1 INFO-VAX	Thu, 27 Feb 2003	Volume 2003 : Issue 114       Contents:1 Re: About VMS on Itanium in the French "Press"... 1 Re: About VMS on Itanium in the French "Press"... ' Re: Anyone out there with a working EVA  Re: Data in shareable image P Re: ECC (was IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM  monopo5 EMC to acquire Legato ?  How about OpenVMS + Legato ?   Re: Fix for EDT emulation in EVE4 Re: HP killing Golden Eggs- future of goose unknown.4 Re: HP killing Golden Eggs- future of goose unknown.H Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopolyD Re: Idea for installing VMS without a 512byte CD-ROM...Will it work?D Re: Idea for installing VMS without a 512byte CD-ROM...Will it work?B Re: Image restore of "old" saveset fails to restore VMS$COMMON.DIR( Re: Is there a way to FTP a locked file?5 Re: Itanium "threw out all the good parts of the x86" 5 Re: Itanium "threw out all the good parts of the x86" 5 Re: Itanium "threw out all the good parts of the x86" / Re: Linus trashes Itanium - he's scared of VMS! / Re: Linus trashes Itanium - he's scared of VMS! / Re: Linus trashes Itanium - he's scared of VMS! / Re: Linus trashes Itanium - he's scared of VMS! / Re: Linus trashes Itanium - he's scared of VMS! / Re: Linus trashes Itanium - he's scared of VMS! B Re: memory errors and bugcheck dump (PROCGONE) with ALPHA 3000/600- MIME utility, was: Re: Mis-set Terminal Types K Re: Mis-set Terminal Types (was: Re: HUGE showstopper *bug* in HELP/MESSAGE ( More Tar problems (was  VMSTAR question)7 Odd [to me] error when trying to map a pathworks share. ; Re: Odd [to me] error when trying to map a pathworks share. ! Re: Official OpenVMS name for the ! Re: Official OpenVMS name for the ! Re: Official OpenVMS name for the P Re: Official OpenVMS name for the =?ISO-8859-1?Q?Itanium=AE_re?= =?ISO-8859-1?Q?P Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha RetaiP Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha RetaiP Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha RetaiM OT but worth posting, virus warning: "FWD: Look at this pack from Microsoft." P Re: OT but worth posting, virus warning: "FWD: Look at this pack from Microsoft.5 Re: PC developers opinion of 64 bit desktop computing 5 RE: PC developers opinion of 64 bit desktop computing ! Re: Pine Email Problem on OpenVMS 2 Re: pseudointerface (TCPIP SET [CONFIG] INTERFACE)2 Re: pseudointerface (TCPIP SET [CONFIG] INTERFACE)2 Re: pseudointerface (TCPIP SET [CONFIG] INTERFACE)2 Re: pseudointerface (TCPIP SET [CONFIG] INTERFACE)2 Re: pseudointerface (TCPIP SET [CONFIG] INTERFACE]9 Re: SET NOTRUNCATE (was: RE: Upcoming VMS improvements ?)  Sun and Java+ Re: Suspected compromise of my home VMS box $ Re: TCPIP and network loss detection Re: The End of OpenVMS Re: The End of OpenVMS RE: The End of OpenVMS Re: The End of OpenVMS Re: Upcoming VMS improvements ?  Re: Upcoming VMS improvements ?  Re: Upcoming VMS improvements ?  Re: Upcoming VMS improvements ?  Re: Upcoming VMS improvements ? 
 VAX Emulators  Re: VAX Emulators  Re: VMS Backup solutions Re: VMS Backup solutions Re: VMS Backup solutions Re: VMS Backup solutions Re: Volume set across RAID5  Re: Volume set across RAID5 > Re: What makes the VMS port to Itanium/IPF/whatever difficult?  Re: Where to get a recent Emacs? Will HP pull an IBM?1 Re: Wow - OpenVMS Job Advert in UK Computer Press 1 Re: Wow - OpenVMS Job Advert in UK Computer Press   F ----------------------------------------------------------------------  % Date: Thu, 27 Feb 2003 08:58:03 +0100 5 From: "Philippe Bocher" <philippe.bocher@euriware.fr> : Subject: Re: About VMS on Itanium in the French "Press"...3 Message-ID: <b3kged$9tp$1@news-reader10.wanadoo.fr>   7     the Url is http://www.01net.com/article/201922.html G "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> a crit dans le message de * news: 3E5D11DB.F1F61EDF@vl.videotron.ca...D > > "Philippe Bocher" <philippe.bocher@euriware.fr> wrote in message > > >     In french : ) > > >     http://minilien.com/?GNoFDdjPyJ  > L > This page consistently crashes netscape on a mac, even with javascript and; > java disabled. I'll have to try lynx or mosaic on my vax. J     the good Url is http://www.01net.com/article/201922.html, I don't know5 why I've tried to use minilien for a so short URL ;-(    ------------------------------    Date: 27 Feb 2003 09:54:23 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) : Subject: Re: About VMS on Itanium in the French "Press"...= Message-ID: <cf15391e.0302270954.5dd34cab@posting.google.com>   c "Philippe Bocher" <philippe.bocher@euriware.fr> wrote in message news:<3e5c90a1$1@news.euriware>... 2 > About VMS on Itanium in the IT "Press" in France ...   >     Translation by Altavista : ... ' > the mythical operating system of fire   * I kind of like this way of describing VMS.   ------------------------------  % Date: Thu, 27 Feb 2003 09:33:07 -0000 * From: "Craig Cooke" <ccooke@beta.dabs.com>0 Subject: Re: Anyone out there with a working EVA5 Message-ID: <coWF8Mk3CHA.1620@juno.intranet.dabs.com>   5 "Nic Clews" <sendspamhere@127.0.0.1> wrote in message # news:3E5CE6E9.9A9E3B80@127.0.0.1...  > Craig Cooke wrote: > G > > Further to my previous posting regarding if anyone was aware of any L > > companies providing training for an EVA - does anyone have one installed / K > > working (in the U.K.) - and ... would they be prepared to 'show it off'  toH > > myself and a couple of work collegues.  We simply wish to understand it'sL > > potential, discuss any pitfalls, discuss how easy it is to manage etc... >   > I'll ask. What's in it for us? > B > I'm not sure how we could get you to physically see it though... >  > --A > Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  > nclews at csc dot com    Hi Nic,   I What's in it for you - welll, we are an I.T. reseller - I'm sure we could  sort something out :-)  3 We would like to see it up and running if possible.    Regards    Craig Cooke    ------------------------------  % Date: Thu, 27 Feb 2003 10:38:47 -0500 & From: jlsue <jlsuexxxz@screaminet.com>$ Subject: Re: Data in shareable image8 Message-ID: <r6cs5v8slj2n3793g04o5sgfm29ovaq0tc@4ax.com>  F On Wed, 26 Feb 2003 22:10:43 GMT, hoff@hp.nospam (Hoff Hoffman) wrote:  d >In article <3e54bf02$1_7@news.teranews.com>, "Johan Nilsson" <johan.nilsson@esrange.ssc.se> writes: >    > J >:So, am I correct in my assumption here? If so, is there any way avoiding9 >:this without resorting to manual padding of the struct?  > G >  Again, you will be fixing this stuff with some regularity as OpenVMS E >  and the application are upgraded -- applications using COMMONs are  >  commonly relinked.  >  ^^^^^^^^^  ( What?  No smiley after this little joke?   ------------------------------  % Date: Thu, 27 Feb 2003 09:30:53 -0500 & From: jlsue <jlsuexxxz@screaminet.com>Y Subject: Re: ECC (was IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM  monopo 8 Message-ID: <on7s5v49q0nk3s2vs58taeun3jl1qen76h@4ax.com>  E On Wed, 26 Feb 2003 13:52:55 +0000, Andrew Harrison SUNUK Consultancy . <Andrew_No.Harrison_No@nospamn.sun.com> wrote:  
 >jlsue wrote: : >> On Mon, 24 Feb 2003 16:06:47 -0000, "Helmut P. Einfalt"" >> <hp.einfalt@t-online.de> wrote: >>   >>  5 >>>>Always cracks me up the way these kids talk about ? >>>>one machine which has 2% better performance than another...  >>> H >>>...and bear in mind that it is *benchmark* perfomance more often than
 >>>not...! >>>  >>   >>  H >> Oh, but we have it on top authority in here that benchmarks are *the*) >> most important measure of performance.  >>   > < >Depends what kind of benchmark. But lets face it benchmarks@ >have to be better than hand wavey "in my experience" statements% >unsupported by any other collateral.  > : >Now who could I be refering to with the hand wavey "in my >experience" quote, I wonder.  >   ? I don't think I ever used the quote "in my experience" to prove D anything like a general claim that the alpha systems in general were better performers or anything.  @ The "in my experience", which you take completely out of contextB because that'd kill your logic, was that your outrageous claims of@ performance problems as a general case didn't wash.  You see, inF logical arguments when you make all-inclusive statements, all it takesC is one counter-example to disprove the statement.  But you've STILL C not validated how wide-spread this supposed performance problem is. 0 You've found - what? - one or two posts in here.  D And what does that mean?  It happens?  It happens a lot?  It happensC every time?  Your claims fall in the "a lot" category at least, and A border on the "every time" claim.  Exactly how are you using this # information to support your claims?   F I don't make claims that there are no applications which will be slow,+ so I don't have to pull out all this proof.   D And in the end, I don't sell hardware or software.  I help customersF find the right fit for their needs, and if they have requirements thatD are extraordinary, I don't rely on industry benchmarks to support myD case.  If, after benchmarking their own app, they aren't interested,C fine with me - well not really, I'd like to say in business selling B the dec/hp stuff, but I'm not married to it:  My main objective is6 helping IT align with, and solve, real business needs.   ------------------------------  + Date: Thu, 27 Feb 2003 08:23:36 -0800 (PST) . From: Fabio Cardoso <fabiopenvms@yahoo.com.br>> Subject: EMC to acquire Legato ?  How about OpenVMS + Legato ?@ Message-ID: <20030227162336.46425.qmail@web20203.mail.yahoo.com>   Click   4 http://news.com.com/2100-1001-986230.html?tag=fd_top   Regards    FC     =====  ========================== Fbio dos Santos Cardoso OpenVMS System Manager Rio de Janeiro - Brazil  fabiopenvms@yahoo.com.br ==========================  2 __________________________________________________ Do you Yahoo!?2 Yahoo! Tax Center - forms, calculators, tips, more http://taxes.yahoo.com/    ------------------------------  % Date: Thu, 27 Feb 2003 11:29:12 -0500 - From: "Peter Weaver" <peter.weaver@stelco.ca> ) Subject: Re: Fix for EDT emulation in EVE 6 Message-ID: <b3led0$1m9mbi$1@ID-141708.news.dfncis.de>   David J. Dachtera wrote: >...F > In EDT, I use GOLD-F to write the current buffer to a file. It wouldB > be nice if EDT could write the current select range to file. I'dE > probably put that on GOLD-S or something that I don't currently use  > for anything else. >...   Does this do what you want?   D define key gold s as "cutsr=writesrbuffer paste=writesrbuffer ext wr2 ?*'Enter the Filename to Write: ' =writesrbuffer."   -- Peter WeaverD Opinions are my own, and do not reflect the opinions of my employer,A nor the company that it sub-contracts to, nor the company that it  sub-contracts to.    ------------------------------  % Date: Thu, 27 Feb 2003 12:44:36 +0200 % From: "Mariuz" <mariuz@stop.spam.org> = Subject: Re: HP killing Golden Eggs- future of goose unknown. : Message-ID: <pan.2003.02.27.10.44.36.387620@stop.spam.org>  7 On Wed, 26 Feb 2003 15:39:10 +0000, Andy Stoffel wrote:   9 > ld convince 'someone' to let me buy it (when it became  B > available to partners/ISV's/etc a few years ago).... back when I > had a use for it..) & Maybe they should make it open source - I'am realy interested in something like that   to work on/with    ------------------------------  % Date: Thu, 27 Feb 2003 12:23:28 +0000 ( From: Nic Clews <sendspamhere@127.0.0.1>= Subject: Re: HP killing Golden Eggs- future of goose unknown. ) Message-ID: <3E5E0340.1A4DB55E@127.0.0.1>   
 VAXVMS wrote:  > 9 > Did anybody else find this tool to be extremely useful?   C We use this tool, but use more of Visio which has DEC/Compaq boxes, C mainly because we are obviously needing multi vendor diagrams. I'll 9 mention it's [possible?] demise to the collective here...    --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot com    ------------------------------  ! Date: Thu, 27 Feb 03 10:33:48 GMT  From: jmfbahciv@aol.com Q Subject: Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopoly + Message-ID: <b3ks43$6jj$3@bob.news.rcn.net>   3 In article <561.187T1319T7364328@kltpzyxm.invalid>, 3    "Charlie Gibbs" <cgibbs@kltpzyxm.invalid> wrote: D >In article <20030226112733.6D6DAB73.NOFFLE@momos.conceptual.net.au>/ >prep@prep.synonet.com (Paul Repacholi) writes:  >  >>jmfbahciv@aol.com writes:  >>@ >>> "They" can demand all they want; they get what they pay for. >>3 >>I think we need to carve this in stone somewhere!  > 3 >Then we can hit politicians over the head with it.  > : <mutter>  You are assuming that those critters have heads.8 I'm currently in the midst of an IF,THEN,ELSE,PUSHJ into7 the middle of a procedure, JRST over there, then JSP to / over here in my state income tax instructions.    9 And, AND somebody doesn't know the importance of using a  9 semi-colon when listing routine calls under the first IF. : They keep using a period and it takes me donkey's years to= figure out I've just fucked up the form.  I do it every year.    /BAH  ' Subtract a hundred and four for e-mail.    ------------------------------  # Date: Thu, 27 Feb 2003 15:41:50 GMT ( From: "Mark E. Levy" <mlevy70@attbi.com>M Subject: Re: Idea for installing VMS without a 512byte CD-ROM...Will it work? - Message-ID: <2lq7a.259338$tq4.6190@sccrnsc01>   < "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message! news:3E57EB29.6F24C5AB@fsi.net...  > Robert Lawrence wrote: > > G > > I have a friend that gave me a VAXstation 3100/M30, which should be I > > delivered Monday. This machine has a hard drive expansion box, but no G > > CD-ROM.  This machine also is missing an OS.  I tossed around a few E > > ideas and came up with the following method of intalling OpenVMS. E > > (Keep in mind that my only exposure to VAX & VMS was as a user at * > > university, so I am learning as I go.) > > F > > * Install Linux on a PC, with a SCSI host adpater and an extra 1GB > > SCSI drive. H > > * From Linux do a: dd if=/dev/cdrom of=/dev/sda1  (copy VMS media to > > raw device) ? > > * Remove the drive, set the SCSI ID to that of a VAX CD-ROM J > > * Install SCSI hard drive in VAXstation and direct it to boot from the
 > > new drive  > > * Continue with VMS install  > > F > > Will this work?  If I dump the CD contents byte for byte to a SCSIG > > hard drive on a PC, will the VAX be able to read and boot from that 
 > > drive?  F No, because you won't be writing a valid boot block on the SCSI drive.  J > This would be better: use another VMS machine and BACKUP/IMAGE the CD to@ > a hard disc, then stick the hard disc in the 3100 and boot it.  K That would work without question. But, if another VMS machine is available, ) other options present themselves as well.    ------------------------------  % Date: Thu, 27 Feb 2003 16:26:30 +0000 + From: John Laird <john@laird-towers.org.uk> M Subject: Re: Idea for installing VMS without a 512byte CD-ROM...Will it work? 8 Message-ID: <cees5vc2p0btigm5i4ucoq75nt0uk2e5kv@4ax.com>  K On Thu, 27 Feb 2003 15:41:50 GMT, "Mark E. Levy" <mlevy70@attbi.com> wrote:    >> Robert Lawrence wrote:  >>  > G >> > Will this work?  If I dump the CD contents byte for byte to a SCSI H >> > hard drive on a PC, will the VAX be able to read and boot from that >> > drive?  > G >No, because you won't be writing a valid boot block on the SCSI drive.   H Uh ?  I don't know why there is quite so much suspicion about this idea.B It'd be the first thing I'd try if I had just these tools to hand.  I If the CD is bootable, a block-for-block copy of it to a hard disk should J retain that capability.  It will look odd when mounted as it will have theE volume size, cluster factor, etc, of a CDROM, but this shouldn't be a  showstopper.     	John    ------------------------------  % Date: Thu, 27 Feb 2003 09:04:06 -0500 & From: David M Smith <dsmit115@csc.com>K Subject: Re: Image restore of "old" saveset fails to restore VMS$COMMON.DIR 8 Message-ID: <sc6s5vov64k6b4c77k0v83u6us84283rfi@4ax.com>  F On Thu, 27 Feb 2003 01:54:46 GMT, hoff@hp.nospam (Hoff Hoffman) wrote:   > B >  Please provide the header output portion of the output from the >  following command:  > # >   $ BACKUP/LIST saveset.bck /SAVE   * Here is the BACKUP /LIST from the saveset:   Listing of save set(s)  $ Save set:          ALPVMS0711H2-GSE. Written by:        xxxxxxxx     " UIC:               [000026,000004]* Date:              27-APR-2001 12:43:45.82I Command:           BACKUP/VERIFY/IMAGE $1$DUA7: DA:[VMS_KITS.ALPVMS0711H2 I _TEMPLATE]ALPVMS0711H2-GSE./SAVE/LIST=DA:[VMS_KITS.ALPVMS0711H2_TEMPLATE]  ALPVMS0711H2-GSE.LIS+ Operating system:  OpenVMS VAX version V6.2  BACKUP version:    V6.2  CPU ID register:   13000202  Node name:         _xxxxxx:: Written on:        _$1$DUA111: Block size:        32256 Group size:        10  Buffer count:      15    Image save of volume set Number of volumes: 1  G The only qualifiers used on the BACKUP command were /VERIFY and /IMAGE. O According to HELP BACKUP /ALIAS on VAX V6.2, the default is /ALIAS, so it would ( have been done with this default option.  A There is a SYSCOMMON.DIR directory entry but it, too, is invalid:    Directory DKA400:[SYS0]   N DIA$TOOLS.DIR;1            1/4        20-APR-2001 16:31:39.52  (RWE,RWE,RE,RE)N MOM$SYSTEM.DIR;1           1/4        20-APR-2001 16:31:40.01  (RWE,RWE,RE,RE)N SYS$I18N.DIR;1             1/4        20-APR-2001 16:31:40.24  (RWE,RWE,RE,RE)N SYS$LDR.DIR;1              1/4        20-APR-2001 16:31:40.99  (RWE,RWE,RE,RE)N SYS$STARTUP.DIR;1          1/4        20-APR-2001 16:31:41.24  (RWE,RWE,RE,RE)N SYSCBI.DIR;1               1/4        20-APR-2001 16:31:41.49  (RWE,RWE,RE,RE) SYSCOMMON.DIR;1    no such file N SYSERR.DIR;1               1/4        20-APR-2001 16:31:41.72  (RWE,RWE,RE,RE)N SYSEXE.DIR;1               2/4        20-APR-2001 16:02:28.68  (RWE,RWE,RE,RE)N SYSHLP.DIR;1               1/4        20-APR-2001 16:31:41.96  (RWE,RWE,RE,RE)N SYSLIB.DIR;1               1/4        20-APR-2001 16:31:42.89  (RWE,RWE,RE,RE)N SYSMAINT.DIR;1             1/4        20-APR-2001 16:31:43.13  (RWE,RWE,RE,RE)N SYSMGR.DIR;1               1/4        20-APR-2001 16:31:43.36  (RWE,RWE,RE,RE)N SYSMSG.DIR;1               1/4        20-APR-2001 16:31:43.59  (RWE,RWE,RE,RE)N SYSTEST.DIR;1              1/4        20-APR-2001 16:31:43.83  (RWE,RWE,RE,RE)N SYSUPD.DIR;1               1/4        20-APR-2001 16:31:44.06  (RWE,RWE,RE,RE)N TNT.DIR;1                  1/4        20-APR-2001 16:31:44.31  (RWE,RWE,RE,RE)    Total of 17 files, 17/64 blocks.  P I was able to repair the disk using ANALYZE /DISK /REPAIR. A collection of filesO ended up in [SYSLOST]. a few PCSI database files and a lot of directories, e.g. N SYSEXE.DIR, SYSLIB.DIR, etc. I created a [VMS$COMMON] directory, RENAMEd theseP files into it, created a SYSCOMMON.DIR synonym entry in [SYS0] and then the disk; looked "right". I did not take time to try to boot from it.   H >  AFAIK, this stuff works just fine -- I've been reading older savesetsI >  for many years now.  If an older BACKUP cannot be restored, then there H >  is a problem with the newer version of BACKUP, or there was a problemG >  with the version that wrote the saveset, or the source disk had some E >  sort of file structure problem and BACKUP dutifully replicated it.   N Well, that has been my experience heretofore, but I am now seeing this problemO on more than one saveset. What I have is a collection of "template" system disk P images, each for a different version of OpenVMS VAX or Alpha. I use these when IO need to confirm how something worked on an older version, or to test code on an  older version.  > >  The first implementation of /[NO]ALIAS was at OpenVMS V6.2.  M Thanks -- I've also found (empirically) that the BACKUP image included in the L Y2K patch for OpenVMS VAX V5.5-2 also contains support for /[NO]ALIAS, as it8 appears to be a "back-ported" version of the V6.2 image:   $ imgid sys$system:backup.exe F Image Header Information for VAX image SYS$COMMON:[SYSEXE]BACKUP.EXE;1           Image Name: "BACKUP"/         Image Identification: "VAX 552R_62-002" /         Link Date/Time:  1-DEC-1999 15:31:17.11 &         Linker Identification: "05-13"I ------------------------------------------------------------------------- I David M. Smith 302.391.8533                       dsmit115 at csc dot com I Computer Sciences Corporation     (Opinions are those of the writer only) I -------------------------------------------------------------------------    ------------------------------    Date: 27 Feb 2003 05:44:26 -0800( From: bob@instantwhip.com (Bob Ceculski)1 Subject: Re: Is there a way to FTP a locked file? < Message-ID: <d7791aa1.0302270544.170b725@posting.google.com>  [ hoff@hp.nospam (Hoff Hoffman) wrote in message news:<nQd7a.532$Kx5.462@news.cpqcorp.net>... f > In article <48hb5vgd4b9raiji32ohu8loiefdf14o4f@4ax.com>, Kiasu Surfer <verykiasu@hotpop.com> writes:R > :Thinking for FTP my Oracle datafile to a local server (same segment) instead of > :the regular DLT Tape drive. > D >   If the file is locked, the FTP has not completed, or the remote G >   system (the one running the FTP server) has not let go of the file. / >   This latter case would be a bug, of course.  > D >   SHOW DEVICE/FILE/NOSYS will likely show which process is holding >   open the particular file.  > E >   If the remote system is looking for the file and finding one that F >   is not (yet) copied over, I'd switch to a scheme which copies overE >   the file using a temporary name, and then renames it to the final ? >   and appropriate filename.  This greatly reduces the window.  > B > :I don't have NFS or Pathworks installed, running OVMS 7.1 ECO9. > F >   Eh?  OpenVMS V7.1 is very old, as is the typical version of TCP/IPF >   Services found on that release -- if you are running what I assumeF >   is a TCP/IP Services ECO9, I will assume that this TCP/IP Services9 >   version is V4.1.   (ECO10 is the ECO for V4.1, BTW.)   > C >   The current releases are OpenVMS Alpha V7.3-1, and TCP/IP V5.3.  > H >   I'd strongly encourage an upgrade to releases slightly more current. > F >   We made significant changes and significant updates in more recent> >   OpenVMS releases, and with TCP/IP Services V5.0 and later. > G >   I'd move to V7.1-2, to V7.2-2, or to V7.3-1.  With mandatory ECOs.  # >   And  to TCP/IP V5.3.  With ECO.  > P >  ---------------------------- #include <rtfaq.h> -----------------------------L >       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comP >  --------------------------- pure personal opinion ---------------------------G >         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com   ? and why is that Mr. H.?  with TCPware I can run the current 5.6 > version with VMS all the way back to 5.5 ... why can't ucx 5.3 run with VMS 7.1?    ------------------------------    Date: 27 Feb 2003 02:26:11 -0800" From: xeio77@hotmail.com (Tsarkon)> Subject: Re: Itanium "threw out all the good parts of the x86"< Message-ID: <baa86a4a.0302270226.7021f89@posting.google.com>  v koehler@eisner.nospam.encompasserve.org (Bob Koehler) wrote in message news:<TJUhLnSBEJd3@eisner.encompasserve.org>...d > In article <3e5c7d0b.55021888@news.eircom.net>, rw@vorpalbunnyeircom.net (Russell Wallace) writes:F > > On 25 Feb 2003 15:05:37 -0800, xeio77@hotmail.com (Tsarkon) wrote: > >  > >>Torvalds wroteE > >>that Intel had made the same mistakes "that everybody else did 15 H > >>years ago" when RISC architecture was first appearing. Itanium triesD > >>to introduce an architecture that is clean and technically pure,I > >>something that just doesn't seem to work in the real world. He claims F > >>that Intel "threw out all the good parts of the x86 because peopleJ > >>thought those parts were ugly. They aren't ugly, they're the 'charming# > >>oddity' that makes it do well."  > > ) > > It seems to me that this is nonsense.  > F >     The press didn't bother including the fact that this whole thing% >    is an appeal to false authority.  > ) >     Linus is not a chip architect guru.   F Sure, he knows nohting of the subject. Riiight. And he doesnt work forB transmeta nor have anything to do with designing the software that2 takes x86 and transmogrifies it into crusoe speak.  @ The guy is a loon tune at times, but he and his cronies are veryF familiar with IA-32. I would take pause at what he thinks. Proof is inE the pudding, and while you might not like the taste, he certainly has  made a lot of pudding.   ------------------------------    Date: 27 Feb 2003 02:27:31 -0800" From: xeio77@hotmail.com (Tsarkon)> Subject: Re: Itanium "threw out all the good parts of the x86"= Message-ID: <baa86a4a.0302270227.505e5de1@posting.google.com>   g rw@vorpalbunnyeircom.net (Russell Wallace) wrote in message news:<3e5c7d0b.55021888@news.eircom.net>... D > On 25 Feb 2003 15:05:37 -0800, xeio77@hotmail.com (Tsarkon) wrote: >  > >Torvalds wrote D > >that Intel had made the same mistakes "that everybody else did 15G > >years ago" when RISC architecture was first appearing. Itanium tries C > >to introduce an architecture that is clean and technically pure, H > >something that just doesn't seem to work in the real world. He claimsE > >that Intel "threw out all the good parts of the x86 because people I > >thought those parts were ugly. They aren't ugly, they're the 'charming " > >oddity' that makes it do well." > G > It seems to me that this is nonsense. I've yet to see anyone point to H > any feature of x86 that's good (other than by comparison with the evenE > worse CISC architectures like 68k and VAX). x86 is fast despite its H > quirks, because of the vast resources that have been thrown at it. And > the same is true of IA64. C Save one thing - no one buys the IS64 and it sucks a lot worse than E IA32., I have owned and used quite a number of systems, and I have to D say, after some limited exerience with both Itanium1 and 2 machines,9 you would be a fruit not to like just about nything else.    ------------------------------    Date: 27 Feb 2003 02:30:21 -0800" From: xeio77@hotmail.com (Tsarkon)> Subject: Re: Itanium "threw out all the good parts of the x86"= Message-ID: <baa86a4a.0302270230.1eee50d8@posting.google.com>   g JF Mezei <jfmezei.spamnot@vl.videotron.ca> wrote in message news:<3E5D06FF.B29A367F@vl.videotron.ca>...  > Russell Wallace wrote:I > > It seems to me that this is nonsense. I've yet to see anyone point to J > > any feature of x86 that's good (other than by comparison with the evenG > > worse CISC architectures like 68k and VAX). x86 is fast despite its J > > quirks, because of the vast resources that have been thrown at it. And > > the same is true of IA64.  > L > The 8086 is perhaps easier to upgrade because it is, in its core, a simpleN > architecture. Thi makes it easier to review/update than some bloated complex) > architecture that tries to do too much.  > O > I.E. 8086 gets its speed through hardware's raw speed (aka: Mhz). Other chips J > get their speed through design that makes each cycle do a lot more work. > P > Alpha seemed to have the right balance of fancy features and clean design thatN > would have allowed it to move much faster had its owner given it the budgets( > and allowed it to be sold more widely. > O > IA64 seems to be a complex design that tries to do a lot, but not enough (the N > compiler has to do a lot of the work) and Intel won't be able to upgrade theP > hardware speed of a bloated/complex chip as fast as it has been able to update > the 8086.   F You can make FP better by throwing transistors at the problem. This isA not true of integer, you cant be nearly as creative with it. This D being the case, IA64 tends to completely suck at integer - and there@ is lot of that to be crunched. Sure, Itanium has a lot of cutsieD things to help move FP along nicely, but start doing 'simple' thingsD (most cpu cyckes are spun in most implementations on integer) and it starts to blow.   E So much for the superchip. Too bad Alpha is a better architecture and  Carly shithead runs HP.    ------------------------------  % Date: Thu, 27 Feb 2003 14:19:46 +0000 ' From: Andrew Harrison SUNUK Consultancy 8 Subject: Re: Linus trashes Itanium - he's scared of VMS!. Message-ID: <3E5E1E82.8080307@nospamn.sun.com>   Bill Todd wrote:? > "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in message + > news:3E5D17C1.EFC0550B@vl.videotron.ca...  > * >>Andrew Harrison SUNUK Consultancy wrote: >>4 >>>IA32 is a speed deamon its design allows it to be6 >>>clocked faster and implimented in a smaller process$ >>>earlier than a brainiac like IPF. >>I >>Oh my god. The end of the world is close because I find myself agreeing  >  > with >  >>Andrew Harrisson.  >  > N > I'm afraid Andrew's not worth agreeing with this time.  Itanic2's problem isA > not lack of competitive performance, but lack of efficiency (in F > power/required cooling, chip area/manufacturing cost, expenditure ofB > development funding and effort, requirement for more compilationF > optimization and updates thereof...).  And of course the lack of theB > ancillary on-chip performance and scaling features that EV7 has. >   : Interesting in fact you are agreeing with me on your first point and the cost points.  7 If IPF could be manufactured in a 130 nanomater process 7 then it would use a lower power budget and assuming the 4 same wafer size a smaller process would produce more" CPU's per wafer reducing its cost.  2 But currently it cannot because of its complexity.   Regards  Andrew Harrison    ------------------------------  % Date: Thu, 27 Feb 2003 14:15:06 +0000 ' From: Andrew Harrison SUNUK Consultancy 8 Subject: Re: Linus trashes Itanium - he's scared of VMS!. Message-ID: <3E5E1D6A.5050609@nospamn.sun.com>   Rob Young wrote:Z > In article <3E5CE423.8030709@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy writes: >  >> >>Rob Young wrote: >>[ >>>In article <3E5CD80F.3000901@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy writes:  >>>  >>>  >>>>Rob Young wrote: >>>  >>> G >>>>>	Now Andrew here is where you are clearly wrong.  No where in this  >>>>>	paragraph: >>>>> J >>>>>"I don't think so. According to Intel [1], the highest clockfrequencyF >>>>>for a 0.18um part is 2GHz (both for Xeon and P4, for Xeon MP it'sJ >>>>>1.5GHz). The highest reported SPECint for a 2GHz Xeon seems to be 701D >>>>>[2]. In comparison, a 1GHz McKinley gets a SPECint of 810 [3]." >>>>>  >>>>>	C >>>>>	Does he mention 2.4 GHz parts and the feature size to go with D >>>>>	it?  To assume that David Mosberger doesn't know about 2.4 GHzE >>>>>	or is ignoring it wouldn't be correct either.  Read the threads D >>>>>	Andrew.  But you aren't interested in that are you?  You're inE >>>>>	a twisty little passage and intend to debate semantics and what B >>>>>	not as you have nothing other than bluster.  Typical Andrew. >>>>>  >>>>@ >>>>No but Rob even you must be able to work out the a statementA >>>>that says "the highest clock frequency available for" implies ? >>>>that CPU's above that frequency of whatever clock speed bet 1 >>>>it 2.1, 2.2, 2.3 or even 2.4 arn't available.  >>>> >>>  >>> 2 >>>	Ah Andrew, you are almost there!!! Let's help: >>> 1 >>>	"the highest clock frequency available for"    >>> 7 >>>	for what?  Did Andrew leave something out?  Andrew  6 >>>	wouldn't do that would he?   Why, methinks he did! >>> 9 >>>	"highest clock frequency for a 0.18 um part is 2 GHz"  >>>  >>>	Does that help?  >>>  >>8 >>No because its a BS debating point it has no impact on >>the actual available parts.  >> >  > ; > 	Can you show me where in the context of that discussion  4 > 	they were debating whether any P4 was faster than? > 	Itanium and Mosberger claimed otherwise?   No.  Whether they . > 	were comparing fastest shipping parts?  No. > > > 	The crux of the debate that Linus so obviously falls short,C > 	was slamming Itanium.  Based on what merits?  Well Linus waffles B > 	and has nothing regardless of direction he swings the debate.   >   : Linus is slamming Itanium because its actually slower than Xeons on SPECint.   : Mosberger is claiming that at a given process size Itanium; is better than Xeon therefore Itanium is a better processor  technically.  ? Its a BS point because the design of the CPU materially effects ; the process size it can be implimented in. And IPF's design > reduces its ability to exploit the latest and greatest process technology.   = For both Xeon a P4 this isn't the case and since they deliver 9 better SPECint performance and TPC-C for that matter they  win.  8 They are also much cheaper than Itanium and will be much9 cheaper for the forseable future because they have a much : smaller die size so Intel at any given process will always3 be able to produce more Xeons per wafer than IPF's.        > 	Mosberger writes: > @ > http://www.ussg.iu.edu/hypermail/linux/kernel/0302.2/2020.html > D >   Linus> So your "apples to apples" comparison isn't exactly that. > D > I never claimed it's an apples to apples comparison. But comparingF > same-process chips from the same manufacturer does make for a fairerH > "architectural" comparison because it factors out at least some of theH > effects caused by volume (there is no reason other than (a) volume andF > (b) being designed as a server chip for Itanium chips to come out on; > the same process later than the corresponding x86 chips).  >   C No it doesn't because it ignore that fact that processor architects ? also make design choices that influence the process size that a  CPU can be manufactured in.   F > Linus> The only thing that is meaningful is "performace at the same ( > Linus> time of general availability".  > D > You claimed that x86 is inherently superior. I provided data that F > shows that much of this apparent superiority is simply an effect of E > the larger volume that x86 achieves today. Please don't claim that  G > x86 wins on technical grounds when it really wins on economic grounds  >   G No you didn't you provided a reference that claimed that IPF was faster A than IA32 at a given process size. This is irrelevant because IPF A cannot curently be build in a less than 180 nanometer process and " Xeons can because of their design.  B As I said earlier no one buys CPU's because of their process size,= well actually thats not entirely true because the smaller the ; process the lowe the power cons but they certainly don't on  a performance basis.   Regards  Andrew Harrison    ------------------------------  % Date: Thu, 27 Feb 2003 14:48:47 +0000 ' From: Andrew Harrison SUNUK Consultancy 8 Subject: Re: Linus trashes Itanium - he's scared of VMS!. Message-ID: <3E5E254F.4080506@nospamn.sun.com>   Zane H. Healy wrote:R > Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote: > 1 >>How about the c librarys only a small amount of ) >>code here !!! ever heard of glibc ?????  >  > I > As a matter of fact I have heard of glibc.  However, take a look at how J > *pathetic* of a state gcc was in with relation to header files and libc,N > prior to Linux.  It was *bleeping* hard to build stuff on Linux in the earlyM > days, in part becuase gcc was such a mess.  As far as I'm concerned GNU has M > benefited *FAR* more from Linux than Linux has benefited from GNU.  Without D > Linux and *BSD how much attention would the GNU stuff have gotten? >   - Without BSD probably not that much but by the . time Linux came along it was well established.       > 4 >>Most of the network stack from routing to dhcp/nfs9 >>nis/bind, ftpd etc all comes from non Linux sources. Or 4 >>at least they do in RedHat I cannot be bothered to >>look at the other distros. >  > K > Networking on Linux can be a bit of a scarry mess between distro's.  I've L > done a lot of hair pulling of late thanks to that.  Make that dealing with1 > the differences between distro's can be a mess.  >  > C >>In reality only about 5% of a Linux disto is actually what people = >>think of as Linux, the rest is aquired from GNU, Apache.org  >>Samba, FreBSD, Sun etc etc.  >  > J > Did I say otherwise?  What I said is that GNU is a very small portion ofL > what makes up a Linux distribution, and other than gcc it isn't that vitalL > of one.  Keep in mind though that most Opensource developement is now doneK > on Linux, and as such getting apps to compile on other platforms (such as $ > OpenBSD or Solaris) can be tricky. >  > ; >>Without the existing OpenSource/Freeware apps Linux would : >>not exist because the amount of work required to develop; >>their equivalents would have killed the project before it  >>even got started.  >  > F > The fact is, without the various Opensource/freeware apps, the wholeL > computing arena would be a lot different.  Take a look at Solaris, how youL > all are getting rid of CDE in favor of Gnome (is Motif being pushed to theL > side as well).  Pretty much everyone is using Apache and Samba these days, > etc.   >   = Yup but Gome is a graphical user interface which most Solaris > systems don't have because they are servers. Gnome isn't going& to make much difference in this space.  ? And yes Sun supplies Apache but we also have our own WebServer.      Regards  Andrew Harrison    ------------------------------    Date: 27 Feb 2003 09:42:38 -0600+ From: young_r@encompasserve.org (Rob Young) 8 Subject: Re: Linus trashes Itanium - he's scared of VMS!3 Message-ID: <kcLUVGKrmxsP@eisner.encompasserve.org>   X In article <3E5E1D6A.5050609@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy writes: >  >  >>  < >> 	Can you show me where in the context of that discussion 5 >> 	they were debating whether any P4 was faster than @ >> 	Itanium and Mosberger claimed otherwise?   No.  Whether they/ >> 	were comparing fastest shipping parts?  No.  >>  ? >> 	The crux of the debate that Linus so obviously falls short, D >> 	was slamming Itanium.  Based on what merits?  Well Linus wafflesC >> 	and has nothing regardless of direction he swings the debate.    >>   > < > Linus is slamming Itanium because its actually slower than > Xeons on SPECint.  > < > Mosberger is claiming that at a given process size Itanium= > is better than Xeon therefore Itanium is a better processor  > technically. > A > Its a BS point because the design of the CPU materially effects = > the process size it can be implimented in. And IPF's design @ > reduces its ability to exploit the latest and greatest process
 > technology.  >   A 	Nonsense.  IA64 will be trailing Xeon process for quite a while. 7 	Mosberger addressed process generation in the threads.    >  >> 	Mosberger writes:  >>  A >> http://www.ussg.iu.edu/hypermail/linux/kernel/0302.2/2020.html  >>  G >> Linus> The only thing that is meaningful is "performace at the same  ) >> Linus> time of general availability".   >>  E >> You claimed that x86 is inherently superior. I provided data that  G >> shows that much of this apparent superiority is simply an effect of  F >> the larger volume that x86 achieves today. Please don't claim that H >> x86 wins on technical grounds when it really wins on economic grounds >>   > I > No you didn't you provided a reference that claimed that IPF was faster % > than IA32 at a given process size.    ; 	Linus and Mosberger claiming and counter-claiming.  Again, A 	it appears that Mosberger has the upper hand on any given point. ; 	Can you show otherwise?  Any lift quotes?  Betcha can't!     G 	Anyhow. . . that point (IPF faster than IA32 at a given process size)  ? 	is addressed in the threads.  (Focus Andrew:  "process", don't * 	dig up your 2.4 GHz Xeon.  Stay focused).  > http://www.ussg.iu.edu/hypermail/linux/kernel/0302.2/2020.html  5   Linus> On Sun, 23 Feb 2003, David Mosberger wrote:       >> 2 GHz Xeon: 701 SPECint    >> 1 GHz Itanium 2: 810 SPECint     % >> That is, Itanium 2 is 15% faster.      /   Linus> Ehh, and this is with how much cache?      =   Linus> Last I saw, the Itanium 2 machines came with 3MB of  @   Linus> integrated L3 caches, and I suspect that whatever 0.13 @   Linus> Itanium numbers you're looking at are with the new 6MB    Linus> caches.      F Unfortunately, HP doesn't sell 1.5MB/1GHz Itanium 2 workstations, but " we can do some educated guessing:   )   1GHz Itanium 2, 3MB cache: 810 SPECint  -   900MHz Itanium 2, 1.5MB cache: 674 SPECint      B Assuming pure frequency scaling, a 1GHz/1.5MB Itanium 2 would get E around 750 SPECint. In reality, it would get slightly less, but most  $ likely substantially more than 701.      ----  ; 	Get it?  Is there a single lift quote from that discussion E 	you can trot out to prove your point?  Any point?!?  Betcha can't!     	Because there aren't any there.    > This is irrelevant because IPF? > cannot curently be build in a less than 180 nanometer process     - http://www.eetimes.com/story/OEG20020909S0083   ; 	"Intel is sampling "Madison," with shipments due in 2003."   ? 	Elsewhere we find, Madison and Deerfield are at 130 nanometer:   M Both Deerfield and Madison are 0.13-micron versions of McKinley, Fister said. L The Deerfield 64-bit processor is a low-power chip geared for dual-processorO applications. The chip is only 70-80 Watts; in comparison, McKinley runs at 130  Watts.     	So when you say:       > This is irrelevant because IPF? > cannot curently be build in a less than 180 nanometer process   ; 	You really mean to say, they aren't shipping.  To say they B 	can't build it, is incorrect.  But the fact that IPF is currently; 	shipping at .18 and Xeon is at .13 is natural.  Just like  9 	Prescott (desktop) will beat Xeon and IPF to .09 micron.    				Rob    ------------------------------  + Date: Thu, 27 Feb 2003 15:22:39 +0000 (UTC) , From: lewis@PROBE.mitre.org (Keith A. Lewis)8 Subject: Re: Linus trashes Itanium - he's scared of VMS!- Message-ID: <b3lafv$3v$1@newslocal.mitre.org>   } Andrew Harrison SUNUK Consultancy writes in article <3E5E1D6A.5050609@nospamn.sun.com> dated Thu, 27 Feb 2003 14:15:06 +0000: ; >Linus is slamming Itanium because its actually slower than  >Xeons on SPECint.  ? If SPECint is designed to measure performance of 32-bit integer @ manipulation, the 32-bit CPUs should always win or at least tie.  9 >They are also much cheaper than Itanium and will be much : >cheaper for the forseable future because they have a much; >smaller die size so Intel at any given process will always 4 >be able to produce more Xeons per wafer than IPF's.  E Your facts are correct, but I think you may have the cause and effect L reversed. Xeon isn't cheaper because it's smaller; it's smaller because it's cheaper.  H The trend in server CPUs is to load them up with huge amounts of cache. J Intel is aiming the Itanium 2 at a higher-priced market than Xeon, so theyK put 6 times as much cache on it.  Of course it's going to have a larger die  size.     + --Keith Lewis              klewis$mitre.org > The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Thu, 27 Feb 2003 17:30:13 +0000 ' From: Andrew Harrison SUNUK Consultancy 8 Subject: Re: Linus trashes Itanium - he's scared of VMS!. Message-ID: <3E5E4B25.8070509@nospamn.sun.com>   Keith A. Lewis wrote:  > Andrew Harrison SUNUK Consultancy writes in article <3E5E1D6A.5050609@nospamn.sun.com> dated Thu, 27 Feb 2003 14:15:06 +0000:  > < >>Linus is slamming Itanium because its actually slower than >>Xeons on SPECint.  >  > A > If SPECint is designed to measure performance of 32-bit integer B > manipulation, the 32-bit CPUs should always win or at least tie. >   2 I am not quite sure what your point is here. Alpha2 has always been 64 bit and it doesn't ever seem to- have suffered in the SPEC performance stakes.   4 AMD claim that the x86-64 native compilers producing3 64 bit code will improve Hammer performance by ~20% % over the current code which is 32bit.   1 Moving from Solaris 2.6 (32bit) to Solaris 7 or 8 4 on the same hardware platform and then running Forte3 64 bit compilers has generally improved SPECint and , SPECfp performance rather than decreased it.  : >>They are also much cheaper than Itanium and will be much; >>cheaper for the forseable future because they have a much < >>smaller die size so Intel at any given process will always5 >>be able to produce more Xeons per wafer than IPF's.  >  > G > Your facts are correct, but I think you may have the cause and effect N > reversed. Xeon isn't cheaper because it's smaller; it's smaller because it's
 > cheaper. >   9 I am note sure what you point is. Xeon has a much smaller : transitor count than IPF, this means that it has a smaller< die size on the same process technology as IPF and given the1 same wafer size it will have a much higher yield.   8 This higher yield plus a much smaller package results in4 a lower cost. Xeon is cheaper because it is smaller.  : When XeonMP was designed it was the most complex expensive; chip produced by Intel and their premium product. It wasn't 2 designed to be cheap though it is relative to IPF.  J > The trend in server CPUs is to load them up with huge amounts of cache. L > Intel is aiming the Itanium 2 at a higher-priced market than Xeon, so theyM > put 6 times as much cache on it.  Of course it's going to have a larger die 	 > size.    >   < Well part of the IPF die is obviously consumed by the larger@ cache but that doesn't account for the whole of the differential between IPF and Xeon.        regards  Andrew Harrison    ------------------------------  + Date: Thu, 27 Feb 2003 18:52:45 +0100 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> K Subject: Re: memory errors and bugcheck dump (PROCGONE) with ALPHA 3000/600 ; Message-ID: <01KSXX21E98Y9FNHX8@sysdev.deutsche-boerse.com>   C > :I have a nice DEC 3000/600 (21064 ALPHA) with 192 MB and (still) D > :VMS 7.2-1.  It works fine, except occasionally after power up the$ > :self-test gives a memory error.   > C >   I'd find and fix the error -- the default power-up self-test is * >   not an exhaustive memory test, either.  H OK, I've gotten a few tips on what to look for with an extensive memory I test.  Actually, I have some replacement memory for this box, so I could  1 perhaps actually replace some SIMMs if necessary.   C >   PROCGONE can arise for various reasons -- basically, one of the D >   key processes has failed.  There is a code left in R0 that mightA >   help, but these errors can range from hardware errors to disk < >   fragmentation to configuration problems to OpenVMS bugs.  D The disk is not fragmented (I run an online defragger), I'm sure the@ configuration is OK, the same version of VMS is running on otherD systems.  So I suspect it is a hardware problem; maybe just a loose  connection.   B >   Show the error message text, and I'd be more willing to hazardD >   an answer beyond a generic bad memory, bad processor, bad cache,E >   or a bad motherboard.  :-)  (And we both know that you had really A >   intended to include the actual error message text within your  >   posting, too. :-)   H Well, cut-and-paste from the VT320 console is a bit difficult!  I still 7 have a pencil somewhere, though, so I'll write it down.   K >   As you are apparently -- and no offense is intended -- not particularly L >   comfortable with hardware troubleshooting, I'd tend to leave the systems >   powered on.     A I think that is the best bet.  I am not a hardware guy, but I do  F actually enjoy opening up boxes and doing stuff.  However, as this is F one of my only two ALPHAs at home, I don't want to risk making things E worse.  The box is about 1994 vintage and perhaps just opening it up  D would kick up enough dust to cause more problems than my opening it  could solve.  " >   Again, no offense is intended.   None taken.  :-)  E I have a lot of hardware manuals.  It would be nice if, in each one,  F there is a time indicated such that if the system is not needed for a H longer time, then it is better to switch it off.  (Like with cars there J is the rule of thumb that one cold start is about the same load as 300 km  of normal driving.)   K >   Hardware bugs are hardware bugs.  They can stay flakey and intermittent J >   for years (I really dislike these), or they can suddenly fail hard (onI >   the balance, this class of hardware bug is preferable), but -- once a K >   hardware bug appears -- it then seldom really disappears.  I'd first be I >   looking for the specific cause of the memory error that was reported,  >   in other words.   I OK.  Maybe when I open the box I'll discover that I can actually ADD and  E not just REPLACE memory.  Can't use those SIMMs for anything else, I  H suppose (in particular not for my 255/233, which only has 64 instead of  192 like the 3000/600).    ------------------------------    Date: 27 Feb 2003 12:51:26 -0600B From: clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley)6 Subject: MIME utility, was: Re: Mis-set Terminal Types3 Message-ID: <r3y3Z$g$sP5O@eisner.encompasserve.org>   f In article <b3lisi$1ninle$1@ID-141708.news.dfncis.de>, "Peter Weaver" <peter.weaver@stelco.ca> writes: > I > Don't forget MCR MIME, at least it will give a MIME> prompt, but little  > else.  >   E Judging by the types of problems fixed in the MIME ECO kits, giving a 9 MIME> prompt appears to be about all it's good for... :-)   K On a more serious note, how did a piece of software of this, erm, "quality"  become a part of VMS ?   Simon.   --  B Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP       L VMS advocate: One who makes a Mac advocate look like a beginner at advocacy.   ------------------------------  % Date: Thu, 27 Feb 2003 12:45:47 -0500 - From: "Peter Weaver" <peter.weaver@stelco.ca> T Subject: Re: Mis-set Terminal Types (was: Re: HUGE showstopper *bug* in HELP/MESSAGE6 Message-ID: <b3lisi$1ninle$1@ID-141708.news.dfncis.de>   Hoff Hoffman wrote:  >...4 >   LSEDIT, EVE, and TPU text editors will all fail. > ? >   EDT will stay in line-mode; no (functional) CHANGE command.  > 8 >   HELP will dump output without regard to page length. > C >   SHOW CLUSTER/CONTINOUS is mildly entertaining, if you're bored.  > ; >   And MONITOR output is, well, not what you might expect.  >...  G Don't forget MCR MIME, at least it will give a MIME> prompt, but little  else.    -- Peter WeaverD Opinions are my own, and do not reflect the opinions of my employer,A nor the company that it sub-contracts to, nor the company that it  sub-contracts to.    ------------------------------  % Date: Thu, 27 Feb 2003 11:18:05 -0500 1 From: "Farrell, Michael" <MFarrell@voltdelta.com> 1 Subject: More Tar problems (was  VMSTAR question) P Message-ID: <025766C9BBC5D511A4ED00B0D0F08C23163A1E@zny_exchange1.maintech1.com>  E Thanks for everyone's help yesterday.  We solved the looping problem.   H Now I have a new problem.  The files we want to archive in TAR contain a- great deal of binary data.  When we TAR them, L The resulting TAR file is generally less than 1% of the size of the originalF file.  (E.G. a 50000 block data file ends up producing a 360 block TARL file).  Obviously, the data is being lost somehow.  The commands I have used are:  5 $ tar/create/verbose <tar file name> <data file name>    Or  ; $ tar/create/force/verbose <tar file name> <data file name>   # The results are the same with both.   " Any ideas are greatly appreciated.   TIA    Mike Farrell    -----Original Message----- 8 From: 	Farrell, Michael [mailto:MFarrell@voltdelta.com] + Sent:	Wednesday, February 26, 2003 12:30 PM  To:	Info-VAX@Mvb.Saic.Com  Subject:	VMSTAR question  1 I'm hoping someone can help us out with this one.   ; We have to supply data to one of our customers using TAR.   5 So we have installed VMSTAR and it is VMS TAR V3.3-4. " When I give the following command:  : tar -cwvf output.tar PB12_020301_01R_CAP_XXXX_LD_0001.CAD   G the program just loops constantly repeating the input filename over and  over.   J When I kill this run, and look at the tar file using DUMP, it appears thatJ header information for the data file is written over and over again.  None2 of the input file's data is found in the TAR file.  1 This is running on an ES45 running OpenVMS 7.3-1.    Has anyone else seen this?   TIA    Mike Farrell   ------------------------------  % Date: Thu, 27 Feb 2003 10:33:57 -0000 - From: "Steve Spires" <Steve.Spires@torex.com> @ Subject: Odd [to me] error when trying to map a pathworks share.E Message-ID: <91947A84607D9D48B8E674A5FAB54DA63CB208@tahiti.tinuk.com>   = When users try to map the share, they get this error message;   = "the following error has occurred, whilst re-connecting E: to F \\STK000\FOCUS the local device type and the network resource type are not the same."  $ Can anyone give me a pointer or two?  H ISTR when I last used Pathworks seriously (4.1 perhaps?) that one of theB manuals had sections containing all the various error messages andA possible resolutions. Has this manual disappeared over succeeding 	 versions?   F Not the best reintroduction, trying to fix a bodged upgrade from V5 toF V6 [or should I say MIGRATION! ;^D ] and no one to blame but myself...   Thanks   Steve Spires Technical Consultant Torex Health [T] +44 (0) 1295 274200  [F] +44 (0) 1295 275131 
 www.torex.com    ------------------------------  % Date: Thu, 27 Feb 2003 08:11:10 -0500 5 From: "Brad McCusker" <brad.mccuskerNosp@Mcompaq.com> D Subject: Re: Odd [to me] error when trying to map a pathworks share./ Message-ID: <v5s3jl7ktf3k9f@corp.supernews.com>   H Have you been able to make any connections?  Like to the \\PWUTIL share?  C Do you have licenses correctly installed, and the correct versions?   K I think this error might occur if you don't have a license available.  Does  ADMIN/ANAL show you anything?   K Assuming this is the only PATHWORKS Server, don't get bogged down trying to G run the license server.  In ADMIN/CONFIG, choose to NOT run the license L server.  By doing that, you will then use server based licensing and as longI as you have a PWLMXXXCA06.00 or higher (For V6.0*) or PWLMXXXCA07.03 (for $ V6.1) then you should eb good to go.      8 "Steve Spires" <Steve.Spires@torex.com> wrote in message? news:91947A84607D9D48B8E674A5FAB54DA63CB208@tahiti.tinuk.com... = When users try to map the share, they get this error message;   = "the following error has occurred, whilst re-connecting E: to F \\STK000\FOCUS the local device type and the network resource type are not the same."  $ Can anyone give me a pointer or two?  H ISTR when I last used Pathworks seriously (4.1 perhaps?) that one of theB manuals had sections containing all the various error messages andA possible resolutions. Has this manual disappeared over succeeding 	 versions?   F Not the best reintroduction, trying to fix a bodged upgrade from V5 toF V6 [or should I say MIGRATION! ;^D ] and no one to blame but myself...   Thanks   Steve Spires Technical Consultant Torex Health [T] +44 (0) 1295 274200  [F] +44 (0) 1295 275131 
 www.torex.com    ------------------------------  % Date: Thu, 27 Feb 2003 11:43:56 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>* Subject: Re: Official OpenVMS name for the/ Message-ID: <3E5E323B.8770E201@vl.videotron.ca>    Robert Deininger wrote: O > The customers might like some way to distinguish between the various flavors.   M Then let the customer call it what THEY want. There is no need for Digital/HP J to officially give VMS different names. VMS is VMS, especially if the codeN base will be identical between Alpha and that other thing. Identical products,K but different names ? You'll get plenty "how difficult is the port from the , real VMS to that industry standard system ?"  J (I was going to use the word "upgrade", but since IA64 isn't an upgrade, I  won't use the word "upgrade" :-)  N Furthermore, when you look in the long term, (if VMS survives), there won't beJ VAX or Alpha anymore. So there is no point in attaching yet another set ofN chains to VMS's feet, chains that will remain, just like the stupid Open seems) to remain even against customer's wishes.   K In Qubec, the wife keeps her mainden name. The kids first get a hyphenated L family name (father's surname-mother's surname, and once old enough, the kidE gets to choose between one OR the other, otherwise, you'd end up with * EXTREMELY long names in a vew generations.  K VMS must learn to drop the useless advectives otherwise, it will really get  tedious over time.   ------------------------------  % Date: Thu, 27 Feb 2003 11:58:02 -0500 & From: John Reagan <john.reagan@hp.com>* Subject: Re: Official OpenVMS name for the/ Message-ID: <nns7a.585$kf6.22@news.cpqcorp.net>    JF Mezei wrote:  > O > Then let the customer call it what THEY want. There is no need for Digital/HP * > to officially give VMS different names.   ) I like that.  Make it a SYSGEN parameter.   I Kinda like the Joe Walsh album many years ago titled "You bought it, you  	 name it".      --   John Reagan ' Compaq Pascal/{A|I}MACRO Project Leader  Hewlett-Packard Company    ------------------------------    Date: 27 Feb 2003 10:47:40 -07001 From: nothome@spammers.are.scum (Malcolm Dunnett) * Subject: Re: Official OpenVMS name for the- Message-ID: <nQMBET9XYhYZ@malvm7.mala.bc.ca.>   0 In article <nns7a.585$kf6.22@news.cpqcorp.net>, +    John Reagan <john.reagan@hp.com> writes:    > JF Mezei wrote:  >>  P >> Then let the customer call it what THEY want. There is no need for Digital/HP+ >> to officially give VMS different names.   > + > I like that.  Make it a SYSGEN parameter.  > K > Kinda like the Joe Walsh album many years ago titled "You bought it, you   > name it".  > ;     Might as well. We've had clustering here since 1986 and I nothing but Alphas in the cluster for the last 3 or 4 years, but everyone L in the organization still refers to our Alpha-only VMS cluster as "The VAX".G I suppose they'll still be doing that when it's all IA64s ( if that day  ever really comes to pass ).   ------------------------------  % Date: Thu, 27 Feb 2003 14:25:49 +0000 ' From: Andrew Harrison SUNUK Consultancy Y Subject: Re: Official OpenVMS name for the =?ISO-8859-1?Q?Itanium=AE_re?= =?ISO-8859-1?Q? . Message-ID: <3E5E1FED.1020700@nospamn.sun.com>   JF Mezei wrote: * > Andrew Harrison SUNUK Consultancy wrote: > A >>I have no idea if Sun is considering using Opteron, though from B >>its SPECs we would have to consider it for our x86 Solaris/Linux >>boxes. >  > O > Considerin the raucus when Sun announced that it wouldn't continue to upgrade L > the 8086 version of Solaris, is it fair to state that Sun will continue to$ > keep the 8086 version at 32 bits ? > N > Or do you see the day coming shortly that Sun would go Hammer at 64 bits and1 > then truly drop the 32 bit version of Solaris ?   7 AMD have tested Solaris x86 on Hammer and it apparently 2 runs unmodified. So we could simply stick with the 32 bit port.  3 However Solaris is a 64 bit OS and so porting it to 5 AMD shouldn't be that hard particularly as it is only 1 an extension of the existing 32 bit architecture.   1 Solaris also has good support for mixed 32/64 bit 1 apps so this also wouldn't be a big issue either.   3 I doubt that we would drop x86 32bit quickly simply 0 because its unrealistic to expect a total switch5 to Hammer by all Intels 32 bit OEMS. And also because 4 we would still have to maintain the 32 bit libraries3 etc as we do with Solaris on UltraSPARC in order to ( provide full 32bit binary compatibility.   Regards  Andrew Harrison    ------------------------------  % Date: Thu, 27 Feb 2003 09:34:50 -0500 & From: jlsue <jlsuexxxz@screaminet.com>Y Subject: Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retai 8 Message-ID: <eq6s5vsgqeagsnsqbjoibqtq8hbfbd1jdd@4ax.com>  E On Wed, 26 Feb 2003 14:01:13 +0000, Andrew Harrison SUNUK Consultancy . <Andrew_No.Harrison_No@nospamn.sun.com> wrote:   >  > 
 >jlsue wrote: H >> On Mon, 24 Feb 2003 13:14:20 +0000, Andrew Harrison SUNUK Consultancy1 >> <Andrew_No.Harrison_No@nospamn.sun.com> wrote:    >>  G >> I see, so what you're saying now is that customers should make their 7 >> buying decisions based on these kinds of benchmarks?  >>  E >> Sorry, I don't advise customers that way.  We try to get their own I >> apps benchmarked to demonstrate the value they will receive from their  >> investment. >>   > < >But you cannot provide any actual example of this or havn't >been able to so far.  >   9 Exactly what am I saying that you're asking for proof on? D Is it that "there are certainly customers who have GS160/320 servers: who are very happy with the performance of their systems?"  = What?  Is this claim so extraordinary that it requires proof?   C Your statements are more of a general "it's slower than gs140" as a E way to warn folks away from it.  You state implicitly (and iirc, even F explicitly at times) that it's  That does qualify, imho, as an exampleF of an extraordinary claim; therefore I ask for proof that this is trueE in a general sense.  And so far you've only proven that it's true for D *some* applications.  If you can't support your claim that it's true; in general, then withdraw it, or at least re-qualify it.      E There are definitly posts in this very group which claim how slow Sun > servers are on their applications.  Lots of them.  Go look for	 yourself.    ------------------------------  % Date: Thu, 27 Feb 2003 11:58:20 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>Y Subject: Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retai / Message-ID: <3E5E359A.79CE39C7@vl.videotron.ca>    jlsue wrote:F > Is it that "there are certainly customers who have GS160/320 servers< > who are very happy with the performance of their systems?"  J It all has to do with expectations. Singapore Airlines dumped their AirbusM A340-300 because they didn,t quite perform as promised (had to fly slowler to K achieve the promised range). But other airlines are perfectly happy with it L because they are able to operate it in a way that is more effiecent and thusK saves them money even though the plane is slower than 747 and 777s. The new F batch of 340s (500s and 600s) have had their speed increased to regain competitive edge).  B I think that the wildfire stuff is probably similar. Those who hadN promises/expectations set high enough would be disapointed. But those who came$ without expectations would be happy.  L If someone purchased the 340 with full knowledge of its slower-than-expectedM cruise speed, he is more likley to be happy. If someone purchased it prior to K it being produced, based on promises made by the vendor, then if the vendor E didn't deliver on certain key metrics, the customer would be unhappy.    ------------------------------  % Date: Thu, 27 Feb 2003 16:55:28 +0000 ' From: Andrew Harrison SUNUK Consultancy Y Subject: Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retai . Message-ID: <3E5E4300.8090106@nospamn.sun.com>   jlsue wrote:G > On Wed, 26 Feb 2003 14:01:13 +0000, Andrew Harrison SUNUK Consultancy 0 > <Andrew_No.Harrison_No@nospamn.sun.com> wrote: >  >  >> >>jlsue wrote: >>H >>>On Mon, 24 Feb 2003 13:14:20 +0000, Andrew Harrison SUNUK Consultancy1 >>><Andrew_No.Harrison_No@nospamn.sun.com> wrote:  >> > G >>>I see, so what you're saying now is that customers should make their 7 >>>buying decisions based on these kinds of benchmarks?  >>> E >>>Sorry, I don't advise customers that way.  We try to get their own I >>>apps benchmarked to demonstrate the value they will receive from their  >>>investment. >>>  >>= >>But you cannot provide any actual example of this or havn't  >>been able to so far. >> >  > ; > Exactly what am I saying that you're asking for proof on? F > Is it that "there are certainly customers who have GS160/320 servers< > who are very happy with the performance of their systems?" > ? > What?  Is this claim so extraordinary that it requires proof?  >   8 Yes because your own engineers say that there are issues6 your own benchmark results illustrate these issues and? customers on this newsgroup have benchmarked their applications # and discovered them for themselves.   = At the moment the score is 5 for the argument that GS160/320s E do in fact have performance issues because of their NUMA architecture  and 0 against.  < Unless that is you can substantiate your claim to never have9 seen any issues with results from the benchmarks that you < apparently advise your customers to do the score will remain	 as it is.   E > Your statements are more of a general "it's slower than gs140" as a G > way to warn folks away from it.  You state implicitly (and iirc, even H > explicitly at times) that it's  That does qualify, imho, as an exampleH > of an extraordinary claim; therefore I ask for proof that this is trueG > in a general sense.  And so far you've only proven that it's true for F > *some* applications.  If you can't support your claim that it's true= > in general, then withdraw it, or at least re-qualify it.     >   9 Hang on you are the parson who claims that you have never 7 seen an issue. I have simply pointed out that there are 8 customers and benchmark results that show that there are issues.   ! Yours is the extraordinary claim.   7 So how about answering this question. Did you know that 7 your own engineering teams were advising your customers 6 that ther were issues transitioning from UMA to NUMA ?  8 And if you did know what was the basis of your engineers advice ?  2 This shouldn't be too difficult for you to answer.   regards  Andrew Harrison    ------------------------------  % Date: Thu, 27 Feb 2003 08:23:10 -0800 $ From: Shane Smith <ssmith@icius.com>V Subject: OT but worth posting, virus warning: "FWD: Look at this pack from Microsoft."0 Message-ID: <01C2DE39.96543B80@sulfer.icius.com>  @ I just got some clueless skript kiddie send me "q414183_exe.vir"B directly to this address, and the mailing list was quite obviouslyB harvested from one or more of info-vax's newsgroups. McAfee didn'tD identify the virus or trojan, but it's pretty obviously some kind ofA crack attempt, so I thought I'd just send out a quick heads up. I G realise probably nobody on this list is dumb enough to fall for it, but  everyone has off moments.   E He tried to hide the origin, but the relay put the genuine IP address H next to the server name he was claiming, so I've reported him in detail.   Shane   >  #####   -----------------------------------------------------? #-O-O-# | Arthur: "It's times like this I wish I'd listened   | ? #  L  # | to what my Mother used to say." Ford: Why, what did | ?  #===#  | she say?" Arthur: "I don't know, I didn't listen."  | >   ###    -----------------------------------------------------   ------------------------------    Date: 27 Feb 2003 10:49:42 -0600+ From: young_r@encompasserve.org (Rob Young) Y Subject: Re: OT but worth posting, virus warning: "FWD: Look at this pack from Microsoft. 3 Message-ID: <0AC83DwPON4a@eisner.encompasserve.org>   W In article <01C2DE39.96543B80@sulfer.icius.com>, Shane Smith <ssmith@icius.com> writes: B > I just got some clueless skript kiddie send me "q414183_exe.vir"D > directly to this address, and the mailing list was quite obviouslyD > harvested from one or more of info-vax's newsgroups. McAfee didn'tF > identify the virus or trojan, but it's pretty obviously some kind ofC > crack attempt, so I thought I'd just send out a quick heads up. I I > realise probably nobody on this list is dumb enough to fall for it, but  > everyone has off moments.  > G > He tried to hide the origin, but the relay put the genuine IP address J > next to the server name he was claiming, so I've reported him in detail. >   > 	A bunch of us on Eisner were on that email list too.  Viruses> 	don't run real well on VMS.  Eisner got it and ate it anyhow.  E 	Looking at the email list, it was an InfoVAX/Usenet harvest coupled   	with some other source.                   		Rob    ------------------------------  % Date: Thu, 27 Feb 2003 14:35:19 +0000 ' From: Andrew Harrison SUNUK Consultancy > Subject: Re: PC developers opinion of 64 bit desktop computing. Message-ID: <3E5E2227.3080405@nospamn.sun.com>   Rob Young wrote:Y > In article <01C2DD96.8AFC0400@sulfer.icius.com>, Shane Smith <ssmith@icius.com> writes:  > 7 >>http://slashdot.org/comments.pl?sid=54835&cid=5371889  >>I >>Tim Sweeny of Epic (big game developer) about Intel saying 64bit on the ' >>desktop is "end of the decade" stuff:  >>H >>"Intel's claims are wholly out of touch with reality. On a daily basisE >>we're running into the Windows 2GB barrier with our next-generation A >>content development and preprocessing tools. If cost-effective, G >>backwards-compatible 64-bit CPU's were available today, we'd buy them D >>today. We need them today. It looks like we'll get them in April." >> >  > I > 	But nothing for the masses to run it on.  Unless there is some hidden  E > 	market for Unreal 2003 on Linux based Opteron servers.  After all, , > 	ClodHammer isn't due out until September: > * > http://www.theinquirer.net/?article=7868 >   ; Well given that all the current Wintel games are 32 bit and 8 use 32 bit content development tools it is going to take. a little time to produce customer ready games.  6 However developers can get Betas of Windows for Hammer1 and even if Althon64 isn't available Opteron will 1 be which will allow the developers to get started 0 allowing AMD to launch the desktop CPU with apps1 in September, if they do then Intel could well be  dead in the gaming space.   9 And one has to assume that the game developers are rather / closer to their market than you are and if they / intend to develop for Hammer they will be doing % so based on a sensible business case.    regards  Andrew Harrison    ------------------------------  % Date: Thu, 27 Feb 2003 08:45:33 -0800 $ From: Shane Smith <ssmith@icius.com>> Subject: RE: PC developers opinion of 64 bit desktop computing0 Message-ID: <01C2DE3C.A5AEDE20@sulfer.icius.com>  H Read the original post, Andrew. There's already at least one 64 bit game nearly ready for release.    Shane    -----Original Message-----' From: Andrew Harrison SUNUK Consultancy . [mailto:Andrew_No.Harrison_No@nospamn.sun.com]) Sent: Thursday, February 27, 2003 6:35 AM  To: Info-VAX@Mvb.Saic.Com > Subject: Re: PC developers opinion of 64 bit desktop computing         Rob Young wrote:> > In article <01C2DD96.8AFC0400@sulfer.icius.com>, Shane Smith <ssmith@icius.com> writes: > 7 >>http://slashdot.org/comments.pl?sid=54835&cid=5371889  >>I >>Tim Sweeny of Epic (big game developer) about Intel saying 64bit on the ' >>desktop is "end of the decade" stuff:  >>H >>"Intel's claims are wholly out of touch with reality. On a daily basisE >>we're running into the Windows 2GB barrier with our next-generation A >>content development and preprocessing tools. If cost-effective, G >>backwards-compatible 64-bit CPU's were available today, we'd buy them D >>today. We need them today. It looks like we'll get them in April." >> >  > I > 	But nothing for the masses to run it on.  Unless there is some hidden  E > 	market for Unreal 2003 on Linux based Opteron servers.  After all, , > 	ClodHammer isn't due out until September: > * > http://www.theinquirer.net/?article=7868 >   ; Well given that all the current Wintel games are 32 bit and 8 use 32 bit content development tools it is going to take. a little time to produce customer ready games.  6 However developers can get Betas of Windows for Hammer1 and even if Althon64 isn't available Opteron will 1 be which will allow the developers to get started 0 allowing AMD to launch the desktop CPU with apps1 in September, if they do then Intel could well be  dead in the gaming space.   9 And one has to assume that the game developers are rather / closer to their market than you are and if they / intend to develop for Hammer they will be doing % so based on a sensible business case.    regards  Andrew Harrison    ------------------------------  % Date: Thu, 27 Feb 2003 11:27:34 -0500   From: John Santos <JOHN@egh.com>* Subject: Re: Pine Email Problem on OpenVMS6 Message-ID: <1030227112621.26948A-100000@Ives.egh.com>  * On Tue, 25 Feb 2003, Barry Skidmore wrote:  G > I just installed Pine 3.91-2 on OpenVMS 7.2 and receive the following K > error when trying to send to an Internet (non-local) email address (local  > email works fine): >=20$ > "Mail not sent.=A0 User not local" >=20A > What configuration change do I need to change to enable sending  > non-local email? >=20	 > Thanks,  > Barry Skidmore    Have you defined an smtp-server?  ( Maybe you should post your .pinerc file.   (Posted with Pine 3.91-2)      --=20  John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  + Date: Thu, 27 Feb 2003 18:01:32 +0100 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> ; Subject: Re: pseudointerface (TCPIP SET [CONFIG] INTERFACE) ; Message-ID: <01KSXWBSF2JA9GVT85@sysdev.deutsche-boerse.com>   C > Well, if 207 is right, yes, but I always use the logical OR of IP E > address and network mask for the broadcast mask. There used to be a I > section in the UCX docs describing how to calculate it like this, but I J > just had a look and it doesn't appear to be in the TCP/IP Services docs. > J > If the network mask is correct you have a very "small" network with onlyH > 3 bits for local addresses. Is this what you intend, Phillip? Then setJ > broadcast mask to 213.252.154.7, and amend allocated IP addresses to fitD > in with this. 207 is a very odd bit pattern to use for this IMHO.   = If the network mask ends in 200, then 207 is all ones, right?   E > Or are you playing super smart tricks with broadcasting? If so, I'd  > appreciate an explaination.   E No.  Well, not intentionally.  Things have been OK for years with an  E ISDN router on my 8-address network (yes, three bits; the network is  F 213.252.154.200, the ISDN router is 213.252.154.206, the broadcast is I 213.252.154.207 and I can use the addresses 201--205 (I actually use 201  2 as a cluster alias)).  I'm pretty sure this is OK.  D This has all worked fine for years.  These addresses are public and H static.  Now, I am moving my hobbyist cluster to DSL.  I will then have H a dynamic IP on the WAN side of the DSL router and private addresses on D the LAN side.  This all works fine (as I verified with my OTHER VMS F cluster, which at the moment is not reachable from the outside world, H but will be soon (and has been briefly for test purposes)).  The router B does NAT and PAT, I have an account at dyndns.org etc.  All works 
 perfectly.  < The only reason I want to use the pseudointerface is, for a H transitionary time, to have both the public and private addresses (with I a router on each network).  If that's not what pseudointerfaces are for,  B then I don't know what they are for.  If the DSL router is on the F private-IP, pseudointerface network, I can connect to it from VMS and C vice versa, but it won't forward incoming connections (of course I  H configured it to do so) to the pseudointerface!  This is true even if I F change the default route to use the pseudointerface (which works fine 3 for outgoing connections and their return traffic).   G (Just to confuse things even more, apparently there is another problem  F with TCPIP 5.1 and perhaps slightly older versions, but not with some E old UCX versions and apparently not with 5.3 (I have 5.1 now): TCPIP  G tries to route all outgoing traffic through the default route, even if  G it is return traffic from an incoming connection via another interface. 3 This is a problem even if all interfaces are real.)    ------------------------------  + Date: Thu, 27 Feb 2003 18:07:25 +0100 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> ; Subject: Re: pseudointerface (TCPIP SET [CONFIG] INTERFACE) ; Message-ID: <01KSXWUHEXUQ9GVT85@sysdev.deutsche-boerse.com>   G > Back in 1996 we configured an Alphastation to use two IP addresses on G > the same interface, using pseudointerfaces.  I don't remember many of A > the technical details such as the command we used to create the E > interface.  I think we created WEA0 and WEB0 and configured both of $ > those (and didn't configure WE0).  > I > Since then I have found that the "cluster alias IP" also works for that H > kind of thing (even if you don't have a cluster) and is more intuitive
 > to use.   C I use a cluster alias is well, no problem, but its not like a real  > pseudointerface (oxymoron?) since it only applies to incoming = connections; new outgoing connections don't use this address.    ------------------------------  + Date: Thu, 27 Feb 2003 18:08:07 +0100 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> ; Subject: Re: pseudointerface (TCPIP SET [CONFIG] INTERFACE) ; Message-ID: <01KSXWVVPSTK9GVT85@sysdev.deutsche-boerse.com>   I > > > TCPIP> SET INTERFACE FFA0 /HOST=KESTREL /NETWORK_MASK=255.255.0.0 - ( > > > _TCPIP> /BROADCAST_MASK=128.30.0.0 > I > I have found that if you do not specify the broadcast_mask or broadcast G > address, it defines a broadcast address  for you based on the network A > mask and interface address. (the last address in your subnet).   >  > is that a big mistake ?   H It does this, and sometimes (always) does this EVEN IF YOU SPECIFY THEM.   ------------------------------  + Date: Thu, 27 Feb 2003 18:05:55 +0100 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> ; Subject: Re: pseudointerface (TCPIP SET [CONFIG] INTERFACE) ; Message-ID: <01KSXWPY6BYK9GVT85@sysdev.deutsche-boerse.com>   B > Well, when you make an outgoing call that matches neither of theC > interface's subnets, how is the TCPIP stack supposed to know what  > interface to use ?   > J > I think that you'd have to be looking at the routes database to put someF > logic that decides where you route certain traffic. You can't reallyI > have 2 default gateway. You'll have to pick specific networks you route 4 > through one modem and send the rest to the other.   H No, if I have two interfaces, each on a different network, and a router H on that network, then, yes, only one default gateway for outgoing stuff B is fine.  However, if an INCOMING connection comes in through one D interface, then its "return traffic" should go out through the same I interface.  It doesn't.  That's the bug.  Second-hand information I have  5 says that it worked in 4.x and in 5.3 but not in 5.1.   @ There is a bit (ACK?) in the protocoll which signifies that the G connection is return traffic to another connection.  Even dumb routers  D can handle this properly; TCPIP for VMS certainly should be able to.   ------------------------------  % Date: Thu, 27 Feb 2003 19:43:54 +0100 " From: "Hans Vlems" <hvlems@iae.nl>; Subject: Re: pseudointerface (TCPIP SET [CONFIG] INTERFACE] 6 Message-ID: <b3lm9k$1nk7g4$1@ID-143435.news.dfncis.de>  ? "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> schreef in bericht ) news:3E5D5D26.45BE9D77@vl.videotron.ca... I > > > TCPIP> SET INTERFACE FFA0 /HOST=KESTREL /NETWORK_MASK=255.255.0.0 - ( > > > _TCPIP> /BROADCAST_MASK=128.30.0.0 > I > I have found that if you do not specify the broadcast_mask or broadcast L > address, it defines a broadcast address  for you based on the network mask and 7 > interface address. (the last address in your subnet).  >  > is that a big mistake ?   J No it isn't anymore. Since TCPIP V5.1 HPQ assumes all ones for a broadcastL and all zeroes specify the network. A couple of years ago this still was notL standardized and UCX had to ask which of the two conventions you were using.L The documentation may still show examples of both conventions. For a regularB (classfull) class B network the mask is 255.255.0.0, the net, say,, 128.30.0.0 and the broadcast 128.30.255.255.E In other words, the broadcast has all ones where the net mask has all  zeroes. J So a network with mask 255.255.255.248 uses three bits for the hosts in itK and thus has a broadcast mask that has the lower three bits set to 1. Hence  the 200+7=207.   ------------------------------  + Date: Thu, 27 Feb 2003 15:44:04 +0000 (UTC) , From: lewis@PROBE.mitre.org (Keith A. Lewis)B Subject: Re: SET NOTRUNCATE (was: RE: Upcoming VMS improvements ?)- Message-ID: <b3lbo4$9l$1@newslocal.mitre.org>    spamsink2001@yahoo.com (Alan E. Feldman) writes in article <b096a4ee.0302261959.5712253b@posting.google.com> dated 26 Feb 2003 19:59:12 -0800:Q >> Besides, the ability to use TPU in a decwindows environment makes this a clear Q >> winner for long lines since you need not define keys, you use use the mouse to  >> scroll left and right.  > F >Guess what? I don't have DECwindows. Well, I have the software, but IE >can't use it and don't need it anyway. And I don't consider it to be ? >any advantage to be able to use a mouse. Why is that any good?   K Can't use it?  What do you have on your desktop?  You can probably download 6 a Linux distribution for it, which includes X-windows.  F You don't have to use the mouse to control the scrollbars in the Motif? version of TPU/EVE.  Pageup and Pagedown move vertically; while 6 control-Pageup and control-Pagedown move horizontally.  K There are a lot of other nice features in the Motif version; it's very easy K to cut and paste between windows (not just other TPU windows but just about ) any X-windows app which displays text).     J While we're bragging about what we've done with TPU, I'll remind everybodyK that I wrote a newsreader in it called TNR.  It's what I use to participate K in this forum.  It's very simple-minded (less than 1000 lines of code), but H it works, even has a killfile function, and I get to use EVE to write myE articles.  Nyaah, Nyaah, Nyaah!  (Drop me a line if you want a copy.)   + --Keith Lewis              klewis$mitre.org > The above may not (yet) represent the opinions of my employer.   ------------------------------    Date: 27 Feb 2003 12:36:25 -0600+ From: young_r@encompasserve.org (Rob Young)  Subject: Sun and Java 3 Message-ID: <597g3TEeiwkF@eisner.encompasserve.org>   X In article <3E5E4300.8090106@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy writes: >  > : > Yes because your own engineers say that there are issues8 > your own benchmark results illustrate these issues andA > customers on this newsgroup have benchmarked their applications % > and discovered them for themselves.  >   $ 	So?  Solaris sucks when using Java:  ? http://www.internalmemos.com/memos/memodetails.php?memo_id=1321   K - Large footprint of applications when run on Solaris. A simple application L ("hello world" type) has a total footprint of 35-40 megs on Solaris 9 (buildN 48, using Java 1.4 build 82) on both Intel and Sparc machines. Sparc machines,L by far, have a much higher resident footprint then Intel machines (~30 megs,F compared to ~11 megs). The same program run on a Windows machine has a8 footprint of ~5 megs, resident footprint being ~3.5 megs  B 	So maybe you want to run your Java-centric app on Windows, right? 	Appears that way:  O Following is an excerpt from Kevin Tay's e-mail to three Java aliases regarding O a customer installation of a third-party product written in Java called Vitria. H We see typical very large RSS numbers compared to a WinNT implementationA combined with increased resource usage from Solaris7 to Solaris8:  ------- O Customer said they have something like 450+ container servers and 80+ automator N server for the Vitria system. So the estimation for the hardware RAM is aroundB 9GB for USII machine and 14-15GB for the USIII machine. Questions:  0 1. Why is Sun systems using so much more memory?L 2. Why is the UltraSPARC III/Solaris 8 system using a lot more memory than aD UltraSPARC II/Solaris 7 system (with every other thing being equal)?H 3. How can I reduce the memory utilization of the UltraSPARC III system?   ---   B 	Maybe the Java team gets Java healthy under Solaris 8 and SolarisG 	9.  Or maybe the dirty little secret is Windows NT / 2000 is a better  A 	OS in comparison?  Intel better than Sparc?  Well, that's no big  	secret.  
 	Actually:  N Imagine what happens if our current implementation of Java were ubiquitous andK all 150 users on a SunRay server were running one and only one Java program K equivalent to Component Manager above. The twenty-four gigabytes of RAM the I server would have to supply exclusively to these users is well beyond the H typical configuration. RAM is cheap but performance is what we sell, allM customers on that SunRay server would see significant performance degradation I even with the maximum amount of RAM installed as all other processes were  forced to reside on swap.   M The resident set size required by the JRE makes it impractical to run Java in H an initial Solaris install environment. It is impractical to run it as aJ non-terminating daemon. A Java daemon could be started from inetd run longL enough to do its job and then quit but the rpc protocol required to pass theE socket port to the daemon is very complex and not Java-friendly. Java H applications cannot be executed at boot time since the loading of the VMO introduces an unacceptable performance degradation. If the Java runtime were as O small as that of Python, it is likely that the Java daemon would become popular I and could provide basic services to applications written in any number of 
 languages.    B 	Seems from the tone of that article, a much better solution would> 	be to ditch Java and start developing around Python?  Is thatA 	too much of a stretch or too late to ditch Java?  Very positive   	article on Python there!    				Rob    ------------------------------  # Date: Mon, 24 Feb 2003 03:42:19 GMT 1 From: Michael Austin <maustin@firstdbasource.com> 4 Subject: Re: Suspected compromise of my home VMS box1 Message-ID: <3E599462.40444C3@firstdbasource.com>    Paul Sture wrote:  > 0 > I cannot log onto my home VMS system any more. > D > I logged in this morning with no problem, but later in the day all) > three usernames I use no longer worked.  > F > Setting UAFALTERANATE and rebooting still doesn't get me in, so I am+ > thinking I have been hacked by an expert.  > C > OK, I'm on the big bad Internet, but my router points telnet at a / > non-existant address, so one guess is that it @ > was done via Apache or some other route of which I am unaware. > = > Obviously the best course is to revert to backups here, but : > suggestions please to make sure it doesn't happen again. > @ > It's firmly switched off for now, but my checklist once I boot > from a CD goes like this:  > : > Grab the accounting and audit files for later inspection7 > Grab my other monitoring files for similar inspection  > Grab the Apache logs too > Operator.log also  > A DIR/MOD/SINCE=whenever > ? > Then restore from last trusted backup, change passwwords etc,  > BEFORE I put the cables back.  > " > Have I forgotten anything there? >  > -- > Paul Sture  H The very first thing you want to do once you get this box back online isE to install SSL with SSH support and use something like TeraTERM w/the H TTSSH update as your emulator.  This will encrypt the sessions such thatG people utilizing some sort of sniffer can't find your plain text telnet H u/p.  I never use telnet port 23 to this box fronm the internet if I can help it.  H I have installed Apache and it has TMP/NETBMX  only and it works without3 any problems -- so far... Current system uptime is: D OpenVMS V7.2-1  on node ALPHA1  23-FEB-2003 21:36:52.33  Uptime  128 03:22:27    E for those interested, I finally got zoneedit's ddclient running on my - VMS box. It is not 100% perl, but it works...      --   Regards,  6 Michael Austin            OpenVMS User since June 19847 First DBA Source, Inc.    Registered Linux User #261163    ------------------------------  + Date: Thu, 27 Feb 2003 16:32:50 +0000 (UTC) , From: lewis@PROBE.mitre.org (Keith A. Lewis)- Subject: Re: TCPIP and network loss detection - Message-ID: <b3leji$qg$1@newslocal.mitre.org>    JF Mezei <jfmezei.spamnot@vl.videotron.ca> writes in article <3E5C4986.104D2450@vl.videotron.ca> dated Wed, 26 Feb 2003 00:58:47 -0400:  >McEagle wrote:  >>  0 >> How about PINGing the switch every x minutes? > J >I was about to ask a very similar question.  PING every X minutes doesn't* >provide instant notification of problem.  > K >There should be some sort of service you can establish a connection to and G >leave "idle" 7/24, with an AST triggered if the connection is broken.   > - >Apart from pinging, is there any other way ?   G No.  A tcp connection cannot always detect if it is broken.  Maybe this  anecdote will illustrate why.   G Back in 1996 I was doing live tests of air traffic control software.  I B would wheel an X-terminal into the control room on a cart, and theJ controller doing the test would turn away from the existing screen and use the one on the cart.  L These X-terminals ran at about 25 MHz, and they took a while to boot.  WhileK they were booting, they displayed dark text on a white screen.  This really L annoyed the controllers, as it was the brightest thing in the room.  I couldD have "solved" the problem by turning off the monitor during the bootK process, but then I wouldn't know when it was ready for me to go downstairs - to the Alphaserver and start the application.   J So I put a UPS on the cart and booted it, then wheeled it into the controlH room and plugged it in.  At the end of the testing session, my colleagueF brought it back out, but he hadn't bothered to turn it off (or to callK downstairs to tell me to shut down the app), and the UPS was still powering H the X-terminal.  The application X-windows were still on the screen, butB since it wasn't connected to the network, they weren't updating.    G I decided to see what would happen if I connected the thing again, so I L plugged it into the network connection there.  The windows updated!  After a7 few seconds of activity, everything was back to normal.   H X-windows applications use open tcp connections, which (as I proved) canH survive being plugged and unplugged, as long as it's on the same network
 segment.    F Another experiment that will demonstrate this is the power-down test. K Establish a telnet connection from machine A to machine B.  Then turn B off J with extreme prejudice (i.e. don't let it run the normal shutdown stuff). K The telnet connection will stay open.  Turn B back on and boot it.  Usually K when it starts its tcp/ip, the connection will finally break.  It's the new K incarnation of B's tcp/ip that tells A that the connection doesn't exist as / far as it's concerned, which is why A gives up.   L The only good way to test an IP connection is to send some data on an actual& round-trip, which is what PING does.    + --Keith Lewis              klewis$mitre.org > The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Thu, 27 Feb 2003 12:46:41 +0000 ( From: Nic Clews <sendspamhere@127.0.0.1> Subject: Re: The End of OpenVMS ) Message-ID: <3E5E08B1.37E07D7F@127.0.0.1>    "David J. Dachtera" wrote: >  > Nic Clews wrote: ...   1 Gone to email to say a few things I can't here...    --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot com    ------------------------------  # Date: Thu, 27 Feb 2003 14:55:30 GMT # From: "John Smith" <a@nonymous.com>  Subject: Re: The End of OpenVMS B Message-ID: <CFp7a.85$em1.76@news04.bloor.is.net.cable.rogers.com>  5 "Nic Clews" <sendspamhere@127.0.0.1> wrote in message # news:3E5E08B1.37E07D7F@127.0.0.1...  > "David J. Dachtera" wrote: > >  > > Nic Clews wrote: > ...  > 3 > Gone to email to say a few things I can't here...      They're listening anyway. 1 http://news.zdnet.co.uk/story/0,,s2079845,00.html    ------------------------------  % Date: Thu, 27 Feb 2003 15:39:45 -0000 - From: "Steve Spires" <Steve.Spires@torex.com>  Subject: RE: The End of OpenVMS E Message-ID: <91947A84607D9D48B8E674A5FAB54DA63CB20D@tahiti.tinuk.com>   C I believe GCHQ are looking for VMS and Tru64 staff at the moment if  anyone is looking for a job...  J http://www.it.jobserve.com/jobserve/JobDetail.asp?jobid=3DD284310CA7E759F= B    Steve Spires Technical Consultant Torex Health [T] +44 (0) 1295 274200  [F] +44 (0) 1295 275131 
 www.torex.com    >>-----Original Message------ >>From: John Smith [mailto:a@nonymous.com]=20  >>Sent: 27 February 2003 14:56 >>To: Info-VAX@Mvb.Saic.Com ! >>Subject: Re: The End of OpenVMS  >> >> >>: >>"Nic Clews" <sendspamhere@127.0.0.1> wrote in message=20% >>news:3E5E08B1.37E07D7F@127.0.0.1...  >>> "David J. Dachtera" wrote: >>> >  >>> > Nic Clews wrote: >>> ...  >>> 5 >>> Gone to email to say a few things I can't here...  >> >> >>They're listening anyway.=205 >>>>http://news.zdnet.co.uk/story/0,,s2079845,00.html  >> >> >> >> >>   ------------------------------  % Date: Thu, 27 Feb 2003 16:34:48 +0000 ( From: Nic Clews <sendspamhere@127.0.0.1> Subject: Re: The End of OpenVMS ) Message-ID: <3E5E3E28.FB1325CE@127.0.0.1>    John Smith wrote:  > 7 > "Nic Clews" <sendspamhere@127.0.0.1> wrote in message % > news:3E5E08B1.37E07D7F@127.0.0.1...  > > "David J. Dachtera" wrote: > > >  > > > Nic Clews wrote: > > ...  > > 5 > > Gone to email to say a few things I can't here...  >  > They're listening anyway. 3 > http://news.zdnet.co.uk/story/0,,s2079845,00.html   E I figured, so I only said things that could get me fired, not shot or 
 banged up.   --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot com    ------------------------------  # Date: Sun, 23 Feb 2003 00:20:07 GMT ' From: Don Sykes <anonymous@pacbell.net> ( Subject: Re: Upcoming VMS improvements ?+ Message-ID: <3E58148E.4B1FC938@pacbell.net>    "Alan E. Feldman" wrote: > \ > Don Sykes <anonymous@pacbell.net> wrote in message news:<3E569B38.2FBEE3FD@pacbell.net>... > I > > I often hear the "EDT is better" stuff, but I can't think of a single H > > thing that EDT can do that I can't do with my customized TPU editor. > G > I don't know what you can do with your customized TPU editor. I don't G > have that. But I can pretty much be sure you don't have an equivalent < > for the EDT line mode command, SET NOTRUNCATE (see below).F No, but I can do a PF1+(right-arrow) and shift the screen 'n' bytes to? the right each time. Where 'n' is defined by pressing <Ctrl>+B.    > K > > Whereas, there are a host of things that I can't do with EDT that I can  > > do in TPU. > G > True. But there are things you have to put with in TPU that you don't  > in EDT. See below.  C Remember, I said my "customized TPU editor". Since TPU is a texting @ language my version is an extension of what standard TPU/EVE is.   >  > > What am I missing ? = > > (EDT'ers please do not flame. This is an honest question)  > C > [This is not intended as a flame, but I have very strong feelings A > about this. Please don't misinterpret this as flamage. Thanks.]  > D > OK, yes, there are advantages to EVE/TPU. When I need box editing,E > wildcard searches, long lines, etc., I use EVE/TPU. But most of the  > time I use EDT.  > E > One of these days I am going to make a list of EDT vs EVE features.  > But for right now, here goes:  > G > Well, SET NOTRUNCATE can't be done with TPU. Based on previous posts, H > I remember being told that to implement its equivalent using TPU wouldC > be very difficult and/or very slow or awkward in some way. Please > > don't mention SET WRAP. It is *not* the same thing at all!!! > G > Simple things are easier in EDT -- especially when trying to script a C > command. For example, in EDT, to substitute, you press control/z,  >  >     S/OLDSTRING/NEWSTRING/ WH  > H > and press Return. You can put that in an .EDT file verbatim and run itA > and it works. You can't do that with EVE. You have to learn yet  > another language. B I conceed EDTs line mode editing capabilities. But I think I couldF expand my version to save the learned key expressions in much the same way.   > B > Simple things take less typing. I want to go to line 34. In EDT,G > control/z, 34, Return, I'm there! EVE: Do, LINE 34, Return. Move down G > 25 lines, *+25, I'm there! EVE: I don't even know. OK I could use the 
 > Repeat key.   H Or in TPU, just press the "Do" key and enter "set keypad edt" to get all the EDT keypad functionality.    > 4 > Another simple thing that is much easier with EDT: > 5 > In EDT I can easily write out the paste buffer with  > 0 > control/Z                   ! Enter line mode.A > *WRITE FILE.TYP =PASTE      ! Write the PASTE buffer to a file. 1 > *C                          ! Get back to work.  > F > In EVE I have to do the following: Press Do, Command: SHOW SYS, moveG > the cursor up to the line that says "Insert Here", press enter, press H > Do again, Command: WRITE FILE.TYP, press Do *yet* again, Command: showE > buff, move the cursor to the main buffer, press Return. And writing D > the paste buffer to a file is something I actually do quite often.  F > Finally I stumbled upon a cool EVE INIT file that was posted in thisH > newsgroup that contains a key defined to write the select region to a + > file. Cool, but someone had to write it.  @ Someone had to write EDT too. So what it's been written already.   > And this function alwaysB > writes to TMP.TMP. With EDT I choose the name right in the WRITE4 > command. This is a simple common basic function!!! > E With mine you just press PF1+A enter a filename and a separate buffer G appears on a split screen. To write out that separate buffer just enter H PF1+W. You can also  use PF1+B,C,D...through K and get 10 other separate% buffers to work with in the same way. E And, unlike EDT (24 lines) the screen can be set to any size you like F and the buffer portion of the screen can be varied in size a line at a time with just a <Ctrl>+A.   > G > Yeah, it takes a little more memorizing here to use EDT, but not much C > at all, and I can handle it (you have to memorize =PASTE, not too E > tough!). With EVE it's more like using Windows, of which I am not a H > fan. And for other equivalents you actually have to learn and memorizeC > more because you have to write and debug programs. (OK, with long C > buffer names I like being able to move the cursor to the name and 9 > press return, but not for the paste buffer, see above.)   F I think the issue is that TPU commands & functions aren't really meantG for interactive mode. I almost never use it that way. Rather I employed H it to create an editor that had easy features to remember. The advantageE of TPU is that it is a powerful language of functions. IMO you're not > meant to memorize them any more that you would memorize all ofC DECWindows functions. You're meant to explore and employ only those E functions needed to create easy, friendly functions for regular use.     > G > You can't do case-independent searches with uppercase search strings. C > I like my DCL in uppercase and comments in mixed case. So you can H > imagine what it's like when I type in SEARCH_STRING, press Return, andF > get nothing for a string that I know is in the file. No sir, I don't> > like it. ---So work in lowercase.--- I don't want to work in > lowercase!< Again with the TPU editor I use I have a choice of searches:* Find(E1)  Gives me a case sensitive search- PF1+PF3	  Gives me a case INsensitive search.    > G > Much smoother scrolling and cursor movement with EDT. Also, opening a G > new file takes about three seconds with spasmodic blinking and flying D > around of the cursor, messages flashing and inverting too fast tooE > read, and then finally the status bar (It's because of my init file D > which only 11 lines long. Now I know someone once posted somethingD > like "Well, it takes time for EVE/TPU to interpret INIT commands."D > We're talking 11 lines! I could almost read them that fast myself!@ > (OK, I exaggerate.) My EDT script is 262 lines (some lines are2 > comments) and it runs almost instantaneously!!!)  G I use TPU Section files and have no such problem. It comes up instantly + without any of of those quirks you mention.    > E > With EDT, the cursor stays on the same row with repeated presses of G > the Find Next key. In EVE it sometimes goes top, bottom, top, bottom, E > .... Sometimes it bounces around in other ways. I hate that. Pick a  > row and go with it!   G Again I see no such problems with my TPU editor. I use PF3 to do a find - next and it just takes me to the line I seek.    > F > You can go to, and stay in, line mode. You can't do that with EVE. IC > don't see the advantage of having to hit the Do key over and over  > again.F With mine I enter <Ctrl>+Z to go to an EDT prompt where I stay until I
 enter "C".H Or, I can press the Do key and go to a TPU prompt that works as you say.   > F > Nokeypad mode: Cool for writing editable customized startup scripts.B > You can display any individual key definition and it displays inH > nokeypad commands. I don't know how to do that in EVE. How do you lookD > at your key definitions in EVE? Well, I need to research the threeF > methods, /init, /command and the /section variations. I suppose withF > at least the /init version you can do it. But not if you define yourE > keys interactively and save them in a section file(!), I think. How G > would you do that short of writing down your keystrokes on a piece of G > paper *as you define them* for any keys you define interactively with * > "LEARN" and then save in a section file?F With my Learn (Ctrl+K) I can actually see what I'm doing as it's being learned.C e.g. if I say go down 3 lines, over 2 bytes and change the case. It D happens as I learn it. Also, I can enter text of my choice, not just from my patse buffer, anywhere.   E I can also "learn" my mouse buttons to do things EDT can't even dream D about, like set the text cursor to the mouse location and change the case of the text there.    > E > I can edit the first few lines of a huge file much faster and don't G > have to worry about running out of virtual memory in the process. And = > I don't have to consume all that memory in the first place!  > F True, but you can only see 24 lines at a time and I have never run out of virtual memory.  E > EDT has the ALL range-specifier. You can do something like TYPE ALL F > 'HELLO' and it displays all lines that contain the word "HELLO". You > can also do stuff like > & >     S/STRING1/STRING2/ ALL 'STRING3' > G > which will change string1 to string2 in all lines containing string3.  > G > There are other cool range-specifiers that let you specify a range of F > lines by number, all lines before (or after) the cursor, first line, > end, and more.H Again, if it were important to me to do these things they can be done inF TPU with all the options you want, because TPU is a language. True you have to code them.   > H > With EDT you can edit a file that has fixed-width records and when youD > exit, it saves it as a fixed-width-records file. You can't do thatE > with EVE, and probably not even with a TPU program (well, maybe you  > could; I don't know).  > C > And if you can do any of these things in TPU you probably have to D > write a program to do it. I have work to do. I don't want to spendD > hours writing code that does simple things that EDT can do without > having to write code. G I'm not trying to get you to convert, but I don't think you've explored  TPU enough to pooh-pooh it.    > C > In EDT you can edit your journal files. This is great for undoing D > mistakes. If you use keyboard journaling (which, BTW, is *not* the5 > default in EVE/TPU), you can do it but it's harder.  >  Ok.   G > In EDT you can use ^U to abort almost any command that you are in the F > process of typing and you get a quick, clean abort. Not so with EVE. >  Ok.    D > EVE/TPU for normal use is a little like using a rocket ship to get > across the street.' Does that mean EDT is like a bicycle?:)    >  > OK, that's enough for now. >  > > Have VMS, Will Travel  >  > Cool.  >  > > Wire paladin, San Francisco  >  > Huh?9 That's part of Paladin's calling card. (the 50's TV show)    >  > > (paladinATalphaseDOTcom) > [...]  >  > Disclaimer: JMHO > Alan E. Feldman    --     Have VMS, Will Travel  Wire paladin, San Francisco    (paladinATalphaseDOTcom)   ------------------------------    Date: 27 Feb 2003 08:23:42 -0800. From: spamsink2001@yahoo.com (Alan E. Feldman)( Subject: Re: Upcoming VMS improvements ?= Message-ID: <b096a4ee.0302270823.669a2b0e@posting.google.com>   Z Don Sykes <anonymous@pacbell.net> wrote in message news:<3E58148E.4B1FC938@pacbell.net>... > "Alan E. Feldman" wrote: > > ^ > > Don Sykes <anonymous@pacbell.net> wrote in message news:<3E569B38.2FBEE3FD@pacbell.net>... > > K > > > I often hear the "EDT is better" stuff, but I can't think of a single J > > > thing that EDT can do that I can't do with my customized TPU editor.  D Right here you're already in trouble. I don't think anyone ever said? that EDT is better than your customized editor. No one has your  customized editor but you!   > > I > > I don't know what you can do with your customized TPU editor. I don't I > > have that. But I can pretty much be sure you don't have an equivalent > > > for the EDT line mode command, SET NOTRUNCATE (see below).H > No, but I can do a PF1+(right-arrow) and shift the screen 'n' bytes toA > the right each time. Where 'n' is defined by pressing <Ctrl>+B.   E Sorry, but that's not good enough. When I read log files that contain C many lines with width .gt. 132 chars, I have the choice with EDT to D use either SET NOTRUNCATE or SHIFT LEFT and SHIFT RIGHT. Since it isF much easier to read said lines with SET NOTRUNCATE, that's what I use.; So that proves it's better. And in my opinion, much better.   C There was a time where I had to read database files in an editor. I ? don't remember exactly why, but since the lines were too long I D couldn't use EDT and it was a real pain in the ass not to be able to; use SET NOTRUNCATE. At first I thought "I'll just use SHIFT A LEFT,RIGHT", but for some reason it just stunk. It might have had D something to do with how the cursor moved. I can't remember for sure1 though. But I do remember for sure that it stunk.   M > > > Whereas, there are a host of things that I can't do with EDT that I can  > > > do in TPU. > > I > > True. But there are things you have to put with in TPU that you don't  > > in EDT. See below. > E > Remember, I said my "customized TPU editor". Since TPU is a texting B > language my version is an extension of what standard TPU/EVE is.  @ Remember, our choice is not between EDT and your "customized TPUF editor." It's the choice between our already customized EDT and having? to totally recustomize EVE via TPU (or whatever the technically C correct expression would be) which includes changes just to make it B like EDT, changes to duplicate our EDT customizations, and it alsoD means having to put up with stuff that can't be customized away. AndE learning to customize TPU is more work than the same for EDT. I don't # know how to make it any more clear.    > > > What am I missing ? ? > > > (EDT'ers please do not flame. This is an honest question)  > > E > > [This is not intended as a flame, but I have very strong feelings C > > about this. Please don't misinterpret this as flamage. Thanks.]  > > F > > OK, yes, there are advantages to EVE/TPU. When I need box editing,G > > wildcard searches, long lines, etc., I use EVE/TPU. But most of the  > > time I use EDT.  > > G > > One of these days I am going to make a list of EDT vs EVE features. ! > > But for right now, here goes:  > > I > > Well, SET NOTRUNCATE can't be done with TPU. Based on previous posts, J > > I remember being told that to implement its equivalent using TPU wouldE > > be very difficult and/or very slow or awkward in some way. Please @ > > don't mention SET WRAP. It is *not* the same thing at all!!! > > I > > Simple things are easier in EDT -- especially when trying to script a E > > command. For example, in EDT, to substitute, you press control/z,  > > ! > >     S/OLDSTRING/NEWSTRING/ WH  > > J > > and press Return. You can put that in an .EDT file verbatim and run itC > > and it works. You can't do that with EVE. You have to learn yet  > > another language. D > I conceed EDTs line mode editing capabilities. But I think I couldH > expand my version to save the learned key expressions in much the same > way.   OK.   D > > Simple things take less typing. I want to go to line 34. In EDT,I > > control/z, 34, Return, I'm there! EVE: Do, LINE 34, Return. Move down I > > 25 lines, *+25, I'm there! EVE: I don't even know. OK I could use the  > > Repeat key.  > J > Or in TPU, just press the "Do" key and enter "set keypad edt" to get all > the EDT keypad functionality.   C What does keypad functionality have to do with typing 34 instead of D LINE 34? What does it have to do with pressing Return (in line mode)= to see what line you're on and typing WHAT LINE with EVE? The C EDT-keypad emulation in EVE is okay. I still don't like things like A "Found in reverse direction.  Go there? " I don't like the editor D treating me like an idiot and it's an extra worthless thing to boot.  6 > > Another simple thing that is much easier with EDT: > > 7 > > In EDT I can easily write out the paste buffer with  > > 2 > > control/Z                   ! Enter line mode.C > > *WRITE FILE.TYP =PASTE      ! Write the PASTE buffer to a file. 3 > > *C                          ! Get back to work.  > > H > > In EVE I have to do the following: Press Do, Command: SHOW SYS, moveI > > the cursor up to the line that says "Insert Here", press enter, press J > > Do again, Command: WRITE FILE.TYP, press Do *yet* again, Command: showG > > buff, move the cursor to the main buffer, press Return. And writing F > > the paste buffer to a file is something I actually do quite often. >   H > > Finally I stumbled upon a cool EVE INIT file that was posted in thisJ > > newsgroup that contains a key defined to write the select region to a - > > file. Cool, but someone had to write it.  B > Someone had to write EDT too. So what it's been written already.  D Sorry, I'm totally missing your point here. Someone has to write any program. Please clarify.   > > And this function alwaysD > > writes to TMP.TMP. With EDT I choose the name right in the WRITE6 > > command. This is a simple common basic function!!! > > G > With mine you just press PF1+A enter a filename and a separate buffer I > appears on a split screen. To write out that separate buffer just enter J > PF1+W. You can also  use PF1+B,C,D...through K and get 10 other separate' > buffers to work with in the same way.   @ I don't have your editor. And I don't want to waste time and getC aggravated trying to write it when I already have something I like.   G > And, unlike EDT (24 lines) the screen can be set to any size you like H > and the buffer portion of the screen can be varied in size a line at a > time with just a <Ctrl>+A.  E Yes, I am painfully aware of the 24-line limitation. I *still* prefer 0 EDT. Again, I don't have your customized editor.  I > > Yeah, it takes a little more memorizing here to use EDT, but not much E > > at all, and I can handle it (you have to memorize =PASTE, not too G > > tough!). With EVE it's more like using Windows, of which I am not a J > > fan. And for other equivalents you actually have to learn and memorizeE > > more because you have to write and debug programs. (OK, with long E > > buffer names I like being able to move the cursor to the name and ; > > press return, but not for the paste buffer, see above.)  > H > I think the issue is that TPU commands & functions aren't really meantI > for interactive mode. I almost never use it that way. Rather I employed   6 We're talking about editing files! That's interactive!  J > it to create an editor that had easy features to remember. The advantageG > of TPU is that it is a powerful language of functions. IMO you're not   F I never disputed that. But it's often TOO powerful meaning you have toA worry about too many things just to do or write something that is  basically very simple.  @ > meant to memorize them any more that you would memorize all ofE > DECWindows functions. You're meant to explore and employ only those G > functions needed to create easy, friendly functions for regular use.    F I've already done that with EDT. Would you like to convert my 262-line? EDT startup file to TPU? And you still can't do SET NOTRUNCATE.   I > > You can't do case-independent searches with uppercase search strings. E > > I like my DCL in uppercase and comments in mixed case. So you can J > > imagine what it's like when I type in SEARCH_STRING, press Return, andH > > get nothing for a string that I know is in the file. No sir, I don't@ > > like it. ---So work in lowercase.--- I don't want to work in > > lowercase!> > Again with the TPU editor I use I have a choice of searches:, > Find(E1)  Gives me a case sensitive search/ > PF1+PF3	  Gives me a case INsensitive search.    I don't have your editor.   I > > Much smoother scrolling and cursor movement with EDT. Also, opening a I > > new file takes about three seconds with spasmodic blinking and flying F > > around of the cursor, messages flashing and inverting too fast tooG > > read, and then finally the status bar (It's because of my init file F > > which only 11 lines long. Now I know someone once posted somethingF > > like "Well, it takes time for EVE/TPU to interpret INIT commands."F > > We're talking 11 lines! I could almost read them that fast myself!B > > (OK, I exaggerate.) My EDT script is 262 lines (some lines are4 > > comments) and it runs almost instantaneously!!!) > I > I use TPU Section files and have no such problem. It comes up instantly - > without any of of those quirks you mention.    I don't have your editor.   G > > With EDT, the cursor stays on the same row with repeated presses of I > > the Find Next key. In EVE it sometimes goes top, bottom, top, bottom, G > > .... Sometimes it bounces around in other ways. I hate that. Pick a  > > row and go with it!  > I > Again I see no such problems with my TPU editor. I use PF3 to do a find / > next and it just takes me to the line I seek.   B But the position of that line jumps all over the screen (with each? press of the Find Next key) in an irritating way. If it doesn't ( irritate you, then fine. But it does me.  H > > You can go to, and stay in, line mode. You can't do that with EVE. IE > > don't see the advantage of having to hit the Do key over and over 
 > > again.H > With mine I enter <Ctrl>+Z to go to an EDT prompt where I stay until I > enter "C".J > Or, I can press the Do key and go to a TPU prompt that works as you say.   I DON'T HAVE YOUR EDITOR.   H > > Nokeypad mode: Cool for writing editable customized startup scripts.D > > You can display any individual key definition and it displays inJ > > nokeypad commands. I don't know how to do that in EVE. How do you lookF > > at your key definitions in EVE? Well, I need to research the threeH > > methods, /init, /command and the /section variations. I suppose withH > > at least the /init version you can do it. But not if you define yourG > > keys interactively and save them in a section file(!), I think. How I > > would you do that short of writing down your keystrokes on a piece of I > > paper *as you define them* for any keys you define interactively with , > > "LEARN" and then save in a section file?H > With my Learn (Ctrl+K) I can actually see what I'm doing as it's being
 > learned.E > e.g. if I say go down 3 lines, over 2 bytes and change the case. It  > happens as I learn it.    E How do you display it later? I can define keys interactively with EDT  and later display them:    *SHOW KEY CONT N	 +V+V+V+C.  *   E I can then write that down, type it into a file in another window, or @ if I'm in a rare mouse mood, I can copy and paste it into an EDTF command file. I don't know how'd you do that with EVE or TPU. Yes, youF could save it in a section file, but then you can't display what it isD later. You can display it later with EDT with the same command given% above or just look at your init file!   / > Also, I can enter text of my choice, not just ! > from my patse buffer, anywhere.   A Could you be a little more vague? I have no idea at all what this C sentence means. I can type text into EDT, too. What's the big deal?   G > I can also "learn" my mouse buttons to do things EDT can't even dream F > about, like set the text cursor to the mouse location and change the > case of the text there.   F I have no interest in the mouse. I avoid it as much as possible. I can/ see how some people would like that though. OK.   G > > I can edit the first few lines of a huge file much faster and don't I > > have to worry about running out of virtual memory in the process. And ? > > I don't have to consume all that memory in the first place!  > > H > True, but you can only see 24 lines at a time and I have never run out > of virtual memory.  A What's the biggest file you ever edited? Maybe you never ran on a 
 small system.   G > > EDT has the ALL range-specifier. You can do something like TYPE ALL H > > 'HELLO' and it displays all lines that contain the word "HELLO". You > > can also do stuff like > > ( > >     S/STRING1/STRING2/ ALL 'STRING3' > > I > > which will change string1 to string2 in all lines containing string3.  > > I > > There are other cool range-specifiers that let you specify a range of H > > lines by number, all lines before (or after) the cursor, first line, > > end, and more.J > Again, if it were important to me to do these things they can be done inH > TPU with all the options you want, because TPU is a language. True you > have to code them.   Yes, you have to code them.   J > > With EDT you can edit a file that has fixed-width records and when youF > > exit, it saves it as a fixed-width-records file. You can't do thatG > > with EVE, and probably not even with a TPU program (well, maybe you  > > could; I don't know).  > > E > > And if you can do any of these things in TPU you probably have to F > > write a program to do it. I have work to do. I don't want to spendF > > hours writing code that does simple things that EDT can do without > > having to write code. I > I'm not trying to get you to convert, but I don't think you've explored  > TPU enough to pooh-pooh it.   $ I've wasted plenty of time with TPU.  E > > In EDT you can edit your journal files. This is great for undoing F > > mistakes. If you use keyboard journaling (which, BTW, is *not* the7 > > default in EVE/TPU), you can do it but it's harder.  > >  > Ok.  > I > > In EDT you can use ^U to abort almost any command that you are in the H > > process of typing and you get a quick, clean abort. Not so with EVE. > >  > Ok.  >   F > > EVE/TPU for normal use is a little like using a rocket ship to get > > across the street.) > Does that mean EDT is like a bicycle?:)   D It means it's a more appropriate tool for normal day-to-day editing,B at least for me. If you do well by EVE and/or TPU, fine. Would youE take a rocket ship down a path in the park? No. It's overkill. If you D need to go to the moon, fine. Get a rocket ship. I just want to editD files and make it as easy as possible to do so, without having to go" to said moon just to make it easy!   [...]    Disclaimer: JMHO Alan E. Feldman    ------------------------------    Date: 27 Feb 2003 09:30:59 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) ( Subject: Re: Upcoming VMS improvements ?= Message-ID: <cf15391e.0302270930.34290dc3@posting.google.com>   d "Bill Todd" <billtodd@metrocast.net> wrote in message news:<UDudne40z9xY1sCjXTWc3w@metrocast.net>...@ > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message9 > news:cf15391e.0302251203.5debe97a@posting.google.com... * >  including fork() (and I know that's notH > > an easy one; it wouldn't make sense to even even think about that if+ > > VMS development were to cease in 2005).  > + > Unless it was part of the COE commitment.   B VMS can meet COE standards without this -- it already has a fork()? that works correctly; it's just been slow because of the design 5 differences between process creation on VMS and Unix.   E This new feature is another example of how VMS is aiming for the long < term -- beyond just COE -- with the broader Unix PortabilityE Initiative.  This project will achieve a fast fork() -- a significant B performance improvement in an area where VMS has historically beenE weak when running code which came from the *nix world.  Being able to : run *nix code seamlessly helps immensely with the issue of! applications availability on VMS.    ------------------------------  % Date: Thu, 27 Feb 2003 12:56:48 -0500 A From: "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com> ( Subject: Re: Upcoming VMS improvements ?. Message-ID: <3e5e5161$1_3@hpb10302.boi.hp.com>  3 "jlsue" <jlsuexxxz@screaminet.com> wrote in message 2 news:3t6q5vk3j8ueaqejopmf58rmtf5ev9n3ou@4ax.com.... > On Tue, 25 Feb 2003 21:23:52 -0400, JF Mezei* > <jfmezei.spamnot@vl.videotron.ca> wrote: >  > >Keith Parris wrote:B > >> VMS development were to cease in 2005).  LAN HW support shows4 > >> 10-gigabit Ethernet in the latter half of 2005. > > K > >Pardon my ignorance, but what is such a big deal in suppoprting a faster L > >ethernet card ? Is 10Gb ethernet such a radically different beast that itL > >requires 3 yuears worth of work to design a new network philosophy, or is itL > >just a faster ethernet with tewaking of exsiting drivers to change timing > >information ? > F > Depends on the engineering standards.  Do they accept the exact sameH > standards as the 10/100/1000 e-net standards?  Or are there additional: > extensions (e.g., larger frames, etc) than merely speed? >   E I expect this to be more than just a bit that says "Ludicrous Speed".    ------------------------------  % Date: Thu, 27 Feb 2003 12:54:57 -0500 A From: "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com> ( Subject: Re: Upcoming VMS improvements ?. Message-ID: <3e5e50f2$1_2@hpb10302.boi.hp.com>  K The fact it's on the roadmap doesn't indicate that development starts today J and ends at that date.  That is the delivery date for when the driver willI be completed.  As 10GB is verrry expensive backbone hardware today, it is L lower on the priority list than other things, and hence the driver work, and5 decision on what to support, won't start for a while.       = "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in message ) news:3E5C170A.30A671EA@vl.videotron.ca...  > Keith Parris wrote: A > > VMS development were to cease in 2005).  LAN HW support shows 3 > > 10-gigabit Ethernet in the latter half of 2005.  > J > Pardon my ignorance, but what is such a big deal in suppoprting a fasterK > ethernet card ? Is 10Gb ethernet such a radically different beast that it K > requires 3 yuears worth of work to design a new network philosophy, or is  itK > just a faster ethernet with tewaking of exsiting drivers to change timing  > information ?    ------------------------------  % Date: Thu, 27 Feb 2003 07:02:00 -0800 # From: "Tom Linden" <tom@kednos.com>  Subject: VAX Emulators9 Message-ID: <CIEJLCMNHNNDLLOOGNJICEEHGLAA.tom@kednos.com>   ; Does any have data on comparison of VAX emulator running on = Pentium to Alpha?  In other words, what must the frequency of 9 an Alpha be for it offer comparable performace to, say, a ; 2.7GHz Pentium (with scsi and all that to offer a more fair  comparison)  --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.449 / Virus Database: 251 - Release Date: 1/27/2003   ------------------------------   Date: 27 Feb 2003 12:27 CDT ' From: carl@gerg.tamu.edu (Carl Perkins)  Subject: Re: VAX Emulators- Message-ID: <27FEB200312275440@gerg.tamu.edu>   ' "Tom Linden" <tom@kednos.com> writes... < }Does any have data on comparison of VAX emulator running on> }Pentium to Alpha?  In other words, what must the frequency of: }an Alpha be for it offer comparable performace to, say, a< }2.7GHz Pentium (with scsi and all that to offer a more fair }comparison)  B Your question is not clear, but there is no pentium fast enough toA run a VAX emulator on to make that emulator as fast as a DS10/466 C (a 466MHz EV6 - the slowest Alpha they still make, unless they have B stopped making the 466MHz version). Not even close - and by close,& I mean "within an order of magnitude".  H The fastest P4 currently available won't run an emulator that is as fastI as the fastest real VAX processor was, as far as I know. (I recall seeing F an estimate of somthing like 100MHz => 1 VUP for one of the emulators,I thus a 3.06GHz P4 should produce something around 30 VUPs, which is a bit 6 under half as fast as the fastest real VAX processor.)  F I think that it might be almost comparable to a DEC 3000 model 400, orF maybe one of the variants of the the model 300. (Anyhow, that would beG about a 9 year old 1st generation Alpha.) The DEC 3000m600 (175MHz EV4) H has an extimated VUP raing somewhere in the 90s. You'd have faster SCSI,J since the DEC 3000 series had 10MB/sec Fast Narrow SCSI and you'd probablyK have at least 80MB/sec Ultra2, which would help even things out even though ; your emulated VAX processor would be slower at most things.    --- Carl   ------------------------------  % Date: Wed, 26 Feb 2003 21:32:57 -0500 0 From: "Alan Boyles" <alan.boyles@mindspring.com>! Subject: Re: VMS Backup solutions / Message-ID: <v5s2cmo4d3t762@corp.supernews.com>   K We have been looking into the same sorts of questions and found a couple of H answers.  The new Legato solution that the VMS group is sponsoring looksJ quite promising, we also have been working with Tivoli Storage Manager andL the Archive Backup Client from a company called StorServer.  Our environmentG uses Oracle so we have some special requirements  but both of these are  Enterprise solutions.    Alan Boyles   & <brandon@dalsemi.com> wrote in message+ news:03022615422724@dscis6-0.dalsemi.com... E > Our "newly" formed operations group is looking at enterprise backup + solutions for our NT, UNIX, and VMS groups.  >  > Our VMS setup: > . > GS160, 4 each TL89-2 (35/70), 520-GB of data > L > Our VMS solution is to use BACKUP with DCL command files.  Automated using7 ISAM files and scripts with logging, e-mail, reporting, H > WWW, paging, etc.  All in-house created and maintained.  I have worked3 with SLS in the past and incorporated many of those - > concepts into the in-house backup solution.  > L > We keep our tapes at a 95% to 98% capacity so we make efficient use of our$ tapes, no multi-volume tapes either. > K > Unlike the NT or UNIX group, the VMS group has maintained a stable backup 5 solution incorporated with DR practices and policies.  > E > I am greatly concerned with this new drive to centralize the backup = solution onto a UNIX platform.  They want to NFS mount drives ' > and backup the data, etc.  Gasp-Argh!  > I > (?) I would like to know if anyone has a similar environment (NT, UNIX, 7 VMS) that has a successful enterprise solution and what  > they are using.  >  > (?) Any horror stories?  > 0 > (?) What about Disaster Recovery implications? >  >  >  > John Brandon > VMS Systems Administrator  > Dallas Semiconductor > john.brandon@dalsemi.com > 972.371.4172 wk  > 972.371.4003 fx    ------------------------------  % Date: Thu, 27 Feb 2003 08:11:24 -0600  From: brandon@dalsemi.com ! Subject: Re: VMS Backup solutions 1 Message-ID: <03022708112475@dscis6-0.dalsemi.com>   
 >>> BAD IDEA.  >>> M >>> If you back up VMS to U**X, you lose file attributes, ACEs and much, much 	 >>> more. ! >>> Have a look at ABS from HP.   ) >>> VMS server, VMS, NT and U**X clients.     3 I know it is a BAD IDEA.  Add explicits as desired.     A >>> The learning curve is a bit steep but once you get it running  >>> it's rock-solid. >>> 4 >>> Feel free to give me a ring if you want to talk.    * Sorry, I already gave my wife a ring.  ;-)  % OK, thanks, maybe in the near future.     J >>> Point out to them that backing up data is an annoying extra procedure. >>> 4 >>> The real goal is being able to restore the data.   The idea to have to convert, export, ZIP, TAR, or other some methodology (dependent on the file system), to a NFS mounted drive is NUTS.      The REAL GOAL?  Disaster Recovery (or even as you say restore).  I can see nothing good about three groups attempting to restore their drives (system, data, user, app) after an incident.  Who take priority?  Hell, while they are just getting around to getting? their tape system restored I could have my servers back online.   < >>> Compared to Veritas Netbackup, it can save you many $$$.B >>> Compared to Legato Networker, it can boost backup performance.K >>> For more information, please visit http://support.mti.com and press the M >>> OpenVMS button, call me at (800) 999.9684 or direct at (714) 481-5535, or $ >>> send email to VMSSupport@mti.com    > I do not mind a solicited sales pitch.  I will look into it.    $ This is a quote from our UNIX admin:  I    Next quarter we're going to be looking at expanding our current setup  K    (bigger, faster tape drives) to handle our data growth and the addition  K    of NT and ???? clients.  We'll probably continue to use our same backup  K    software, Reliaty (www.reliaty.com).  But before we can do any of that,  G    we'll have to upgrade our NetApp filers to a newer OS, so this will  A    probably all go nowhere because that will cost a lot of money.    I looked at the company    http://www.reliaty.com/    and did a search.         Results for: VMS )    No results were found for your search. P    Try changing some of the words in your query. (Why not try getting off UNIX?)    G >>> Well, alright, since you opened the Pandorda's box, are you talking 8 >>> about the followup product to OASIS and TapeControl?    ` No.  I thought OASIS was a Car Wash and TapeControl was the gizmo that switched your 8-tracks... Nope, my own code.    J >>> Like you, I wrote (2 years' work, 4400 lines of DCL) a backup solutionL >>> using OASIS and TapeControl with a STK-9710 jukebox. That outfit laid me9 >>> off three years ago. My code is still running, AFAIK.        	 Bastards.n    O >>> We have been looking into the same sorts of questions and found a couple ofsL >>> answers.  The new Legato solution that the VMS group is sponsoring looksN >>> quite promising, we also have been working with Tivoli Storage Manager andP >>> the Archive Backup Client from a company called StorServer.  Our environmentK >>> uses Oracle so we have some special requirements  but both of these aret >>> Enterprise solutions.f     Thanks, will look into it also.g     I look at all this and wonder how hard it would be to create our own backup tools that will run across the NT, UNIX platforms and ] onto the VMS.  Keeping in mind the DR implications and all.  Then, OMG, make it freeware?!?!!S       John Brandon VMS Systems AdministratorV Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fx    ------------------------------    Date: 27 Feb 2003 09:13:35 -0600+ From: young_r@encompasserve.org (Rob Young)N! Subject: Re: VMS Backup solutionsy3 Message-ID: <MKVYCITJfrhi@eisner.encompasserve.org>n  b In article <v5s2cmo4d3t762@corp.supernews.com>, "Alan Boyles" <alan.boyles@mindspring.com> writes:M > We have been looking into the same sorts of questions and found a couple of J > answers.  The new Legato solution that the VMS group is sponsoring looksL > quite promising, we also have been working with Tivoli Storage Manager andN > the Archive Backup Client from a company called StorServer.  Our environmentI > uses Oracle so we have some special requirements  but both of these arei > Enterprise solutions.  >   < 	I'm about 80-90% done writing a script interfacing with ABC 	that will:   5 		1)  Automatically archive files based on threshold.   1 		        Current arbitrary threshold I will use:i  + 			o  File hasn't been modified in 6 monthsi2 				( trial sweep without /modified shows I would * 				have been archiving INDEXF.SYS - grin)% 			o  File is larger than 1000 blocksp) 			o  Exclude *.mai ( trial sweep withoute2 			    /modified shows I would have been archiving 				INDEXF.SYS - grin)  < 		2)  Via captive account allow end users to restore files (, 			with appropriate username/password check 6 			- you have to have an account - and audit trail to  			wave at abusers).  G 	Basically, users are abusers.  I politely send out email saying pleasenA 	remove those 50000+ block reports and associated fluff.  Without E 	a formal policy, not much you can do.  And everyone thinks they neednG 	to keep things.  I can't delete it and it isn't worth the aggravation gC 	to fight for the space.  I have one abuser that has 600000 blocks V5 	of old reports, etc.  Several others doing the same.e  2 	The script should be slicker than this mechanism:  M http://www.esm.ucsb.com/fac_staff/fac/mahowald/webpage/insidestuff/legato.htms  B 	The captive account will give menu choices to abstract (hide) the& 	commands and eliminate service calls.  B 	Using a 5 year retention on the archives (arbitrary, if they moan 	I'll up it to 10 years):a  9 $ abc archive x.x/summary/description="February 27, 2003"e  ; Archive Backup Client for ADSM on OpenVMS, Version V3.1.0.1-8 Copyright 1996-2001, Storage Solutions Specialists, Inc.? %ABC-I-SUMM, 27-FEB-2003 09:57:01.13 Summary for ARCHIVE of X.XrI Files scanned:                     1    Buffered I/O count:            33FI Files sent:                        1    Direct I/O count:             706 I Files deactivated:                 0    Page faults:                  115.I Catalog data sent:           0.08 Kb    Elapsed CPU time:   0 00:00:00.08DI Catalog data rcvd:           0.11 Kb    Data rate:            698.23 Kb/srI Data transferred:            7.50 Kb    Data xfer time:     0 00:00:00.01DI Total:                       7.70 Kb    Elapsed time:       0 00:00:00.42a   $ abc show archive x.x/fullI  ; Archive Backup Client for ADSM on OpenVMS, Version V3.1.0.1 8 Copyright 1996-2001, Storage Solutions Specialists, Inc.  $ DISK$SITE_USERS:[USERS.JOEUSER]X.X;1   ADSM server metadata:-%     Server node:  ADSMSERVER.SITE.COM0@     Mgmt class:   DEFAULT                         Owner: JOEUSERH     Media class:  Fixed                           State: <Not specified>M     Backup:       27-FEB-2003 09:57:00          Expires: 26-FEB-2008 09:57:00iN     Object ID:    0x00000000 0x08690020     Restore seq: 0x00006ACE 0x00000000N     Copy Group:   0x00000001                             0x0000001F 0x00000000#     Description:  February 27, 2003iN   ABC file metadata:                                                0x00000000     Owner:      [JOEUSER] D     Size:       14/17                           Created: 30-NOV-2001 17:17:57.45cD     Revision:   (1)                             Revised: 30-NOV-2001 17:17:57.48aH     File ID:    (629,202,0)                     Expires: <Not specified>D     Shelved state:      Online                   Backup:  1-DEC-2001 03:49:14.20rA     File protection:    System: RWED, Owner: RWED, Group:, World: "     File organization:  SequentialI     File attributes:    Allocation: 17, Extend: 0, Global buffer count: 0h)                         Version limit: 30y'     Record format:      Variable lengths9                         Maximum 0 bytes, Longest 17 bytesp(     Record attributes:  Carriage control     Access cntrl list:  None     				Rob?   ------------------------------    Date: 27 Feb 2003 12:52:05 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) ! Subject: Re: VMS Backup solutions 3 Message-ID: <rU7mUPNNPtt7@eisner.encompasserve.org>   M In article <03022615422724@dscis6-0.dalsemi.com>, brandon@dalsemi.com writes:l > Our VMS solution is to use BACKUP with DCL command files.  Automated using ISAM files and scripts with logging, e-mail, reporting,| > WWW, paging, etc.  All in-house created and maintained.  I have worked with SLS in the past and incorporated many of those- > concepts into the in-house backup solution.2 > q > We keep our tapes at a 95% to 98% capacity so we make efficient use of our tapes, no multi-volume tapes either.n >  > Unlike the NT or UNIX group, the VMS group has maintained a stable backup solution incorporated with DR practices and policies.r  H    So they're trying to solve a problem that doesn't exist.  NT and UNIXG    backup utilties are crap, or non-existant.  So they're assuming theyu-    need to include VMS in the solution.  PHM.c   > I am greatly concerned with this new drive to centralize the backup solution onto a UNIX platform.  They want to NFS mount drives3' > and backup the data, etc.  Gasp-Argh!s >  > (?) I would like to know if anyone has a similar environment (NT, UNIX, VMS) that has a successful enterprise solution and what2 > they are using.@  K    A mixture of Legato on and native utilities non-VMS systems with little c    success other than on VMS.a   > (?) Any horror stories?w  C    Normally when we loose a file, it has never been backed up.  The<F    responsible folks have procedures that they follow rigourously, but    those procedures don't work.n  0 > (?) What about Disaster Recovery implications?      Sigh.   ------------------------------  % Date: Thu, 27 Feb 2003 10:36:40 -0500 & From: jlsue <jlsuexxxz@screaminet.com>$ Subject: Re: Volume set across RAID58 Message-ID: <gtas5v81eruoldi6vd0n032h1a0tmnp24h@4ax.com>  7 On Wed, 26 Feb 2003 21:33:34 -0600, "David J. Dachtera"l <djesys.nospam@fsi.net> wrote:   >Bill Todd wrote:w >> "? >> "David J. Dachtera" <djesys.nospam@fsi.net> wrote in messagel$ >> news:3E5C272E.EFFEA73E@fsi.net... >> > jlsue wrote:e >>   >> ... >> cJ >> > > It's just simply amazing the simplification that the EVA can bring. >> >J >> > I tend to view that rather like the choice between burning garbage orA >> > burying it in a landfill. The bad stuff has to go SOMEplace.c >> >M >> > Rather like outsourcing: that heat may be taken off of you, but how muchuK >> > do you trust those upon whom the burden now falls? ...and how will yous= >> > respond should that trust be given and then be breached?s >>  L >> Don't you think that you really ought to learn something about EVA before- >> presuming to voice such opinions about it?e >rG >My point was that the management effort *HAS* to go someplace. You cannF >either manage the storage farm, or manage the stuff that manages yourI >storage farm for you, but either way, you're gonna have your hand in the 2 >pie. The question is just how deep, when and why. >  >TANSTAAFL, as they say.  ; Well, sure, but that's why they put the intelligence in theoB controller.... to save YOU time and energy.  The management of the= storage is now automated, and optimized for performance *and*c> reliability.  (For example, when you create a VRAID-1 LUN, the> controller takes care that LBN 'x' is mirrored on a completely different drive in the group.)  A It appears that you have your doubts about it and don't trust it,t@ which I can empathize with.  But for me, the controllers alreadyE manage the RAID allocation of blocks, and I trust them for that, this F is just an evolution to more intelligence.  And I've seen it in action6 many times, so I'm a bit more comfortable with it now.  F If (like me) you've been burned before by technology claims, then it's; understandable that you'd be hesitant to jump in with newer1F technology.  However, in this case, it's no longer "new", or "bleeding@ edge" stuff - it's been in use for quite awhile in many customerC locations (we probably install 10-20 a month); and it has worked asm advertised.   ? What it's done for databases is eliminate the need to regularly=F re-partition/position data on LUNs to address I/O bottlenecks - a very- manual, time-consuming, often off-hours task.i  E Couple this with the CASA solution that allows you to virtualize youraE LUNs across controllers, even over IP WAN, and you've got some prettyrD cool capabilities for high availability solution as well (far better! than the EMA DRM solution, imho).c  D I think we have the technology available in different locations thatA you can test-drive.  I think we also have some travelling "truck"i< that's full of gear that goes around the country demoing our technology.  Also, if you go to'B DECUS/HPETS/whatever-its-called-this-year, you'll probably have an  opportunity for hands-on access.   ------------------------------    Date: 27 Feb 2003 10:34:14 -0600+ From: young_r@encompasserve.org (Rob Young)v$ Subject: Re: Volume set across RAID53 Message-ID: <bPR48vtsVMkr@eisner.encompasserve.org>h  a In article <gtas5v81eruoldi6vd0n032h1a0tmnp24h@4ax.com>, jlsue <jlsuexxxz@screaminet.com> writes:-   > A > What it's done for databases is eliminate the need to regularlypH > re-partition/position data on LUNs to address I/O bottlenecks - a very/ > manual, time-consuming, often off-hours task.s >    	EVA is a winner all around.  G > Couple this with the CASA solution that allows you to virtualize yourtG > LUNs across controllers, even over IP WAN, and you've got some pretty.F > cool capabilities for high availability solution as well (far better# > than the EMA DRM solution, imho).i  ; 	CASA is kind of an odd bird.  It is neat in some respects,-G 	troublesome in others.  RAID1 across datacenters is good, you would ber8 	beyond full merge problem with shadowing.  However, VMS 	isn't supported yet:s  D http://h18000.www1.hp.com/products/quickspecs/11539_na/11539_na.html  O Windows 2000 with service pack 2 - requires AutoPath version 1.05 for failover oG Windows NT 4.0 with service pack 4 - requires AutoPath version 1.05 ford	 failover e; Solarisd 2.6, 2.7, 2.8 - requires Veritas DMP for failover cL HP-UX K, L, R, V class servers - HP-UX 10.20, 11, 11i - requires Pvlinks for	 failover e& IBM AIX 4.3.3 - no failover available L Linux - Red Hat 7.1/Linux Kernel 2.4 - requires Native Red Hat for failover  NetWare 5.1 and 6   A 	Secondly, others have pointed out (traderags, let me duck) that  G 	CASA in largish situations could be a choke point as it is constrained9) 	to 80000 IOPS and 800 MB/sec throughput.i  ; 	This brochure shows on page 7, you may want 2 of them for 2 	availability:  c http://www.hp.com/products1/storage/pdfs/virtualization_appliances/CASA_Whitepaper_FINAL_110602.pdf   7 	Thirdly, CASA is not cheap.  This is my biggest issue.   	 	Page 14:   N "In a CASA SAN with additional switches, up to approximately 20 hosts and fourK midrange storage systems or a single enterprise-class storage system is theh# recommended maximum configuration."l  ? 	Single EVA?  One EVA per CASA?  Well sure as two EVAs (or moredA 	technically 4 EVA controllers) would be constrained by the CASA x  	chokepoint.  There is an issue.  : 	Divide that out per server and it is a pricey solution.  H 	"Yes, but what price availability?  Can you afford to be without it?"  3 	Good points, but still a large issue cost.   ($$$)c  = 	It is a nice product and I'm sure you are selling it, but torC 	me it would make a whole lot more sense if the out-of-band portionh+ 	of virtualization was quite a bit cheaper.L  > 	Finally, rock solid new method of Shadow mini-merge (storage F 	independent, tying into a new filesystem?) would in many ways render  	CASA irrelevent for VMS.    				Robd   ------------------------------  % Date: Thu, 27 Feb 2003 13:15:30 -0500 A From: "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com>VG Subject: Re: What makes the VMS port to Itanium/IPF/whatever difficult?h. Message-ID: <3e5e55d2$1_2@hpb10302.boi.hp.com>  5 "Andreas Davour" <ante@update.uu.se> wrote in message * news:cs9fzqb2jby.fsf@Tempo.Update.UU.SE...' > Arne Vajhj <arne@vajhoej.dk> writes:K >  > > Alan E. Feldman wrote:I > > > A Unix Admin co-worker asked me what the big deal was about porting)K > > > VMS to Itanium. He says that NetBSD has been ported to numerous chipsBL > > > and that it's no big deal. I told him that I didn't really know. WouldK > > > any of the chip experts reading this care to comment? What is it thatt  > > > makes this port difficult? > >I2 > > VMS was never intended for multi-platform and:9 > >    - uses several constructs which was present in the ? > >      original VAX hardware and need to be emulated on other  > >      architectures> > >    - VMS still has a lot of Macro-32 and Bliss code (while# > >      NetBSD is probably 100% C)i >aE > This last point is interesting. Is this also true of the IA64 port?t
 > Amazing. >   L The truth is that when VMS was written, a lot of it was written in Macro-32.K A *lot*.  Transposing that code into something else is a time consuming and  very error prone process.n  L That said, we have recently made the decision that not only will no new codeL be written in Macro-32 (C is the default for any new code), but that we willK take every "reasonable" opportunity to rewrite Macro-32 source into C.  ButeC it will be case by case... since again, this can be tricky and timeFJ consuming.  There are some very specific areas of the OS where dumping theC Macro-32 code would not only make maintenance easier, but will helpnK performance.  By and large compilers optimize *much* better than hand coded K assembler.  And even though Macro-32 is a compiler, because it is emulatingn@ an assembler it doesn't do as good a job optimizing as it could.  J We are not rewriting Bliss code at this point.  But it has been raised (by* me) as somthing to consider in the future.   ------------------------------  + Date: Thu, 27 Feb 2003 13:11:47 +0000 (UTC)n9 From: Roar =?iso-8859-1?Q?Thron=E6s?= <roart@nvg.ntnu.no> ) Subject: Re: Where to get a recent Emacs? - Message-ID: <b3l2qj$gtp$1@tyfon.itea.ntnu.no>e  ) Galen Tackett <gtackett@yahoo.com> wrote:-E : Is there somewhere within the US where I can download an Emacs thattE : will work with OpenVMS V5.3-1 on Alpha? I can't get to the officialoA : mirrors at MIT and in Sweden(?) and Japan, for some reason, but@3 : perhaps there's another stateside site I can try..  > If you want a recent emacs (21), you may not(?) find it there.= If you want it, you may also have some developement yourself.    --  
 -Roar Thronse   ------------------------------  # Date: Thu, 27 Feb 2003 15:50:07 GMTf# From: "John Smith" <a@nonymous.com>P Subject: Will HP pull an IBM? E Message-ID: <Psq7a.1499$em1.430@news04.bloor.is.net.cable.rogers.com>   2 http://www.eweek.com/article2/0,3959,809041,00.asp  D ....Three sentences and a bullet list later, one finds the disarming< suggestion to "deploy new e-business technology applicationsF concurrently with existing OS/2 applications until platform neutrality8 has been achieved and then change the operating system."  3 This has been the Digital/Compaq/HP line for years.e    ' Lessons for VMS (from the same article)r ============================@ No. 1: If you're only incrementally better than the competition,= prospective customers have to be sure you'll do them no harm.L  E No. 2: If developers don't love you, users will never even get to trye you.   ------------------------------    Date: 27 Feb 2003 06:45:16 -0600B From: clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley): Subject: Re: Wow - OpenVMS Job Advert in UK Computer Press3 Message-ID: <7fHb8G9lBg$9@eisner.encompasserve.org>R  _ In article <v5p7c8st7ssab9@corp.supernews.com>, Roy Omond <Roy.Omond@BlueBubble.UK.Com> writes:( > B > The position is in Cambridge, which would have been nice for me. >  > Roy OmondE > Blue Bubble Ltd. >   G Actually, I'm more interested in knowing what the VMS market is like in  the North... :-)   Simon.   --  B Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP       L VMS advocate: One who makes a Mac advocate look like a beginner at advocacy.   ------------------------------  # Date: Thu, 27 Feb 2003 16:50:34 GMTR4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk>: Subject: Re: Wow - OpenVMS Job Advert in UK Computer Press0 Message-ID: <3E832BCD.740CC1C6@blueyonder.co.uk>   Simon Clubley wrote: > a > In article <v5p7c8st7ssab9@corp.supernews.com>, Roy Omond <Roy.Omond@BlueBubble.UK.Com> writes:0 > >3D > > The position is in Cambridge, which would have been nice for me. > > 
 > > Roy OmondW > > Blue Bubble Ltd. > >  > I > Actually, I'm more interested in knowing what the VMS market is like in  > the North... :-)  T There is a VMS Basic Programmer position in West Yorks that has been knocking around" on jobserve for months and months.   regards,   >  > Simon. >  > --= > Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFPcN > VMS advocate: One who makes a Mac advocate look like a beginner at advocacy.   -- t tim.llewellyn@blueyonder.co.uk D  H * PLEASE NOTE tim.llewellyn@cableinet.co.uk address is NO LONGER VALID *   ------------------------------   End of INFO-VAX 2003.114 ************************