1 INFO-VAX	Tue, 21 May 2002	Volume 2002 : Issue 279       Contents:' A tape drive, a cluster, and a question + Re: A tape drive, a cluster, and a question + Re: A tape drive, a cluster, and a question + Re: A tape drive, a cluster, and a question 8 Re: Comments on ITUG/DECUS joint Euro conference in Lyon8 Re: Comments on ITUG/DECUS joint Euro conference in Lyon& Re: Compaq TCPIP folks killing OpenVMS& Re: Compaq TCPIP folks killing OpenVMS& Re: Compaq TCPIP folks killing OpenVMS& Re: Compaq TCPIP folks killing OpenVMS& Re: Compaq TCPIP folks killing OpenVMS& Re: Compaq TCPIP folks killing OpenVMS& Re: Compaq TCPIP folks killing OpenVMS& Re: Compaq TCPIP folks killing OpenVMS& Re: Compaq TCPIP folks killing OpenVMS' Copying Authorized and Default privs... + Re: Copying Authorized and Default privs... $ Re: DCL symbol substitution question* Re: DECUS Lyon: Another VMS summary (long) Re: disk geometry and shadowing  Re: Diskeeper Usage  Re: Diskeeper Usage  Re: Diskeeper Usage $ Does UK ATC use VMS, if not why not?( Re: Does UK ATC use VMS, if not why not?( Re: Does UK ATC use VMS, if not why not?( Re: Does UK ATC use VMS, if not why not?( Re: Does UK ATC use VMS, if not why not?( Re: Does UK ATC use VMS, if not why not?( Re: Does UK ATC use VMS, if not why not?( Re: Does UK ATC use VMS, if not why not? Re: End of an era. Re: End of an era. Re: End of an era.- Re: Forced migration to HPHUX - Storm Warning - Re: Forced migration to HPHUX - Storm Warning - Re: Forced migration to HPHUX - Storm Warning  Re: freeVMS * Re: Geting Data off a DEC Microvax 3100-95* Re: Geting Data off a DEC Microvax 3100-95 GhostView on VMS Re: GhostView on VMS Re: HP startup screen anyone?  Re: LISP for OpenVMS Alpha ? Re: LISP for OpenVMS Alpha ?* Re: LISP for OpenVMS Alpha ? (RuleWorks ?)D Re: Most unusual resident in a system ?, was: Re: No new Alpha salesD Re: Most unusual resident in a system ?, was: Re: No new Alpha salesD Re: Most unusual resident in a system ?, was: Re: No new Alpha salesD Re: Most unusual resident in a system ?, was: Re: No new Alpha salesD Re: Most unusual resident in a system ?, was: Re: No new Alpha salesD Re: Most unusual resident in a system ?, was: Re: No new Alpha salesD Re: Most unusual resident in a system ?, was: Re: No new Alpha salesD Re: Most unusual resident in a system ?, was: Re: No new Alpha salesD Re: Most unusual resident in a system ?, was: Re: No new Alpha sales Re: No new Alpha salesC Re: OpenVMS (not) on DS20L (was: re: Sayonara DS10, new org charts) C Re: OpenVMS (not) on DS20L (was: re: Sayonara DS10, new org charts) C Re: OpenVMS (not) on DS20L (was: re: Sayonara DS10, new org charts) & OpenVMS v7.3 and Password Dict Problem* Re: OpenVMS v7.3 and Password Dict ProblemB Re: Other HP printers and USB (was: Newest version of DCPS and ourL Re: Other HP printers and USB (was: Newest version of DCPS and our  problem)L Re: Other HP printers and USB (was: Newest version of DCPS and our  problem)G Other HP printers and USB (was: Newest version of DCPS and our problem) K Re: Other HP printers and USB (was: Newest version of DCPS and our problem) K Re: Other HP printers and USB (was: Newest version of DCPS and our problem) K Re: Other HP printers and USB (was: Newest version of DCPS and our problem) P Re: Other HP printers and USB (was: Newest version of DCPS and our problem) prob& Re: Please help: Hobbes failed to boot- Re: Queue info via Web Server (Apache & WASD)  Reading unreadable tapes...  Re: Reading unreadable tapes...  Re: Reading unreadable tapes...  Re: Reading unreadable tapes... ! Re: Sayonara DS10, new org charts  Re: scsi cluster Re: Stallards smoking gun! Re: System manager available Re: System manager available* TCP/IP intermittent dropping of connection. Re: TCP/IP intermittent dropping of connection9 Re: Traditional VMS NFS names vs. Extended Filename Parse 9 Re: Traditional VMS NFS names vs. Extended Filename Parse  Unix on VaxStation 4000 90A 4 Re: vax/alpha print to hp laser printers help needed4 Re: vax/alpha print to hp laser printers help needed4 Re: vax/alpha print to hp laser printers help needed# Re: Volume shadowing and a disaster # Re: Volume shadowing and a disaster # Re: Volume shadowing and a disaster # Re: Volume shadowing and a disaster # Re: Volume shadowing and a disaster   What is the exact role of DCPS ?? Re: Which Unix for VMS users ?, was: Re: Stallards smoking gun! ? Re: Which Unix for VMS users ?, was: Re: Stallards smoking gun! ? Re: Which Unix for VMS users ?, was: Re: Stallards smoking gun! ? Re: Which Unix for VMS users ?, was: Re: Stallards smoking gun! ? Re: Which Unix for VMS users ?, was: Re: Stallards smoking gun! ? Re: Which Unix for VMS users ?, was: Re: Stallards smoking gun!  Re: Who cares about marketing! Re: Who cares about marketing!  F ----------------------------------------------------------------------    Date: 20 May 2002 21:12:20 -0700! From: russ@gordon.edu (manrubble) 0 Subject: A tape drive, a cluster, and a question= Message-ID: <53e4993e.0205202012.31cf3cbe@posting.google.com>   A I installed a TZ87 SCSI tape drive on the boot node of a two node C cluster.  I can see the drive on the boot node, however when logged ' onto the satellite node, no tape drive.    My question is...   9 What do I need to change to make the tape drive available 
 cluster-wide?    thanks,    ------------------------------  # Date: Tue, 21 May 2002 04:54:28 GMT ) From: Andrew Balaam <abalaam@yahoo.co.uk> 4 Subject: Re: A tape drive, a cluster, and a question3 Message-ID: <20020521.4542800.935019137@imagnu.geo>   I I only have experience with <=3D VMS6.2, and tape drives are not a clust=  er=20  wide resource.  5 To use the tape, you must set host to the other node.     6 >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<  I On 21/05/02, 05:12:20, russ@gordon.edu (manrubble) wrote regarding A tap=  e=20! drive, a cluster, and a question:     C > I installed a TZ87 SCSI tape drive on the boot node of a two node E > cluster.  I can see the drive on the boot node, however when logged ) > onto the satellite node, no tape drive.    > My question is...   ; > What do I need to change to make the tape drive available  > cluster-wide?   	 > thanks,    ------------------------------   Date: 21 May 2002 05:13:59 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)4 Subject: Re: A tape drive, a cluster, and a question* Message-ID: <accl2n$f3h$1@web1.cup.hp.com>  _ In article <20020521.4542800.935019137@imagnu.geo>, Andrew Balaam <abalaam@yahoo.co.uk> writes: I :I only have experience with [somebody's software ate a character or two  E :here -hoff] VMS6.2, and tape drives are not a cluster-wide resource.   H   TMSCP series tape drives have been cluster-wide devices for some time B   now (V5.5), and served SCSI tape support was introduced in V6.2.  6 :To use the tape, you must set host to the other node.  +   That works, but so does serving the tape.   7 :>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<  : J :On 21/05/02, 05:12:20, russ@gordon.edu (manrubble) wrote regarding A tape" :drive, a cluster, and a question: :  : D :> I installed a TZ87 SCSI tape drive on the boot node of a two nodeF :> cluster.  I can see the drive on the boot node, however when logged* :> onto the satellite node, no tape drive. :> :> My question is... :>< :> What do I need to change to make the tape drive available :> cluster-wide?    @   When posting, please include the OpenVMS version and platform.@   (Details on the sorts of information that can be useful are inA   the FAQ.)  Details on exactly how the TZ87 is connected to the  =   particular (unspecified) host would be interesting for this D   particular case, too.  Also, you will want to acquire and install C   the mandatory ECO kits for the particular OpenVMS release in use.     N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Tue, 21 May 2002 05:47:04 GMT ) From: Andrew Balaam <abalaam@yahoo.co.uk> 4 Subject: Re: A tape drive, a cluster, and a question3 Message-ID: <20020521.5470400.228009603@imagnu.geo>   I I stand corrected. I have a small mixed version cluster, with only one=20 I node at 6.2, another at 6.1 and the rest at 5.5-2, and I was not aware=20 I that SCSI tape drives were cluster wide in 6.2. I will have to read the =   : release notes, and get another node running 6.2 I suppose.  6 >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<  I On 21/05/02, 06:13:59, hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) wrot=  e=206 regarding Re: A tape drive, a cluster, and a question:    F > In article <20020521.4542800.935019137@imagnu.geo>, Andrew Balaam=20 <abalaam@yahoo.co.uk> writes: I > :I only have experience with [somebody's software ate a character or t=  woG > :here -hoff] VMS6.2, and tape drives are not a cluster-wide resource.   I >   TMSCP series tape drives have been cluster-wide devices for some tim=  e D >   now (V5.5), and served SCSI tape support was introduced in V6.2.   ------------------------------  % Date: Mon, 20 May 2002 15:08:12 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca> A Subject: Re: Comments on ITUG/DECUS joint Euro conference in Lyon , Message-ID: <3CE9499C.4E2F8A7D@videotron.ca>   Alan Greig wrote: E > When Compaq bought Digital, HP distributed a sales 'coacher' to key 6 > sales staff to go after VMS sites in particular. ...H > accounts (and Tru64). HP know all about VMS and the perception of only* > grudging minimal support from its owner.  1 Is the HP of 4 years ago the same as Carly's HP ?   N I get the distinct impression that HP has transformed itself into a wintel boxL maker that also makes printers, and it seems, cameras, with a "legacy" thingQ with Unix. This is very different from the serious equipment maker HP used to be.   L Are there still many high end managers/VPs at HP who were part of the "real"; HP ? Or was a new slate brought in full of wintel weenies ?    ------------------------------  # Date: Mon, 20 May 2002 21:35:37 GMT  From: lbohan@spamless..dbc.comA Subject: Re: Comments on ITUG/DECUS joint Euro conference in Lyon 8 Message-ID: <6uqieusspfa371dulsrhgu4rniunc6t270@4ax.com>  C On Mon, 20 May 2002 11:25:02 +0100, Alan Greig <a.greig@virgin.net>  wrote:  G >CA have just sold a division which markets and supports MANMAN on VMS. A >Just a few months after assuring us that it was still strategic.   = I missed hearing/reading about this;  Who did CA sell it to ?    ------------------------------  + Date: Mon, 20 May 2002 20:01:09 +0100 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> / Subject: Re: Compaq TCPIP folks killing OpenVMS ; Message-ID: <01KHYODC146096VU4K@sysdev.deutsche-boerse.com>   I > you can't really blame the ucx team as their hands have been tied since J > Palmer, and they are trying to play catchup ... that's why until they doG > you should be using TCPware or Multinet, although I would use TCPware J > permanently and it is the only IP stack for VMS based on the VMS kernel,! > and therefore runs crisper ...     Would you like fries with that?    ------------------------------    Date: 20 May 2002 11:40:38 -0700- From: jodonnell@hrblock.com (Jason O'Donnell) / Subject: Re: Compaq TCPIP folks killing OpenVMS = Message-ID: <9059bf6b.0205201040.5a9281c5@posting.google.com>   n P.Young@unsw.EDU.AU (Patrick Young) wrote in message news:<55f85d77.0205200426.48754157@posting.google.com>...B > At this point in time there are two _VERY SERIOUS_ problems with5 > TCPIP services "OpenVMS Alpha Version V5.1 - ECO 3"  >   > (1) The shared socket problem.F > (2) The socket leak problem (mentioned earlier this year as a memoryH >     leak problem, which it is - as it leaks non paged dynamic memory). >  > No public patches available. > G > I'd like to say to the TCP/IP folks - You are doing a _LOT_ of damage E > by not getting these patches out. IMHO, Not a good game plan (can I  > help ?) ECO 4 ?  > 	 > Thanks.   F I am not aware of the problems about which you wrote.  We have writtenC routines in C that share sockets between processes (single listener E process - multiple worker processes configuration).  We have had some D problems, but they seem to occur before they get to the hosts.  What are the symptoms?    ------------------------------  # Date: Mon, 20 May 2002 19:47:57 GMT 0 From: "Matt Muggeridge" <Matt.Muggeridge@hp.com>/ Subject: Re: Compaq TCPIP folks killing OpenVMS @ Message-ID: <NpcG8.157590$o66.465170@news-server.bigpond.net.au>   > No public patches available.  G So you have access to the ECOs, but you would like to access individual C patches too?  If that is the case, then you need CSC support, since K sometimes the rules and dependencies for applying individual patches become  complex.  ! Have you tried your friendly CSC?    Matt.    --= -------------------------------------------------------------  OpenVMS TCP/IP Engineering Enterprise Computing Group Hewlett-Packard Company  Gold Coast, AUSTRALIA = -------------------------------------------------------------     6 "Patrick Young" <P.Young@unsw.EDU.AU> wrote in message7 news:55f85d77.0205200426.48754157@posting.google.com... B > At this point in time there are two _VERY SERIOUS_ problems with5 > TCPIP services "OpenVMS Alpha Version V5.1 - ECO 3"  >   > (1) The shared socket problem.F > (2) The socket leak problem (mentioned earlier this year as a memoryH >     leak problem, which it is - as it leaks non paged dynamic memory). >  > No public patches available. > G > I'd like to say to the TCP/IP folks - You are doing a _LOT_ of damage E > by not getting these patches out. IMHO, Not a good game plan (can I  > help ?) ECO 4 ?  > 	 > Thanks.    ------------------------------  # Date: Mon, 20 May 2002 19:53:15 GMT ? From: "Non John Gemignani, Jr." <John.non.Gemignani@non.hp.com> / Subject: Re: Compaq TCPIP folks killing OpenVMS * Message-ID: <3CE91A67.647B74F0@non.hp.com>   Patrick Young wrote: > B > At this point in time there are two _VERY SERIOUS_ problems with5 > TCPIP services "OpenVMS Alpha Version V5.1 - ECO 3"  >   > (1) The shared socket problem.F > (2) The socket leak problem (mentioned earlier this year as a memoryH >     leak problem, which it is - as it leaks non paged dynamic memory). >  > No public patches available.  F The memory leak problem has been fixed and will be in ECO 4.  That ECOF has been through our SQA cycle and goes to TIMA this week.  It usuallyG takes up to two weeks for them to come back with an approval so that we , can post the kit for the whole world to see.  H I believe that the shared socket problem is the Apache accept problem onF the shared BG device?  This particular problem had a fix but it failedF QA in that it appeared to have a side effect.  That fix was backed outB of ECO 4.  A better fix was engineered and has just begun the testG cycle.  I am told that the point fix will be available at the beginning  of June.   > G > I'd like to say to the TCP/IP folks - You are doing a _LOT_ of damage E > by not getting these patches out. IMHO, Not a good game plan (can I  > help ?) ECO 4 ?  > 	 > Thanks.   G In the past we (TCP/IP Services for OpenVMS, nee UCX) had not performed G the degree of SQA testing that we now do prior to making fixes publicly G available.  As a result, the product suffered from very poor press (not D to mention me personally wearing egg on my face).  The SQA group hasH also been busy qualifying V5.3 SSB prior to its final submission, so all< of the personnel and equipment resources had to be scheduled accordingly.  E Below is the SQA report dated May 17 concerning ECO 4.  I removed the C listing of bug numbers (120 of them) as they are meaningless to the  public.   ) Patrick, I am sure you will be satisfied.    -John   @ >                 SQA Test Report for TCP/IP Version 5.1 - ECO 4 > . >                                 May 17, 2002 > L > Periodically, the TCP/IP Maintenance group will build an ECO kit comprisedJ > of the fixes they have ported to the product over a period of time.  TheJ > ECO 4 kit contains 120 fixes which have been ported to the product since > the release of the ECO 3 kit.a > K > The ECO 4 kit underwent a four-week SQA test cycle during which more than J > 800 individual tests were conducted to demonstrate that the kit was freeK > of regressions and, where possible, that the fixes properly corrected thee( > defects they were designed to correct. > K > The four week SQA test cycle for TCP/IP Version 5.1 - ECO 4 was conductedeL > between April 18 and May 15, 2002.  The test cycle went very smoothly withK > only three significant bugs encountered.  The first involved system hangsgI > while an FTP test script was running.  The maintenance group determinedlK > that recent changes to the tcpip$bgdriver.exe, tcpip$inetacp.exe, and thedE > tcpip$internet_services.exe images, were the cause of this problem.  > H > The changes to those images corrected a bug that was identified by theI > Apache SQA group approximately a month before the start of the SQA test,I > cycle for ECO 4.  The maintenance group determined that the best course I > of action would be to back out the fix for the Apache bug for the ECO 4sK > kit.  They also determined that the Apache bug was a low severity bug andeK > that a fix for it would not be required for the ECO 4 kit.  A new fix foro6 > the Apache bug will be implemented in the ECO 5 kit. > J > The second bug was not unique to the ECO 4 kit.  It was noticed that theK > SNMP tests under tcpip$config.com was taking an inordinate amount of timerL > to complete.  It was confirmed that this was also the case in all versionsJ > of 5.1.  This bug was identified because the SQA team had just completedL > the test process for TCP/IP version 5.3, and the SNMP test on that version5 > of the product completed in a fraction of the time.l > J > The problem was examined by a member of the development group and it was; > determined that the test was looping in a tcp subroutine.r > L > Since the test completed without encountering an error, this bug was givenM > a low severity level and will be corrected in a later release.  There isn't,5 > any need to correct this problem for the ECO 4 kit.  > K > The third and final bug caused the test systems to crash if a sequence of F > steps were executed to verify a series of UCP commands that modify aH > system's network interface card with cluster aliases.  The sequence ofG > steps included deleting the primary interface card then recreating it2H > with a cluster alias.  Doing so resulted in the crash.  If a secondaryL > interface card was deleted and then recreated, the system would not crash. > L > The series of UCP commands that were used are not likely to be repeated byL > a customer.  One would not delete one's primary network interface card andJ > then recreate it.  This series of commands were developed to demonstrateH > that the UCP "set interface" command functions properly, which it did. > L > The maintenance team determined that it is highly unlikely that a customerH > would execute such a series of commands.  Therefore, they chose not toJ > address this issue at this time.  The bug was recorded in a QAR and sentJ > to the development group for resolution.  The fix that results from this: > process will be ported to all future releases of TCP/IP. > H > The release notes for ECO 4 contains the PTR and tcpip_bug numbers forJ > all of the PTR and tcpip_bugs that have been corrected and ported to theI > product since the initial release of TCP/IP version 5.1.  Each containstJ > a description of the bug, the image modified to correct the problem, theI > ECO letter version of the image, and the ECO version of the product theb > fix was first ported to. >e   ------------------------------  # Date: Mon, 20 May 2002 19:59:36 GMTp? From: "Non John Gemignani, Jr." <John.non.Gemignani@non.hp.com>l/ Subject: Re: Compaq TCPIP folks killing OpenVMSx* Message-ID: <3CE91BE4.47D21A69@non.hp.com>   Bob Ceculski wrote:  >  >> [CHIP CHIP CHIP]s > I > you can't really blame the ucx team as their hands have been tied sincehG > Palmer, and they are trying to play catchup ... that's why until theysJ > do you should be using TCPware or Multinet, although I would use TCPwareJ > permanently and it is the only IP stack for VMS based on the VMS kernel,  > and therefore runs crisper ...   What?   ' Where do you get your information from?r  F Seriously, you should stop posting information that you obtained while waiting  on line at the supermarket.?   ------------------------------  # Date: Mon, 20 May 2002 20:44:54 GMTR0 From: Paul Anderson <paul.r.anderson@compaq.com>/ Subject: Re: Compaq TCPIP folks killing OpenVMSh; Message-ID: <200520021639316834%paul.r.anderson@compaq.com>e  B In article <3CE91BE4.47D21A69@non.hp.com>, Non John Gemignani, Jr.& <John.non.Gemignani@non.hp.com> wrote:  ) > Where do you get your information from?  > H > Seriously, you should stop posting information that you obtained while% > waiting on line at the supermarket.e  6 Does the Weekly World News have an OpenVMS column now?   Paul   --    Paul Anderson   OpenVMS Engineeringd   Hewlett-Packard Company    ------------------------------  # Date: Mon, 20 May 2002 21:20:30 GMT  From: lbohan@spamless..dbc.com/ Subject: Re: Compaq TCPIP folks killing OpenVMSr8 Message-ID: <cspieu0tg3gg5eipdin0j7scqg1k9ig22s@4ax.com>  ; On Mon, 20 May 2002 19:53:15 GMT, "Non John Gemignani, Jr."m& <John.non.Gemignani@non.hp.com> wrote:  F >Below is the SQA report dated May 17 concerning ECO 4.  I removed theD >listing of bug numbers (120 of them) as they are meaningless to the >public.  : We saw a problem, early April, seemingly related to ECO3, 9 that went away, with rolling back to TCPIP 5.1 + ECO2.   > (axp v7.3 + current eco's)  4 Tried this back && forth, several times to convince # myself that it was related to ECO3.A  5 The problem being sporadic reception (transmission?)  ( of low-bandwidth  udp multicast streams.  8 I didn't report it to the CSC, figuring I would have to * look at, and capture sniffer traces of the0 low-bandwidth streams, and IGMP traffic first.    . (joins, responses to router queries, and such)  % But I've not had the time to do so.     0 Had any another customers reported udp multicast, problems in ECO3, (IGMP troubles, my guess)  that might be fixed in ECO4 ?a   ------------------------------  % Date: Mon, 20 May 2002 17:37:17 -04002- From: JF Mezei <jfmezei.spamnot@videotron.ca>1/ Subject: Re: Compaq TCPIP folks killing OpenVMSa, Message-ID: <3CE96C83.B2DD5E30@videotron.ca>   Paul Anderson wrote:J > > Seriously, you should stop posting information that you obtained while' > > waiting on line at the supermarket.e > 8 > Does the Weekly World News have an OpenVMS column now?  L No, but what they say about virgin teenage girls giving birth to alien blackL sheep triplets  is more reliable than what we hear from HP about VMS :-) ;-) :-) :-) :-) :-) ;-) :-)>   ------------------------------  % Date: Mon, 20 May 2002 23:09:53 +0200: From: Dirk Munk <munk@home.nl>/ Subject: Re: Compaq TCPIP folks killing OpenVMSq& Message-ID: <3CE96621.1090306@home.nl>  ? It is good to hear that the memory leak problem has been fixed sI officially now.  I was wondering why it took so long because there was a oF fix more then half a year ago. I have the ECO_AI_A_V51 fixes for this H problem, and they are dated 9 october 2001. I'm also happy to hear that H the testing is more extensive now. TCPIP could do with a little quality H boost. Not that it is bad, but as I stated before, for instance the CLI I could be much more coherent and logical. Most of all, the Unix / VMS CLI lE hodgepodge should be sorted out, so that we can do all things in VMS tD style and don't have to look for obscure Unix utilities/commands to F perform vital TCPIP settings. But I'm glad that engineering is taking  the TCPIP things seriously !!!   Regards,   Dirk   Non John Gemignani, Jr. wrote:   >Patrick Young wrote:m > B >>At this point in time there are two _VERY SERIOUS_ problems with5 >>TCPIP services "OpenVMS Alpha Version V5.1 - ECO 3"e >>  >>(1) The shared socket problem.F >>(2) The socket leak problem (mentioned earlier this year as a memoryH >>    leak problem, which it is - as it leaks non paged dynamic memory). >> >>No public patches available. >> >2G >The memory leak problem has been fixed and will be in ECO 4.  That ECOeG >has been through our SQA cycle and goes to TIMA this week.  It usually H >takes up to two weeks for them to come back with an approval so that we- >can post the kit for the whole world to see.e > I >I believe that the shared socket problem is the Apache accept problem oniG >the shared BG device?  This particular problem had a fix but it failed.G >QA in that it appeared to have a side effect.  That fix was backed out C >of ECO 4.  A better fix was engineered and has just begun the test7H >cycle.  I am told that the point fix will be available at the beginning	 >of June.e > G >>I'd like to say to the TCP/IP folks - You are doing a _LOT_ of damage.E >>by not getting these patches out. IMHO, Not a good game plan (can Is >>help ?) ECO 4 ?e >>	 >>Thanks.  >> > H >In the past we (TCP/IP Services for OpenVMS, nee UCX) had not performedH >the degree of SQA testing that we now do prior to making fixes publiclyH >available.  As a result, the product suffered from very poor press (notE >to mention me personally wearing egg on my face).  The SQA group has I >also been busy qualifying V5.3 SSB prior to its final submission, so allt= >of the personnel and equipment resources had to be scheduledf
 >accordingly.M >iF >Below is the SQA report dated May 17 concerning ECO 4.  I removed theD >listing of bug numbers (120 of them) as they are meaningless to the >public. >f* >Patrick, I am sure you will be satisfied. >  >-John >t@ >>                SQA Test Report for TCP/IP Version 5.1 - ECO 4 >>. >>                                May 17, 2002 >>L >>Periodically, the TCP/IP Maintenance group will build an ECO kit comprisedJ >>of the fixes they have ported to the product over a period of time.  TheJ >>ECO 4 kit contains 120 fixes which have been ported to the product since >>the release of the ECO 3 kit.- >>K >>The ECO 4 kit underwent a four-week SQA test cycle during which more thancJ >>800 individual tests were conducted to demonstrate that the kit was freeK >>of regressions and, where possible, that the fixes properly corrected the1( >>defects they were designed to correct. >>K >>The four week SQA test cycle for TCP/IP Version 5.1 - ECO 4 was conductednL >>between April 18 and May 15, 2002.  The test cycle went very smoothly withK >>only three significant bugs encountered.  The first involved system hangs-I >>while an FTP test script was running.  The maintenance group determinedhK >>that recent changes to the tcpip$bgdriver.exe, tcpip$inetacp.exe, and the E >>tcpip$internet_services.exe images, were the cause of this problem.- >>H >>The changes to those images corrected a bug that was identified by theI >>Apache SQA group approximately a month before the start of the SQA testuI >>cycle for ECO 4.  The maintenance group determined that the best coursesI >>of action would be to back out the fix for the Apache bug for the ECO 4VK >>kit.  They also determined that the Apache bug was a low severity bug andhK >>that a fix for it would not be required for the ECO 4 kit.  A new fix forr6 >>the Apache bug will be implemented in the ECO 5 kit. >>J >>The second bug was not unique to the ECO 4 kit.  It was noticed that theK >>SNMP tests under tcpip$config.com was taking an inordinate amount of timeeL >>to complete.  It was confirmed that this was also the case in all versionsJ >>of 5.1.  This bug was identified because the SQA team had just completedL >>the test process for TCP/IP version 5.3, and the SNMP test on that version5 >>of the product completed in a fraction of the time.e >>J >>The problem was examined by a member of the development group and it was; >>determined that the test was looping in a tcp subroutine.t >>L >>Since the test completed without encountering an error, this bug was givenM >>a low severity level and will be corrected in a later release.  There isn't 5 >>any need to correct this problem for the ECO 4 kit.W >>K >>The third and final bug caused the test systems to crash if a sequence ofeF >>steps were executed to verify a series of UCP commands that modify aH >>system's network interface card with cluster aliases.  The sequence ofG >>steps included deleting the primary interface card then recreating it H >>with a cluster alias.  Doing so resulted in the crash.  If a secondaryL >>interface card was deleted and then recreated, the system would not crash. >>L >>The series of UCP commands that were used are not likely to be repeated byL >>a customer.  One would not delete one's primary network interface card andJ >>then recreate it.  This series of commands were developed to demonstrateH >>that the UCP "set interface" command functions properly, which it did. >>L >>The maintenance team determined that it is highly unlikely that a customerH >>would execute such a series of commands.  Therefore, they chose not toJ >>address this issue at this time.  The bug was recorded in a QAR and sentJ >>to the development group for resolution.  The fix that results from this: >>process will be ported to all future releases of TCP/IP. >>H >>The release notes for ECO 4 contains the PTR and tcpip_bug numbers forJ >>all of the PTR and tcpip_bugs that have been corrected and ported to theI >>product since the initial release of TCP/IP version 5.1.  Each containsnJ >>a description of the bug, the image modified to correct the problem, theI >>ECO letter version of the image, and the ECO version of the product the' >>fix was first ported to. >>   ------------------------------  % Date: Mon, 20 May 2002 13:28:36 -0600M% From: "Dave Rich" <drich@nucorar.com>p0 Subject: Copying Authorized and Default privs...5 Message-ID: <acbf2e$ofcag$1@ID-136223.news.dfncis.de>r  I Whats the simpliest, most fail proof method of copying the Authorized andC6 Default privilages from one user to another via UAF> ?   Thanks in advance>   ------------------------------    Date: 20 May 2002 14:08:58 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)p4 Subject: Re: Copying Authorized and Default privs...3 Message-ID: <p4tvHH1Ri2A5@eisner.encompasserve.org>n  ] In article <acbf2e$ofcag$1@ID-136223.news.dfncis.de>, "Dave Rich" <drich@nucorar.com> writes:aK > Whats the simpliest, most fail proof method of copying the Authorized anda8 > Default privilages from one user to another via UAF> ?   	Simplest and most fail proof:  5 		Minimize those privileges to the default privilegest   	Runner-up:h  = 		UAF> COPY user1 user2/PASSWORD=x/UIC=x/DEVICE=x/DIRECTORY=x   9 		(if this is a newly privileged user, they should have ae5 		 separate account for their privileged work anyway)l   ------------------------------    Date: 20 May 2002 14:29:15 -0700/ From: chris@applied-synergy.com (Chris Scheers)A- Subject: Re: DCL symbol substitution questionh= Message-ID: <754a27c1.0205201329.54bbe3ad@posting.google.com>a  \ carl@gerg.tamu.edu (Carl Perkins) wrote in message news:<18MAY200201324721@gerg.tamu.edu>...e > In article <0KQE8.81048$vm6.15241244@ruti.visi.com>, "Mark Jensen" <markje@vistainfo.com> writes... M > }Was able to get it to work by skipping the double quotes in the write line $ > }$ write WIZARCHIVE ''TEMPSTRING2' > I > Do you have any idea what the various single quotes in that command do?o > H > My suggestion: unless there is some specific reason that you are awareD > of for putting in quotes, either single or double, in any specific" > case, you should leave them out. > A > And if you were wondering, the net result of the quotes in yournE > command above is exactly zero - but you wast a few processor cyclese > to get it.    C Actually, that is not correct.  The net result of the quotes is not  zero.c  B The single quotes cause the contents of TEMPSTRING2 to be insertedD into the command line.  Then command line parsing is applied against the result.   @ For example, if TEMPSTRING2 contains double quotes, they will be@ parsed and removed from the resultant string which is written to WIZARCHIVE.   5 Worse, embedded double quotes will also be processed.i  ) It is likely that the desired command is:h  "     $ write WIZARCHIVE TEMPSTRING2  C But if double quotes were added to TEMPSTRING2 to make it work with B the single quotes, those quotes should NOT be added if the command above is used.   ------------------------------   Date: 21 May 2002 00:38:17 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)3 Subject: Re: DECUS Lyon: Another VMS summary (long) * Message-ID: <acc4tp$4j4$5@web1.cup.hp.com>  w In article <01KHTXOPFQJC96VU4K@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:e  F   I have not received Didier's original post, so this reply may (will)C   miss any topics not included in the response from Phillip Helbig.e  E :> As said, Rich Marcello and Mark Gorham have been amazing (are theyo :> former DECcies?)  :hH :I think Gorham comes from Compaq, i.e. was there before DEC was bought.  E   Mark Gorham worked in OpenVMS Engineering as a Product Manager for  F   a number of years -- yes, Mark arrived at HP via Compaq and DIGITAL.  G :> I also asked: "About the French FreeVMS open source project, will HPsH :> support, fight or just not care about it?" Unfortunately, I have beenG :> disturbed during Rich's short answer, so I cannot reproduce it here.  :> John Smith, please?  F   Didier and Rich were unfortunately not having related conversations.D   When I provided the necessary background and provided Rich with a D   rephrased version of the question Didier had intended to ask, RichA   indicated that he had no particular concerns and no particular fB   problems with folks working on an (unencumbered) variant of the    hp OpenVMS operating system.    K :> I asked Christian about no more MOP boot on Itanium. He said "you should E :> not care about the way your Itanium satellite will remote boot, ith :> will!"... :uJ :Does this mean that booting will require DECnet?  Currently, one can use + :MOP instead of DECnet to boot satellites. t  H   DECnet is one way to acquire MOP support for OpenVMS, but as has been F   occasionally mentioned, the two are not particularly related -- mostG   folks are used to using DECnet to register for the MOP downloads, buttF   the DECnet entries are simply used as storage for the details neededF   by the MOP server -- MOP is also available in the LANCP component ofG   OpenVMS V6.2 and later, and using LANCP means you do not need to loadn;   DECnet Phase IV or DECnet-Plus "just" to get MOP support.i   : Yes, I know that DECnet J :(Phase IV as well) is being ported to Itanium.  Still, it is possible to J :run a cluster with no DECnet and it would be overkill to install it just @ :to enable satellites to boot.  OK, perhaps there will be a new H :mechanism, but we need some way to boot VAX satellites from an Itanium  :boot server.  :-)  G   I expect LANCP will be ported.  That said, we have not yet identified,K   customers that want to cluster VAX systems with IA-64 systems -- support eH   for clustering Alpha and IA-64 has been announced.  (This is a case ofG   customer identification and qualification more than anything else -- dK   given that IA-64 will cluster with Alpha, clustering with VAX is largely t8   a matter of testing and of formal support statements.)  G   We expect to be bootstrapping IA-64 boxes using the modified version d4   of bootp and tftp that is used by the EFI console.  ?   The following text is included here, out of original order...n   D :>... But he also stated that "yes, no PAL code means no more ^P and :> console actions..." :-(    B   The Itanium Processor Family has a completely different console D   environment.  ACPI is the interface to the "halt" button and such.    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   Date: 20 May 2002 23:54:01 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)( Subject: Re: disk geometry and shadowing* Message-ID: <acc2ap$4j4$3@web1.cup.hp.com>  w In article <01KHOX5JIRH895MSV4@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes: : :Members of a shadow set have to have the same geometry.    H   The geometry restriction of shadowing was lifted a couple of years agoI   now, please upgrade and/or please apply current ECO kits for shadowing.mJ   (The geometry information itself is now often fictional, but I digress.)  ; :They can otherwise be completely different disks, right?  r  D   So long as the devices are reported to the OpenVMS host as having C   identical total block counts, yes -- and if you have a controlleryC   that can be programmed to report a particular disk capacity valuei   to the host...  2 :To determine if two disks have the same geometry I :for the purpose of shadowing, is it enough if these things are the same?  :  :   Total blocks   :n :   Sectors per track      :k :   Total cylinders        :    :   Tracks per cylinder  : E :Is there a "blocks per sector"?  If so, why isn't it listed in SHOW mB :DEVICE?  If it is fixed, then "Total blocks" is redundant, right?  )   "Total blocks" is all that matters now.   5   Ask The Wizard topic (4715), probably a few others.n    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Mon, 20 May 2002 12:20:38 -0700t" From: Brad Hughes <brad@tgsmc.com> Subject: Re: Diskeeper Usage& Message-ID: <3CE94C86.50709@tgsmc.com>   Roy Omond wrote:   [...]rH > I'm genuinely curious to know why.  Can you expand a bit on why OracleJ > should care if a file is moved (and made more contiguous) ?  I'd presumeE > that the file was not open at the time (otherwise DFO wouldn't haveo > touched it).    J ISTR (fuzzily) that Ingres bypassed RMS and instead manipulated raw blocksF directly;  if Oracle does that as well I can see how it would be upset if the blocks get moved.   Brad   ------------------------------   Date: 20 May 2002 22:51:23 GMT4 From: "Jim Strehlow" <JimStrehlowNoSpam@data911.com> Subject: Re: Diskeeper Usage0 Message-ID: <acbulb$1cm@dispatch.concentric.net>  . Diskeeper recommends that you run it 24x7; butJ I often recommend to clients to only run it on specific disks when needed.  G Before building large Oracle tablespaces, run Diskeeper to help providee4 as much contiguous available disk space as possible.> Turn off diskeeper. Build Oracle tablespace(s) or extend them.H If a tablespace is highly fragmented, then at your earliest convenience:   - shut down the Oracle instanceA  - copy/contiguous or run Diskeeper to defragment the tablespace.p  - shutdown DiskeeperE  - startup the Oracle instance.o   Jim Strehlow, Data911 Systemsr Alameda, CA, USA    ; "Jack Trachtman" <Jack.Trachtman@vmmc.org> wrote in messagee7 news:69d784c4.0205171640.1d1e3529@posting.google.com... F > We have used the Diskeeper defragmentor for years on single node VMS > systems with good results. >hF > I am now looking at using it in a Cluster, and where the application > uses Oracle 8i.i >aE > Does anyone have any experience in this usage that they can commenth
 > on?  Thanksa   ------------------------------  # Date: Tue, 21 May 2002 02:08:58 GMT 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>  Subject: Re: Diskeeper Usage' Message-ID: <3CE9AF80.84BF9B3C@fsi.net>I   Brad Hughes wrote: >  > Roy Omond wrote: >  > [...]CJ > > I'm genuinely curious to know why.  Can you expand a bit on why OracleL > > should care if a file is moved (and made more contiguous) ?  I'd presumeG > > that the file was not open at the time (otherwise DFO wouldn't haveO > > touched it). > L > ISTR (fuzzily) that Ingres bypassed RMS and instead manipulated raw blocksH > directly;  if Oracle does that as well I can see how it would be upset > if the blocks get moved.  = I'm told that Oracle cannot handle multi-spindle volume-sets.    Draw your own conclusions.   -- e David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/o   ------------------------------  # Date: Mon, 20 May 2002 19:18:40 GMTt4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk>- Subject: Does UK ATC use VMS, if not why not?a0 Message-ID: <3CE94AF2.E7E8328E@blueyonder.co.uk>  A Does anyone know what technology the new UK ATC system at Swanick. uses.t   This is scary,  > http://www.guardian.co.uk/airlines/story/0,1371,718325,00.html  $ Air travellers 'face year of chaos'      -- e tim.llewellyn@blueyonder.co.uk T  F * tim.llewellyn@cableinet.co.uk address will cease to work June 2002 *   ------------------------------  % Date: Mon, 20 May 2002 23:26:56 +0200n From: Dirk Munk <munk@home.nl>1 Subject: Re: Does UK ATC use VMS, if not why not?m& Message-ID: <3CE96A20.1040502@home.nl>   Tim Llewellyn wrote:  B >Does anyone know what technology the new UK ATC system at Swanick >uses. >  >This is scary,  >pH I'm sure it isn't Windooz, otherwise it would be raining Jumbo Jets all  over the UK. :-)   >  >i? >http://www.guardian.co.uk/airlines/story/0,1371,718325,00.htmlo >e% >Air travellers 'face year of chaos'   >  >l   ------------------------------  % Date: Mon, 20 May 2002 22:35:54 +0100g4 From: John Laird <john@laird-towers.freeserve.co.uk>1 Subject: Re: Does UK ATC use VMS, if not why not?b8 Message-ID: <ouqieu0bilq04ej5uj2178oub84eogtshn@4ax.com>  / On Mon, 20 May 2002 19:18:40 GMT, Tim Llewellyn ' <tim.llewellyn@blueyonder.co.uk> wrote:   B >Does anyone know what technology the new UK ATC system at Swanick >uses. >f >This is scary,a >p? >http://www.guardian.co.uk/airlines/story/0,1371,718325,00.htmlt >s% >Air travellers 'face year of chaos' e  I Every time I read of another holdup (what is it, 6-10 years late ?), UnixnC boxes were mentioned.  Don't know what flavour of vendor/os though.l   Shambles of the highest order.     	John    ------------------------------  # Date: Mon, 20 May 2002 22:17:32 GMTl1 From: LESLIE@JRLVAX.HOUSTON.RR.COM (Jerry Leslie)o1 Subject: Re: Does UK ATC use VMS, if not why not?I< Message-ID: <0CeG8.107586$Q42.5791735@typhoon.austin.rr.com>  5 Tim Llewellyn (tim.llewellyn@blueyonder.co.uk) wrote:n : C : Does anyone know what technology the new UK ATC system at Swanick  : uses.n :  : This is scary, : @ : http://www.guardian.co.uk/airlines/story/0,1371,718325,00.html : & : Air travellers 'face year of chaos'  :  : 4 It looks like a variety of systems are used by NATS:  (    http://www.ukatts.org.uk/gradnet1.htm>    Graduate Opportunities in National Air Traffic Services Ltd  F   "...a variety of platforms (Sun, DEC, HP and PC), operating systems D    (Unix, Ultrix, OpenVMS, HP-UX and Windows) and languages (C, C++     and Ada)...  6 The system that had the problem is an IBM 0/S390, per:  $    http://www.vnu.co.uk/News/11307927    vnunet.com Computer problems hit air traffic - againo  ,   "Computer problems hit air traffic - again    By Steve Ranger [10-04-2002]i      Twice in two weeksi  H    Thousands of travellers were delayed this morning (Wednesday) after aC    key air traffic control system failed for the second time in twoaE    weeks.The Flight Data Processing System (FDPS) at its West Drayton E    centre went down between 6am and 6.40am this morning. Although theAH    system was up and running by 6.40am, air traffic control was expectedE    to run at 70 per cent of capacity until 10am.The system prints theoE    paper strips that controllers use to plot the movements of planes.mH    When the system fails the strips have to be written out manually. The;    same system failed two weeks ago for 20 minutes, causing E    delays.Although the hardware for the FDPS was upgraded from an IBMtC    4381 to an IBM 0/S390 in Autumn last year, the application stilll@    contains some code created in the 1970s. National Air TrafficD    Services, which runs air traffic control, said there are softwareF    upgrades to the system around once a month. The same system crashedF    four times in 2000 causing massive delays, and is not scheduled forH    replacement until 2007. Airspace over London and the south east up toD    20,000 feet is still controlled from the site at West Drayton.TheB    control of high level airspace is now done from the air trafficF    control centre at Swanwick, which opened in January five years late    and 180m over budget."      H --Jerry Leslie   leslie@clio.rice.edu  (my opinions are strictly my own)9   Note: leslie@jrlvax.houston.rr.com is invalid for emaile   ------------------------------  % Date: Tue, 21 May 2002 00:06:26 +0100c' From: Elliott Roper <elliott@yrl.co.uk>e1 Subject: Re: Does UK ATC use VMS, if not why not?f2 Message-ID: <210520020006261922%elliott@yrl.co.uk>  C In article <ouqieu0bilq04ej5uj2178oub84eogtshn@4ax.com>, John Lairdt* <john@laird-towers.freeserve.co.uk> wrote:  1 > On Mon, 20 May 2002 19:18:40 GMT, Tim Llewellyn ) > <tim.llewellyn@blueyonder.co.uk> wrote:e > D > >Does anyone know what technology the new UK ATC system at Swanick > >uses. > >p > >This is scary,e > >rA > >http://www.guardian.co.uk/airlines/story/0,1371,718325,00.htmlP > >e' > >Air travellers 'face year of chaos' o > K > Every time I read of another holdup (what is it, 6-10 years late ?), UnixnE > boxes were mentioned.  Don't know what flavour of vendor/os though.  >   > Shambles of the highest order.  C I did quite a few googles. As you might expect, everyone is keeping  their heads below the parapet.  G On slim evidence, (A few CVs of people who were working on it) my money2E goes to AIX for the workstations. Could not find anything to say whatp the servers are. One site 4  http://www.shef.ac.uk/~pc1jm/hci98cc/HCI98CC36.htmlB discussed the GUI. Baroque and Bizzare were the words that come to? mind. They seem to have made a software imitation of the littlelE cardboard slips the controllers would slide into wooden racks back in  Biggles day.  G There was also a parliamentary report prepared by experts claiming thatc@ software replication was inappropriate for such a complex system: because most of the errors in such systems was in the userG requirements/ specifications that both teams would work to. Without toon5 much discussion of how they checked for errors in thee requirements/specifications.  H http://www.rvs.uni-bielefeld.de/publications/Reports/NERC/HMG-evidence.h tml   A The author of the GUI paper above alluded to the unwillingness tooE prototype and refine user specs in the face of fairly major usability-B errors. No doubt brought on by spectacular cost and time overruns.   Oh my sainted aunt!j  / I say Ginger, Algy's^h^h^h Byers' bought it! **-  G **Stephen Byers is currently the UK Minister for Transport, EnvironmentuG and some other quaint concept. He'll be out of a job in less than three.F weeks. In addition to the technical errors, he has to wear the botchedE privatisation of the Air Traffic Control system, their running out ofpD money, and the railways literally falling apart at the seams. Not toE mention being conned into leaking some damaging news about the timingtG of Euro zone entry to the press. If ever a politician were being set upo for the fall, it is he.f  C It should be noted that of the three recent ATC failures in the UK,eF only the most recent is laid on the new Swanwick system. The other twoD were the ancient West Drayton facility and the link between the two.  3 So *that's* all right then Best Beloved do you see?l   ------------------------------  % Date: Tue, 21 May 2002 00:27:09 +0100h' From: Elliott Roper <elliott@yrl.co.uk>P1 Subject: Re: Does UK ATC use VMS, if not why not?n2 Message-ID: <210520020027096471%elliott@yrl.co.uk>  B In article <0CeG8.107586$Q42.5791735@typhoon.austin.rr.com>, Jerry, Leslie <LESLIE@JRLVAX.HOUSTON.RR.COM> wrote:  7 > Tim Llewellyn (tim.llewellyn@blueyonder.co.uk) wrote:e > : E > : Does anyone know what technology the new UK ATC system at Swanickh	 > : uses.r > :  > : This is scary, > : B > : http://www.guardian.co.uk/airlines/story/0,1371,718325,00.html > : ( > : Air travellers 'face year of chaos'  > :  > : 6 > It looks like a variety of systems are used by NATS: > * >    http://www.ukatts.org.uk/gradnet1.htm@ >    Graduate Opportunities in National Air Traffic Services Ltd > H >   "...a variety of platforms (Sun, DEC, HP and PC), operating systems F >    (Unix, Ultrix, OpenVMS, HP-UX and Windows) and languages (C, C++  >    and Ada)...  D I'd take that with a grain of salt. ukatts.org.uk is the air traffic? controllers travel service, some kind of staff association. ThehF advertisement you mention is very old and is for a development sectionD unrelated to the NATS itself. It is a bit puzzling why it was on the travel site at all.m > 8 > The system that had the problem is an IBM 0/S390, per: > & >    http://www.vnu.co.uk/News/11307929 >    vnunet.com Computer problems hit air traffic - againv > . >   "Computer problems hit air traffic - again! >    By Steve Ranger [10-04-2002]t  D No, check the date again. That was last month's at West Drayton, not Friday's at Swanwick.b  D So far it is very difficult to glean details of what really happened from the general press.r  @ The best I could find was by Andrew Clark in Saturday's Guardian  G "National Air Traffic Services (Nats) said half of the terminals in theeG centre's control room would not start after developing hitches during ah# routine software upgrade overnight"n   ------------------------------  % Date: Mon, 20 May 2002 19:40:44 -0400o- From: JF Mezei <jfmezei.spamnot@videotron.ca>I1 Subject: Re: Does UK ATC use VMS, if not why not? , Message-ID: <3CE9896A.66D41A4D@videotron.ca>   Elliott Roper wrote:D > discussed the GUI. Baroque and Bizzare were the words that come toA > mind. They seem to have made a software imitation of the littlenG > cardboard slips the controllers would slide into wooden racks back inr > Biggles day.  L Remember seing some documentary of US aircraft carriers. And the crew in oneM of the control room explained that they had tried various software solutions, J but they always returned to the big table on which they moved the aircraftN around and get a much better situation awareness because computers could never match this.n  L Computers are good for pre-programmed stuff. But if an aicraft deviates fromM the normal stuff, how easy is it for the controller to type commands or use asD GUI to reflect a situation the systems wasn't programmed to handle ?  N Perhaps what is needed is far more than sticking a screen, keyboard and mouse.I It may require multiple screens, including a touch screen that allows theyN controller to "move" aircraft in a stack etc etc. And in the case of the ship,H it might require a table sized touch screen colour flat panel that woudlG project the ship's current situation with location of aircraft etc etc.e   ------------------------------  # Date: Tue, 21 May 2002 01:01:59 GMTi# From: "John Smith" <a@nonymous.com>t1 Subject: Re: Does UK ATC use VMS, if not why not?sH Message-ID: <b0hG8.57239$ah_.26887@news01.bloor.is.net.cable.rogers.com>  L "Although the hardware for the FDPS was upgraded from an IBM  4381 to an IBML 0/S390 in Autumn last year, the application still contains some code createdL in the 1970s. National Air Traffic Services, which runs air traffic control,D said there are software upgrades to the system around once a month."  L I had an app that we wrote in APL around 1980 that we tried on a 43xx seriesI machine. Took over a minute to do the calcs and post a transaction. MovedtD the APL to a 750, fiddled with a bit of it to take into account someH differences under VMS but the calc code and txn posting remained more or> less identical. Same unit of work took 4-5 seconds on the VAX.      K "The same system **crashed** four times in 2000 causing massive delays, andM. is not scheduled for  replacement until 2007."  J Not exactly the choice of words I'd use in an article about ATC for publicI consumption, unless I was trying to make a point or get more funding in ai hurry.   ------------------------------  % Date: Mon, 20 May 2002 19:19:09 +0100s; From: "Leigh G. Bowden" <LGBowden@bowdenfamily.fsnet.co.uk>s Subject: Re: End of an era.m/ Message-ID: <acbem7$10h$1@newsg1.svr.pol.co.uk>r  K Spoke too soon. The training and test systems are going as well. EverythingoG except about ten machines are staying, for hostorical purposes, so it'sg) going to be nearer a hundred been junked.o  K Everything else has been outsourced so technically "not our problem" but itn will be somebodies.    ------------------------------  % Date: Mon, 20 May 2002 15:24:06 -0400n- From: JF Mezei <jfmezei.spamnot@videotron.ca>e Subject: Re: End of an era. , Message-ID: <3CE94D55.5FAB73C0@videotron.ca>   David Mathog wrote:.N > I hope your company has hired the extra support personnel who will be neededM > once the Win2K machines start replacing the OpenVMS ones. Otherwise you cann. > look forward to some very, very long nights.  J Not if yo cut a CD image of each machine, and when a machine start to haveI problems, you just wide the disk and reinstall from the CD image. :-) :-)"' Isn't that standard Windows debugging ?T   ------------------------------  # Date: Tue, 21 May 2002 02:51:58 GMTp  From: cjt <cheljuba@prodigy.net> Subject: Re: End of an era.u+ Message-ID: <3CE9B654.A38F9AF0@prodigy.net>    JF Mezei wrote:  >  > David Mathog wrote:-P > > I hope your company has hired the extra support personnel who will be neededO > > once the Win2K machines start replacing the OpenVMS ones. Otherwise you canf0 > > look forward to some very, very long nights. > L > Not if yo cut a CD image of each machine, and when a machine start to haveK > problems, you just wide the disk and reinstall from the CD image. :-) :-)o) > Isn't that standard Windows debugging ?u  5 With most recent bios, you can boot right off the CD.    ------------------------------    Date: 20 May 2002 11:08:15 -0700( From: bob@instantwhip.com (Bob Ceculski)6 Subject: Re: Forced migration to HPHUX - Storm Warning= Message-ID: <d7791aa1.0205201008.7f405375@posting.google.com>a   Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> wrote in message news:<acaqr8$c6d$1@new-usenet.uk.sun.com>...- > Alan Greig wrote:0 > D > > On 19 May 02 15:06:23 +0200, p_sture@elias.decus.ch (Paul Sture)
 > > wrote: > >  > >  > > I > >>And the flaw in that is that V7.3 is _not_ a requirement for a GS320.o > >> > > H > > To be fair I thought it was highly recommended because of all the MP > > locking improvements.  > >  >  > > > Exactly, if you are to stand even a tiny chance of getting a@ > 32 way GS32 to deliver 4 x the throughput of an 8-10 way GS140 > then you need 7.3e > 	 > Regardst > Andrew Harrison   A however, that will not be true for EV7 Marvel ... you can run VMSl6 7.2-2 and blow away everything, and I mean everything!   ------------------------------  % Date: Tue, 21 May 2002 01:42:45 +0200nB From: Michiel Erens <I.dont.want.spam@this.mailaddress.is.invalid>6 Subject: Re: Forced migration to HPHUX - Storm Warning7 Message-ID: <3CE989F5.14E7@this.mailaddress.is.invalid>    Alan Greig wrote:s > F > As I recall we had a cluster uptime of around five years on our dualH > admin VAX 4600 cluster (with quorum disk). No crashes and  two rolling+ > upgrades. The systems were UPS protected.a  D Speaking about five years uptime, there was an article in the Dutch C weekly "Automatiseringsgids" about the Amsterdam police. Five yearsMC ago they migrated their VAX machines to a four node Alpha cluster. D@ Since then they have had 100% availability. Apart from this the F article mentions other strong points of VMS like 'upgrades in stages, D because OpenVMS allows two versions of the OS in the same cluster', E and 'the standard management tools of VMS are excellent'. The police t< also have Unix systems, but 'these could not reach the high A availability of the VMS system'. All in all a much more positive  D article then one a couple of months ago where OpenVMS was mentioned  as very reliable but old.l   -- s ME Posted by news://news.nb.nu    ------------------------------  # Date: Tue, 21 May 2002 01:18:52 GMTu# From: "John Smith" <a@nonymous.com>D6 Subject: Re: Forced migration to HPHUX - Storm WarningH Message-ID: <0ghG8.72509$t8_.20201@news01.bloor.is.net.cable.rogers.com>  " Is this article available on-line?    G "Michiel Erens" <I.dont.want.spam@this.mailaddress.is.invalid> wrote inn9 message news:3CE989F5.14E7@this.mailaddress.is.invalid...  > Alan Greig wrote:. > >2H > > As I recall we had a cluster uptime of around five years on our dualJ > > admin VAX 4600 cluster (with quorum disk). No crashes and  two rolling- > > upgrades. The systems were UPS protected.e >'E > Speaking about five years uptime, there was an article in the DutchiE > weekly "Automatiseringsgids" about the Amsterdam police. Five years-D > ago they migrated their VAX machines to a four node Alpha cluster.A > Since then they have had 100% availability. Apart from this the G > article mentions other strong points of VMS like 'upgrades in stages, E > because OpenVMS allows two versions of the OS in the same cluster',aF > and 'the standard management tools of VMS are excellent'. The police= > also have Unix systems, but 'these could not reach the high B > availability of the VMS system'. All in all a much more positiveE > article then one a couple of months ago where OpenVMS was mentioned- > as very reliable but old., >  > -- > ME > Posted by news://news.nb.nu<   ------------------------------   Date: 20 May 2002 23:46:00 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: freeVMS* Message-ID: <acc1ro$4j4$2@web1.cup.hp.com>  K In article <newscache$8gq1wg$4g7$1@news.emea.compuware.com>, "Axel Haringa"t' <axel.haringa@nl.compuware.com> writes:k :a free VMS clone!  J   Um, so?   (No offense, but the FreeVMS effort sees regular discussions.)  $ :http://freevms.free.fr/indexGB.html  K   Grumble.  These FreeVMS sites are hard to track, and I've been trying to UH   keep current pointers in the OpenVMS Frequently Asked Questions (FAQ)    document.o  J   The pointer in the current OpenVMS FAQ (to http://www.free-vms.org/) is K   apparently now stale -- at least, the site is now down.  The most recent eL   pointer to the FreeVMS program that I have -- queued for the next edition    of the FAQ -- is:o  )     http://freshmeat.net/projects/freevmsl#     Mailing list freevms@ml.free.fro  ?   I'll add http://freevms.free.fr/indexGB.html to the next FAQ.   J   There are also pointers to the OpenVMS Hobbyist program available in theK   FAQ.  The Hobbyist program provides free licenses and cheap distributionsl'   for non-commercial users of OpenVMS. y  J   There are also at least three VAX emulators around for PC systems -- theK   SRI Charon-VAX package, Bob's Trailing-Edge emulator, and one other whoseC=   name escapes me -- but all have been discussed here before.-    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Mon, 20 May 2002 17:49:17 GMT # From: "John Smith" <a@nonymous.com>l3 Subject: Re: Geting Data off a DEC Microvax 3100-95aH Message-ID: <xGaG8.53324$ah_.50140@news01.bloor.is.net.cable.rogers.com>  % JF had some good suggestions for you.   J As I recall, there's nothing particularly special about the 3100-95 serialK ports. A breakout box and a couple minutes fiddling will get you connected.p  E Suggest you use a vt2xx emulator with kermit as the simplest means ofeH getting the data off. If the amount of data is relatively small and yourL terminal emulator software supports 'log to file', you may be able to simplyI 'type' the data and save it that way, eg.  $ type foo.txt and log it to ai local file.   0 Here's what the asynch port pin-out probably is:   1    protective gnda 2    Txr 3    Rd  4    RTS 5    CTS 6    DSR 7    GND 8    CD  9    --  10    -- 11    -- 12    SD 13    -- 14    -- 15    -- 16    -- 17    --$ 18    Receiver signal element timing 19    --	 20    DTRs 21    -- 22    calling indicator2 23    Data signal rate selector  24    --
 25    test      0 "Mark Gowans" <mark@gowans.org> wrote in message7 news:287c86c0.0205200232.389f8bf2@posting.google.com...r > Hi,a >tE > I've got a Digital Microvat 3100-95 which I need to get a couple ofr? > files off from and onto a Solaris machine. Unfortunately, I'm@G > completely unfamiliar with all things VMS/VAX so haven't really got ao > clue where to start. > G > I've had a good trawl through the compaq/hp website and cant find any2H > real specs for the 3100-95 which hasn't helped. I was hoping to find aD > serial spec for a port on the machine and hookup a laptop then useD > some form of terminal package to hook files off that way.. FailingD > that, I can get a network connection to the machine - but wouldn't( > really know how to procede from there. >tB > Any thoughts / suggestions on this would be greatly appreciated, >c > Many thanks, >A > Mark   ------------------------------  % Date: Mon, 20 May 2002 17:16:40 -0400i, From: David Michaels <michaedi@email.uc.edu>3 Subject: Re: Geting Data off a DEC Microvax 3100-95 , Message-ID: <3CE967B8.9BB71020@email.uc.edu>   John Smith wrote:? > ' > JF had some good suggestions for you.  > L > As I recall, there's nothing particularly special about the 3100-95 serialM > ports. A breakout box and a couple minutes fiddling will get you connected.o > G > Suggest you use a vt2xx emulator with kermit as the simplest means ofiJ > getting the data off. If the amount of data is relatively small and yourN > terminal emulator software supports 'log to file', you may be able to simplyK > 'type' the data and save it that way, eg.  $ type foo.txt and log it to ao
 > local file.e > 2 > Here's what the asynch port pin-out probably is: >  > 1    protective gnd 	 > 2    Tx 	 > 3    Rd 
 > 4    RTS
 > 5    CTS
 > 6    DSR
 > 7    GND	 > 8    CDa	 > 9    --r
 > 10    --
 > 11    --
 > 12    SD
 > 13    --
 > 14    --
 > 15    --
 > 16    --
 > 17    --& > 18    Receiver signal element timing
 > 19    -- > 20    DTRe
 > 21    -- > 22    calling indicatora! > 23    Data signal rate selectorh
 > 24    -- > 25    test > 2 > "Mark Gowans" <mark@gowans.org> wrote in message9 > news:287c86c0.0205200232.389f8bf2@posting.google.com.../ > > Hi,n > > G > > I've got a Digital Microvat 3100-95 which I need to get a couple ofoA > > files off from and onto a Solaris machine. Unfortunately, I'mlI > > completely unfamiliar with all things VMS/VAX so haven't really got a  > > clue where to start. > >iI > > I've had a good trawl through the compaq/hp website and cant find anyeJ > > real specs for the 3100-95 which hasn't helped. I was hoping to find aF > > serial spec for a port on the machine and hookup a laptop then use> > > some form of terminal package to hook files off that way..    C On Solaris you can tip the serial port eg... `tip -9600 /dev/cua/x`iH where x is a or b I've never had any luck transfering anything but plain@ text with it but I haven't tried very hard either... Ethernet is# probably your best bet.... GoodLucki     > > FailinghF > > that, I can get a network connection to the machine - but wouldn't* > > really know how to procede from there. > >oD > > Any thoughts / suggestions on this would be greatly appreciated, > >i > > Many thanks, > >e > > Mark   ------------------------------  # Date: Mon, 20 May 2002 18:06:35 GMT-  From: AVIGLH <AVIGLH@boeing.com> Subject: GhostView on VMSr* Message-ID: <3CE93B2B.80244FA0@boeing.com>  7 I have Ghostview 3.5.8 on an alpha at OpenVMS V7.1-2.     G The GV/GS window will open, but I cannot open a postscript file nor any7" other kind of file.  The error is   A %DCL-W-NOCOMD, no command on line - reenter with alphabetic firste	 charactery  % In the Ghostscript message popup box.n  < I carefully copied the files into place.  I do not have fullH instructions on installing.  When I copied the files, the GV program ranH right away.  I have tried many files and all sorts of detours.  No files will open.     Thanx for looking,   Greg Hewitte   ------------------------------  # Date: Mon, 20 May 2002 20:32:31 GMT ' From: Rick Dyson <Rick-Dyson@UIowa.EDU>r Subject: Re: GhostView on VMS.) Message-ID: <3CE95D5F.BA550ADA@UIowa.EDU>)  
 AVIGLH wrote:m > 7 > I have Ghostview 3.5.8 on an alpha at OpenVMS V7.1-2.a > I > The GV/GS window will open, but I cannot open a postscript file nor anyg# > other kind of file.  The error isr > C > %DCL-W-NOCOMD, no command on line - reenter with alphabetic firstt > characterl > ' > In the Ghostscript message popup box.  > > > I carefully copied the files into place.  I do not have fullJ > instructions on installing.  When I copied the files, the GV program ranJ > right away.  I have tried many files and all sorts of detours.  No files > will open.  K Ghostview is only part of the equation.  You also have to have GhostScript. J They are separate packages.  Do you know if you specifically installed theG GS package?  How are you installing GV and/or GS?  From the freeware oroI from the original source code sites?  Are you compiling them from scratcht) or are you getting pre-compiled binaries?a   rick   ------------------------------   Date: 20 May 2002 17:30:49 GMT2 From: "Zane H. Healy" <healyzh@shell1.aracnet.com>& Subject: Re: HP startup screen anyone?, Message-ID: <acbbs90304q@enews1.newsguy.com>  : Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> wrote:A >> > Bah!  I prefer a nice d|i|g|i|t|a|l on my startup screen!!!!  >> CN >> I hate to disppoint you, but upgrading to V7.3 / DWMOTIF V1.2-6 replaces it >> with Compaq.   H > How easy is it to save the old version, and use it instead of the new 
 > default?  ! Should be able to make a copy of tK CDE$SYSTEM_COMMON:[APPCONFIG.ICONS.C]DECDTLOGO.PM from a pre 7.3 system andaB then replace whatever they're using to get the Compaq logo in 7.3.   			Zaner   ------------------------------   Date: 20 May 2002 23:33:53 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)% Subject: Re: LISP for OpenVMS Alpha ? * Message-ID: <acc151$4j4$1@web1.cup.hp.com>  g In article <7lwD8.1363$k06.26617@news.kpnqwest.fi>, "Kari Keronen" <kari.keronen@radiolinja.fi> writes:e9 :Are there any free (Common) LISP products for OpenVMS ?    C   You can run a Google search just as easily as I can, of course...f   :This is just forsE :hobbyist purposes and has nothing to do with my current job. AnotheroI :interesting thing would be Prolog for VMS, but that might be hopeless...e  G   As I am sure you are aware, there are various flavors and derivativesfE   and cousins of LISP around, and the RuleWorks (OPS5) package is NOTtC   LISP.  That said, RuleWorks is available on the OpenVMS Freeware.p  @     http://www.openvms.compaq.com/freeware/freeware50/ruleworks/  I   As to whether or not RuleWorks or OPS5 is sufficiently similar to LISP m8   to meet your particular hobbyist needs, I do not know.  D   There have certainly been various LISP ports around, and emacs (of   course) contains LISP.  9   Google found the following, which might be of interest:1  A     http://www.idiom.com/free-compilers/ECATEGORY/lispfami-1.htmls  D   Google also found references to c-Prolog, which include references   to a VMS (OpenVMS) port.  0   Google is available at http://www.google.com/.  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Mon, 20 May 2002 22:04:26 -0700 % From: Dean Woodward <deanw@rdrop.com>a% Subject: Re: LISP for OpenVMS Alpha ?d) Message-ID: <3CE9D55A.1BC7F7C3@rdrop.com>x   Hoff Hoffman wrote:d > 2 >   Google is available at http://www.google.com/.  G Whew! Glad you threw that in.  I mean, it'd be damn tough to use Google 5 to find Google if you don't know where Google is. :-/2   what?    --  : Dean Woodward   | Portland, OR- worst motorcycling weather0 deanw!rdrop,com | in the continental US, there. 4                 |  - Someone on rec.moto, circa 1994A ----------------+------------------------------------------------o@ '66 Duc 250 - '85 FJ1100 - '00 KLR650 - '01 Falco - '03 FJR1300*   (*Soon, but not soon enough!)u   ------------------------------    Date: 20 May 2002 20:19:40 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)-3 Subject: Re: LISP for OpenVMS Alpha ? (RuleWorks ?):3 Message-ID: <28f67JDLoVyF@eisner.encompasserve.org>o  _ In article <acc151$4j4$1@web1.cup.hp.com>, hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) writes:w  I >   As I am sure you are aware, there are various flavors and derivativesbG >   and cousins of LISP around, and the RuleWorks (OPS5) package is NOTeE >   LISP.  That said, RuleWorks is available on the OpenVMS Freeware.f > B >     http://www.openvms.compaq.com/freeware/freeware50/ruleworks/  ; Hmmm.  I have had a distant interest in OPS5 for some time, # but I had never hear of RuleWorks !    ------------------------------    Date: 20 May 2002 12:55:37 -0600+ From: young_r@encompasserve.org (Rob Young).M Subject: Re: Most unusual resident in a system ?, was: Re: No new Alpha sales 3 Message-ID: <AD91itxY1svU@eisner.encompasserve.org>    In article <acb9om$ghq$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> writes: >  >  > Rob Young wrote: >  >> In article <acb37s$ep0$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> writes:w >> h >> k* >>>>	No.  They are still very competitive. >>>> >>>tB >>>So if you are still maintaining that they are still competitive= >>>produce something that justifies your claim that they are.  >>>y >>>Put up or shut up.r >>>o >>>t >> i8 >> http://www.tpc.org/tpcc/results/tpcc_perf_results.asp+ >> http://www.compaq.com/hpc/news/news.html5 >> 8 >  > 7 > Sun hasn't done a TPC-C benchmark on either the F6800c9 > or the F15000. So you don't have any Sun to compare thes > GS320 result with. >   & 	Okay.  So I found a weak spot.  Good.  ? > However Sun has done TPC-H, better result per CPu and overallP; > than a GS320 on a Sun F15000 and a better cost per query.  >  > Oracle Apps ditto. > SAP ditto.  ; 	Ah, you found a strength.  Never denied they didn't exist.h   > : > In addition the result you mention is awfull relative to9 > SuperDome, IBM P Series etc all of which Sun outperforma& > on the 3 benchmarks I covered above. > 9 > Terrible example Rob how about coming up with somethinge	 > better.a >   # http://www.top500.org/list/2001/11/i  1 Click here to show the explaination of the fields.  K List Rank Manufacturer Computer Rmax(GFlops) Installation Site Country Year-% Processors Rpeak (GFlops) Nmax Nhalf 4  N 11/2001 1 IBM ASCI White,SP Power3 375 MHz 7226.00 Lawrence Livermore National- Laboratory USA 2000 8192 12288 518096 179000 i  L 11/2001 2 Compaq AlphaServer SC ES45/1 GHz 4059.00 Pittsburgh Supercomputing Center USA 2001 3024 e   	a >  > > >> 	Your CPU is so under-powered , you haven't won a large HPCB >> 	bid in quite some time.  Couple that with the fact the AussiesA >> 	took a big hit with Zinc Whiskers and your reputation in thathB >> 	space among scientists took a big public and behind-the-scenes >> 	hit too. >> c >  > @ > The Aussies replaced a E10000 which is as you may have noticed7 > a Sun which cam out arround the same time as an 8400.p >   ? 	Nice try at spinning it.  Those E10000s never went production,p0 	failing acceptance testing.  If they had passed? 	acceptance testing, they would have had to honor the purchase.e: 	As it was, they could simply make you eat them on return.  G 	It wasn't as if they had used them for 2 or 3 years and tired of them.t  4 http://www.sun.com.au/news/localpress/2000/08/1.html  N Canberra August 2, 2000. In a major boost to the nation's research sector, SunK Microsystems has joined forces with the Australian Partnership for Advanced F Computing (APAC) to install a powerful computing system for Australian( researchers and industry professionals.   L The system will be installed in the APAC National Facility at the Australian= National University, which is the host institution for APAC. E  J The Sun system will be employed in scientific endeavours such as molecularL modeling for new drugs, simulation of airflow for engineering product designG and pattern discovery for fraud detection. It will open the way for new = research results and product development across the country. r  L The three-year agreement will see an initial commissioning of a 200 GigaflopL system in September 2000, comprising a cluster of four E10000 compute nodes,F and will progressively upgrade this to over one Teraflop by mid-2002.    	--h  = 	What a happy group of gents in that picture.  Sure the faces08 	got longer a few months later.  They are no doubt happy 	again running on Alphas.i    C > But HPC hardly makes a dent on the overall server market which is D > primarely integer and dominated by DBMS's, where are your examples< > to back up the claims made for Alpahservers in this space.   	Spin spin spin.   > , > Ahh we had them TPC-C, not a good example. >   C 	Because they are no longer important OR you can't do decent Oraclen> 	numbers?   Seems IBM still thinks tpmC is important and oftenC 	uses them to show that you are no longer competitive in that very	 @ 	important metric.  But at one time Sun was all the rage to talk; 	up tpmC.  Little Microsoftian/Orwellian in the spin there.j    > > Incedentally I hope the Aussies didn't buy the Alphas on the; > basis of SPECfp results. After Freddies claim that SPECfpc4 > optimisations would not be usefull for users code.  < 	Ha.  You are really struggling to diss them.  No doubt your> 	competitors in the HPC space got a lot of mileage out of that 	table turning event.a   > @ >>>Buzzt, Rob large Sun's deliver better per CPU throughput thanB >>>AlphaServers on all the DBMS based benchmarks published by both >>>Sun and Compaq. >>>:A >>>You know this, we all know this and Oracles own Apps benchmarki= >>>illustrates this so why keep making the same old claim. Weo/ >>>don't suffer on a per-CPU basis Compaq does.  >>>t >>>Q >> sG >> 	No, you do indeed.  But you haven't shipped tpmCs in quite sometimes@ >> 	due to the beatings.  You have a loan TPC-C version 5 entry: >> eD >> http://www.tpc.org/tpcc/results/tpcc_results.asp?orderby=hardware@ >> http://www.tpc.org/tpcc/results/tpcc_results.asp?orderby=dbms >> n >  > @ > As I said earlier bad example try something else, incedentally= > why are you trumpeting a result where the GS320 (the worlds-; > fastest server if you believe Compaq) is actually 3rd and ; > less than 5% faster than a much smaller, much cheaper IBMd@ > which didn't have to use OPS to get a decent result. You would* > seem to be running out of feet to shoot. >   < 	I said competitive, I didn't say leader.  That comes later.   >> 	Please.  Give me a break.- >> p >  > 6 > If you prove that yopu deserve one then you will get > one at the moment you don't. >    	Prove?l  8 > You need to do much better than the tired TPC-C mantra' > that you and Kerry keep trotting out.    	Works though.   				Robj   ------------------------------  # Date: Mon, 20 May 2002 18:13:48 GMT4* From: "Bill Todd" <billtodd@metrocast.net>M Subject: Re: Most unusual resident in a system ?, was: Re: No new Alpha salespA Message-ID: <w1bG8.62725$e66.5831706@bin6.nnrp.aus1.giganews.com>   8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:4J$JIc5UqK8a@eisner.encompasserve.org...h   ...e  ? > My joke is the fact that a SPARC box is powered by squirrels.vC > Your employer is shipping nearly the weakest CPUs in the industry,B > and when going head to head with others on large Oracle RFPs youB > are at a decided disadvantage as your per-CPU costs times number1 > of required CPUs often leave you uncompetitive.   K I suppose it's possible that you're simply out of date with your data.  But J USIII is no longer 'nearly the weakest CPU in the industry'.  I don't knowG its stream numbers, but given the on-chip memory support they should be J respectable.  And the 610 SPECint performance (not to be confused with theB more impressive, but arguably compiler-specific, SPECfp number) is2 respectable as well, even if not industry-leading.  L Sun has traditionally managed to produce at least average-performing serversL using below-average-performing processors.  Unless they suddenly have becomeL inflicted with Compaq-head-office-level incompetence, the USIII should allowL them to produce better-performing servers that should satisfy customer needsJ more than adequately, even if (again) not as industry leaders (whether EV7K will qualify as the industry leader depends on what minimal market share ishJ required to play in this league rather than be relegated to niche status).  I FUD is insulting regardless of its source:  it indicates that the FUDsterlI believes his audience is sufficiently gullible to be misled by it.  WhilesG the c.o.v. audience may well incline toward Alpha jingoism, I'd like to-@ think it still has the ability to distinguish fact from fiction.   - bill   ------------------------------    Date: 20 May 2002 13:50:12 -0600+ From: young_r@encompasserve.org (Rob Young)7M Subject: Re: Most unusual resident in a system ?, was: Re: No new Alpha salesr3 Message-ID: <zEz34brd6DX7@eisner.encompasserve.org>l  n In article <w1bG8.62725$e66.5831706@bin6.nnrp.aus1.giganews.com>, "Bill Todd" <billtodd@metrocast.net> writes: > : > "Rob Young" <young_r@encompasserve.org> wrote in message/ > news:4J$JIc5UqK8a@eisner.encompasserve.org...o >  > ...  > @ >> My joke is the fact that a SPARC box is powered by squirrels.D >> Your employer is shipping nearly the weakest CPUs in the industryC >> and when going head to head with others on large Oracle RFPs you C >> are at a decided disadvantage as your per-CPU costs times numberv2 >> of required CPUs often leave you uncompetitive. > M > I suppose it's possible that you're simply out of date with your data.  But > > USIII is no longer 'nearly the weakest CPU in the industry'.  B 	Sure it is.  Check out a summation of best case SpecInt200  base ) 	numbers for various archs/manufacturers:    SpecInt200      Base   AMD             720  Sun             537a Alpha           621  Dell            790e Hewlett         5682 IBM             804g Intel           819  SGI             410D  6 	SGI trailing the pack, Sun near the back of the pack.   >  I don't knowaI > its stream numbers, but given the on-chip memory support they should beiL > respectable.  And the 610 SPECint performance (not to be confused with theD > more impressive, but arguably compiler-specific, SPECfp number) is4 > respectable as well, even if not industry-leading. >   @ 	Respectable and a jump through hoops.  They have obviously beenC 	up to their compiler tricks as the disparity between base and peakoF 	is quite a bit and I don't believe anyone comes close.  For instance,( 	IBM shows theses peak and base numbers:  =                                                  Base    Peakt< IBM Corporation IBM eServer pSeries 690 Turbo    804     839    @ 	I suppose I could take the time to figure out which application< 	they cooked that will be tossed out of the next Spec suite.H 	(Applications are kicked out all the time when an optimization renders @ 	that application useless as a measurement, maxtrix300, etc. and0 	Spec2000's purpose is to measure CPU strength).  N > Sun has traditionally managed to produce at least average-performing serversN > using below-average-performing processors.  Unless they suddenly have becomeN > inflicted with Compaq-head-office-level incompetence, the USIII should allowN > them to produce better-performing servers that should satisfy customer needsL > more than adequately, even if (again) not as industry leaders (whether EV7M > will qualify as the industry leader depends on what minimal market share is.L > required to play in this league rather than be relegated to niche status). >   = 	No... the point is using known tactics because Sun has a CPUoA 	weaker than others it requires higher CPU counts to gain similarw= 	tpmC numbers, driving up Oracle costs considerably as Oracle2 	charges per-CPU.m  K > FUD is insulting regardless of its source:  it indicates that the FUDsterDK > believes his audience is sufficiently gullible to be misled by it.  While5I > the c.o.v. audience may well incline toward Alpha jingoism, I'd like to B > think it still has the ability to distinguish fact from fiction.    C 	True.  And collectively we have been deflecting Andrew's FUD long i 	before you showed up.   				Robe   ------------------------------  # Date: Mon, 20 May 2002 20:44:47 GMT * From: "Bill Todd" <billtodd@metrocast.net>M Subject: Re: Most unusual resident in a system ?, was: Re: No new Alpha salesv9 Message-ID: <2fdG8.1$oF3.495@bin4.nnrp.aus1.giganews.com>i  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:zEz34brd6DX7@eisner.encompasserve.org... I > In article <w1bG8.62725$e66.5831706@bin6.nnrp.aus1.giganews.com>, "Billa& Todd" <billtodd@metrocast.net> writes:   ...a  5 > Check out a summation of best case SpecInt200  basep* > numbers for various archs/manufacturers: >e > SpecInt200      Base >k > AMD             720- > Sun             537M > Alpha           621r > Dell            790a > Hewlett         568> > IBM             804u > Intel           819g > SGI             410s >e7 > SGI trailing the pack, Sun near the back of the pack.r  ? Hmmm.  You seem to have left out your favorite Alpha successor:t  G Itanic   379   (no peak figure obtained there; best listed peak is 365)W  J Perhaps that's because its owner chose not to bother to submit any SPECintI results (hardly surprising:  what's surprising is that HP bothered, but IaC guess that Dell 314 figure was just too embarrassing to let stand).m  J And you also seem to have missed SGI's 483 (500 peak) entry (though ItanicH would still have been trailing the pack even with the SGI 410 figure you quoted).  A When you throw out the IA32 results (not really applicable to the:K battle-of-the-servers discussion here), you're left with (in base SPECint2K  order)   IBM/POWER4   804   (839 peak)q% cHumPaq/Alpha EV68   621   (679 peak)r" HP/PA-RISC 8700   568   (604 peak)% Sun/UltraSPARC III   537   (610 peak)V  SGI/MIPS R14K   483   (500 peak) Intel/Merced   379   (365 peak)   F Change the ordering to SPECint2K peak, and the Sun and HP entries flipL positions.  In either case, USIII falls squarely in the middle of the pack -J and the only *real* outliers are POWER4 (arguably an artificial figure dueF to its use of far more than a single chip's share of cache) and Itanic& (unfortunately not artificial at all).   >a > >  I don't knowaK > > its stream numbers, but given the on-chip memory support they should beeJ > > respectable.  And the 610 SPECint performance (not to be confused with theyF > > more impressive, but arguably compiler-specific, SPECfp number) is6 > > respectable as well, even if not industry-leading. > >, >dA > Respectable and a jump through hoops.  They have obviously beengD > up to their compiler tricks as the disparity between base and peakG > is quite a bit and I don't believe anyone comes close.  For instance,l) > IBM shows theses peak and base numbers:p >M? >                                                  Base    Peak-> > IBM Corporation IBM eServer pSeries 690 Turbo    804     839  J Why not look at the Alpha base/peak numbers?  621 vs. 679 is closer to theG Sun difference than to the POWER4 difference:  want to accuse Compaq of-  playing compiler tricks as well?   >  >fA > I suppose I could take the time to figure out which application-= > they cooked that will be tossed out of the next Spec suite. H > (Applications are kicked out all the time when an optimization rendersA > that application useless as a measurement, maxtrix300, etc. ando1 > Spec2000's purpose is to measure CPU strength).g  F But instead you'd rather just spread more FUD.  Typical, Rob:  make anK accusation without a shred of evidence just because you don't like the tale  the numbers tell.r   ...o  E > > FUD is insulting regardless of its source:  it indicates that thet FUDstereF > > believes his audience is sufficiently gullible to be misled by it. While-K > > the c.o.v. audience may well incline toward Alpha jingoism, I'd like to0D > > think it still has the ability to distinguish fact from fiction. >- >-C > True.  And collectively we have been deflecting Andrew's FUD longi > before you showed up.y  D Since I showed up at DEC in early 1976, I rather doubt that - thoughJ longevity is no guarantee of competence (as you demonstrate so well).  ButI my memory may not be quite what it used to be, so please help me remember K the last (or indeed *any*) time when we differed on anything of a technical ' nature that you turned out to be right.e   - bill   ------------------------------    Date: 20 May 2002 15:51:24 -0600+ From: young_r@encompasserve.org (Rob Young)eM Subject: Re: Most unusual resident in a system ?, was: Re: No new Alpha salesn3 Message-ID: <yD8uliiCrsgb@eisner.encompasserve.org>o  f In article <2fdG8.1$oF3.495@bin4.nnrp.aus1.giganews.com>, "Bill Todd" <billtodd@metrocast.net> writes: > : > "Rob Young" <young_r@encompasserve.org> wrote in message/ > news:zEz34brd6DX7@eisner.encompasserve.org...-J >> In article <w1bG8.62725$e66.5831706@bin6.nnrp.aus1.giganews.com>, "Bill( > Todd" <billtodd@metrocast.net> writes: >  > ...y > 6 >> Check out a summation of best case SpecInt200  base+ >> numbers for various archs/manufacturers:b >> >> SpecInt200      Baseo >> >> AMD             720 >> Sun             537 >> Alpha           621 >> Dell            790 >> Hewlett         568 >> IBM             804 >> Intel           819 >> SGI             410 >>8 >> SGI trailing the pack, Sun near the back of the pack. > A > Hmmm.  You seem to have left out your favorite Alpha successor:  > I > Itanic   379   (no peak figure obtained there; best listed peak is 365)/ > L > Perhaps that's because its owner chose not to bother to submit any SPECintK > results (hardly surprising:  what's surprising is that HP bothered, but IrE > guess that Dell 314 figure was just too embarrassing to let stand).o > L > And you also seem to have missed SGI's 483 (500 peak) entry (though ItanicJ > would still have been trailing the pack even with the SGI 410 figure you
 > quoted). >   " 	It was a very quick scan, thanks.  C > When you throw out the IA32 results (not really applicable to the M > battle-of-the-servers discussion here), you're left with (in base SPECint2K  > order) >  > IBM/POWER4   804   (839 peak)e' > cHumPaq/Alpha EV68   621   (679 peak)t$ > HP/PA-RISC 8700   568   (604 peak)' > Sun/UltraSPARC III   537   (610 peak)e" > SGI/MIPS R14K   483   (500 peak)! > Intel/Merced   379   (365 peak)m > H > Change the ordering to SPECint2K peak, and the Sun and HP entries flipN > positions.  In either case, USIII falls squarely in the middle of the pack -L > and the only *real* outliers are POWER4 (arguably an artificial figure dueH > to its use of far more than a single chip's share of cache) and Itanic( > (unfortunately not artificial at all).  < 	It is one off the bottom, if looking at base and most of us= 	do not play with compiler flags.  Can we quote McKinley soonk 	and forget about Merced?r   >  >> >> >  I don't knowL >> > its stream numbers, but given the on-chip memory support they should beK >> > respectable.  And the 610 SPECint performance (not to be confused withn > the G >> > more impressive, but arguably compiler-specific, SPECfp number) is 7 >> > respectable as well, even if not industry-leading.f >> > >>B >> Respectable and a jump through hoops.  They have obviously beenE >> up to their compiler tricks as the disparity between base and peakyH >> is quite a bit and I don't believe anyone comes close.  For instance,* >> IBM shows theses peak and base numbers: >>@ >>                                                  Base    Peak? >> IBM Corporation IBM eServer pSeries 690 Turbo    804     839s > L > Why not look at the Alpha base/peak numbers?  621 vs. 679 is closer to theI > Sun difference than to the POWER4 difference:  want to accuse Compaq ofy" > playing compiler tricks as well? >   @ 	Yes.  Most likely the same fellow figured it out is now at Sun.  2 	The blade cuts in all directions in this case :-)   >> >>B >> I suppose I could take the time to figure out which application> >> they cooked that will be tossed out of the next Spec suite.I >> (Applications are kicked out all the time when an optimization rendersaB >> that application useless as a measurement, maxtrix300, etc. and2 >> Spec2000's purpose is to measure CPU strength). > H > But instead you'd rather just spread more FUD.  Typical, Rob:  make anM > accusation without a shred of evidence just because you don't like the tale  > the numbers tell.0 >    	Nope.  Try again.   > ...  > F >> > FUD is insulting regardless of its source:  it indicates that the	 > FUDsteroG >> > believes his audience is sufficiently gullible to be misled by it.n > WhileoL >> > the c.o.v. audience may well incline toward Alpha jingoism, I'd like toE >> > think it still has the ability to distinguish fact from fiction.n >> >>D >> True.  And collectively we have been deflecting Andrew's FUD long >> before you showed up. > ( > Since I showed up at DEC in early 1976   	comp.os.vms, not DEC.   				Robl   ------------------------------  # Date: Mon, 20 May 2002 22:54:49 GMTl* From: "Bill Todd" <billtodd@metrocast.net>M Subject: Re: Most unusual resident in a system ?, was: Re: No new Alpha salessA Message-ID: <Y8fG8.66692$e66.6083222@bin6.nnrp.aus1.giganews.com>   8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:yD8uliiCrsgb@eisner.encompasserve.org...iG > In article <2fdG8.1$oF3.495@bin4.nnrp.aus1.giganews.com>, "Bill Todd"g  <billtodd@metrocast.net> writes:   ...e  E > > When you throw out the IA32 results (not really applicable to theeE > > battle-of-the-servers discussion here), you're left with (in baseu	 SPECint2KA
 > > order) > >l! > > IBM/POWER4   804   (839 peak)d) > > cHumPaq/Alpha EV68   621   (679 peak)k& > > HP/PA-RISC 8700   568   (604 peak)) > > Sun/UltraSPARC III   537   (610 peak)w$ > > SGI/MIPS R14K   483   (500 peak)# > > Intel/Merced   379   (365 peak)C > > J > > Change the ordering to SPECint2K peak, and the Sun and HP entries flipI > > positions.  In either case, USIII falls squarely in the middle of thes pack -J > > and the only *real* outliers are POWER4 (arguably an artificial figure due J > > to its use of far more than a single chip's share of cache) and Itanic* > > (unfortunately not artificial at all). >n= > It is one off the bottom, if looking at base and most of ush" > do not play with compiler flags.  H I'm having difficulty following the above:  exactly what is 'one off theK bottom'?  Not Itanic:  it's at the bottom (no matter how you slice it, baseeH or peak).  Not USIII, which is in the middle (+/- 1/2, and either 2 or 3< 'off the bottom') no matter how you slice it (base or peak).     Can we quote McKinley soon > and forget about Merced?  I I'd *love* to quote McKinley numbers (though I won't be willing to forget H about Merced until you forget about out-of-date US performance):  please- provide a credible SPECint2K source for them.s   ....  J > > Why not look at the Alpha base/peak numbers?  621 vs. 679 is closer to theMK > > Sun difference than to the POWER4 difference:  want to accuse Compaq ofi$ > > playing compiler tricks as well? > >a >pA > Yes.  Most likely the same fellow figured it out is now at Sun.   H Ah - a conspiracy emerges.  And it looks like he has friends at AMD, andB Fujitsu, and even IBM:  significant base/peak spreads aren't rare.   >S3 > The blade cuts in all directions in this case :-)s >i > >> > >>D > >> I suppose I could take the time to figure out which application@ > >> they cooked that will be tossed out of the next Spec suite.K > >> (Applications are kicked out all the time when an optimization renderseD > >> that application useless as a measurement, maxtrix300, etc. and4 > >> Spec2000's purpose is to measure CPU strength). > >CJ > > But instead you'd rather just spread more FUD.  Typical, Rob:  make anJ > > accusation without a shred of evidence just because you don't like the tale > > the numbers tell.t > >o >i > Nope.  Try again.l  F Oh, are you saying that you *do* have evidence?  Could you provide it?   >r > > .... > >uH > >> > FUD is insulting regardless of its source:  it indicates that the > > FUDster-I > >> > believes his audience is sufficiently gullible to be misled by it. 	 > > While-K > >> > the c.o.v. audience may well incline toward Alpha jingoism, I'd liker toG > >> > think it still has the ability to distinguish fact from fiction.  > >> > >>F > >> True.  And collectively we have been deflecting Andrew's FUD long > >> before you showed up. > >k* > > Since I showed up at DEC in early 1976 >a > comp.os.vms, not DEC.   E It's convenient to be able to 'clarify' your statements after they'resI questioned, I guess.  But preferable to say what you mean the first time.s   - bill   ------------------------------  % Date: Mon, 20 May 2002 19:07:09 -0400i- From: JF Mezei <jfmezei.spamnot@videotron.ca>nM Subject: Re: Most unusual resident in a system ?, was: Re: No new Alpha sales0, Message-ID: <3CE9818D.5BCDC6DA@videotron.ca>  ' I noticed dust bunnies on my VAX 3100.    J Is there a Digital manual on the care and feeding of dust bunnies ? If so, what is its order number ?  0 I don't want to kill them if I don't have to....   ------------------------------    Date: 20 May 2002 21:10:48 -0600+ From: young_r@encompasserve.org (Rob Young)gM Subject: Re: Most unusual resident in a system ?, was: Re: No new Alpha sales 3 Message-ID: <ppGoULVPAa9+@eisner.encompasserve.org>r  n In article <Y8fG8.66692$e66.6083222@bin6.nnrp.aus1.giganews.com>, "Bill Todd" <billtodd@metrocast.net> writes: >    	More Sun CPU weakness:e  N http://www.ideasinternational.com/benchmark/Oracle/stand%20V11i.html#Sorted by	 NotesMarkr  I Rank System User Count Average Response Time (sec) Benchmark Version Dateu
 Submitted   , 5 Sun Fire V880 3416 1.900 11.5.3 30-Nov-01 * 6 Compaq ES45 3304 1.660 11.5.3 14-Mar-02     3 	Running the same version of the Oracle applicationS; 	suite and 8 UltraSparc III CPUs manage to barely squeek by  	4 Alpha CPUs.  E 	There are countless other examples cross-platform, whether comparing E 	IBM, Intel, HP, or Alpha CPUs.  There are several Sun strengths too.eD 	It is still one of the weakest, and will probably stay that way, ifB 	it doesn't disappear altogether.  You have a picture of a company 	in total disarray internally:  - http://www.eetimes.com/story/OEG20020510S00621  N Sun Microsystems Inc. recently announced its president, Ed Zander, will retireK in July. He joins several other execs, including the head of Sun's computerdN systems unit, who will leave and not be replaced. Another Sun executive, AnantO Agarwal, who helped build the company's Sparc microprocessor business, left the O company quietly a month ago when a new unit he was trying to establish, focusedeL on communications and security, failed to get sufficient corporate funding.     ? 	Here is an official prognastication.  Sun adopts IA64 in a big I 	way.  UltraSparc V will never happen.  They certainly can't go to AMD.  nF 	AMD is going to have their oxygen supply ripped out of them over the  	next year:c    4 http://news.com.com/2100-1040-917021.html?tag=fd_top  $ Intel chipsets bring lower-cost PCs   O The chipsets also offer built-in graphics and newer features, such as UniversalrM Serial Bus 2.0. A system for connecting devices like cameras to a PC, USB 2.0oM supports faster data transfer, speeding up tasks such as downloading pictures  from a digital camera. e  O The integration of the graphics chip, while it might sound like small potatoes,rO is a big event in the component world. Typically, a low-end graphics card costs L between $25 and $30. By removing the chip, PC makers can eliminate about $50O from the end price of a PC. Manufacturers can use that savings to cut prices or G to add more memory or better components while keeping prices the same. t    D 	It isn't just the CPU.  Intel will suck everything into the chipset7 	that can go there.  AMD doesn't have much of a chance.s     >> >>a >> >>OG >> >> True.  And collectively we have been deflecting Andrew's FUD long: >> >> before you showed up.e >> >+ >> > Since I showed up at DEC in early 1976e >> >> comp.os.vms, not DEC. > G > It's convenient to be able to 'clarify' your statements after they're K > questioned, I guess.  But preferable to say what you mean the first time.  >    	"deflecting Andrew's FUD"  A 	About when did that begin?  1976 or much later?  And where would-C 	that have taken place?  But now we are getting to the pissy stage.r   				Robo   ------------------------------  # Date: Tue, 21 May 2002 04:27:49 GMTt* From: "Bill Todd" <billtodd@metrocast.net>M Subject: Re: Most unusual resident in a system ?, was: Re: No new Alpha salesm@ Message-ID: <91kG8.53519$th.5156257@bin2.nnrp.aus1.giganews.com>  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:ppGoULVPAa9+@eisner.encompasserve.org...eI > In article <Y8fG8.66692$e66.6083222@bin6.nnrp.aus1.giganews.com>, "Billt& Todd" <billtodd@metrocast.net> writes: > >a >r > More Sun CPU weakness: >a > K http://www.ideasinternational.com/benchmark/Oracle/stand%20V11i.html#Sorted  by > NotesMark9 > K > Rank System User Count Average Response Time (sec) Benchmark Version Date  > Submitted1 > - > 5 Sun Fire V880 3416 1.900 11.5.3 30-Nov-01o+ > 6 Compaq ES45 3304 1.660 11.5.3 14-Mar-02o >y > 4 > Running the same version of the Oracle application< > suite and 8 UltraSparc III CPUs manage to barely squeek by > 4 Alpha CPUs.o  L Of course, you neglect to mention that the ES45 was running the latest (1001H MHz) Alphas while the Sun box was running 750 MHz USIIIs rather than theK current top-of-the-line 1050 MHz USIIIs.  700ish MHz Alphas would have beenI; about right if you wanted an apples-to-apples comparison...    >iF > There are countless other examples cross-platform, whether comparingF > IBM, Intel, HP, or Alpha CPUs.  There are several Sun strengths too.E > It is still one of the weakest, and will probably stay that way, ifBC > it doesn't disappear altogether.  You have a picture of a company  > in total disarray internally:) >u/ > http://www.eetimes.com/story/OEG20020510S00624  H Er, I think you may have provided the wrong URL above:  the company that6 appears to be in major disarray in that article is HP.  = As for Sun's grievous straits, try this one (watch for wrap):h  L http://www.silicon.com/bin/bladerunner?30REQEVENT=&REQAUTH=21046&14001REQSUB =REQINT1=53435  E 17% year-over-year shipment growth in a down market, no less:  64% ofeL *total* Unix shipments and 54% of *total* Unix revenue (i.e., more shipmentsI and revenue than all the other Unix competition combined) on 3.3% revenue  growth.P  F Oh, wait:  Windows is going to rip out the soft underbelly of the Unix market.  Right.r   ...i  @ > Here is an official prognastication.  Sun adopts IA64 in a bigH > way.  UltraSparc V will never happen.  They certainly can't go to AMD.F > AMD is going to have their oxygen supply ripped out of them over the > next year:  G Ah.  Another prognostication along the lines of "Wildfire will blow theBF doors off everything in sight!" and "McKinley may well debut with evenI faster than 1.4 GHz clocking!"  Thanks, but I think most people will look & for a seer with a better track record.  E Meanwhile, let's see how Hammer adoption goes:  certainly seems to ber? gathering more and more steam these days (unlike, say, Itanic).l   >  >p6 > http://news.com.com/2100-1040-917021.html?tag=fd_top >e% > Intel chipsets bring lower-cost PCs  >h+ > The chipsets also offer built-in graphics   L Just like those Cyrix (IIRC) processors that took over the world a few years ago.  &  and newer features, such as UniversalK > Serial Bus 2.0. A system for connecting devices like cameras to a PC, USBg 2.0uF > supports faster data transfer, speeding up tasks such as downloading pictures > from a digital camera.  C Wow!  That'll sure be a must-have for the server market, all right.    > G > The integration of the graphics chip, while it might sound like small 	 potatoes,sK > is a big event in the component world. Typically, a low-end graphics card. costsdJ > between $25 and $30. By removing the chip, PC makers can eliminate about $50sG > from the end price of a PC. Manufacturers can use that savings to cute	 prices orhH > to add more memory or better components while keeping prices the same.  G Hoo, hah!  Can't hardly wait.  Sun and IBM servers are toast!  (Alphas,s too.)r   >  >cE > It isn't just the CPU.  Intel will suck everything into the chipsete8 > that can go there.  AMD doesn't have much of a chance.  K Of course not:  they have nowhere near the integration resources that CyrixaL had.  Assuming that consumers will adopt the on-chip video, etc., route withF the same rampant enthusiasm they've exhibited for, say, on-board audio (not).  L Do you have any idea of the range of just video cards that exists out there?L That's because of competition that depends on product cycles about 1/10th asK long as processor product cycles:  come out with even high-end video (whicheJ would have its own economic drawbacks) on the processor chip, and it'll beF obsolete within a few months (and even Intel doesn't really want to be) changing its processor masks that often).l  K And while lowest-common-denominator video might fly, there's no real profita0 down there.  No win there, Rob:  clueless again.   - bill   ------------------------------  % Date: Mon, 20 May 2002 15:17:10 -04007- From: JF Mezei <jfmezei.spamnot@videotron.ca>6 Subject: Re: No new Alpha salesw, Message-ID: <3CE94BB6.25400345@videotron.ca>  ( Andrew Harrison SUNUK Consultancy wrote:5 > If the latest market data for the UK (emerging fromp6 > recession) are indicative then Alpha is worse placed4 > than any other platform in the datacenter space to > benefit from the upswing.t  L Andrew, it is a given that Alpha is dying. It was dealt a rather deadly blowC last June 25, and it is no surprise that sales would be going down.i  N I think you may see spikes whenever they make one sale and they can then claimK huge growth in sales, but this is somewhat similar to Tandem who can go for K month with very little in sales and then they get one big sale and it makes 0 Tandem look really really good for that quarter.   ------------------------------   Date: 21 May 2002 00:09:08 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)L Subject: Re: OpenVMS (not) on DS20L (was: re: Sayonara DS10, new org charts)* Message-ID: <acc374$4j4$4@web1.cup.hp.com>  \ In article <87r8klqqw3.fsf@prep.synonet.com>, Paul Repacholi <prep@prep.synonet.com> writes:< :Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes: : O :> > Interesting that thew new DS20 has a SCISI controller  with a mention "forr :> > Linux only".n :  :> This is the DS20L.a :                 ^i4 :> > Can anyone care to explain what this is about ? : L :> There was a thread here a while back about this being a very specialised B :> box and not supported on VMS---or do you have another question? :a/ :So is this a Not-Blue box, or a Not-White box?b  M   The AlphaServer DS20L is a box specifically targeting a particular market, kK   and the folks in that market wanted a particular operating system on the u   box.    I   OpenVMS is supporting the AlphaServer DS20E, AlphaServer DS10 (both the6J   older box and the newer one with the front-mount disks) and AlphaServer J   DS10L series, and planning to support the dual-processor variant of the    EV7 systems.    K   Write HP a check for enough of these boxes specifically running OpenVMS, vD   and I'm certain we can port OpenVMS over to the AlphaServer DS20L.  I   (The AlphaServer DS20L was discussed at length in this newsgroup, and I J   am certain that even a cursory search will turn up some of the threads.)  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Tue, 21 May 2002 01:44:48 GMTb1 From: "David J. Dachtera" <djesys.nospam@fsi.net> L Subject: Re: OpenVMS (not) on DS20L (was: re: Sayonara DS10, new org charts)' Message-ID: <3CE9A9D5.8B19EB24@fsi.net>e   Hoff Hoffman wrote:l > [snip]L >   Write HP a check for enough of these boxes specifically running OpenVMS,F >   and I'm certain we can port OpenVMS over to the AlphaServer DS20L.  E O.k. I'll bite - why does OpenVMS-Alpha need to be ported to an Alphar machine?   -- R David J. Dachterau dba DJE Systemss http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/.   ------------------------------  % Date: Mon, 20 May 2002 23:02:06 -0400e2 From: rdeininger@mindspring.com (Robert Deininger)L Subject: Re: OpenVMS (not) on DS20L (was: re: Sayonara DS10, new org charts)L Message-ID: <rdeininger-2005022302060001@11cust170.tnt2.nashua.nh.da.uu.net>  ; In article <3CE9A9D5.8B19EB24@fsi.net>, "David J. Dachtera"u <djesys.nospam@fsi.net> wrote:   >Hoff Hoffman wrote:	 >> [snip]bM >>   Write HP a check for enough of these boxes specifically running OpenVMS,vG >>   and I'm certain we can port OpenVMS over to the AlphaServer DS20L.2 >aF >O.k. I'll bite - why does OpenVMS-Alpha need to be ported to an Alpha	 >machine?   E Because every alpha system, or family of systems, requires a layer of J software that understands the hardware details.  The SRM console covers upD most of the hardware oddities, but some remain.  (In the case of theF DS20L, the SRM console has somewhat different origins from other alphaI systems as well.)  So VMS needs this layer of hardware-dependent softwareu8 to boot and run.  It does not currently exist for DS20L.  J Most of this code is packaged in SYS$CPU_ROUTINES_xxyy.EXE, with, I think,G a teeny bit in APB.EXE.  If you have access to the source listings, you/F will notice that the CPU-specific code often contains work-arounds forG certain platform "features" that were not it the specs.  I would expect0H that the DS20L "features" would be mostly new and interesting, since the) box came from a different hardware group.r  J So the box needs one image in VMS, probably some firmware additions, and aE qualification cycle.  The qual would probably uncover new "features".n  H There are probably better 2P alpha VMS systems on the roadmap for almostC all applications, which leaves the DS20L a very small market niche.    ------------------------------  # Date: Mon, 20 May 2002 20:46:06 GMTn' From: Rick Dyson <Rick-Dyson@UIowa.EDU>g/ Subject: OpenVMS v7.3 and Password Dict Problem ) Message-ID: <3CE9608E.8977597A@UIowa.EDU>t  L I have a couple v7.3 boxes (ALP) I have recently installed from scratch, andN then fully patched (I believe!).  I went to update my password recently on oneE of them and it would not let me use the value I entered!  It claimed:o  N %SYSTEM-F-PWDWEAK, password is too easy to guess; please choose another string  H On another v7.3 box I have it did not complain at all!  I was successfulG in setting it to the same value on it.  I have checked both systems andoE the UAF entry for the account names DO NOT have "/Flags = DisPwdDic".   P Does anyone know what might be different with this one node?  These systems are N generally new and mostly stock installs for testing purposes.  Installed apps,= compilers, etc. but nothing very special that I can think of.   N I have specifically searched the Sys$Library:VMS$Password_Dictionary.data fileL for the password I tried to use and it was not found.  I also searched for aL few "obvious" segments of the password to see if it tested some combinationsN and one word was found in the dictionary.  It is also in the dictionary on the" node that let me use the password.  L Neither node has had the VMS$Password_Dictionary.data modified by me (that I/ am aware of!).  Unless it was done by some ECO.'   Rick   ------------------------------  # Date: Mon, 20 May 2002 21:19:47 GMTp' From: Rick Dyson <Rick-Dyson@UIowa.EDU> 3 Subject: Re: OpenVMS v7.3 and Password Dict Problem ) Message-ID: <3CE96873.ADDCC970@UIowa.EDU>e   Rick Dyson wrote:e > N > I have a couple v7.3 boxes (ALP) I have recently installed from scratch, andP > then fully patched (I believe!).  I went to update my password recently on oneG > of them and it would not let me use the value I entered!  It claimed:e > P > %SYSTEM-F-PWDWEAK, password is too easy to guess; please choose another string > J > On another v7.3 box I have it did not complain at all!  I was successfulI > in setting it to the same value on it.  I have checked both systems andeG > the UAF entry for the account names DO NOT have "/Flags = DisPwdDic".   B 	I was reminded by someone that combinations of some words and theK system nodename are tested and it turns out this combination would apply iny@ this case.  I was too close and missed the forest for the trees!   Rick   ------------------------------  + Date: Tue, 21 May 2002 00:20:01 +0000 (UTC)7, From: lewis@PROBE.mitre.org (Keith A. Lewis)K Subject: Re: Other HP printers and USB (was: Newest version of DCPS and oure. Message-ID: <acc3rh$h2q$1@newslocal.mitre.org>   JF Mezei <jfmezei.spamnot@videotron.ca> writes in article <3CE97E59.1083C088@videotron.ca> dated Mon, 20 May 2002 18:53:27 -0400:,< >How many years away is VMS-capable hardware with USB port ?   Since 1999 at least.  	 PWS 500au- DS20E- DS10L    to name a few.    D >Frankly, I don't see  GS class machines with a USB port on them :-)  1 Sorry I don't have any of those so I can't check.e  N >Considering cable limitations of USB, how many would actually be able to plug% >in a printer near enough a server ? 1  G If you can afford a "server", you can afford a printer with an ethernetuJ interface.  :^)  Seriously, the only reason a USB printer would be cheaperK than 10baseT is the manufacturer doesn't have to implement a tcp/ip stack.  7 Cable routers have those, and they're under $100 now.  i  + >> 2. There is a trend towards Postscript. E > L >is there really ? I though it was the opposite, just like a trend away from& >SCSI and towards the cheap IDE stuff.  K When Apple ditches Postscript, I'll start to worry.  And I really don't seeaI that happening.  They got out of SCSI because something better came alongTI (Firewire), but it's hard to see why you'd want to invent another printerw* language when you could extend Postscript.  N >Besiodes, is it true to state that HP doesn't have real postscript, just some >imitation PS ?w  F It's an open standard; HP has as much of a shot at getting it right as anybody.  + --Keith Lewis              klewis$mitre.org > The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Tue, 21 May 2002 01:11:22 +0200A From: Dirk Munk <munk@home.nl>U Subject: Re: Other HP printers and USB (was: Newest version of DCPS and our  problem)n& Message-ID: <3CE9829A.2040903@home.nl>   JF Mezei wrote:i   >Dirk Munk wrote:  >l= >>Indeed , I would like to use a Postscript printer over USB.i >> >n< >How many years away is VMS-capable hardware with USB port ? >nG It seems USB cards with a Lucent or a Symbios chipset have been tested h with VMS 7.3-1. G Look at http://www.asdtech.com for their Quad card for an example of a 1 Lucent chipset card.  @ Modems, mice, keyboards, printers have been tested and approved.   >u >-D >Frankly, I don't see  GS class machines with a USB port on them :-) >t= You could be wrong !! It will be there, just as serial ports.r   >s > N >Does anyone know if HP intends to produce Alpha workstations until IA64 stuffI >is viable ? Would they bother adding USB support to Alpha workstations ?  >l3 Not only workstations, also small 'normal' systems.T   >l > N >Considering cable limitations of USB, how many would actually be able to plug% >in a printer near enough a server ? e >p" 5 meters, and there are extenders.   >s >n* >>2. There is a trend towards Postscript.  >> >cL >is there really ? I though it was the opposite, just like a trend away from& >SCSI and towards the cheap IDE stuff. >aE All high-end printing / type setting is done in Postscript. PDF is a OG kind of Postscript. Postscript is being developed all the time, PCL is :K quite dead in that regard afaik. Postscript can do things PCL can never do.1     >> >>N >Besiodes, is it true to state that HP doesn't have real postscript, just some >imitation PS ?r > C They don't have the Adobe engine, that's true. But other Postcript   engines can be very good too !!'     >  >AO >It seems to me that HP is pushing HPCL far more, when you consider the volumesn) >of low cost HPCL-only printers it ships.y >aH You're wrong there. Take a good look at the printer specs, and you will H see that most of them have Postscript and PCL. Just as with Lexmark and  other companies.   >i > M >Don't get me wrong, I love postscript, but I seem to be cursed because every2M >technology I choose to become proficient in seems to quickly lose popularitys4 >once my skills become good enough to market myself. >n   ------------------------------  % Date: Tue, 21 May 2002 01:14:54 +0200d From: Dirk Munk <munk@home.nl>U Subject: Re: Other HP printers and USB (was: Newest version of DCPS and our  problem)m& Message-ID: <3CE9836E.6090104@home.nl>   Dirk Munk wrote:   >k >s > JF Mezei wrote:a >R >> Dirk Munk wrote:n >>? >>> Indeed , I would like to use a Postscript printer over USB.1 >>>  >>> >> How many years away is VMS-capable hardware with USB port ? >>I > It seems USB cards with a Lucent or a Symbios chipset have been tested I > with VMS 7.3-1.eI > Look at http://www.asdtech.com for their Quad card for an example of a e > Lucent chipset card. e    $ Sorry, typo: http://www.adstech.com/   >n > B > Modems, mice, keyboards, printers have been tested and approved. >. >> >>F >> Frankly, I don't see  GS class machines with a USB port on them :-) >>? > You could be wrong !! It will be there, just as serial ports.p >f >> >>F >> Does anyone know if HP intends to produce Alpha workstations until 
 >> IA64 stuff K >> is viable ? Would they bother adding USB support to Alpha workstations ?d >>5 > Not only workstations, also small 'normal' systems.n >  >> >>I >> Considering cable limitations of USB, how many would actually be able 6
 >> to plug& >> in a printer near enough a server ? >b$ > 5 meters, and there are extenders. >m >> >>+ >>> 2. There is a trend towards Postscript.g >> >>E >> is there really ? I though it was the opposite, just like a trend b >> away from( >> SCSI and towards the cheap IDE stuff. >>G > All high-end printing / type setting is done in Postscript. PDF is a %I > kind of Postscript. Postscript is being developed all the time, PCL is iD > quite dead in that regard afaik. Postscript can do things PCL can  > never do.  >w >P >> >>G >> Besiodes, is it true to state that HP doesn't have real postscript, t >> just some >> imitation PS ?n >>E > They don't have the Adobe engine, that's true. But other Postcript w! > engines can be very good too !!a >  >t >> >>F >> It seems to me that HP is pushing HPCL far more, when you consider  >> the volumes+ >> of low cost HPCL-only printers it ships.o >>E > You're wrong there. Take a good look at the printer specs, and you aC > will see that most of them have Postscript and PCL. Just as with   > Lexmark and other companies. >  >> >>B >> Don't get me wrong, I love postscript, but I seem to be cursed  >> because everyE >> technology I choose to become proficient in seems to quickly lose a
 >> popularityp6 >> once my skills become good enough to market myself. >> >    ------------------------------  % Date: Mon, 20 May 2002 23:23:02 +0200n From: Dirk Munk <munk@home.nl>P Subject: Other HP printers and USB (was: Newest version of DCPS and our problem)& Message-ID: <3CE96936.8000602@home.nl>  C A bit of another topic, but since you signed with "hp.com" in your   address ...e  A Will DCPS support all HP postscript printers, incl. Inkjet etc. ?rF And will we be able to use a supported USB interface as well for DCPS  (from VMS 7.3-1)?    Regards,  	 Dirk Munk    Paul Anderson wrote:  0 >[[ This message was both posted and mailed: see< >   the "To," "Cc," and "Newsgroups" headers for details. ]] >rH >In article <aus-92BCBE.23015917052002@wrzx08.rz.uni-wuerzburg.de>, Hans- >Magnus Aus <aus@vim.uni-wuerzburg.de> wrote:C >aG >>What is the newest version of DCPS ? We currently use DCPS 1.8, OVMS y? >>7.2-1, TCPIP 5.1 and our own home brew   HP4050$DEVCTL plus   ) >>"ip_rawtcp/n.n.n.n:9100" on the hp4100.e >> > A >The latest released version is V2.0.  V2.1 is now in field test.s > E >>We're sporadically plagued by hung DCPS queues which look like the c >>listing below. >>J >>1) Could the problem be that the DCPS queue is told that the printer is K >>busy with a job from another PC and never gets a signal that the printer h >>is available?t >>- >>2) Does the newest DCPS solve this problem?  >> >>$show queue:F >> Printer queue HP4100_1, busy, on V99::"ip_rawtcp/141.27.96.6:9100",- >>  mounted form DCPS$DEFAULT (stock=DEFAULT)c >>6 >>  Entry  Jobname         Username     Blocks  Status6 >>  -----  -------         --------     ------  ------8 >>    707  B2002011184_7   SCHUBERT          4  Printing7 >>    708  B2002011185_7   SCHUBERT          4  Pendingt7 >>    703  B2002011082_7   SCHUBERT          4  Holdinge7 >>    704  B2002011084_7   SCHUBERT          4  Holdingl >> >eE >You're currently running the LaserJet 4100 printer as an unsupported C >printer, but this shouldn't cause the problem you are seeing.  I'd:E >suggest you run the DCPS V2.1 field test to see if the problem stillR >occurs. >h( >I'll send you the field test agreement. >o >Paul> >    ------------------------------  # Date: Mon, 20 May 2002 21:51:07 GMT 0 From: Paul Anderson <paul.r.anderson@compaq.com>T Subject: Re: Other HP printers and USB (was: Newest version of DCPS and our problem); Message-ID: <200520021745435206%paul.r.anderson@compaq.com>a  F In article <3CE96936.8000602@home.nl>, Dirk Munk <munk@home.nl> wrote:  C > Will DCPS support all HP postscript printers, incl. Inkjet etc. ?w  G We've made no decisions about expanding DCPS support from the networkeduD laser lines (Color LaserJet and LaserJet) that we now support.  OnceG DCPS V2.1 ships, we'll review what to do and add for the next version.  D Expanding support to as many HP printers as possible certainly makes# sense now that DCPS is HP software.r  H > And will we be able to use a supported USB interface as well for DCPS  > (from VMS 7.3-1)?a  E Although DCPS Engineering was consulted in the original USB design by B OpenVMS Engineering a few years ago, there are no plans to add USBG support to DCPS.  We are unsure if there is a demand for such a thing. s7 Are you asking because you have a need for DCPS on USB?a   Paul   -- o  Paul Anderson   OpenVMS Engineeringo   Hewlett-Packard CompanyD   ------------------------------  % Date: Tue, 21 May 2002 00:07:26 +0200a From: Dirk Munk <munk@home.nl>T Subject: Re: Other HP printers and USB (was: Newest version of DCPS and our problem)& Message-ID: <3CE9739E.7060601@home.nl>  ; Indeed , I would like to use a Postscript printer over USB.l   Consider the following:e  G 1. Laser and Inkjet printers get cheaper all the time, but the network wG variants are still relatively expensive. Sometimes you need a seperate p? expensive printserver. (hobbiest would like to print to :-)  ).eF 2. There is a trend towards Postscript. All HP laser printers (except F the cheapest pure Windows printers) have Postscript now as a standard G feature. It used to be a rather expensive option. PCL is at its end it i seems.C 3. Serial and parallel connections are superseded by USB, many new l6 printers don't even have a parallel connector anymore.I 4. DCPS did support serial connections, so migrating to USB is a natural a
 way to go.H 5. There is printer support in the new USB drivers for VMS 7.3-1 I have 
 been told.   Regards,  	 Dirk Munk            Paul Anderson wrote:  G >In article <3CE96936.8000602@home.nl>, Dirk Munk <munk@home.nl> wrote:p > C >>Will DCPS support all HP postscript printers, incl. Inkjet etc. ?a >> >nH >We've made no decisions about expanding DCPS support from the networkedE >laser lines (Color LaserJet and LaserJet) that we now support.  Once H >DCPS V2.1 ships, we'll review what to do and add for the next version. E >Expanding support to as many HP printers as possible certainly makes $ >sense now that DCPS is HP software. >tH >>And will we be able to use a supported USB interface as well for DCPS  >>(from VMS 7.3-1)?n >> >pF >Although DCPS Engineering was consulted in the original USB design byC >OpenVMS Engineering a few years ago, there are no plans to add USBcH >support to DCPS.  We are unsure if there is a demand for such a thing. 8 >Are you asking because you have a need for DCPS on USB? >- >Paul- >-   ------------------------------    Date: 20 May 2002 20:11:03 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)RT Subject: Re: Other HP printers and USB (was: Newest version of DCPS and our problem)3 Message-ID: <OF8YdI3NOeIX@eisner.encompasserve.org>.  G In article <3CE9739E.7060601@home.nl>, Dirk Munk <munk@home.nl> writes:y  H > 2. There is a trend towards Postscript. All HP laser printers (except H > the cheapest pure Windows printers) have Postscript now as a standard I > feature. It used to be a rather expensive option. PCL is at its end it   > seems.   That is great news !   ------------------------------  % Date: Mon, 20 May 2002 18:53:27 -0400r- From: JF Mezei <jfmezei.spamnot@videotron.ca>aY Subject: Re: Other HP printers and USB (was: Newest version of DCPS and our problem) probo, Message-ID: <3CE97E59.1083C088@videotron.ca>   Dirk Munk wrote: > = > Indeed , I would like to use a Postscript printer over USB.l  ; How many years away is VMS-capable hardware with USB port ?o  C Frankly, I don't see  GS class machines with a USB port on them :-)c  M Does anyone know if HP intends to produce Alpha workstations until IA64 stuffEH is viable ? Would they bother adding USB support to Alpha workstations ?  M Considering cable limitations of USB, how many would actually be able to plugh$ in a printer near enough a server ?   * > 2. There is a trend towards Postscript.   K is there really ? I though it was the opposite, just like a trend away from % SCSI and towards the cheap IDE stuff.a  M Besiodes, is it true to state that HP doesn't have real postscript, just some  imitation PS ?  N It seems to me that HP is pushing HPCL far more, when you consider the volumes( of low cost HPCL-only printers it ships.  L Don't get me wrong, I love postscript, but I seem to be cursed because everyL technology I choose to become proficient in seems to quickly lose popularity3 once my skills become good enough to market myself.3   ------------------------------  % Date: Tue, 21 May 2002 01:43:08 +0100e From: nic <junk@127.0.0.1>/ Subject: Re: Please help: Hobbes failed to boot ) Message-ID: <3CE9981C.20EE0B99@127.0.0.1>    Scott Squires wrote: > E > Hobbes is a MicroVAX 3100/40, running VMS 7.2.  I rebooted him, andoD > when he came back up, OPCOM was spitting out this message over and > over:i > 7 > %%%%%%%%%%  OPCOM  19-MAY-2002 19:49:06:11 %%%%%%%%%%e) > Message from use AUDIT$SERVER on HOBBEStD > Security alarm (SECURITY) and security audit (SECURITY) on HOBBES, > system id: 1025a5 > Auditable event:     Detached process login failure . > Event time:          19-MAY-2002 19:49:06:08 > PID:                 20400078  > Process name:        EVL > Username:            DECNETt > Process owner:       [SYSTEM]eF > Image name:          $1$DKB300:[SYS0.SYSCOMMON.][SYSEXE]LOGINOUT.EXEB > Status:              %LIB-F-ACTIMAGE, error activating image !AS  H Amazing. Had this same issue with a system just before I left work for a
 few days.   C Boot minimum and rename AUDIT$SERVER.DAT. Reboot full, reboot againn full.o  D Should be OK. I think what happens is this file gets corrupted. I'veB seen it on an Alpha where it is documented in DSNlink, but you see3 similar symptoms with a different pattern on a VAX.-  ? Another troubleshooting tip is to get STARTUP_P2 to "V" or "VD"n  (unfortunate!) to log to a file.   -- e Regards, Nic Clews (from home), nic at python dot demon dot co dot uk (play) nclews at csc dot com (work)   ------------------------------  # Date: Tue, 21 May 2002 04:06:54 GMT,) From: rob.buxton@wcc.govt.nz (Rob Buxton) 6 Subject: Re: Queue info via Web Server (Apache & WASD)/ Message-ID: <3ce9c6dc.6274191@news.wcc.govt.nz>t  C On 20 May 2002 06:29:47 GMT, ccburgess@uqstu.jdstory.uq.edu.au (Iana Burgess) wrote:   ^ >In article <3ce850bf.444120842@news.wcc.govt.nz>, rob.buxton@wcc.govt.nz (Rob Buxton) writes:	 >>Hi All,e >>F >>I've got a Web Page for our Frontline Staff that allows them to show= >>the full details of Printer Queues (either selected or all)m >>C >>The problem I have is that the /Description label done during the C >>Queue Init places the description within a couple of <> brackets.s >>8 >>So, on VMS show queue /full shows the description like >><Queue For Rob>- >>B >>Alas, when this comes to display on the Web Page the Description6 >>disappears, because it sees it as an HTML statement. >>: >>So the nice description gets translated to a blank line! >>H >>I've tried with <PRE> and without but the display doesn't cough up the >>description. >>F >>I'm sending this via a GET then just running a Command Procedure via >>CGI. >> >>Any hints? >> >>TIA as always. >> >>Rob. >  >How about this? >a1 >  $ Set queue fred /description=">Fred Printer<"  >   > >Show queue, and your browser, will both show <>Fred Printer<>  D That's a nice bit of lateral thinking! And it does indeed display as you've suggested.t   Rob. >,
 >Ian Burgess   >University of Queenslandi >I.Burgess[at]its.uq.edu.auv >www.its.uq.edu.au   ------------------------------    Date: 20 May 2002 18:57:29 -0700- From: contracer11@uol.com.br (Shiva MahaDeva)n$ Subject: Reading unreadable tapes...= Message-ID: <ddf392ea.0205201757.1b8b0913@posting.google.com>a  0 Many times, in my job, when I issue the command:8 $ backup/log/noassi/sele=(login.com) mka500:bkp110502 [] I get this message:o  -error positioning mka500:    parity error , (or)    excessive error , etc...  C Id like to know if is there any software to enable access data in I fault DAT tapes...	 Thanks...i   ------------------------------  # Date: Tue, 21 May 2002 02:26:07 GMT)1 From: "David J. Dachtera" <djesys.nospam@fsi.net>t( Subject: Re: Reading unreadable tapes...' Message-ID: <3CE9B384.BBC179D5@fsi.net>o   Shiva MahaDeva wrote:v > 2 > Many times, in my job, when I issue the command:: > $ backup/log/noassi/sele=(login.com) mka500:bkp110502 [] > I get this message:t >  -error positioning mka500:r >   parity error , (or)  >   excessive error , etc... > D > Id like to know if is there any software to enable access data in > fault DAT tapes... > Thanks...m  F DATs are known to be fragile and especially to stretch easily. If thisA is your enterprise backup solution, you should seriously consider  investing in DLT.   H ...IMHO, YMMV, no purchase necessary, void where prohibited by law, someG restrictions apply, see store display for details, some assembly (or C)SG required, batteries not included, subject to change without notice, ...b   -- t David J. Dachtera1 dba DJE Systems/ http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/b   ------------------------------  # Date: Tue, 21 May 2002 04:42:38 GMT') From: Andrew Balaam <abalaam@yahoo.co.uk> ( Subject: Re: Reading unreadable tapes...4 Message-ID: <20020521.4423800.2276748688@imagnu.geo>  I If you have the disk space, you could try mounting the tape, and using=20iE the copy command to copy the save set to disk, then do a selective=20lI restore from the copy. If there is a problem copying the save set, try=20iG again (and again...), and if you're lucky you may get the data back.=20 I Don't forget to run a cleaning tape through first. It may be a good idea=  =20 I to run a cleaning tape after the second re-try, if it still won't read=20IC the tape. If the tape is old, it may be shedding some of its oxide.o  % Set def [somewhere.with.enough.space]e# mount/over=3Did/block=3D8192 mka500  copy mka500:*.* []/log   Geoff.  6 >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<  G On 21/05/02, 02:57:29, contracer11@uol.com.br (Shiva MahaDeva) wrote=20e& regarding Reading unreadable tapes...:    2 > Many times, in my job, when I issue the command:< > $ backup/log/noassi/sele=3D(login.com) mka500:bkp110502 [] > I get this message:s >  -error positioning mka500:o >   parity error , (or)I >   excessive error , etc...  F > I=B4d like to know if is there any software to enable access data in > fault DAT tapes... > Thanks...g   ------------------------------  % Date: Mon, 20 May 2002 23:34:18 -0500 , From: "Rich Jordan" <rjordan@mindspring.com>( Subject: Re: Reading unreadable tapes...2 Message-ID: <accihd$e39$1@slb0.atl.mindspring.net>  E Get a good, preferably new DAT cleaning tape for DATA (not audio) DATsL drives.  Run it through the drive three or four times in sequence.  Then tryF reading your tape again.  If that fails, try it on another system/tapeK drive.  We've found that there is often enough variance between tape driveseL to sometimes allow a failing tape to be read on a different drive (though ofJ course that can lead to apparently good tapes that can only be read on the3 original drive and nowhere else... the joys of DAT)o    # Shiva MahaDeva wrote in message ... 1 >Many times, in my job, when I issue the command:e9 >$ backup/log/noassi/sele=(login.com) mka500:bkp110502 []  >I get this message: > -error positioning mka500: >  parity error , (or) >  excessive error , etc...  >dC >Id like to know if is there any software to enable access data int >fault DAT tapes... 
 >Thanks...   ------------------------------  % Date: Mon, 20 May 2002 22:40:49 -0400>2 From: rdeininger@mindspring.com (Robert Deininger)* Subject: Re: Sayonara DS10, new org chartsL Message-ID: <rdeininger-2005022240500001@11cust170.tnt2.nashua.nh.da.uu.net>  C In article <3CE90603.71B5E82D@clarityconnect.com>, "Mark D. Jilson">! <jilly@clarityconnect.com> wrote:n  C >The DS20L is a box that was built and targeted for a very specificsE >market and customer.  If someone has a requirement for a boatload ofsE >systems running VMS for which the DS20L would be an exact fit I sure G >that your sales person would be happy to get folks involved that could - >discuss the cost to qualify VMS on this box.g  E In the case of the DS20L, it's not just a question of qualifying VMS.t  E This box is sufficiently different from the others that VMS has never I booted on it, AFAIK.  Non-trivial coding would have to be done to supporto the box.  I Yes, a customer with a big enough checkbook could change this, but so fart none have come forward.    ------------------------------  # Date: Tue, 21 May 2002 03:58:47 GMTt) From: rob.buxton@wcc.govt.nz (Rob Buxton)s Subject: Re: scsi clusteri/ Message-ID: <3ce9c5ea.6032093@news.wcc.govt.nz>n  E On Sun, 19 May 2002 21:00:37 -0400, rdeininger@mindspring.com (Robert! Deininger) wrote:   P >In article <3ce8448f.441000284@news.wcc.govt.nz>, rob.buxton@wcc.govt.nz wrote: > H >>On Thu, 16 May 2002 11:00:51 GMT, "John N." <JNixon@cfl.rr.com> wrote: >>I >>>Trying to create a scsi cluster without documentation. (thru trial andv
 >>>error). >>>Openvms 7.2-1.i >>> G >>>I have a DS10 with one scsi controller and an AS2100 with three scsi D >>>controllers. They both have ethernet, so that will be the clusterK >>>interconnect, but I want to connect to the DS10 to the HSZ70 that is nowoI >>>connected only to the 2100.  The HSZ70 is connected to the second scsi-N >>>controller on the 2100, therefore the 2100 sees all the disks on the HSZ as
 >>>$1$dkB* >>>cO >>>But if I connect the only scsi adapter in the DS10 into the HSZ, it sees theg >>>disks as $1$dkA*0M >>>This won't work;  will it?  I would rather not re-configure the 2100, so trO >>>looks like I need to put a second scsi controller in the DS10 and connect toc1 >>>the HSZ that way.  Is this a valid conclusion?t >>" >>As far as I understand it - yes.C >>The only SCSI Cluster I set up was two AS2100s and, yes, each HSZsE >>connected to the same controller (PKA to PKA and PKB to PKB on eachp >>machine).rF >>You certainly don't want a device seen as $1$DKA0 on one machine and >>$!$DKB0 on the other!u >>E >>Others have mentioned Allocation Classes, but that is NOT the issueo* >>here. The allocation class has been set. >.K >I mentioned PORT allocation classes, which are not the same as traditionaln >node allocation classes.  > O >I'm looking at "Guidelines for OpenVMS Cluster Configurations", version 7.2-1.r >pJ >Figure 6-14 shows an example where port PKB on node EDGAR and port PKC onI >node POE are connected to the same SCSI bus.  This is achieved by givingtJ >each port the same port alloclass.  The alloclass in the example is 4, soK >the disk on the shared SCSI bus is $4$DKA500.  Note the drive gets the "A" B >controller letter, even though the host adapters are PKB and PKC. Apologies.... Rob1   ------------------------------  # Date: Tue, 21 May 2002 01:46:58 GMTa1 From: "David J. Dachtera" <djesys.nospam@fsi.net>-# Subject: Re: Stallards smoking gun!-' Message-ID: <3CE9AA58.A9CF08E6@fsi.net>    Alan Greig wrote:  > . > On Sat, 18 May 2002 16:16:02 -0400, JF Mezei' > <jfmezei.spamnot@videotron.ca> wrote:  >  > >4 > >dN > >Some time ago, there were many calls for Compaq to start to take enterpriseK > >systems seriously. Marcello then let it be known to some of us to expect P > >Compaq to change significantly and start to take enterprise systems seriouslyQ > >and that there would be a advertising campaign on TV to try to change Compaq'sdP > >image from a PC manufacturer to an enterprise company. Our hopes were all way) > >up. Imagine seing a VMS ad on TV !!!!!o > D > Agreed. Rich led us to believe we would see certain things happen.H > WIth a few exceptions they did not. One of the exceptions was the fullG > page ads for VMS appearing in leading UK trade publications featuringDH > Northern Light and I was told by senior folk within Compaq UK that theF > ads brought a noticeable response and so would be repeated.  The ads > have never run again.w  F Of course not! Heaven forbid anyone should actually make a buck off of* VMS! Can't let *THAT* cat outta the bag...   -- d David J. Dachterao dba DJE Systemsu http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/n   ------------------------------  % Date: Mon, 20 May 2002 15:04:17 -0700o% From: Dean Woodward <deanw@rdrop.com>r% Subject: Re: System manager availablee) Message-ID: <3CE972E1.902336E6@rdrop.com>g   Bill Gunshannon wrote: > 2 > In article <3CE8F798.85C87D28@blueyonder.co.uk>,9 >  Tim Llewellyn <tim.llewellyn@blueyonder.co.uk> writes:  > |> > |> > |> Paul Sture wrote: > |> >h > |> > In article <FmnXGxf21SVQ@eisner.encompasserve.org>, kuhrt@encompasserve.org (Marty Kuhrt) writes:d > |> > > In article <3CE1D332.9592C5D2@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes: > |> > >> Bill Sticker wrote: 
 > |> > >>>8 > |> > >>> I'd switch to anything if I could find a job.	 > |> > >> J > |> > >> I've often said that I wished I'd gotten that job on the garbage > |> > >> truck... > |> > >D > |> > > Which is shorthand for NT/W2K developer/system mangler?  :) > |> >L > |> > I'd prefer to be a real garbage collector. Early start, early finish,- > |> > physical exercise, can sleep at night.d > |> > > |>! > |> but can u live on the wages?  > G > My garbageman owns a house worth easily 8x what my house is worth andIG > I struggle to make the mortgage payment every month just like most of.G > my neighbors.  We may live by the rule "Garbage in - Garbage out" buteC > it looks like they live by the rule "Garbage in - Gold out".  :-(   C Like dentistry, hauling away peoples trash pays well because it's aT' nasty [and occasionally dangerous] job.i   ------------------------------  # Date: Tue, 21 May 2002 01:58:00 GMT 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>0% Subject: Re: System manager availablew' Message-ID: <3CE9ACEE.53760691@fsi.net>s   Bill Gunshannon wrote: > 2 > In article <3CE8F798.85C87D28@blueyonder.co.uk>,9 >  Tim Llewellyn <tim.llewellyn@blueyonder.co.uk> writes:  > |> > |> > |> Paul Sture wrote: > |> >h > |> > In article <FmnXGxf21SVQ@eisner.encompasserve.org>, kuhrt@encompasserve.org (Marty Kuhrt) writes:d > |> > > In article <3CE1D332.9592C5D2@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes: > |> > >> Bill Sticker wrote:B
 > |> > >>>8 > |> > >>> I'd switch to anything if I could find a job.	 > |> > >>eJ > |> > >> I've often said that I wished I'd gotten that job on the garbage > |> > >> truck... > |> > >D > |> > > Which is shorthand for NT/W2K developer/system mangler?  :) > |> >L > |> > I'd prefer to be a real garbage collector. Early start, early finish,- > |> > physical exercise, can sleep at night.  > |> > > |>! > |> but can u live on the wages?o > G > My garbageman owns a house worth easily 8x what my house is worth and G > I struggle to make the mortgage payment every month just like most ofaG > my neighbors.  We may live by the rule "Garbage in - Garbage out" butmC > it looks like they live by the rule "Garbage in - Gold out".  :-(s  E ...and I'll bet your garbage man doesn't wear a pager or get calls top0 pick up trash at all hours of the day and night.   -- x David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/2   ------------------------------    Date: 20 May 2002 16:17:58 -0700+ From: garywill@eircom.net (Gary Williamson)o3 Subject: TCP/IP intermittent dropping of connection5= Message-ID: <f0a0960f.0205201517.6f436b6b@posting.google.com>w  F We have a VMS cluster running OpenVMS V7.2-2 on Alpha V7.2 on MicroVAXA and Compaq TCP/IP Services for OpenVMS Alpha Version V5.1 - ECO 3c   COILL - AlphaServer 2100 4/200 OAK - AlphaServer 1000 4/266   GIS - MicroVAX 3100-90  C We have been experiencing intermittent TCP/IP connectivity dropout.gB Several times a day TCP/IP connectivity will be lost between a VMSC server and one or more Windows 2000 or VMS or Unix servers. This is E observable by request timeouts when pinging the VMS server from theset? other servers. The problem may only last a few minutes and theny< resolve itself or it may require a stop and start of TCP/IP.  ? One hunch that we are trying to investigate is that it could be E resulting from LPD queues that are servicing remote LPD printers over D dialup connections. We have Compaq T1010 terminals which have an LPDE server. There is a Compaq LA30W line printer attached to these. Thesee connect over ISDN dialup.o  F We do not have any local LPD printers. We have LPD service stopped and; disabled. (we have also tried with this on - no difference)-  D Errors we have received if we drop the connection when printing to a remote printer are:e    # %TCPIP-I-LPD_LOGSUC, using log file-% SYS$SPECIFIC:[TCPIP$LPD]BALLYTEST.LOGc= %TCPIP-I-LPD_SYMBRUN, symbiont is running the queue BALLYTEST ( lpd$ast_handler save_entry_number = 1518* %TCPIP-F-LPD_BADPORT, error getting a port( lpd$ast_handler save_entry_number = 1520* %TCPIP-F-LPD_BADPORT, error getting a port( lpd$ast_handler save_entry_number = 1537* %TCPIP-F-LPD_BADPORT, error getting a port    ? I can understand a queue getting messed up if the connection is D dropped while printing, but surely this shouldn't cause connectivity# to local servers to be interrupted.2  C BTW we have changed switches, network cards and cables to no avail. E Originally we had all printer queues on COILL, we have now split themSE so that all the dialup printer queues are on OAK and queus over fixed D lines (Frame Relay) are on COILL. We use TELNETSYM printing on COILLC also for some printers. We are still experienceing problems on bothm
 OAK and COILLD  C Any ideas, we are completely stumped and naturally getting a lot ofp flack from irrate users.   Thanks in advance    Gary Williamsont   ------------------------------  # Date: Tue, 21 May 2002 01:02:23 GMT - From: "John E. Malmberg" <wb8tyw@qsl.network>i7 Subject: Re: TCP/IP intermittent dropping of connection * Message-ID: <3CE9B6BA.9020304@qsl.network>   Gary Williamson wrote:H > We have a VMS cluster running OpenVMS V7.2-2 on Alpha V7.2 on MicroVAXC > and Compaq TCP/IP Services for OpenVMS Alpha Version V5.1 - ECO 3o >   > COILL - AlphaServer 2100 4/200 > OAK - AlphaServer 1000 4/266 v > GIS - MicroVAX 3100-90 > E > We have been experiencing intermittent TCP/IP connectivity dropout.cD > Several times a day TCP/IP connectivity will be lost between a VMSE > server and one or more Windows 2000 or VMS or Unix servers. This isxG > observable by request timeouts when pinging the VMS server from thesefA > other servers. The problem may only last a few minutes and theni> > resolve itself or it may require a stop and start of TCP/IP. <snip> > E > Any ideas, we are completely stumped and naturally getting a lot ofl > flack from irrate users.  H This is where having an Ethernet packet sniffer, and a line analyzer is  very useful.  ) Possible causes for such outages include:s  I Insufficient resources on the OpenVMS system.  Possibly a tuning problem eE where not enough buffers or virtual memory are available for a short t period of time.u  B IIRC: The installation manual for TCP/IP recommends that you have H 300,000 pages free in npagedyn when your system is at it's typical load.  H Also look for pagefile exhaustion, or processes near their quota limits.  E Use AUTOGEN testfiles to get a SYS$SYSTEM:AGEN$PARAMS.REPORT on your y system tuning.     Other things to check for:   Damaged cable.  G Eletromagnetic interference from a radio transmitter or a spark source.o   Broadcast storm.  7 Browser election war from misconfigured LANMAN servers.v   -Johns wb8tyw@qsl.network Personal Opinion Only-   ------------------------------  # Date: Mon, 20 May 2002 17:39:33 GMT & From: Joshua Cope <Joshua.Cope@hp.com>B Subject: Re: Traditional VMS NFS names vs. Extended Filename Parse& Message-ID: <3CE9352E.2B61E3B2@hp.com>  A TCP/IP Services does have an NFS automount feature. I'm not sure 0: about synchronizing UNIX/VMS users and passwords, though.    David Mathog wrote:  >  > Joshua Cope wrote:E > > A small, single-disk workstation running VMS 7.3-1 and TCP/IP 5.3 0 > > should run nicely in a UNIX/NFS environment. > F > Last time I looked VMS would do NFS client/server ok (not perfectly,D > because of RMS) with Unix machines,  but that's as far as it wouldI > integrate into a UNIX/NFS environment.  For instance, there was neither J > a NIS server nor a client for VMS and so no easy way to unify/centralizeK > the usernames and their passwords.  There also wasn't an automount on VMSnJ > which was slightly less serious if the users directories could be servedM > from the VMS side, but which was messy otherwise (static mounts of all user + > directories).  Has any of this changed???e > 
 > Regards, >  > David Mathog > mathog@caltech.edu   ------------------------------  # Date: Mon, 20 May 2002 19:57:28 GMTn0 From: "Matt Muggeridge" <Matt.Muggeridge@hp.com>B Subject: Re: Traditional VMS NFS names vs. Extended Filename Parse@ Message-ID: <IycG8.157592$o66.465397@news-server.bigpond.net.au>  5 Last time I looked, NIS was not on the event horizon.a  ' >and so no easy way to unify/centralizew# > the usernames and their passwordsl  F In a NIS-only environment that is true.  With External Authentication,K windows domains, and probably other mechanisms I don't know about, unifyingaL and centralizing usernames and their passwords is supported.  Unfortunately, just not with NIS.   Matt.a   --= -------------------------------------------------------------p OpenVMS TCP/IP Engineering Enterprise Computing Group Hewlett-Packard Companyi Gold Coast, AUSTRALIAe= -------------------------------------------------------------i    4 "David Mathog" <mathog@caltech.edu> wrote in message% news:3CE91610.A1CFD24D@caltech.edu...i > Joshua Cope wrote:E > > A small, single-disk workstation running VMS 7.3-1 and TCP/IP 5.3 0 > > should run nicely in a UNIX/NFS environment. >tF > Last time I looked VMS would do NFS client/server ok (not perfectly,D > because of RMS) with Unix machines,  but that's as far as it wouldI > integrate into a UNIX/NFS environment.  For instance, there was neither<J > a NIS server nor a client for VMS and so no easy way to unify/centralizeK > the usernames and their passwords.  There also wasn't an automount on VMSeJ > which was slightly less serious if the users directories could be servedH > from the VMS side, but which was messy otherwise (static mounts of all user+ > directories).  Has any of this changed???  >4
 > Regards, >i > David Mathog > mathog@caltech.edu   ------------------------------  % Date: Mon, 20 May 2002 23:33:02 -0400-, From: David Michaels <michaedi@email.uc.edu>$ Subject: Unix on VaxStation 4000 90A* Message-ID: <3CE9BFEE.2C71EF@email.uc.edu>   Hi,   $ WHich versions of Unix run on VAX???  G I would Guess Ultrix and linux/ BSD but what I was really wondering wasdD are there DecUnix versions that run on VAX and what they might be???    G Sorry, If this is off topic,, Perhaps I should have said which versionss@ of Unix can I run on a Vax with "VMS' and Multiple hard disks...       Thanks Dave   ------------------------------  % Date: Mon, 20 May 2002 23:32:23 +0200g From: Dirk Munk <munk@home.nl>= Subject: Re: vax/alpha print to hp laser printers help needede$ Message-ID: <3CE96B67.20809@home.nl>  I Sounds good, but is also worrying. It could lead to a HP-only policy, so s( no more support for non-HP printers etc.   Fabio Cardoso wrote:  . >Maybe in a few months OpenVMS 7.3-X will have4 >embedded HP drivers. Printers, CD-RW, Scanners, etc+ >...all those Software stuff for OpenVMS!!!r >n >Any idea about this ? s >a >Regards >t >FC  >s >  >T >===== >==========================- >Fbio dos Santos Cardoso0 >OpenVMS System Manager  >Rio de Janeiro - Brazil >fabiopenvms@yahoo.com.br: >==========================h >s3 >__________________________________________________t >Do You Yahoo!?u& >LAUNCH - Your Yahoo! Music Experience >http://launch.yahoo.com >    ------------------------------   Date: 21 May 2002 01:31:50 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)= Subject: Re: vax/alpha print to hp laser printers help neededd* Message-ID: <acc826$7fu$1@web1.cup.hp.com>  q In article <20020520124052.38149.qmail@web20209.mail.yahoo.com>, Fabio Cardoso <fabiopenvms@yahoo.com.br> writes:t. :Maybe in a few months OpenVMS 7.3-X will have4 :embedded HP drivers. Printers, CD-RW, Scanners, etc+ :...all those Software stuff for OpenVMS!!!t :o :Any idea about this ?    I   OpenVMS Alpha V7.3-1 will not include special embedded printer drivers pG   for HP or other (Postscript) printers, DCPS -- which is licensed witheG   OpenVMS and is seperately installed -- is (still) required.  Various -H   HP printers are supported by DCPS.  (I would also prefer to completelyH   avoid the Windows "driver" scheme in preference for a more generic set4   of interfaces if such is feasible, but I digress.)  E   For IP-based printers and for HP JetDirect access, TCP/IP Services  G   or a third-party IP stack is required.  Various HP printers function.   H   CDRECORD is expect to be included in V7.3-1, and this software should H   work with HP CD-R drives -- I haven't specifically tried an HP widget.5   (LD is also appearing on the OpenVMS distribution.)S  D   OpenVMS does not presently provide integrated scanner support, so -   this project would be, um, mildly involved.e  E   HP IA-64 systems will be bootstrapping OpenVMS as part of the port.   I   I would expect to see better support for HP printers and other devices eK   over time (both as part of OpenVMS or OpenVMS ECO kits, and also as part oK   of the distributions that accompany many of the HP devices), but OpenVMS t4   V7.3-1 is simply too far along for this inclusion.    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Tue, 21 May 2002 02:04:01 GMTc1 From: "David J. Dachtera" <djesys.nospam@fsi.net>c= Subject: Re: vax/alpha print to hp laser printers help needed ' Message-ID: <3CE9AE53.F4CDAB63@fsi.net>a   Paul Anderson wrote: > H > In article <20020520124052.38149.qmail@web20209.mail.yahoo.com>, Fabio+ > Cardoso <fabiopenvms@yahoo.com.br> wrote:r > 1 > > Maybe in a few months OpenVMS 7.3-X will haveg7 > > embedded HP drivers. Printers, CD-RW, Scanners, etci. > > ...all those Software stuff for OpenVMS!!! > >o > > Any idea about this ?r > F > It sounds like *you* have the idea.  It's not a bad one, but if suchF > things did not appear in OpenVMS when Digital had both the operatingG > systems and many of the products you mention, what makes you think iti > will be different now? > E > The only part I can address is printer support.  I believe the onlyoH > thing that will change is more timely support for HP printers in DCPS.  D I'm still hoping they'll "front the money" to make DCPS sufficientlyB modular that adding a printer variant would be as simple as addingD libraries / modules / shared sections / etc. rather than the current need to upgrade DCPS.o  E > OpenVMS, the operating system, has no printer drivers in it anyway.  > G > I agree that more and better intergration with other HP products is ae > good idea.   Fer sure...o   --   David J. Dachterag dba DJE Systemsn http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Mon, 20 May 2002 15:36:13 -0400e- From: JF Mezei <jfmezei.spamnot@videotron.ca>e, Subject: Re: Volume shadowing and a disaster, Message-ID: <3CE9502B.D42357E6@videotron.ca>   Bill Todd wrote:M > 'just keep processing and depend upon careful updates' ODS-2 approach.  ButrL > ODS-2's handling of metadata makes the likelihood of structural corruptionN > very low, and applications have at least the opportunity to incorporate userI > data checks (similar to those that would be applied on restart after an . > unclean interruption) should they choose to.  N In the case of volume shadowing, is this not done almost at the physical levelD (replace this block with this contents) or is ODS-2 still involved ?  H If I change 5 bytes in the middle of a block and write it back, will theN remote node get a 512 byte buffer and be just told to write it, or will it getM the changed 5 bytes and told to read the block, replace the 5 bytes and writeE	 it back ?6  G If the volume shadowing is done at the block/physical level, then is it2G correct to state that the failing node can theoretically send shadowingr/ commands that would corrupt the remote drives ?>  , Perhaps I should have rephrased my question:N What mechanisms exist in the software volume shadowing to ensure the sanity of all participating nodes ?   L If the failing node is the one holding quorum, does the remote node have theN ability to either freeze itself if it thinks the quorum node is going nuts, orN must it wait until it has lost total contact before declaring the failing node to be gone ?   ------------------------------  % Date: Tue, 21 May 2002 01:24:08 +0100  From: nic <junk@127.0.0.1>, Subject: Re: Volume shadowing and a disaster) Message-ID: <3CE993A8.DBB098BC@127.0.0.1>x   JF Mezei wrote:t > I > Lets say you have a disaster tolerant cluster with fibre link between 2  > distant buildings. > O > Big explosion in or near building one happens and the electrical power supplyoP > gets a spike followed by sporadic failures and final powerfailure, smoke, fire
 > etc etc. > N > Should one be concerned about the node in building 1 having the time to sendK > shadowing commands to echo erroneous data and data corruption to the disks< > drives in building 2 ?, or is that a very unlikely event ?  : I'm not sure what the problem is you're trying to foresee.  F Volume shadowing uses the same basic VMS IO structures, with all their  data integrity features, period.  E In such a situation, you may need an AMDS (or AM now) to force quorumrH manually, cancel mount verifications and carry on. There is a whitepaper@ somewhere on the OpenVMS website about fibrechannel and disasterF tolerant clusters written (I believe) by John AtoZ. Well worth a read.   --   Regards, Nic Clews (from home), nic at python dot demon dot co dot uk (play) nclews at csc dot com (work)   ------------------------------    Date: 21 May 2002 06:27:14 +0800, From: Paul Repacholi <prep@prep.synonet.com>, Subject: Re: Volume shadowing and a disaster0 Message-ID: <878z6eo3p9.fsf@k9.prep.synonet.com>  / JF Mezei <jfmezei.spamnot@videotron.ca> writes:s  ? > Lets say you have a disaster tolerant cluster with fibre linkc > between 2 distant buildings.  ,B > Big explosion in or near building one happens and the electricalC > power supply gets a spike followed by sporadic failures and finalh$ > powerfailure, smoke, fire etc etc.   F > Should one be concerned about the node in building 1 having the time< > to send shadowing commands to echo erroneous data and dataB > corruption to the disk drives in building 2 ?, or is that a very > unlikely event ?  F If it is MSCP over fibre, the packets will be NAKed for bung checksum.C Foo knows what SAN stuff could do, but I suspect some one will have B the joy of `But that should not happen...' one day. More risk from9 some one doing a `fame and visibility' though I'd expect.-   -- -< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.n@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------    Date: 20 May 2002 21:25:04 -0600+ From: young_r@encompasserve.org (Rob Young)2, Subject: Re: Volume shadowing and a disaster3 Message-ID: <uSuP0aeat2Lh@eisner.encompasserve.org>g  F In article <3CE993A8.DBB098BC@127.0.0.1>, nic <junk@127.0.0.1> writes: > JF Mezei wrote:W >> sJ >> Lets say you have a disaster tolerant cluster with fibre link between 2 >> distant buildings.f >> hP >> Big explosion in or near building one happens and the electrical power supplyQ >> gets a spike followed by sporadic failures and final powerfailure, smoke, firet >> etc etc.  >> aO >> Should one be concerned about the node in building 1 having the time to sendoL >> shadowing commands to echo erroneous data and data corruption to the disk= >> drives in building 2 ?, or is that a very unlikely event ?h > < > I'm not sure what the problem is you're trying to foresee. > H > Volume shadowing uses the same basic VMS IO structures, with all their" > data integrity features, period. > G > In such a situation, you may need an AMDS (or AM now) to force quorumaJ > manually, cancel mount verifications and carry on. There is a whitepaperB > somewhere on the OpenVMS website about fibrechannel and disasterH > tolerant clusters written (I believe) by John AtoZ. Well worth a read. >   $ 	The paper in question resides here:  > http://www.openvms.compaq.com/openvms/fibre/fc_hbvs_dtc_wp.pdf  7 	Anyone following this.. a suggestion, make pointers to @ 	old links and move the other links up a notch.  In other words, 	I constantly try to type in:e  $ 	http://www.openvms.compaq.com/fibre   	Instead of the "correct":  - 	http://www.openvms.compaq.com/openvms/fibre/n   				Robs   ------------------------------  % Date: Tue, 21 May 2002 01:35:38 -0400u- From: JF Mezei <jfmezei.spamnot@videotron.ca>l, Subject: Re: Volume shadowing and a disaster, Message-ID: <3CE9DCA6.9FCF1DE4@videotron.ca>   Rob Young wrote:@ > http://www.openvms.compaq.com/openvms/fibre/fc_hbvs_dtc_wp.pdf  G Thanks for the link. However, this doesn't quite deal with the damage ap@ failing node could potentially do to the shadow set as it fails.  L One of the recommendation in that document is that the backup node not mountM the shadow set so that the main node has full control over the shadow set and J a single FC path to it without worrying about backup node MSCP serving it.G However in this scenario, wouldn't there be even less "sanity" checks ?e  I In a FC scenario, is there some specific protocol used to send disk drivecN commands ? Does the master node send MSCP commands and there is a MSCP-capableN controller in the disk array ? Or does the communication inside the FC contain direct SCSI commands ?  N Either way, will the backup node be advised if the master nodes starts to sendM funky commands over the FC ? Or will its only notification of the master nodeeH going nuts happen when the master node is take out of the cluster due to failure at the SCS link level ?e   ------------------------------  % Date: Tue, 21 May 2002 00:56:48 -0400r- From: JF Mezei <jfmezei.spamnot@videotron.ca> ) Subject: What is the exact role of DCPS ?n, Message-ID: <3CE9D38F.57C370F6@videotron.ca>  0 There have been a lot of discussions about DCPS.  > What is the mission/role of DCPS in terms of printing on VMS ?  H Does DCPS actually provide the communications layer ? (TCPIP, serial are provided by VMS, right ?)e  K I know that DCPS provides conversion from various file types to postscript.d  K I know that DCPS initiates a dialogue with the printer to initialise it and G get page count before and after a job to count number of pages printed..  L But what else does DCPS actually do that isn't provided by the vanilla queue	 manager ?   > As I understand it, DCPS doesn't do font management, correct ?P DCPS doesn't handle ppd files to learn about a printer's capabilities, correct ?  K I am just trying to understand why DCPS would have to be "printer specific"wN instead of "we can support any postscript printer provided you setup the right( ppd file or library modules" philosophy.   ------------------------------    Date: 20 May 2002 11:05:37 -0700( From: bob@instantwhip.com (Bob Ceculski)H Subject: Re: Which Unix for VMS users ?, was: Re: Stallards smoking gun!= Message-ID: <d7791aa1.0205201005.64db450a@posting.google.com>u  } clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley) wrote in message news:<NWO5hGB5XIis@eisner.encompasserve.org>... N > In article <ac899l$cvu$1@aquila.mdx.ac.uk>, david20@alpha1.mdx.ac.uk writes: > >>R > > HP-UX may have a large market share but it is not a good Unix. Tru64 is a much+ > > better Unix from a technical viewpoint.  > >  > I > Since discussion of porting VMS systems to a Unix variant is becoming aFM > more frequent topic these days (and, depending on the way things go, a moreoM > likely possibility), I would be interested in hearing, from a technical anddK > applications availability viewpoint, which Unix would be the best for VMS  > people to port to. > I > I am interested in areas such as reliability, robustness, functionalitymL > and security. I am especially interested in hearing about which Unixes areK > robust enough to just have the power plug pulled out of the back and then-F > recover without a corrupted filesystem/data, which is something that$ > ODS-2/RMS is _very_ good at doing. > N > I am aware of the negative comments made about HP-UX when I previously askedH > why we should take HP's advice to consider HP-UX as a VMS replacement. > P > > I don't think this will change. Application availability will push people to= > > SUN. Performance and other factors will push them to IBM.e< > > HP-UX appears to be secondbest in just about every area. > >  >  > Thanks for any information,  >  > Simon.  E talk is cheap, and that is all some on this board do, even though the G truth has just hit them in the head ... VMS will be fine, and after the-F itanium port, beyond 2011 when alpha support stops, I predict that vmsF will still live whether it be w/hp or some other buyer ... forget unixD as Capellas said it was going to be eviscerated by linux, neither ofB which have any security ... even the unix linux controlling boardsF admitted recently that they have more patches than they can handle andF will not be able to complete them ... the fact remains that VMS is theA only true, secure environment to be on today, esp. on the web ...c; otherwise, when hackers day comes, you will not survive ...r   ------------------------------    Date: 20 May 2002 12:33:24 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) H Subject: Re: Which Unix for VMS users ?, was: Re: Stallards smoking gun!3 Message-ID: <MbkkS7A$LuY9@eisner.encompasserve.org>   w In article <01KHYMTU2RZW96VU4K@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:eM >> > (hey - if they can write to it they should be able to delete it, right?)  > G > Not necessarily.  One could imagine some sort of log file opened for -K > shared write which many users can write to but which none should be able h > to delete. >  >> This is doable under Unix 0 >  > And under VMS. > 1 >> and it doesn't require ACL's.  A writable fileo> >> in a non-writable directory is editable but not deletable.  > C > Right, but since unix has no versions, one can edit away all the VK > contents, which in practice might be just as good (or bad) as being able u > to delete it.h  G If you are willing to live with those restrictions, VMS can do the same.G style of editing as Unix, but it is harder for newcomers to get starteda at doing it that way.    ------------------------------  # Date: Mon, 20 May 2002 19:08:43 GMTp* From: "Bill Todd" <billtodd@metrocast.net>H Subject: Re: Which Unix for VMS users ?, was: Re: Stallards smoking gun!B Message-ID: <%QbG8.196688$M7.20228403@bin7.nnrp.aus1.giganews.com>  G "Simon Clubley" <clubley@remove_me.eisner.decus.org-Earth.UFP> wrote ine5 message news:NWO5hGB5XIis@eisner.encompasserve.org...h   ...t  I > Since discussion of porting VMS systems to a Unix variant is becoming aFH > more frequent topic these days (and, depending on the way things go, a moreI > likely possibility), I would be interested in hearing, from a technicalm and K > applications availability viewpoint, which Unix would be the best for VMS_ > people to port to. >uI > I am interested in areas such as reliability, robustness, functionality-L > and security. I am especially interested in hearing about which Unixes areK > robust enough to just have the power plug pulled out of the back and thenrF > recover without a corrupted filesystem/data, which is something that$ > ODS-2/RMS is _very_ good at doing. >0H > I am aware of the negative comments made about HP-UX when I previously asked H > why we should take HP's advice to consider HP-UX as a VMS replacement.  7 I'm no Unix expert, so take what follows appropriately.!  F Asynchronous I/O may be a major differentiator:  my impression is thatI support varies significantly among major Unixes (and their file systems -rJ e.g., while the third-party Veritas file system is reportedly good in many? respects, I'm not sure it supports asynchronous access at all).D  K As John said (I get his posts as text attachments, which makes them awkwardnC to respond to), the various Unix journaling file systems (VxFS fromlL Veritas - IIRC the HP-UX fs and available on Solaris and AIX, JFS on AIX andG Linux, XFS on IRIX and Linux, journaled UFS on Solaris, AdvFS on Tru64,tD ext3fs on Linux, reiserfs on Linux, possibly others - including *BSDI options) offer the combination of file system integrity and fast restart.tJ The BSD 'soft update' approach is more like ODS-2's careful updates than a< log but obtains potentially better-than-ODS-2 performance inL metadata-intensive update activity, and also supports fast restart with good
 integrity.  H Fail-over clustering for user data (and usually metadata as well) is theH Unix norm, and virtually requires a file system supporting fast restartsD (since that's what fail-over effectively is).  Partitioning data and@ scripting resource fail-over cascades is more awkward than a VMSI shared-storage environment but probably not a major obstacle.  Clustering-J support also varies significantly among Unixes (Tru64's seems to be in theJ lead for now, especially insofar as it includes a VMS DLM clone that couldI make porting cluster-aware VMS applications much easier - though IIRC AIXn has a VMS DLM clone as well).   K AFAIK no Unix vendor supports anything much like RMS, but third parties do. I I'm vaguely aware of things like Berkeley db and C-ISAM, and there may be 1 actual RMS clones such as those available on PCs.e  J Hope others more knowledgeable hop in with more details:  I'm curious too.   - bill   ------------------------------   Date: 20 May 2002 20:39:00 GMT From: phn@icke-reklam.ipsec.nuH Subject: Re: Which Unix for VMS users ?, was: Re: Stallards smoking gun!) Message-ID: <acbmt4$4ms$1@nyheter.crt.se>e  ) Bill Todd <billtodd@metrocast.net> wrote:   I > "Simon Clubley" <clubley@remove_me.eisner.decus.org-Earth.UFP> wrote in 7 > message news:NWO5hGB5XIis@eisner.encompasserve.org...n   > ...   J >> Since discussion of porting VMS systems to a Unix variant is becoming aI >> more frequent topic these days (and, depending on the way things go, a  > moreJ >> likely possibility), I would be interested in hearing, from a technical > andnL >> applications availability viewpoint, which Unix would be the best for VMS >> people to port to.a >>J >> I am interested in areas such as reliability, robustness, functionalityM >> and security. I am especially interested in hearing about which Unixes are L >> robust enough to just have the power plug pulled out of the back and thenG >> recover without a corrupted filesystem/data, which is something thato% >> ODS-2/RMS is _very_ good at doing.  >>I >> I am aware of the negative comments made about HP-UX when I previouslyo > askedfI >> why we should take HP's advice to consider HP-UX as a VMS replacement.-  9 > I'm no Unix expert, so take what follows appropriately.1  H > Asynchronous I/O may be a major differentiator:  my impression is that  + %% some stuff deleted to save electrons %%%2M > AFAIK no Unix vendor supports anything much like RMS, but third parties do.dK > I'm vaguely aware of things like Berkeley db and C-ISAM, and there may bem3 > actual RMS clones such as those available on PCs.l  C The equivalent of RMS is not considered to be part of kernel at all > in unux - it's a different design attitude. A file is a stream8 of bytes, it's up to the application to make structure.     @ Anyone is free to implement RMS as a library, and i'm not aware 2 of anything "not possible to do" by that approach.    L > Hope others more knowledgeable hop in with more details:  I'm curious too.   > - bill         --   Peter Hkanson         e7         IPSec  Sverige      ( At Gothenburg Riverside ) J            Sorry about my e-mail address, but i'm trying to keep spam out,; 	   remove "icke-reklam" if you feel for mailing me. Thanx.y   ------------------------------  # Date: Tue, 21 May 2002 01:17:51 GMTr# From: "John Smith" <a@nonymous.com>aH Subject: Re: Which Unix for VMS users ?, was: Re: Stallards smoking gun!H Message-ID: <3fhG8.57465$ah_.33414@news01.bloor.is.net.cable.rogers.com>  5 "Bill Todd" <billtodd@metrocast.net> wrote in messagei< news:%QbG8.196688$M7.20228403@bin7.nnrp.aus1.giganews.com... >e >bI > AFAIK no Unix vendor supports anything much like RMS, but third partiesr do.tK > I'm vaguely aware of things like Berkeley db and C-ISAM, and there may beo3 > actual RMS clones such as those available on PCs.  >,L > Hope others more knowledgeable hop in with more details:  I'm curious too. >d > - bill >i  L Faircom has (or maybe had by now) c-isam implementation that was portable toD VMS... but that's a memory from the 'way, way back machine Sherman'.   ------------------------------    Date: 20 May 2002 19:17:31 -0700( From: bob@instantwhip.com (Bob Ceculski)H Subject: Re: Which Unix for VMS users ?, was: Re: Stallards smoking gun!= Message-ID: <d7791aa1.0205201817.537a3664@posting.google.com>i  k John Eisenschmidt <jweisen@eisenschmidt.org> wrote in message news:<20020520143346.C66@eisenschmidt.org>...s > M > Security is a whole other ball of wax. My biggest complaint is not having aaM >  delete big (hey - if they can write to it they should be able to delete itsM > , right?) I've shot myself in the foot more than once wishing I could writecM >  to but not delete a file. Most commercial Unixes now support ACLs, which hcM > elps (but is far from perfect). I think like VMS if you understand the secueM > rity model thoroughly, you can build a secure system. Some commercial UnixeeM > s also support a "Secure" version of their OS (Trusted Solaris for example)i+ >  like the secure layered product for VMS.c  B Security for unix/linux is an oxymoron ... and there is a lot more? to a security model than acl's ... it starts as the base of theeB os ... you can't root vms, and you can't bufferoverflow easily, ifB at all ... because vms was built right from the ground up, and putB in its 22 years of dues to prove itself ... all the other "install4 and patch" models are timebombs, waiting to explode!   > M > > I am aware of the negative comments made about HP-UX when I previously asg >  kedJ > > why we should take HP's advice to consider HP-UX as a VMS replacement. > > M > > > I don't think this will change. Application availability will push peop  >  le to? > > > SUN. Performance and other factors will push them to IBM. > > > > HP-UX appears to be secondbest in just about every area. >   D vms apps are slowly on the increase, and itanium will push that evenF more ... and what isn't for sale or freeware can be written ... no oneH in their right mind would ever port to unix/linux w/a no security model,I no 99.9999 uptime and clustering, no anything ... as for IBM, been there,cD done that, AS400, one month, 2 crashes, unknowns reasons, nightmare  environments, no thank you!l   ------------------------------  % Date: Tue, 21 May 2002 01:12:39 +0100  From: nic <junk@127.0.0.1>' Subject: Re: Who cares about marketing!e) Message-ID: <3CE990F7.A624C4FF@127.0.0.1>l   Bill Todd wrote: > 6 > "Elliott Roper" <elliott@yrl.co.uk> wrote in message   > - > > And that shit-for-brains remark from Nic?: > K > Was an entirely appropriate and measured response to his calling Olson anoL > idiot.  Had he simply made a naive statement about marketing, I would have > responded more 'gently'.  H You don't have to be gentle, however resorting to profanity demonstratesG a clear inability to communicate using the English language, and such apC failure often stops readers from looking further. It is a marketingr
 statement.  F Perhaps I was a little strong saying he was an idiot, perhaps fool was more appropriate.t  C Followers of this list are probably aware that media was a previousoC chosen career of mine, yet I decided not to embark on it, because IiB disliked what it had become, and I have said as much here but I'll refrain from the details again.t  E As far as I'm concerned, you can throw the whole thing into a bucket, F probably a fairly large one, but it doesn't have to be strong, or evenD watertight. No-one can move these days, have any form of social lifeC without something in this bucket infiltrating your existence, ofteneG almost subconsciously. Welles' 1984 has been and its still here, socialc7 conditioning, not by the authorities, but by the media.a  E Success good enough to keep a major organization in profit means thataC this media engine is used effectively, because if you aren't there,s someone else will be.s  B This I would surmise was the issue with DEC under Olson. The mediaF engine wasn't as advanced as now, but this was the time it was turningH into something I didn't like, and that snake oil crept under the carpet.F No matter what you think of snake oil, it allowed the rug to be pulled from beneath DEC's feet.  D To believe that marketing is not important is naive, perhaps this is what I should stated about Ken.3  G It is perhaps ironic, that when I decided that computing was the bettermE career for me, the entity I decided to avoid became one part of DEC'sr	 downfall.n   -- e Regards, Nic Clews (from home), nic at python dot demon dot co dot uk (play) nclews at csc dot com (work)   ------------------------------  # Date: Tue, 21 May 2002 03:25:23 GMTw* From: "Bill Todd" <billtodd@metrocast.net>' Subject: Re: Who cares about marketing!hA Message-ID: <D6jG8.86067$Ze4.9201822@bin5.nnrp.aus1.giganews.com>   K "nic" <junk@127.0.0.1> wrote in message news:3CE990F7.A624C4FF@127.0.0.1...n > Bill Todd wrote: > >f8 > > "Elliott Roper" <elliott@yrl.co.uk> wrote in message >t > >d/ > > > And that shit-for-brains remark from Nic?  > >oJ > > Was an entirely appropriate and measured response to his calling Olson anI > > idiot.  Had he simply made a naive statement about marketing, I woulde have > > responded more 'gently'. >oJ > You don't have to be gentle, however resorting to profanity demonstrates= > a clear inability to communicate using the English languagea  I Really?  I think I communicate quite well, thank you - and certainly said,K *exactly* what I considered appropriate in the above case.  If people don't J like it, or even stop reading because of it, that's their problem:  peopleC who choose to be uninformed are not the target of anything I write.h  K Of course, were I a politician I'd be more circumspect, since my goal wouldlJ be less to communicate than to persuade.  A difference that, if indeed youF are someone who had considered a career in the media, should be (and IL suspect is, which says a lot about the aim of your response) obvious to you.   - bill   ------------------------------   End of INFO-VAX 2002.279 ************************what else does DCPS actually do that isn't provided by the vanilla queue	 manager ?   > As I understand it, DCPS doesn't do font management, correct ?P DCPS doesn't handle ppd files to learn about a printer's capabilities, correct ?  K I am just trying to understand why DCPS would have to be "printer specific"wN instead of "we can support any postscript printer provided you setup the right( ppd file or library modules" philosophy.   ------------------------------                                         	    
            
                                                                                 !    "    #    $    %    &    '    (    )    *    +    ,    -    .    /    0    1    2    3    4    5    6    7    8    9    :    ;    <    =    >    ?    @    A    B    C    D    E    F    G    H    I    J    K    L    M    N    O    P    Q    R    S    T    U    V    W    X    Y    Z    [    \    ]    ^    _    `    a    b    c    d    e    f    g    h    i    j    k    l    m    n    o    p    q    r    s    t    u    v    w    x    y    z    {    |    }    ~                                                                                                                                                                                                                                                                                µ    õ    ĵ    ŵ    Ƶ    ǵ    ȵ    ɵ    ʵ    ˵    ̵    ͵    ε    ϵ    е    ѵ    ҵ    ӵ    Ե    յ    ֵ    ׵    ص    ٵ    ڵ    ۵    ܵ    ݵ    ޵    ߵ    ു    ᵁ    ⵁ    㵁    䵁    嵁    流    絁    赁    鵁    굁    뵁    쵁            ﵁    𵁪    񵁪    򵁪    󵁪                                                    