/6497.pdf thanks to Didier. To answer one ofG >>> the questions I received off-line, a .PDF is a lot easier to search  >>> than HTML. >> >>A >> Actually, I spoke too soon, that document is the "Applications A >> Installation and Advanced Configuration" manual not the DECdts  >> Management. Still looking.  >> >  > G > Which IP stack are you using?  If you are using TCPIP (5.1 and later) = > you can use UTC$TIME_SETUP and add NET$DISABLE_DTSS to your D > SYLOGICALS.COM and then configure NTP to keep the systems in sync.> > Defining the logical will ignore DTSS and use the UTC setup. > F > Since almost everyone supports NTP, in a very large shop we found it@ > easier to use.  And... if you are at 7.3-1 and above, definingB > AUTO_DLIGHT_SAV = 1 in sysgen will automagically change the timeC > during daylight savings changes (not a dynamic parameter).  Works H > great!  Also, be warned, there is an NTP patch out for EV79 processors8 > (ES47/80/GS1280) so that it calculates time correctly. >  > Michael Austin.   : Thanks, but NTP is not an option for this particular site.   --   Peter Weaver Weaver Consulting Services Inc.  Canadian VAR for CHARON-VAX  www.weaverconsulting.ca    ------------------------------  % Date: Tue, 08 Jun 2004 15:27:11 +0000 - From: David B Sneddon <dbsneddon@bigpond.com> * Subject: Re: DECnet-Plus DECdts Management* Message-ID: <40C5DACF.4010805@bigpond.com>  " Peter Weaver was overheard to say: > < > Thanks, but NTP is not an option for this particular site. >    why not?   Regards, Dave.  --  I David B Sneddon (dbs)    VMS Systems Programmer     dbsneddon@bigpond.com I Sneddo's quick guide ...          http://www.users.bigpond.com/dbsneddon/ I DBS freeware at ...   http://www.users.bigpond.com/dbsneddon/software.htm I "Life is what happens to you while you're busy making other plans" Lennon    ------------------------------   Date: 8 Jun 2004 00:54:20 -0700 / From: stuie_norris@yahoo.com.au (Stuart Norris) $ Subject: Exit and Exception Handlers= Message-ID: <51262235.0406072354.50555f5d@posting.google.com>   
 Dear Readers,   D I am trying to understand exit and exception handlers with a FORTRANE program under Open VMS 6.2 on Alpha.  I have a program which uses TCP F (UCX) sockets and this program can be run as both detached process and interactive process.  E How do I catch stopping the program with stop/id=<PID> in the program E so I can cancel all the I/O and close any open channels?  What system C service should I reveiew?  What section of the Programming Concepts  Guide is this described in?   E To catch CRTL-C and CTRL-Y for interactive active processes can I use ? the same as above or do I need to use QIO calls to catch these.   B I have searched my-deja  but I have not located something that is% simple enough to remove my confusion.   ' Thanks for any assistance you can give?    Stuart   ------------------------------   Date: 8 Jun 2004 07:19:42 -0600 - From: Kilgallen@SpamCop.net (Larry Kilgallen) ( Subject: Re: Exit and Exception Handlers3 Message-ID: <$EHSYaZCMMVI@eisner.encompasserve.org>   o In article <51262235.0406072354.50555f5d@posting.google.com>, stuie_norris@yahoo.com.au (Stuart Norris) writes:  > Dear Readers,  > F > I am trying to understand exit and exception handlers with a FORTRANG > program under Open VMS 6.2 on Alpha.  I have a program which uses TCP H > (UCX) sockets and this program can be run as both detached process and > interactive process. > G > How do I catch stopping the program with stop/id=<PID> in the program : > so I can cancel all the I/O and close any open channels?  J Of course STOP/ID _will_ cancel all I/O and close open channels by itself.  
 > What system E > service should I reveiew?  What section of the Programming Concepts  > Guide is this described in?   F But if you insist on doing it yourself you will need to write an innerH mode exit handler, which is very tricky.  Given that you asked what partF of Programming Concepts you need to look in, I would say it is not for you.  G > To catch CRTL-C and CTRL-Y for interactive active processes can I use A > the same as above or do I need to use QIO calls to catch these.   B To just "handle" Control/C and Control/Y would likely be an error,B since a user should be able to type Control/C followed by CONTINUE to keep running in the program.   D > I have searched my-deja  but I have not located something that is' > simple enough to remove my confusion.   1 The question you have posed has no simple answer.   B You should proceed by posting your actual requirements, in detail,A considering that you seem confused about a need to close channels B (or else not explaining some intricate detail that would make your case different from others).   ------------------------------  # Date: Tue, 08 Jun 2004 12:24:08 GMT $ From: Ben Carter <bhc@augmentia.com>( Subject: Re: Exit and Exception Handlers, Message-ID: <Idixc.19770$Sw.12368@attbi_s51>   Stuart Norris wrote: > Dear Readers,  > F > I am trying to understand exit and exception handlers with a FORTRANG > program under Open VMS 6.2 on Alpha.  I have a program which uses TCP H > (UCX) sockets and this program can be run as both detached process and > interactive process. > G > How do I catch stopping the program with stop/id=<PID> in the program G > so I can cancel all the I/O and close any open channels?  What system E > service should I reveiew?  What section of the Programming Concepts  > Guide is this described in?  >   I Process deletion on VMS will already take care of this for you.  There is G no need to do anything special here.  However, just FYI, the guaranteed E way to catch this is to write a user-written system service which can & contain a kernel mode rundown handler.  0 I really doubt that you want to do this, though.  G > To catch CRTL-C and CTRL-Y for interactive active processes can I use A > the same as above or do I need to use QIO calls to catch these.  >   G Look at the terminal driver section of the I/O User's reference manual.   
 IO$M_CTRLYAST   D > I have searched my-deja  but I have not located something that is' > simple enough to remove my confusion.  > ) > Thanks for any assistance you can give?  >  > Stuart   --    
 Ben Carter	 Augmentia    http://www.augmentia.com.  866.302.3074   ------------------------------  * Date: Tue, 8 Jun 2004 09:44:20 +0000 (UTC) From: david20@alpha2.mdx.ac.uk$ Subject: Re: Invaild SMTP in My Name) Message-ID: <ca41pk$oe0$1@news.mdx.ac.uk>   k In article <J63xc.6823$MY2.6116@newssvr27.news.prodigy.com>, Product Manager <paladin@mydomain.com> writes: < >Yet another SMTP issue for the administrators in the group.J >I've been getting an increasing number of these messages. They appear to D >  be telling me that an infected mail has been returned to me even D >though I never sent one! Any ideas of how to stop this? I'm afraid / >others may stop accepting legit email from me. : >I'm running HP TCPIP Services 5.1 and NORELAY is defined. >  > G >Header from the latest example follows. I'm not sure I understand who  = >the initiator was (reddot1.securesites.net or aptec.co.ae) :  > ! > From - Mon Jun 07 11:58:56 2004 ' >X-UIDL: dsykes_07-JUN-2004_11:50:57.42  >X-Mozilla-Status: 0001  >X-Mozilla-Status2: 00000000. >Return-Path: DubaiS1/Server/Aptec@aptec.co.ae5 >Received: from reddot1.securesites.net (204.202.2.6) 6 >	 by vms2.alphase.com (V5.1-15D, OpenVMS V7.2 Alpha);& >	Mon, 7 Jun 2004 11:50:54 -0700 (PDT)2 >Received: from notes.aptec.co.ae ([213.42.74.51])J >	by reddot1.securesites.net (8.12.11/8.12.6) with ESMTP id i57IkK69035891C >	for <list1@reddot1.securesites.net>; Mon, 7 Jun 2004 18:46:22 GMT 2 >	(envelope-from DubaiS1/Server/Aptec@aptec.co.ae)" >To: list1@reddot1.securesites.net@ >Sender: DubaiS1/Server/Aptec <DubaiS1/Server/Aptec@aptec.co.ae>F >Subject: Symantec AV detected a violation in a document you authored. >X-Priority: 3 (Normal) % >Date: Mon, 7 Jun 2004 22:47:01 +0400 B >Message-ID: <OF1C422917.95F2D0B9-ON44256EAC.00672E82@aptec.co.ae>@ >From: Notes_Aptec/UAE/Aptec <Notes_Aptec/UAE/Aptec@aptec.co.ae>I >X-MIMETrack: Serialize by Router on DubaiS1/Server/Aptec(Release 5.0.11   >  |July 24, 2002) at  >  06/07/2004 10:47:08 PM  >MIME-Version: 1.0+ >Content-type: text/plain; charset=us-ascii  > * >Please contact your system administrator. >  > < >The infected component in the scanned document was deleted. >  >....  >--  >   D Unfortunately you are the innocent victim of a broken virus scanning implementation. K Tons of systems out there still try to inform the sender when they detect a O virus infected mail message. In the dim and distant past this was an acceptable O thing to do. Unfortunately most viruses now forge the from/return address hence H they are now spamming innocent people and accusing them of sending virus infected mail.N There isn't much you can do about this since the original infected mail almostM certainly never went anywhere near your systems. All you see is the complaint H generated by the misconfigured anti-virus software and this is extremelyJ difficult if not impossible to block since different systems generate such# notifications in different formats.   / About the only thing you can do is complain to     abuse@reddot1.securesites.net  or" postmaster@reddot1.securesites.net  J that they should reconfigure their Symantec AV solution not to try to send= notifications to the supposed sender when it detects viruses.   
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  # Date: Tue, 08 Jun 2004 14:07:16 GMT 1 From: Michael Austin <maustin@firstdbasource.com> $ Subject: Re: Invaild SMTP in My Name< Message-ID: <oKjxc.5136$vx7.4440@newssvr22.news.prodigy.com>   <snip> > F > Unfortunately you are the innocent victim of a broken virus scanning > implementation. M > Tons of systems out there still try to inform the sender when they detect a Q > virus infected mail message. In the dim and distant past this was an acceptable Q > thing to do. Unfortunately most viruses now forge the from/return address hence J > they are now spamming innocent people and accusing them of sending virus > infected mail.P > There isn't much you can do about this since the original infected mail almostO > certainly never went anywhere near your systems. All you see is the complaint J > generated by the misconfigured anti-virus software and this is extremelyL > difficult if not impossible to block since different systems generate such% > notifications in different formats.  > 1 > About the only thing you can do is complain to   >  > abuse@reddot1.securesites.net  > or$ > postmaster@reddot1.securesites.net > L > that they should reconfigure their Symantec AV solution not to try to send? > notifications to the supposed sender when it detects viruses.   G I know of too many people who said by chewing up bandwidth it will put  F pressure on ISP's to stop spam and viruses.. they all live in a dream I world.  At this point it is really up to Symantec and other AV solutions  H to remove this feature from their products. As it causes more harm than # any good it could have ever done...    M.   >  > David Webb > VMS and Unix team leader > CCSS > Middlesex University   ------------------------------   Date: 8 Jun 2004 00:55:15 -0700 ( From: finnjj@telefonica.net (Jerry Finn) Subject: Re: Merci= Message-ID: <31a78418.0406072355.2df38289@posting.google.com>   X Didier Morandi <no@spam.com> wrote in message news:<c9vh7e$1shv$1@biggoron.nerim.net>... > 6-JUN-1944 ~ 6-JUN-2004  >  > Merci. >  > D. > French   Yorktown, October 19, 1781  	 Thank you 
 Jerry Finn   ------------------------------   Date: 8 Jun 2004 07:55:36 -0700 . From: fabiopenvms@yahoo.com.br (Fabio Cardoso) Subject: Re: Merci= Message-ID: <f30679fb.0406080655.5dd7149d@posting.google.com>   y "Dave Weatherall" <djw-nothere@nospam.nohow> wrote in message news:<DTiotGxQ0bj6-pn2-Sd4Xn2xLDkNj@dave2_os2.home.ours>... , > On Sun, 6 Jun 2004 18:12:18 UTC, JF Mezei ' > <jfmezei.spamnot@teksavvy.com> wrote:  >  > > Rich Alderson wrote:  > > > Il n'y a point des mots... > > > - > > > Vive la France!  Vivent les Etats-Unis!  > > P > > Vive le Canada, Vive l'Angleterre. Vive tout autre pays qui a particip? mais > > qui n'est pas mention?.  > > O > > (i.e. some media outlets only mention their own country as participating in @ > > D-day, but there were a lot of nations who did participate). > > Q > > Does anyone know if Australia-New Zealand had sent troups over , or were they R > > too busy defending themselves against the Japanese to contribute to the war in > > europe ? > > Q > > And also, vive la R?sistance. All those french who risked their lives just as @ > > much as soldiers by helping defeat the invadors from within. > G > And Poles, Czechs, Greeks, Belgians, Nowegians, Luxembourg. Heck who   > have I left out?    K And about 25.000 brazilians sent to northern Italy to clean up the Germans   mess ! About 2000 died !     Regards    FC   ------------------------------  % Date: Tue, 08 Jun 2004 12:27:46 +0200 " From: labadie <labadie_g@decus.fr>4 Subject: Re: Merrill Lynch to HP: Time for a breakup2 Message-ID: <ca44b9$3lc$1@news-reader5.wanadoo.fr>   Michael Austin a crit :K > If HP wants to do the right thing for the shareholders, they should tell  I > ML to get lost.  The biggest problem with big business for the last 15  G > years or so hasn't been big business, it is idiots like this on Wall  K > Street trying to tell businesses what is best for them... Horse-hockey..  K > it is only to line their own greedy pockets. If these guys are "so good"  J > then why are they not running companies?   They seem to know how to run  >  every business in America...   I agreed, this reminds me of a french politician  (and former president !)  H having the bright idea that the 2 car makers Renault and Peugeot should ) merge ! Happily his idea was not followed   ? Do you think a merge of GM and Ford is the best idea for them ?    ------------------------------  $ Date: Tue, 8 Jun 2004 08:47:01 +0200( From: "Rudolf Wingert" <win@fom.fgan.de>> Subject: Re: More on Processors from May issues of "PROCESSOR"3 Message-ID: <001901c44d24$66df2d00$994614ac@wat153>    Hello;  
 Andrew wrote:    >>>  Well except for Opteron EE.  <<<   F AFAIK the Opteron EE needs 35W. This is much more then 21W. The killerG Centrino (1.7GHz) will be as good as a 2800+ and do have more 2nd level C cache. This is hard, but the real life. I hope that AMD will have a . killer (may be Geode with 1.7GHz and 14W) too.   Best regards R. Wingert    ------------------------------  % Date: Tue, 08 Jun 2004 09:44:05 +0100 9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com> > Subject: Re: More on Processors from May issues of "PROCESSOR"0 Message-ID: <ca3u8m$cc2$1@new-usenet.uk.sun.com>   Rudolf Wingert wrote:    > Hello; >  > Andrew wrote:  >  >  > Well except for Opteron EE.  > <<<  > H > AFAIK the Opteron EE needs 35W. This is much more then 21W. The killerI > Centrino (1.7GHz) will be as good as a 2800+ and do have more 2nd level E > cache. This is hard, but the real life. I hope that AMD will have a 0 > killer (may be Geode with 1.7GHz and 14W) too. >   D The EE needs 30 watts. That is more than 21 W but then we are hardlyC comparing like fruit. The Opteron is a SMP processor and includes a H set of additional units that you need to source as additional components for the Pentium M.  C It would be fairer to compare the Athlon-64 Mobile with the Pentium 4 M though even then the Athlon-64 is more integrated.   Regards  Andrew Harrison    ------------------------------   Date: 7 Jun 2004 23:25:55 -0700 ' From: icerq4a@spray.se (David Svensson)n Subject: Re: Not adding up?n= Message-ID: <734da31c.0406072225.4a62c5cb@posting.google.com>i  q Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca2g1r$59u$1@new-usenet.uk.sun.com>...o > David Svensson wrote: u > > Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca1eo9$gpa$1@new-usenet.uk.sun.com>...  > >  > >>perfnerd wrote:  > >> > >>i > >>>"Jack Peacock" <peacock@simconv.com> wrote in message news:<8ImdncfsSKbepiLdRVn-uw@mpowercom.net>...e > >>>e > >>>r5 > >>>>>"John Smith" <a@nonymous.com> wrote in message  > >>>>. > >>>>news:<U92dnaFADoTYrSDdRVn-jg@igs.net>... > >>>> > >>>>E > >>>>>>http://www.theregister.co.uk/2004/05/28/itanium_opteron_q104/o > >>>>>>M > >>>>>>....The latest server data from Gartner shows that only 6,281 Itanium  > >>>> > >>>>boxes  > >>>> > >>>>P > >>>>>>and 31,184 Opteron boxes were shipped in the first quarter of this year. > >>>>>fN > >>>>The numbers for Itanium don't seem to be adding up.  I thought Intel hadR > >>>>shipped 100,000+ Itanium CPUs last year?  What are the year on year shippingR > >>>>numbers?  Are all those 6281 boxes last quarter fitted with at least 4 CPUs? > >>>> Jack Peacock. > >>>u > >>>t? > >>>Since the average price for the Itanium box is $45,000 and-H > >>>100K/6281=~16, so I would think that puts the average box in the 16F > >>>CPU range.  While the Opteron sales appear to be in the 1 - 2 CPU > >>>range.e > >> > >>C > >>The cheapest way of buying a 4 way HP Itanium box is to buy the E > >>rx4640 and a 4 way system with 8 GB of RAM lists for 60K dollars.  > >>A > >>The rx5670 which was until recently the only way of getting acC > >>4 way unit from HP starts is about 12K more than the rx4640 forf > >>the same config. > >>B > >>Even the rx4640 is priced at over 2x the price of a equivalent? > >>config but faster Opteron based system even the HP DL585 noeA > >>price beater lists at ~27K under half the price of the rx4640 B > >>and its quicker for most of the things that people care about. > >>B > >>The last numbers I saw said that in fact the average number of2 > >>CPU's in Itanium systems was just less than 3. > >  > > I > > I can buy a 4-way rx4640 with 8GB for 30K on the online store and gets- > > even better deals through other channels.S > > > Not quite you can just about get a rx4640 with 4 x 1.3 GHz 3= > MB cache slower Itanium II's for 31K.......................V >  > With no disks and one PSU. > C > With redundant PSU's and 2 disks you are in fact looking at 33588 ! > but what is 3.5K among friends.n > ? > Even then you are getting a system that is roughly 33% slower ; > than the Opteron system and at least 6.5K more expensive.g > 9 > I say at least because the HP DL585 isn't exactly price-6 > competitive as far as Opteron boxes go 20K is a more6 > reasonable price and leaves the rx4640 13.5K adrift. > 7 > Now is being 10% slower and over twice the price less07 > or more attractive than being 33% slower and 40% moreM > expensive. >   > Sure, Itanium boxes are more expensive than x86 boxes, but for? database performance I have seen other people around me runninghB customer benchmarks, and a 4-way 1.3Ghz Itanium is a little fasterA than a 4-way 2.2GHz Opteron. And when you add everything you needf= around a database server there difference in box price is not 	 critical.-  B The main reason Opteron sells more and is cheaper is because it isB sold in the high-volume x86 market. With your thinking we can skipE Sparc, Itanium and Power and run the whole world on x86 servers whichuB indeed is a viable alternative but not that fun. I like to see howA different architectures can evolve and perhaps make a difference.b   ------------------------------  % Date: Tue, 08 Jun 2004 09:55:14 +0100>9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com>> Subject: Re: Not adding up?o0 Message-ID: <ca3utj$dqf$1@new-usenet.uk.sun.com>   Karsten Nyblad wrote:f  H > "Andrew Harrison" <andrew_._remove_harrison@su_n.com> wrote in message, > news:ca2g1r$59u$1@new-usenet.uk.sun.com... > 9 >>I say at least because the HP DL585 isn't exactly pricer6 >>competitive as far as Opteron boxes go 20K is a more6 >>reasonable price and leaves the rx4640 13.5K adrift. >  > K > The DL585 is a rather expensive design, with each processor placed on itsdD > own rather board.  That allows for having 8 dimm slots for RAM perN > processor, where all other designs, that I have seen, allows for 2 or 4 dimmM > slots per processor.  No wonder DL 585 is rather expensive.  If you want torM > compare DL585 with other opteron servers, then you have to compare machines N > fully equipped with main storage (RAM), because on DL585 you can get 16GB inJ > 1/2GB RAM blocks where you need to use 1 or 2GB blocks on other designs. >  >     @ That only makes the comparison worse because HP charge a premium for memory.p  < The going rate for an 4 way 32 GB Opteron server is ~28K the same config in a DL585 is 52K.   Regardse Andrew Harrisona   ------------------------------  % Date: Tue, 08 Jun 2004 09:58:42 +0100p9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com>s Subject: Re: Not adding up?t0 Message-ID: <ca3v43$e5v$1@new-usenet.uk.sun.com>   David Svensson wrote:,  s > Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca2g1r$59u$1@new-usenet.uk.sun.com>...g >  >>David Svensson wrote:u >>t >>>Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca1eo9$gpa$1@new-usenet.uk.sun.com>... >>>t >>>U >>>>perfnerd wrote:o >>>> >>>> >>>>i >>>>>"Jack Peacock" <peacock@simconv.com> wrote in message news:<8ImdncfsSKbepiLdRVn-uw@mpowercom.net>...i >>>>>n >>>>>e >>>>>i5 >>>>>>>"John Smith" <a@nonymous.com> wrote in messages >>>>>>. >>>>>>news:<U92dnaFADoTYrSDdRVn-jg@igs.net>... >>>>>> >>>>>> >>>>>>E >>>>>>>>http://www.theregister.co.uk/2004/05/28/itanium_opteron_q104/v >>>>>>>>M >>>>>>>>....The latest server data from Gartner shows that only 6,281 Itaniume >>>>>> >>>>>>boxesd >>>>>> >>>>>> >>>>>>P >>>>>>>>and 31,184 Opteron boxes were shipped in the first quarter of this year. >>>>>>>rN >>>>>>The numbers for Itanium don't seem to be adding up.  I thought Intel hadR >>>>>>shipped 100,000+ Itanium CPUs last year?  What are the year on year shippingR >>>>>>numbers?  Are all those 6281 boxes last quarter fitted with at least 4 CPUs? >>>>>>Jack Peacock >>>>>y >>>>>l? >>>>>Since the average price for the Itanium box is $45,000 andiH >>>>>100K/6281=~16, so I would think that puts the average box in the 16F >>>>>CPU range.  While the Opteron sales appear to be in the 1 - 2 CPU >>>>>range.b >>>> >>>>C >>>>The cheapest way of buying a 4 way HP Itanium box is to buy thecE >>>>rx4640 and a 4 way system with 8 GB of RAM lists for 60K dollars.X >>>>A >>>>The rx5670 which was until recently the only way of getting aeC >>>>4 way unit from HP starts is about 12K more than the rx4640 forl >>>>the same config. >>>>B >>>>Even the rx4640 is priced at over 2x the price of a equivalent? >>>>config but faster Opteron based system even the HP DL585 nouA >>>>price beater lists at ~27K under half the price of the rx4640 B >>>>and its quicker for most of the things that people care about. >>>>B >>>>The last numbers I saw said that in fact the average number of2 >>>>CPU's in Itanium systems was just less than 3. >>>  >>> H >>>I can buy a 4-way rx4640 with 8GB for 30K on the online store and get, >>>even better deals through other channels. >>> >>Not quite you can just about get a rx4640 with 4 x 1.3 GHz 3= >>MB cache slower Itanium II's for 31K.......................i >> >>With no disks and one PSU. >>C >>With redundant PSU's and 2 disks you are in fact looking at 33588 ! >>but what is 3.5K among friends.i >>? >>Even then you are getting a system that is roughly 33% slower,; >>than the Opteron system and at least 6.5K more expensive.  >>9 >>I say at least because the HP DL585 isn't exactly price 6 >>competitive as far as Opteron boxes go 20K is a more6 >>reasonable price and leaves the rx4640 13.5K adrift. >>7 >>Now is being 10% slower and over twice the price lesst7 >>or more attractive than being 33% slower and 40% more, >>expensive. >> >  > @ > Sure, Itanium boxes are more expensive than x86 boxes, but forA > database performance I have seen other people around me runningwD > customer benchmarks, and a 4-way 1.3Ghz Itanium is a little fasterC > than a 4-way 2.2GHz Opteron. And when you add everything you needq? > around a database server there difference in box price is not  > critical.f >   ; Odd we have just done exactly the same thing with a 2.2 Ghzs9 Opteron vs a rx5670 with 1.5 GHz I2's and the Opteron boxr4 is faster ~10% and well under half the price in fact9 more like 1/3 of the price its also much smaller. It doesu7 only have 7 PCI slots vs 10 on the rx5670 but no one is 9 prepared to pay 50K for 3 PCI slots trust me on this one.i   Regardsr Andrew Harrisona   ------------------------------   Date: 8 Jun 2004 07:18:08 -0700c' From: icerq4a@spray.se (David Svensson)y Subject: Re: Not adding up?l= Message-ID: <734da31c.0406080618.5c805092@posting.google.com>n  q Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca3v43$e5v$1@new-usenet.uk.sun.com>...l > David Svensson wrote:  > u > > Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca2g1r$59u$1@new-usenet.uk.sun.com>...  > >  > >>David Svensson wrote:o > >>v > >>>Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca1eo9$gpa$1@new-usenet.uk.sun.com>... > >>>, > >>>  > >>>>perfnerd wrote:s > >>>> > >>>> > >>>>k > >>>>>"Jack Peacock" <peacock@simconv.com> wrote in message news:<8ImdncfsSKbepiLdRVn-uw@mpowercom.net>...n > >>>>>s > >>>>>a > >>>>>y7 > >>>>>>>"John Smith" <a@nonymous.com> wrote in messagea > >>>>>>0 > >>>>>>news:<U92dnaFADoTYrSDdRVn-jg@igs.net>... > >>>>>> > >>>>>> > >>>>>>G > >>>>>>>>http://www.theregister.co.uk/2004/05/28/itanium_opteron_q104/ 
 > >>>>>>>>O > >>>>>>>>....The latest server data from Gartner shows that only 6,281 Itanium@ > >>>>>>
 > >>>>>>boxesg > >>>>>> > >>>>>> > >>>>>>R > >>>>>>>>and 31,184 Opteron boxes were shipped in the first quarter of this year.	 > >>>>>>>aP > >>>>>>The numbers for Itanium don't seem to be adding up.  I thought Intel hadT > >>>>>>shipped 100,000+ Itanium CPUs last year?  What are the year on year shippingT > >>>>>>numbers?  Are all those 6281 boxes last quarter fitted with at least 4 CPUs? > >>>>>>Jack Peacock > >>>>>  > >>>>>nA > >>>>>Since the average price for the Itanium box is $45,000 and J > >>>>>100K/6281=~16, so I would think that puts the average box in the 16H > >>>>>CPU range.  While the Opteron sales appear to be in the 1 - 2 CPU
 > >>>>>range.n > >>>> > >>>>E > >>>>The cheapest way of buying a 4 way HP Itanium box is to buy thecG > >>>>rx4640 and a 4 way system with 8 GB of RAM lists for 60K dollars.w > >>>>C > >>>>The rx5670 which was until recently the only way of getting alE > >>>>4 way unit from HP starts is about 12K more than the rx4640 fors > >>>>the same config. > >>>>D > >>>>Even the rx4640 is priced at over 2x the price of a equivalentA > >>>>config but faster Opteron based system even the HP DL585 no C > >>>>price beater lists at ~27K under half the price of the rx4640uD > >>>>and its quicker for most of the things that people care about. > >>>>D > >>>>The last numbers I saw said that in fact the average number of4 > >>>>CPU's in Itanium systems was just less than 3. > >>>  > >>> J > >>>I can buy a 4-way rx4640 with 8GB for 30K on the online store and get. > >>>even better deals through other channels. > >>@ > >>Not quite you can just about get a rx4640 with 4 x 1.3 GHz 3? > >>MB cache slower Itanium II's for 31K.......................  > >> > >>With no disks and one PSU. > >>E > >>With redundant PSU's and 2 disks you are in fact looking at 33588u# > >>but what is 3.5K among friends.e > >>A > >>Even then you are getting a system that is roughly 33% slowers= > >>than the Opteron system and at least 6.5K more expensive.o > >>; > >>I say at least because the HP DL585 isn't exactly pricei8 > >>competitive as far as Opteron boxes go 20K is a more8 > >>reasonable price and leaves the rx4640 13.5K adrift. > >>9 > >>Now is being 10% slower and over twice the price less$9 > >>or more attractive than being 33% slower and 40% more  > >>expensive. > >> > >  > > B > > Sure, Itanium boxes are more expensive than x86 boxes, but forC > > database performance I have seen other people around me runningnF > > customer benchmarks, and a 4-way 1.3Ghz Itanium is a little fasterE > > than a 4-way 2.2GHz Opteron. And when you add everything you needbA > > around a database server there difference in box price is noth
 > > critical.t > >  > = > Odd we have just done exactly the same thing with a 2.2 Ghzr; > Opteron vs a rx5670 with 1.5 GHz I2's and the Opteron boxi6 > is faster ~10% and well under half the price in fact; > more like 1/3 of the price its also much smaller. It does 9 > only have 7 PCI slots vs 10 on the rx5670 but no one iso; > prepared to pay 50K for 3 PCI slots trust me on this one.   ' Odd things happen from time to time. :)r   ------------------------------   Date: 8 Jun 2004 07:22:13 -0700e' From: icerq4a@spray.se (David Svensson)e Subject: Re: Not adding up?p= Message-ID: <734da31c.0406080622.6f76f824@posting.google.com>l  q Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca3utj$dqf$1@new-usenet.uk.sun.com>...h > Karsten Nyblad wrote:e > J > > "Andrew Harrison" <andrew_._remove_harrison@su_n.com> wrote in message. > > news:ca2g1r$59u$1@new-usenet.uk.sun.com... > > ; > >>I say at least because the HP DL585 isn't exactly pricet8 > >>competitive as far as Opteron boxes go 20K is a more8 > >>reasonable price and leaves the rx4640 13.5K adrift. > >  > > M > > The DL585 is a rather expensive design, with each processor placed on itsrF > > own rather board.  That allows for having 8 dimm slots for RAM perP > > processor, where all other designs, that I have seen, allows for 2 or 4 dimmO > > slots per processor.  No wonder DL 585 is rather expensive.  If you want toeO > > compare DL585 with other opteron servers, then you have to compare machines-P > > fully equipped with main storage (RAM), because on DL585 you can get 16GB inL > > 1/2GB RAM blocks where you need to use 1 or 2GB blocks on other designs. > >  > >  >  > B > That only makes the comparison worse because HP charge a premium
 > for memory.n > > > The going rate for an 4 way 32 GB Opteron server is ~28K the  > same config in a DL585 is 52K.  F Tier-1 vendors are usually more expensive than whitebox vendors. HP isE thus far the only major 4-way vendor and is apparently in position tot= set the price for good profit. HPs price may change when more.A competition join in. One major complaint I have heard from people F wanting to buy Opteron motherboards and make their own boxes or buyingA from non-Tier-1 is that the motherboards don't have enough memory 	 capacity.r  D BTW. Suns 2-way Opteron box look very expensive compared to HPs (and IBM) 2-Way Opteron boxes.e   ------------------------------  $ Date: Tue, 8 Jun 2004 07:30:52 -0700# From: "Tom Linden" <tom@kednos.com>r Subject: RE: Not adding up?n9 Message-ID: <NDEMLKKEBOIFBMJLCECIAECNDGAA.tom@kednos.com>t     -----Original Message-----0   From: David Svensson [mailto:icerq4a@spray.se]&   Sent: Tuesday, June 08, 2004 7:18 AM   To: Info-VAX@Mvb.Saic.Comf   Subject: Re: Not adding up?h    >   Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in6   message news:<ca3v43$e5v$1@new-usenet.uk.sun.com>...   > David Svensson wrote:    >nB   > > Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in6   message news:<ca2g1r$59u$1@new-usenet.uk.sun.com>...   > >o   > >>David Svensson wrote:a   > >>C   > >>>Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote ink6   message news:<ca1eo9$gpa$1@new-usenet.uk.sun.com>...   > >>>e   > >>>    > >>>>perfnerd wrote:i   > >>>>   > >>>>   > >>>>>   > >>>>>"Jack Peacock" <peacock@simconv.com> wrote in message0   news:<8ImdncfsSKbepiLdRVn-uw@mpowercom.net>...	   > >>>>>s	   > >>>>>.	   > >>>>> 9   > >>>>>>>"John Smith" <a@nonymous.com> wrote in messagea
   > >>>>>>2   > >>>>>>news:<U92dnaFADoTYrSDdRVn-jg@igs.net>...
   > >>>>>>
   > >>>>>>
   > >>>>>>I   > >>>>>>>>http://www.theregister.co.uk/2004/05/28/itanium_opteron_q104/e   > >>>>>>>>C   > >>>>>>>>....The latest server data from Gartner shows that onlym   6,281 Itaniumc
   > >>>>>>   > >>>>>>boxesa
   > >>>>>>
   > >>>>>>
   > >>>>>>>   > >>>>>>>>and 31,184 Opteron boxes were shipped in the first   quarter of this year.    > >>>>>>>s@   > >>>>>>The numbers for Itanium don't seem to be adding up.  I   thought Intel had.@   > >>>>>>shipped 100,000+ Itanium CPUs last year?  What are the   year on year shippingl@   > >>>>>>numbers?  Are all those 6281 boxes last quarter fitted   with at least 4 CPUs?i   > >>>>>>Jack Peacock	   > >>>>>o	   > >>>>>cC   > >>>>>Since the average price for the Itanium box is $45,000 andgL   > >>>>>100K/6281=~16, so I would think that puts the average box in the 16J   > >>>>>CPU range.  While the Opteron sales appear to be in the 1 - 2 CPU   > >>>>>range.a   > >>>>   > >>>>G   > >>>>The cheapest way of buying a 4 way HP Itanium box is to buy thenI   > >>>>rx4640 and a 4 way system with 8 GB of RAM lists for 60K dollars.a   > >>>>E   > >>>>The rx5670 which was until recently the only way of getting aiG   > >>>>4 way unit from HP starts is about 12K more than the rx4640 for    > >>>>the same config.   > >>>>F   > >>>>Even the rx4640 is priced at over 2x the price of a equivalentC   > >>>>config but faster Opteron based system even the HP DL585 nocE   > >>>>price beater lists at ~27K under half the price of the rx4640tF   > >>>>and its quicker for most of the things that people care about.   > >>>>F   > >>>>The last numbers I saw said that in fact the average number of6   > >>>>CPU's in Itanium systems was just less than 3.   > >>>0   > >>>mL   > >>>I can buy a 4-way rx4640 with 8GB for 30K on the online store and get0   > >>>even better deals through other channels.   > >>B   > >>Not quite you can just about get a rx4640 with 4 x 1.3 GHz 3A   > >>MB cache slower Itanium II's for 31K.......................    > >>    > >>With no disks and one PSU.   > >>G   > >>With redundant PSU's and 2 disks you are in fact looking at 33588w%   > >>but what is 3.5K among friends.r   > >>C   > >>Even then you are getting a system that is roughly 33% slowers?   > >>than the Opteron system and at least 6.5K more expensive.    > >>=   > >>I say at least because the HP DL585 isn't exactly price :   > >>competitive as far as Opteron boxes go 20K is a more:   > >>reasonable price and leaves the rx4640 13.5K adrift.   > >>;   > >>Now is being 10% slower and over twice the price lessd;   > >>or more attractive than being 33% slower and 40% moren   > >>expensive.   > >>   > >e   > >rD   > > Sure, Itanium boxes are more expensive than x86 boxes, but forE   > > database performance I have seen other people around me runninghH   > > customer benchmarks, and a 4-way 1.3Ghz Itanium is a little fasterG   > > than a 4-way 2.2GHz Opteron. And when you add everything you needaC   > > around a database server there difference in box price is not    > > critical.    > >L   >r?   > Odd we have just done exactly the same thing with a 2.2 Ghz =   > Opteron vs a rx5670 with 1.5 GHz I2's and the Opteron boxd8   > is faster ~10% and well under half the price in fact=   > more like 1/3 of the price its also much smaller. It doesa;   > only have 7 PCI slots vs 10 on the rx5670 but no one is =   > prepared to pay 50K for 3 PCI slots trust me on this one.K  )   Odd things happen from time to time. :)a  D But you can't run VMS on the opteron - yet, so I don't know why thisD thread persists, particularly in this group.  What is the connection to VMS?    --- (   Incoming mail is certified Virus Free.<   Checked by AVG anti-virus system (http://www.grisoft.com).A   Version: 6.0.701 / Virus Database: 458 - Release Date: 6/7/2004i   --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).? Version: 6.0.701 / Virus Database: 458 - Release Date: 6/7/2004i   ------------------------------  % Date: Tue, 08 Jun 2004 16:11:18 +0100 8 From: Andrew Harrison <andrew_remove_harrison@__sun.com> Subject: Re: Not adding up?s0 Message-ID: <ca4l6e$ius$2@new-usenet.uk.sun.com>   David Svensson wrote:ts > Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca3v43$e5v$1@new-usenet.uk.sun.com>...i >  >>David Svensson wrote:k >> >>t >>>Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca2g1r$59u$1@new-usenet.uk.sun.com>... >>>- >>>  >>>>David Svensson wrote:v >>>> >>>>v >>>>>Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca1eo9$gpa$1@new-usenet.uk.sun.com>... >>>>>d >>>>>  >>>>>r >>>>>>perfnerd wrote:l >>>>>> >>>>>> >>>>>> >>>>>>k >>>>>>>"Jack Peacock" <peacock@simconv.com> wrote in message news:<8ImdncfsSKbepiLdRVn-uw@mpowercom.net>...t >>>>>>>  >>>>>>>p >>>>>>>  >>>>>>> 7 >>>>>>>>>"John Smith" <a@nonymous.com> wrote in messagel >>>>>>>>0 >>>>>>>>news:<U92dnaFADoTYrSDdRVn-jg@igs.net>... >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>>G >>>>>>>>>>http://www.theregister.co.uk/2004/05/28/itanium_opteron_q104/ 
 >>>>>>>>>>O >>>>>>>>>>....The latest server data from Gartner shows that only 6,281 Itanium- >>>>>>>>
 >>>>>>>>boxes, >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>>R >>>>>>>>>>and 31,184 Opteron boxes were shipped in the first quarter of this year.	 >>>>>>>>>.P >>>>>>>>The numbers for Itanium don't seem to be adding up.  I thought Intel hadT >>>>>>>>shipped 100,000+ Itanium CPUs last year?  What are the year on year shippingT >>>>>>>>numbers?  Are all those 6281 boxes last quarter fitted with at least 4 CPUs? >>>>>>>>Jack Peacock >>>>>>>  >>>>>>>vA >>>>>>>Since the average price for the Itanium box is $45,000 and<J >>>>>>>100K/6281=~16, so I would think that puts the average box in the 16H >>>>>>>CPU range.  While the Opteron sales appear to be in the 1 - 2 CPU
 >>>>>>>range.  >>>>>> >>>>>>E >>>>>>The cheapest way of buying a 4 way HP Itanium box is to buy theoG >>>>>>rx4640 and a 4 way system with 8 GB of RAM lists for 60K dollars., >>>>>>C >>>>>>The rx5670 which was until recently the only way of getting aeE >>>>>>4 way unit from HP starts is about 12K more than the rx4640 fory >>>>>>the same config. >>>>>>D >>>>>>Even the rx4640 is priced at over 2x the price of a equivalentA >>>>>>config but faster Opteron based system even the HP DL585 nooC >>>>>>price beater lists at ~27K under half the price of the rx4640UD >>>>>>and its quicker for most of the things that people care about. >>>>>>D >>>>>>The last numbers I saw said that in fact the average number of4 >>>>>>CPU's in Itanium systems was just less than 3. >>>>>  >>>>>tJ >>>>>I can buy a 4-way rx4640 with 8GB for 30K on the online store and get. >>>>>even better deals through other channels. >>>>@ >>>>Not quite you can just about get a rx4640 with 4 x 1.3 GHz 3? >>>>MB cache slower Itanium II's for 31K.......................m >>>> >>>>With no disks and one PSU. >>>>E >>>>With redundant PSU's and 2 disks you are in fact looking at 33588r# >>>>but what is 3.5K among friends.h >>>>A >>>>Even then you are getting a system that is roughly 33% slowerf= >>>>than the Opteron system and at least 6.5K more expensive.c >>>>; >>>>I say at least because the HP DL585 isn't exactly pricen8 >>>>competitive as far as Opteron boxes go 20K is a more8 >>>>reasonable price and leaves the rx4640 13.5K adrift. >>>>9 >>>>Now is being 10% slower and over twice the price lesss9 >>>>or more attractive than being 33% slower and 40% morei >>>>expensive. >>>> >>>i >>>eA >>>Sure, Itanium boxes are more expensive than x86 boxes, but forDB >>>database performance I have seen other people around me runningE >>>customer benchmarks, and a 4-way 1.3Ghz Itanium is a little faster D >>>than a 4-way 2.2GHz Opteron. And when you add everything you need@ >>>around a database server there difference in box price is not >>>critical. >>>c >>= >>Odd we have just done exactly the same thing with a 2.2 Ghzs; >>Opteron vs a rx5670 with 1.5 GHz I2's and the Opteron box 6 >>is faster ~10% and well under half the price in fact; >>more like 1/3 of the price its also much smaller. It doesT9 >>only have 7 PCI slots vs 10 on the rx5670 but no one isp; >>prepared to pay 50K for 3 PCI slots trust me on this one.  >  > ) > Odd things happen from time to time. :)X  6 Obviously you, you just provided one example yourself.   regardsa Andrew Harrison7   ------------------------------  % Date: Tue, 08 Jun 2004 16:09:55 +0100o8 From: Andrew Harrison <andrew_remove_harrison@__sun.com> Subject: Re: Not adding up?p0 Message-ID: <ca4l3r$ius$1@new-usenet.uk.sun.com>   David Svensson wrote:Ms > Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca3utj$dqf$1@new-usenet.uk.sun.com>...  >  >>Karsten Nyblad wrote:t >> >>I >>>"Andrew Harrison" <andrew_._remove_harrison@su_n.com> wrote in message - >>>news:ca2g1r$59u$1@new-usenet.uk.sun.com...l >>>a >>> ; >>>>I say at least because the HP DL585 isn't exactly pricee8 >>>>competitive as far as Opteron boxes go 20K is a more8 >>>>reasonable price and leaves the rx4640 13.5K adrift. >>>a >>>mL >>>The DL585 is a rather expensive design, with each processor placed on itsE >>>own rather board.  That allows for having 8 dimm slots for RAM per O >>>processor, where all other designs, that I have seen, allows for 2 or 4 dimmlN >>>slots per processor.  No wonder DL 585 is rather expensive.  If you want toN >>>compare DL585 with other opteron servers, then you have to compare machinesO >>>fully equipped with main storage (RAM), because on DL585 you can get 16GB infK >>>1/2GB RAM blocks where you need to use 1 or 2GB blocks on other designs.@ >>>t >>>u >> >>B >>That only makes the comparison worse because HP charge a premium
 >>for memory.S >>> >>The going rate for an 4 way 32 GB Opteron server is ~28K the  >>same config in a DL585 is 52K. >  > H > Tier-1 vendors are usually more expensive than whitebox vendors. HP isG > thus far the only major 4-way vendor and is apparently in position toe? > set the price for good profit. HPs price may change when more C > competition join in. One major complaint I have heard from peopleoH > wanting to buy Opteron motherboards and make their own boxes or buyingC > from non-Tier-1 is that the motherboards don't have enough memory  > capacity.  >     Who said white box pricing ?????  F > BTW. Suns 2-way Opteron box look very expensive compared to HPs (and > IBM) 2-Way Opteron boxes.y  < Rubbish, the IBM and Sun boxes are about the same price, the< HP is cheaper and its cheaper for a very good reason .... it= damn well should be. It only has 1 PCI slot and it uses fixeds= ATA drives rather than removable Ultra320 drives like the Sunt and the IBM.  < The client I work for evaluated the 2 way HP Opteron box andB threw it out, they needed 2 PCI slots for a significant proportion? of their 2 way units and they didn't like the ATA drives. TheirhB comment was that it might make a good HPC cluster node. Ironically< they do have a HPC type clustered app but don't want to have? one unit for HPC and a completely different unit for commercial  apps.r   Regardst Andrew Harrisons   ------------------------------  % Date: Tue, 08 Jun 2004 08:25:27 +0300d4 From: Mike Rechtman <michael.rechtman.nospam@hp.com>/ Subject: Re: OpenVMS Primer for System Managerso& Message-ID: <40C577F7.19F6E3E8@hp.com>   Stanley F. Quayle wrote: > . > On 7 Jun 2004 at 15:26, warren sander wrote:  4 > > If you do like the primer please respond back... > A > The primer is incredibly useful.  I point prospective customers 
 > towards it.   
 <AOL mode=ON>    Me too!a Mike.r   <AOL mode=OFF> -- nE --------------------------------------------------------------------- E Usual disclaimer: All opinions are mine alone, perhaps not even that.e? Mike Rechtman                            *rechtman@tzora.co.il*iF Kibbutz Tzor'a.                          Voice (home): 972-2-9908337  B   "20% of a job takes 80% of the time, the rest takes another 80%"E ---------------------------------------------------------------------- -----BEGIN GEEK CODE BLOCK-----e Version: 3.1: GCM/CS d(-)pu s:+>:- a++ C++ U-- L-- W++ N++ K? w--- V+++$6 PS+ PE-- t 5? X- tv-- b+ DI+ D-- G e++ h--- r+++ y+++@ ------END GEEK CODE BLOCK------    ------------------------------   Date: 8 Jun 2004 06:21:05 -0700T& From: smohapatra@hss.hns.com (Sukanta)( Subject: Start Ingres with VAX time back= Message-ID: <5c75209e.0406080521.1df293d2@posting.google.com>   F I found that if the Ingres license has already been expired then it isC possible to start the Ingres with the VAX reboot by setting the VAXi1 time back to a period when the license was valid.p  F How can we check this type of problem( the validity period) so that ItD should not be possible to start Ingres by setting the VAX time back.   regds  Sukantat   ------------------------------  # Date: Tue, 08 Jun 2004 14:03:31 GMTe1 From: Michael Austin <maustin@firstdbasource.com> , Subject: Re: Start Ingres with VAX time back; Message-ID: <TGjxc.5135$Ms7.893@newssvr22.news.prodigy.com>o   Sukanta wrote:  H > I found that if the Ingres license has already been expired then it isE > possible to start the Ingres with the VAX reboot by setting the VAXe3 > time back to a period when the license was valid.a > H > How can we check this type of problem( the validity period) so that ItF > should not be possible to start Ingres by setting the VAX time back. >  > regds-	 > Sukanta-  I If the license is based on a date and time, then the only real way is to >I have an encrypted file somewhere with the last date used and if the last $H date used is > current date, then exit. But that also does not preclude D one from identifying this file as a part of the license process and I restoring from backup....  And for every way you come up with to prevent rF it, there are a dozen ways to get around it.  Including being able to I modify all of the date fields for any file given the proper tools... and   they do exist.    : $pipe lic list ingres /full |sear sys$pipe termination | - 	(read sys$pipe $tmp ; -9 	$tmp = f$edit(f$element(1,":",$tmp),"TRIM,COMPRESS") ; -n 	define/job $DATE$ &$tmp)d% $ write sys$output f$trnlnm("$DATE$")n  @ Look at f$cvt and date differential... to get a date difference.     Michael Austin.u   ------------------------------  $ Date: Tue, 8 Jun 2004 17:49:28 +0200, From: "Hans Vlems" <hvlems.dotweg@zonnet.nl>, Subject: Re: Start Ingres with VAX time back* Message-ID: <2im5lfFonr9sU1@uni-berlin.de>  5 "Sukanta" <smohapatra@hss.hns.com> schreef in berichte7 news:5c75209e.0406080521.1df293d2@posting.google.com...cH > I found that if the Ingres license has already been expired then it isE > possible to start the Ingres with the VAX reboot by setting the VAXi3 > time back to a period when the license was valid.i >bH > How can we check this type of problem( the validity period) so that ItF > should not be possible to start Ingres by setting the VAX time back. >G > regdsW	 > Sukantae  L If your customers have no use for a correct timestamp then they're hobbyists7 and no license will prevent them for using the product.eJ That said, this trick is not limited to VMS, it'll work on other operating systems as well.L If your company relies on timestamp based licenses make sure the license not onlyH has an expiration date but also a valid starting date. Even if the users don't care aboutH the correct value of a timestamp, having duplicates in the database is a sure way to  mess things up.    Hans   ------------------------------  % Date: Tue, 08 Jun 2004 17:36:42 +0200l9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> , Subject: Re: Start Ingres with VAX time back' Message-ID: <40C5DD0A.9B580DFB@aaa.com>i  0 Besides of the licensing problem, there might be/ some very large problems with fiddling with theu, systems date on a database server. Depending. on how the actual database handles incremental5 backups, you can get into some "interesting" problemso at a possible restore...  . If someone would like to run their system with2 a clock set back in time, they aren't professional enough to care about anyway.  	 Jan-Erik.j     Sukanta wrote: > H > I found that if the Ingres license has already been expired then it isE > possible to start the Ingres with the VAX reboot by setting the VAX23 > time back to a period when the license was valid.n > H > How can we check this type of problem( the validity period) so that ItF > should not be possible to start Ingres by setting the VAX time back. >  > regds 	 > Sukanta"   ------------------------------   Date: 8 Jun 2004 01:41:28 -0700b( From: pulley_p@hotmail.com (Phil Pulley)( Subject: Re: Storing system files in CMS= Message-ID: <7041fc41.0406080041.77c41e3d@posting.google.com>i  m pulley_p@hotmail.com (Phil Pulley) wrote in message news:<7041fc41.0406062359.54f662a8@posting.google.com>...>g > David J Dachtera <djesys.nospam@comcast.net> wrote in message news:<40C38785.E15FC0D3@comcast.net>...  > > David J Dachtera wrote:t > > >  > > > Phil Pulley wrote: > > > >rA > > > > I am trying to store the following system files in CMS...t > > > >n > > > > Sysuaf.dat > > > > Vms$audit_server.dat > > > > vue$profile.vue$datt > > > >y' > > > > ...but it doesn't seem to work.  > > > > L > > > > I can (apparently)create the element OK but, when I try and fetch orM > > > > reserve it CMS terminates & the file is not retrieved (I am using the L > > > > Decwindows front end to access CMS) however - I can store & retrieve > > > > text files OK.L > > > > I suspect that CMS does not handle this type of file but the manuals, > > > > do not confirm that this is the case > > > = > > > I'd be curious as to what the value of this would be...r > > C > > Am I to conclude by the lack of response that this is merely ano+ > > exercise, and holds no practical value?s > > 
 > > D.J.D. >  > Apologies for late responsen > G > My reason for attempting to store these files in CMS is as follows...t > G > We have several stand-alone Alpha workstations at different locations ) > running a Satcomms related application.fE > These all need to be configured the same way in terms of user id's,rG > audit setup and applications available during Decwin sessions - thus,nA > to achieve this,  the files listed need to be installed on each  > workstation. > D > As all the application source code etc. comprising the baseline isF > stored in a set of CMS libraries I thought it would be neat to store% > these system files in the same way.B > E > Obviously this is not essential, they could be stored outside CMS -iA > however I could not see any reason why I should NOT be able to.  >  > Hope this clarifies things.i > C > I shall now try some of the suggestions in earlier posts to try &r* > resolve the problems & post the results. >  > Phil P    % No luck resolving this unfortunately!t  E I have tried increasing quota's as suggested in various posts but theo problem remains unaffected.e  : As a previous poster managed to store/retrieve these filesC successfully it is obviously a problem with the configuration of myi( system or the version of CMS I am using.  C However.... I have no more time to spend on it and will store these, files outside of CMS.o  & Thanks for the interest & suggestions.     Phil P   ------------------------------  % Date: Tue, 08 Jun 2004 09:32:47 +0100w9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com>r# Subject: Re: Sun, SAP and Microsoftc0 Message-ID: <ca3tjg$avm$1@new-usenet.uk.sun.com>   JF Mezei wrote:o  J > Sun has been maming big inroads in canadian schools with its Star OfficeO > product. They got a whole bunch of Ontario schools recently, and they alreayd " > have quite a few Qubec schools. > O > My newphew has a SUN smart card which he inserts into a SUN "terminal" (woulddP > it be just an X terminal, or would that be a bona fide workstation) ? and usesB > star office. He is not overly impressed due to slow performance. >   K Its almost certainly a SunRay running off a Solaris server at the back-end.u  > Generally SunRay performance is reasonably snappy but they canA be adversely affected by network latency and server side loading.@  < The next release of the SunRay server software will help the: latency issue because it includes adaptive compression and improved client side caching.b  O > However, it is interesting that by getting schools, Sun may in fact be sewingsN > seeds that later on will yield business sales and hurt MS Office in the longK > term. And it also shows that if Sun is able to compete against MS Office,hP > there is no reason Digital couldn't have done the same a decade earlier with aL > combination of ALL-IN-1 and its Decvwindows applicatiosn such as DECwrite, > DECCalc etc. >   D Of course but Palmer dropped ALL-IN-1 in favour of a closer allianceA with MS that to Sun's huge amusement saw Digital services helpingeC deploy MS-Office and Exchange on Compaq servers to replace ALL-IN-1o running on VMS.c  B Now that was a sweet deal wasn't it. Almost but not quite as sweet. as the Sybase MS deal that spawned SQL-Server.   >5O > Also, it ahs been revealed that Microsoft had been in negotiations to acquireeM > SAP. Imagine the impact this would have had. (SAP would have been ported toiV > Wintel servers and probably abandonned from other servers). Thankfully, this failed.  C SAP is already ported to Wintel servers, Sun and HP make quite goodh? money out of picking up the platform business for failed SAP onu Wintel projects.  D On the other hand NetWeaver SAP's app/dev environment might not have2 ended up being Java based if MS had had their way.   Regardsk Andrew Harrisont   ------------------------------   Date: 8 Jun 2004 07:31:07 -0700p' From: icerq4a@spray.se (David Svensson):S Subject: Re: The Loon UpChucketh (Was:  Re: Merrill Lynch to HP: Time for a breakuph< Message-ID: <734da31c.0406080631.c3ee624@posting.google.com>  q Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca3ssk$9pb$1@new-usenet.uk.sun.com>...o > Rob Young wrote:V > > In article <SO2dnSDJ1PkhXlnd4p2dnA@igs.net>, "John Smith" <a@nonymous.com> writes: > > 1 > >>http://zdnet.com.com/2100-1104_2-5227760.htmlO > >>+ > >>Merrill Lynch to HP: Time for a breakuph > >>By Martin LaMonica > >>CNET News.comt > >>June 7, 2004, 8:35 AM PT4 > >>URL: http://zdnet.com.com/2100-1104-5227760.html > >> > >>N > >>Longtime Wall Street analyst Steven Milunovich is urging Hewlett-Packard'sE > >>top management to split the company into two separate businesses.> > >> > >  > > : > > http://www.theregister.co.uk/2004/06/07/loon_hp_split/ > > S > > "We think breaking up HP makes financial as well as business and urge the boardW7 > > to show courage and creativity," Milunovich closed.> > > S > > Miloonovich - known here fondly as "The Loon" - has a real knack for timing his L > > recommendations to companies at points when they will generate plenty ofN > > personal press. You might recall a certain open letter issued to Sun's CEOS > > Scott McNealy in which The Loon suggested Sun abandon its SPARC server business O > > - presumably so Sun could receive a later letter about how it had no way tot7 > > differentiate against other Itanium server vendors.r > > M > > Now, the day before HP's analyst meeting, The Loon vomits up a few bullet P > > points on how HP could split into separate companies and expects to be taken > > seriously. > >  > A > The problem with your point is Loon or not the numbers stack up A > in favour of his argument and not yours (and they always have).o > A > Printing and Imaging basically pays for the rest of the companytD > to go to work in the morning. The services business could probablyC > just about survive by itself but the two systems units would havenF > to radically restructure to survive without the toner dollar keeping > them afloat. > C > Now you could argue that the systems businesses leverage a bit ofmD > printing and imaging business but that isn't a strong argument. HPG > is the dominant market player in the printing business and that wouldcE > not change if the two systems units disappeared and just think whateC > the 2+ billion dollar subsidy per year that keeps the two systemsg > units afloat could pay for.p > B > As usual you make the mistake of attacking the messenger not theC > message. What ever you may think of Steven Milunovich his message E > is spot on just read HP's balance sheet if you want that confirmed.u  D As I read it and what I think the point was, is that if the businessD units were on their own they could put full focus on their business.C HP's enterprise and service business could do better if they didn'teE had competition on focus within HP from printers and Microsoft stuff.i! Look at the www.hp.com frontpage.5  C Now, having full focus is not always sufficient it appears, we haveeC one company who has full focus on Solaris and Sparc during the lasta? couple of years but for some strange reasons cannot make profito anyway.r   ------------------------------  # Date: Tue, 08 Jun 2004 15:41:06 GMTr+ From: "Dan Notov" <danno@large.INVALID.com>-S Subject: Re: The Loon UpChucketh (Was:  Re: Merrill Lynch to HP: Time for a breakupc- Message-ID: <m6lxc.62960$eY2.30441@attbi_s02>4  E "Andrew Harrison" <andrew_remove_harrison@__sun.com> wrote in messagee* news:ca4kmn$iq3$1@new-usenet.uk.sun.com... > David Svensson wrote:.H > > Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message, news:<ca3ssk$9pb$1@new-usenet.uk.sun.com>... > >2 > >>Rob Young wrote: > >>> > >>>In article <SO2dnSDJ1PkhXlnd4p2dnA@igs.net>, "John Smith" <a@nonymous.com> writes: > >>>s > >>>>3 > >>>>http://zdnet.com.com/2100-1104_2-5227760.htmln > >>>>- > >>>>Merrill Lynch to HP: Time for a breakup> > >>>>By Martin LaMonica > >>>>CNET News.comk > >>>>June 7, 2004, 8:35 AM PT6 > >>>>URL: http://zdnet.com.com/2100-1104-5227760.html > >>>> > >>>>> > >>>>Longtime Wall Street analyst Steven Milunovich is urging Hewlett-Packard'seG > >>>>top management to split the company into two separate businesses.> > >>>> > >>>w > >>>i; > >>>http://www.theregister.co.uk/2004/06/07/loon_hp_split/> > >>>.J > >>>"We think breaking up HP makes financial as well as business and urge	 the board>8 > >>>to show courage and creativity," Milunovich closed. > >>>nI > >>>Miloonovich - known here fondly as "The Loon" - has a real knack for>
 timing hisJ > >>>recommendations to companies at points when they will generate plenty ofK > >>>personal press. You might recall a certain open letter issued to Sun'su CEO?K > >>>Scott McNealy in which The Loon suggested Sun abandon its SPARC server> businessI > >>>- presumably so Sun could receive a later letter about how it had noi way to8 > >>>differentiate against other Itanium server vendors. > >>>aG > >>>Now, the day before HP's analyst meeting, The Loon vomits up a few  bulletK > >>>points on how HP could split into separate companies and expects to be4 taken  > >>>seriously.  > >>>c > >>C > >>The problem with your point is Loon or not the numbers stack up C > >>in favour of his argument and not yours (and they always have).t > >>C > >>Printing and Imaging basically pays for the rest of the companyoF > >>to go to work in the morning. The services business could probablyE > >>just about survive by itself but the two systems units would have H > >>to radically restructure to survive without the toner dollar keeping > >>them afloat. > >>E > >>Now you could argue that the systems businesses leverage a bit of F > >>printing and imaging business but that isn't a strong argument. HPI > >>is the dominant market player in the printing business and that would G > >>not change if the two systems units disappeared and just think whattE > >>the 2+ billion dollar subsidy per year that keeps the two systemso > >>units afloat could pay for.h > >>D > >>As usual you make the mistake of attacking the messenger not theE > >>message. What ever you may think of Steven Milunovich his messageUG > >>is spot on just read HP's balance sheet if you want that confirmed.o > >i > >eH > > As I read it and what I think the point was, is that if the businessH > > units were on their own they could put full focus on their business.G > > HP's enterprise and service business could do better if they didn'tvI > > had competition on focus within HP from printers and Microsoft stuff. % > > Look at the www.hp.com frontpage.  > >i >1< > The article is about realising value for the shareholders. >tL > > "Our strong intuition is that shareholders will benefit by HP eventuallyK > > breaking up," Milunovich said. "As much as the company would argue thato itseK > > divisions are separately run and optimized, we don't fully believe it."4 >y= > Its hard to argue against when you consider that valuationseA > have put Printing and Imaging as having a higher value than thee& > current market cap of HP as a whole. >  > G > > Now, having full focus is not always sufficient it appears, we have G > > one company who has full focus on Solaris and Sparc during the last C > > couple of years but for some strange reasons cannot make profit/ > > anyway.e >sD > Sun sells servers and server related software. It may have escapedA > your notice but the bit of HP that does the same thing actuallyoB > loses rather more money than Sun does but that loss is masked by% > Printing and Imaging profitability.0I Not true. Both ESG and PSG have been profitable operations. Can't say theN same for Sun as a whole.   ------------------------------  % Date: Tue, 08 Jun 2004 09:20:35 +0100d9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com>tY Subject: Re: The Loon UpChucketh (Was:  Re: Merrill Lynch to HP: Time for a breakup a bred0 Message-ID: <ca3ssk$9pb$1@new-usenet.uk.sun.com>   Rob Young wrote:T > In article <SO2dnSDJ1PkhXlnd4p2dnA@igs.net>, "John Smith" <a@nonymous.com> writes: > / >>http://zdnet.com.com/2100-1104_2-5227760.html2 >>) >>Merrill Lynch to HP: Time for a breakuph >>By Martin LaMonica >>CNET News.com  >>June 7, 2004, 8:35 AM PT2 >>URL: http://zdnet.com.com/2100-1104-5227760.html >> >>L >>Longtime Wall Street analyst Steven Milunovich is urging Hewlett-Packard'sC >>top management to split the company into two separate businesses.l >> >  > 8 > http://www.theregister.co.uk/2004/06/07/loon_hp_split/ > Q > "We think breaking up HP makes financial as well as business and urge the board 5 > to show courage and creativity," Milunovich closed.D > Q > Miloonovich - known here fondly as "The Loon" - has a real knack for timing his J > recommendations to companies at points when they will generate plenty ofL > personal press. You might recall a certain open letter issued to Sun's CEOQ > Scott McNealy in which The Loon suggested Sun abandon its SPARC server businesssM > - presumably so Sun could receive a later letter about how it had no way too5 > differentiate against other Itanium server vendors.h > K > Now, the day before HP's analyst meeting, The Loon vomits up a few bulleteN > points on how HP could split into separate companies and expects to be taken > seriously. >   ? The problem with your point is Loon or not the numbers stack upv? in favour of his argument and not yours (and they always have).v  ? Printing and Imaging basically pays for the rest of the company-B to go to work in the morning. The services business could probablyA just about survive by itself but the two systems units would havenD to radically restructure to survive without the toner dollar keeping them afloat.  A Now you could argue that the systems businesses leverage a bit ofuB printing and imaging business but that isn't a strong argument. HPE is the dominant market player in the printing business and that would@C not change if the two systems units disappeared and just think whatmA the 2+ billion dollar subsidy per year that keeps the two systemse units afloat could pay for.   @ As usual you make the mistake of attacking the messenger not theA message. What ever you may think of Steven Milunovich his message<C is spot on just read HP's balance sheet if you want that confirmed.g   Regardse Andrew Harrison.   ------------------------------  % Date: Tue, 08 Jun 2004 16:02:54 +0100>8 From: Andrew Harrison <andrew_remove_harrison@__sun.com>Y Subject: Re: The Loon UpChucketh (Was:  Re: Merrill Lynch to HP: Time for a breakup a bre 0 Message-ID: <ca4kmn$iq3$1@new-usenet.uk.sun.com>   David Svensson wrote: s > Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message news:<ca3ssk$9pb$1@new-usenet.uk.sun.com>...t >  >>Rob Young wrote: >>U >>>In article <SO2dnSDJ1PkhXlnd4p2dnA@igs.net>, "John Smith" <a@nonymous.com> writes:  >>>o >>> 1 >>>>http://zdnet.com.com/2100-1104_2-5227760.html  >>>>+ >>>>Merrill Lynch to HP: Time for a breakup1 >>>>By Martin LaMonica >>>>CNET News.com  >>>>June 7, 2004, 8:35 AM PT4 >>>>URL: http://zdnet.com.com/2100-1104-5227760.html >>>> >>>>N >>>>Longtime Wall Street analyst Steven Milunovich is urging Hewlett-Packard'sE >>>>top management to split the company into two separate businesses.  >>>> >>>b >>>l9 >>>http://www.theregister.co.uk/2004/06/07/loon_hp_split/k >>>oR >>>"We think breaking up HP makes financial as well as business and urge the board6 >>>to show courage and creativity," Milunovich closed. >>>uR >>>Miloonovich - known here fondly as "The Loon" - has a real knack for timing hisK >>>recommendations to companies at points when they will generate plenty ofeM >>>personal press. You might recall a certain open letter issued to Sun's CEOwR >>>Scott McNealy in which The Loon suggested Sun abandon its SPARC server businessN >>>- presumably so Sun could receive a later letter about how it had no way to6 >>>differentiate against other Itanium server vendors. >>>cL >>>Now, the day before HP's analyst meeting, The Loon vomits up a few bulletO >>>points on how HP could split into separate companies and expects to be taken 
 >>>seriously.e >>>l >>A >>The problem with your point is Loon or not the numbers stack upeA >>in favour of his argument and not yours (and they always have).> >>A >>Printing and Imaging basically pays for the rest of the company>D >>to go to work in the morning. The services business could probablyC >>just about survive by itself but the two systems units would havedF >>to radically restructure to survive without the toner dollar keeping >>them afloat. >>C >>Now you could argue that the systems businesses leverage a bit of D >>printing and imaging business but that isn't a strong argument. HPG >>is the dominant market player in the printing business and that wouldlE >>not change if the two systems units disappeared and just think what>C >>the 2+ billion dollar subsidy per year that keeps the two systems> >>units afloat could pay for.l >>B >>As usual you make the mistake of attacking the messenger not theC >>message. What ever you may think of Steven Milunovich his message E >>is spot on just read HP's balance sheet if you want that confirmed.r >  > F > As I read it and what I think the point was, is that if the businessF > units were on their own they could put full focus on their business.E > HP's enterprise and service business could do better if they didn'teG > had competition on focus within HP from printers and Microsoft stuff.o# > Look at the www.hp.com frontpage.r >   : The article is about realising value for the shareholders.  J > "Our strong intuition is that shareholders will benefit by HP eventuallyM > breaking up," Milunovich said. "As much as the company would argue that its I > divisions are separately run and optimized, we don't fully believe it."   ; Its hard to argue against when you consider that valuationsd? have put Printing and Imaging as having a higher value than theo$ current market cap of HP as a whole.    E > Now, having full focus is not always sufficient it appears, we haveeE > one company who has full focus on Solaris and Sparc during the last A > couple of years but for some strange reasons cannot make profit 	 > anyway.r  B Sun sells servers and server related software. It may have escaped? your notice but the bit of HP that does the same thing actuallym@ loses rather more money than Sun does but that loss is masked by# Printing and Imaging profitability.6   Regardsr Andrew Harrison-   ------------------------------  % Date: Tue, 08 Jun 2004 17:23:12 +0100h8 From: Andrew Harrison <andrew_remove_harrison@__sun.com>Y Subject: Re: The Loon UpChucketh (Was:  Re: Merrill Lynch to HP: Time for a breakup a bre-0 Message-ID: <ca4pd9$ke6$1@new-usenet.uk.sun.com>   Dan Notov wrote:G > "Andrew Harrison" <andrew_remove_harrison@__sun.com> wrote in messageg, > news:ca4kmn$iq3$1@new-usenet.uk.sun.com... >  >>David Svensson wrote:r >>G >>>Andrew Harrison <andrew_._remove_harrison@su_n.com> wrote in message$ > . > news:<ca3ssk$9pb$1@new-usenet.uk.sun.com>... >  >>>>Rob Young wrote: >>>> >>>>> >>>>>In article <SO2dnSDJ1PkhXlnd4p2dnA@igs.net>, "John Smith" >  > <a@nonymous.com> writes: >  >>>>>>3 >>>>>>http://zdnet.com.com/2100-1104_2-5227760.htmle >>>>>>- >>>>>>Merrill Lynch to HP: Time for a breakup  >>>>>>By Martin LaMonica >>>>>>CNET News.comm >>>>>>June 7, 2004, 8:35 AM PT6 >>>>>>URL: http://zdnet.com.com/2100-1104-5227760.html >>>>>> >>>>>>> >>>>>>Longtime Wall Street analyst Steven Milunovich is urging >  > Hewlett-Packard'st > G >>>>>>top management to split the company into two separate businesses.o >>>>>> >>>>>s >>>>>>; >>>>>http://www.theregister.co.uk/2004/06/07/loon_hp_split/> >>>>>>J >>>>>"We think breaking up HP makes financial as well as business and urge >  > the board_ > 8 >>>>>to show courage and creativity," Milunovich closed. >>>>> I >>>>>Miloonovich - known here fondly as "The Loon" - has a real knack for> >  > timing his > J >>>>>recommendations to companies at points when they will generate plenty >  > of > K >>>>>personal press. You might recall a certain open letter issued to Sun'sI >  > CEO> > K >>>>>Scott McNealy in which The Loon suggested Sun abandon its SPARC serverg > 
 > business > I >>>>>- presumably so Sun could receive a later letter about how it had no> >  > way to > 8 >>>>>differentiate against other Itanium server vendors. >>>>>xG >>>>>Now, the day before HP's analyst meeting, The Loon vomits up a fewe >  > bullet > K >>>>>points on how HP could split into separate companies and expects to be> >  > taken  >  >>>>>seriously.> >>>>>a >>>>C >>>>The problem with your point is Loon or not the numbers stack upaC >>>>in favour of his argument and not yours (and they always have).> >>>>C >>>>Printing and Imaging basically pays for the rest of the company F >>>>to go to work in the morning. The services business could probablyE >>>>just about survive by itself but the two systems units would haveeH >>>>to radically restructure to survive without the toner dollar keeping >>>>them afloat. >>>>E >>>>Now you could argue that the systems businesses leverage a bit of F >>>>printing and imaging business but that isn't a strong argument. HPI >>>>is the dominant market player in the printing business and that wouldnG >>>>not change if the two systems units disappeared and just think what4E >>>>the 2+ billion dollar subsidy per year that keeps the two systemst >>>>units afloat could pay for.  >>>>D >>>>As usual you make the mistake of attacking the messenger not theE >>>>message. What ever you may think of Steven Milunovich his message>G >>>>is spot on just read HP's balance sheet if you want that confirmed.u >>>i >>>tG >>>As I read it and what I think the point was, is that if the businesstG >>>units were on their own they could put full focus on their business.tF >>>HP's enterprise and service business could do better if they didn'tH >>>had competition on focus within HP from printers and Microsoft stuff.$ >>>Look at the www.hp.com frontpage. >>>p >>< >>The article is about realising value for the shareholders. >> >>K >>>"Our strong intuition is that shareholders will benefit by HP eventuallyrJ >>>breaking up," Milunovich said. "As much as the company would argue that >  > itss > J >>>divisions are separately run and optimized, we don't fully believe it." >>= >>Its hard to argue against when you consider that valuationsmA >>have put Printing and Imaging as having a higher value than thea& >>current market cap of HP as a whole. >> >> >>F >>>Now, having full focus is not always sufficient it appears, we haveF >>>one company who has full focus on Solaris and Sparc during the lastB >>>couple of years but for some strange reasons cannot make profit
 >>>anyway. >>D >>Sun sells servers and server related software. It may have escapedA >>your notice but the bit of HP that does the same thing actually B >>loses rather more money than Sun does but that loss is masked by% >>Printing and Imaging profitability.. > K > Not true. Both ESG and PSG have been profitable operations. Can't say thew > same for Sun as a whole. >  >   E Sorry but it is you who are wrong. HP moved the R&D costs for ESG and B PSG into a central fund that doesn't appear in each BU's P&L hence their apparent profitability.n  B HP currently spends 4 billion a year on R&D, historically Printing@ and Imaging spent ~1 billion a year on R&D the rest was spent by ESG and PSG.  @ ESG and PSG currently report a profit sans R&D but these numbers> do not remotely add up the the 3 billion they need to to cover the R&D costs for the units.  = Why do you think that the analysis want HP to split out their'= unprofitable systems divisions from their profitable printingK$ division ?? Perhaps they can add !!!   regardsx Andrew Harrison. >    ------------------------------  $ Date: Tue, 8 Jun 2004 18:00:00 +0200, From: "Hans Vlems" <hvlems.dotweg@zonnet.nl>' Subject: Re: Xwindows: SAVE-AS question>* Message-ID: <2im697Fnfo2hU1@uni-berlin.de>  < "JF Mezei" <jfmezei.spamnot@teksavvy.com> schreef in bericht: news:64c27b7d152ef3dc080e8aac382aa8f0@news.teranews.com...G > When adding a SAVE-AS... menu item in an xwindows program, should one>H > a- use a full fledged file selection menu to allow directory traversal etc,G > b- or should one just supply a simple dialog asking for a file name ?s > > > In the case of a file selection dialogue (VMS specific), any recommendationseI > on options that should be turned on/off to simplify the dialogue ? (forrA > instance, would it be right to remove the file filter section ?n >sH > In the case of a simple dialogue that asks for a file name, should the default K > directory be assumed to be sys$login (the home directory for that user) ?i >HJ > (In my case, the data to be stored comes from aome library that could be in aK > system directory, so it isn't obvious that one would want to SAVE-AS intob that directory). >m > Any comments ?   JF,o  C on VMS I'd rather type the name, including the device and directorye information.K The file filter is useful, especially in directories with many entries. Youd could restrict theI file filter for the default filetype and toggle to *.* as an alternative.  That may be easier to code.   K The default directory ought to be the value of default directory when the X  application got started.J Captive applications (or those started from a desktop icon)  would default
 to SYS$LOGIN. I These two rules are in line with non-X application behaviour under VMS, It think.  K Not sure what you mean with "system directory" and what your application iss
 intended for,nD but I'd be dead against *any* application that stores information in SYS$SYSTEM, SYS$SHARE,K SYS$* whatever. Even if the application will be used by system managers, att best it's asking for pollutingB your system directories, worse, you could end up by having a "new"
 SYSUAF.DAT...0   As usual, YMMV   Hans   ------------------------------   Date: 8 Jun 2004 06:22:11 -0700e% From: Bart.Zorn@xs4all.nl (Bart Zorn)C, Subject: Re: [OpenVMS V7.3-2] Mount question= Message-ID: <a98cd882.0406080522.10e45f0c@posting.google.com>   : Does the satelite have shadowing enabled (SHADOWING = 2) ?  	 Bart Zornd  o peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) wrote in message news:<newscache$l4jyyh$4c31$1@news.sil.at>...>J > You might already know, I have a Shadowset of 2 160GB IDE Disks (with an@ > ACARD SCSI-IDE Adapter) running on my PWS for over a year now. > I > Today I tried to boot a satellite from the PWS as bootserver and failedb* > to mount the (user disk) shadow set with > Z > %MOUNT-I-OPRQST, mount or dismount in progress on device, $DSAdevice$IN_SET lock failureE > %MOUNT-F-BATCHNOOPR, no operator available to service batch request- > K > Despite the fact, that no operator was enabled (you know, the old problemWN > where the OPC$OPA0_ENABLE logical is defined as TRUE but OPA0 doesn't enableO > on clustered workstations) does the lock error message tell anything to you ?0 >  > TIA  >  > -Peter > J > PS: As a DSAx isn't serveable and the satellite seems to can't mount the< > shadowset members itself, it is currently short of luck...   ------------------------------  # Date: Tue, 08 Jun 2004 14:53:08 GMTo& From: Jilly <jilly@clarityconnect.com>, Subject: Re: [OpenVMS V7.3-2] Mount question@ Message-ID: <0825bb40083cb6e80ff7900ac556dd98@news.teranews.com>  K Right, the satellite must have it's parameters set to shadow this disk justs= the same as these systems do.  So SHADOWING has to be set and-< SHADOW_SYS_DISK and SHADOW_SYS_UNIT as well must be correct.   Bart Zorn wrote:  < > Does the satelite have shadowing enabled (SHADOWING = 2) ? >  > Bart Zorns > C > peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) wrote in message:/ > news:<newscache$l4jyyh$4c31$1@news.sil.at>...aK >> You might already know, I have a Shadowset of 2 160GB IDE Disks (with antA >> ACARD SCSI-IDE Adapter) running on my PWS for over a year now.v >> aJ >> Today I tried to boot a satellite from the PWS as bootserver and failed+ >> to mount the (user disk) shadow set withb >> t< >> %MOUNT-I-OPRQST, mount or dismount in progress on device,L >> $DSAdevice$IN_SET lock failure %MOUNT-F-BATCHNOOPR, no operator available >> to service batch request  >> dL >> Despite the fact, that no operator was enabled (you know, the old problemH >> where the OPC$OPA0_ENABLE logical is defined as TRUE but OPA0 doesn'tE >> enable on clustered workstations) does the lock error message tellm >> anything to you ? >>   >> TIA >> o	 >> -Peter  >> rK >> PS: As a DSAx isn't serveable and the satellite seems to can't mount thee= >> shadowset members itself, it is currently short of luck...    -- gC Jilly - Working from Home in the Chemung River Valley - Waverly, NY$H       - jilly@clarityconnect.com                      - Brett Bodine fanH       - Mark.Jilson@hp.com                            - since 1975 or soH       - http://www.jilly.baka.com           - http://www.brettbodine.com   ------------------------------   End of INFO-VAX 2004.317 ************************lems.dotweg@zonnet.nl>, Subject: Re: Start Ingres with VAX time back* Message-ID: <2im5lfFonr9sU1@uni-berlin.de>  5 "Sukanta" <smohapatra@hss.hns.com> schreef in berichte7 news:5c75209e.0406080521.1df293d2@posting.google.com...cH > I found that if the Ingres license has already been expired then it isE > possible to start ЬRbTˏTR BjT~r  PSTSSTTUTݬ~ݬ#PRPRRPP  ^4PЬP`QQ\P`	PPQ%Q*P֬P    ^R
  P `UP PUUWT
  P `UP PUUSˏSVVˏSPPPPSTSPPP QTPTV@aSWTSSSPP<^UY|S0ެоV<Fnެ о V<Fެ$о$V<Fެ,о,V<FP 1ZO(     T7P
  V fWP jPWWPxTPPPSTTS1SST׮T7P
  V fWP jPWWPxTPPPSTTSVVnVVVXQQVV PVPSSQThRRPP nRc   PʏRTR5
  P `VP jPVVPxTPPPSTTRSP<BjVPPPVPPPX1wPYVYFY   1f Y4PԮ0Y1MPSSTT,8P
  V fWP jPWWPxTPPPSTT,SVV[о,WW VWVSSWTT$6
  V fWP jPWWPxTPPPSTT$SVVVV   VXQQVV PVPSSQThRRPP nRc   