1 INFO-VAX	Tue, 04 Mar 2003	Volume 2003 : Issue 123       Contents: Announcing Weatherduck for VMS6 Re: Anyone use a Compaq 20/40 dlt tape drive in a vax? Athens Olympics / Re: Backup options (external SCSI) for PWS500au  C-Kermit 8.0.208 Re: C-Kermit 8.0.208 Re: CMSVE04041 ? Re: CMSVE04041 ?6 Re: Editor wars - Was: Re: Upcoming VMS improvements ?6 Re: Editor wars - Was: Re: Upcoming VMS improvements ?6 Re: Editor wars - Was: Re: Upcoming VMS improvements ?6 Re: Editor wars - Was: Re: Upcoming VMS improvements ?6 RE: Editor wars - Was: Re: Upcoming VMS improvements ?6 Re: Editor wars - Was: Re: Upcoming VMS improvements ?6 Re: Editor wars - Was: Re: Upcoming VMS improvements ?6 Re: Editor wars - Was: Re: Upcoming VMS improvements ?6 RE: Editor wars - Was: Re: Upcoming VMS improvements ?P Re: EDT and command tables have invalid format (was: Re: What makes the VMS port  Re: Fix for EDT emulation in EVE7 Re: fork() on VMS, was: Re: Upcoming VMS improvements ? ' Re: HP Reports 1st Quarter 2003 Results ' Re: HP Reports 1st Quarter 2003 Results % I'm going to sleep so well tonight !! ) Re: I'm going to sleep so well tonight !! ) Re: I'm going to sleep so well tonight !! ) Re: I'm going to sleep so well tonight !! ) Re: I'm going to sleep so well tonight !! ) Re: I'm going to sleep so well tonight !! : Re: Intel Itanium 67% slower than Intel Alpha in SAP tests: Re: Intel Itanium 67% slower than Intel Alpha in SAP tests/ Re: Linus trashes Itanium - he's scared of VMS!  MONITOR SYSTEM screen values  Re: MONITOR SYSTEM screen values  Re: MONITOR SYSTEM screen values  Re: MONITOR SYSTEM screen values  Re: MONITOR SYSTEM screen values  Re: MONITOR SYSTEM screen values Re: oracle benchmarks on VMS Re: oracle benchmarks on VMS Re: oracle benchmarks on VMS, Re: OT: America's Cup  [ was Re: I am back ], Re: OT: America's Cup  [ was Re: I am back ]5 Re: PC developers opinion of 64 bit desktop computing E Re: Question: why sometimes asctim returns "17-NOV-1858 00:00:00.00"? 7 Re: RDB "Tech Day" comes to New York on  March 14, 2003 7 Re: RDB "Tech Day" comes to New York on  March 14, 2003 , Re: running autogen on alternate system disk, Re: running autogen on alternate system disk, Re: running autogen on alternate system disk, Re: running autogen on alternate system disk, Re: running autogen on alternate system disk Re: SCS/DECnet over IP Re: SCS/DECnet over IP3 Re: SCSI cluster disk thrashing between connections 3 Re: SCSI cluster disk thrashing between connections 3 Re: SCSI cluster disk thrashing between connections $ simh: ftp access to vms from host os Re: Terminal Servers Re: unix Re: unix Re: unix Re: unix< Re: VAXELN vs Posix Threads (Was: DECthreads problem on VAX) Re: VMS Backup solutions Re: VMS Backup solutions Re: VMS Backup solutions Re: VMS Backup solutions Re: VMS Backup solutions, Re: VMS Backup solutions (alias directories) Re: Volume set across RAID5  Re: Volume set across RAID5   F ----------------------------------------------------------------------   Date: 3 Mar 2003 19:43:10 -0800 $ From: bfarmer@wcfa.com (Bill Farmer)' Subject: Announcing Weatherduck for VMS = Message-ID: <43840735.0303031943.4e2a9568@posting.google.com>   F The Weatherduck is a $179.00 environment monitor that communicates via RS232.  @ Reports temperature, humidity, airflow, lightlevel, three analog inputs + 1-wire bus interface.  F Additional remote sensors available (door switches, remote temperature sensors)  F Attach via terminal server port (LAT or TELNET), CISCO router AUX portE (TELNET), or direct attach to processor serial port.  No wall wart as C power is supplied by serial port (Works anywhere in the galaxy (VMS  pun intentional)).  E VMS software monitors duck, logs data, draws charts (JPG or PNG), and D displays charts via HTTP.  No web server needed, but charts can drop+ directly into your VMS web server directly.   A Order at WWW.WEATHERDUCK.US or in Europe at WWW.WEATHERDUCK.CO.UK   D If you have to, attach the duck to a Win2K machine serial port.  The, VMS software can still get the data via XML.   Enjoy!   Bill   ------------------------------  # Date: Tue, 04 Mar 2003 03:01:59 GMT # From: hoff@hp.nospam (Hoff Hoffman) ? Subject: Re: Anyone use a Compaq 20/40 dlt tape drive in a vax? - Message-ID: <HGU8a.13$gl.11@news.cpqcorp.net>   i In article <_0GdnRgcn5ZNHcOjXTWc3g@brightview.com>, "Jefferson Humber" <matrix01@globalnet.co.uk> writes: B :I have used a Compaq external 20/40 DLT on a VAX 4000-100 running :5.5-2H4..... worked OK for me.  : H :I think if I remember correctly OpenVMS see's it as 'Generic TZ Drive'.    C   Close.  Usually Generic MK, but the device names can and do vary.   @   Releases as old as V5.5-2 are not particularly good at generic?   SCSI device support -- I generally encourage V6.2 and V7.1 as ?   being better choices, given improvements made in SCSI support +   in these (and in later) OpenVMS releases.     / :"Jim Agnew" <jpagnew@vcu.edu> wrote in message " :news:3E5E664C.77A4C2E4@vcu.edu...K :> Version 5.5-2 VMS  they say it's not supported til v7 or so, but if it's I :> a generic scsi tape, that works, that's enuf for me.. I just happen to F :> have one laying around, and would like to see if it works before we! :> cought up for an adapter.. ;-)   E   Which VAX (Jefferson Humber mentioned a VAX 4000 model 100, but Jim H   Agnew has yet to admit to a particular model :-), and which DAT drive?  E   This stuff probably works as a generic device, but there are rather G   few specific details here for anything approaching certainty -- newer H   versions of OpenVMS do expect more from SCSI tape drives, and this canI   cause (harmless) messages to be logged against certain drives.  (Please G   see the OpenVMS V7.3 release notes for "Stricter Requirement for Mode +   Page 1 on SCSI Tape Drives" for details.)   J   If the (unnamed) VAX box does not have SCSI, then the connection can getH   rather more "interesting" -- systems without native SCSI usually mean G   boxes with DSSI or Q-bus or some other connection, and from there out    to SCSI.  If possible.  C   Another obvious option is to cluster with or network with a newer C   OpenVMS VAX or (better) OpenVMS Alpha box (with native SCSI), and D   avoid the whole controller-related problem.  You might be able to B   add a whole (used) box for the same time and effort spent on the*   SCSI controller and support issues here.    N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com    ------------------------------  % Date: Mon, 03 Mar 2003 22:06:36 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca> Subject: Athens Olympics/ Message-ID: <3E640A29.8FA09DFA@vl.videotron.ca>   I I remember hearing some blurb about IBM not wanting to continue to be the - "official supplier of" Olympics after Sydney.   M Does anyone know more about this ? Would HP have gotten the contract ? If so, J now that HP has inherited Digital, would there be any chances of Alpha/VMSL being used by HP to run the olynpics, or would that be just HP-UX and wintel crap ?   ------------------------------  # Date: Mon, 03 Mar 2003 21:42:21 GMT A From: "Colin Butcher" <colin_DOT.butcher_AT@xdelta_DOT.co_DOT.uk> 8 Subject: Re: Backup options (external SCSI) for PWS500au? Message-ID: <1%P8a.13429$rr4.108626010@news-text.cableinet.net>   I Cheat. Buy a SCSI bulkhead connector and cable it up so that the internal J SCSI bus also becomes external. Not great throughput, but it does work. BeJ careful with termination. Multiple controllers is a better way to do it. IJ happen to use a 3 bus SWXCR with 3x internal 9Gb discs, each on a separateG SCSI channel, then cable the internal bus straight to the outside world  using a bulkhead connector.   C See www.videk.co.uk and follow the SCSI internal ribbon cable link.    -- Hope this helps. Cheers, Colin. ' (colinDOT.butcherAT@xdeltaDOT.coDOT.uk)    ------------------------------   Date: 3 Mar 2003 15:27:59 -0500 & From: fdc@columbia.edu (Frank da Cruz) Subject: C-Kermit 8.0.208 1 Message-ID: <b40dsf$aip$1@watsol.cc.columbia.edu>   6 There's a release candidate for C-Kermit 8.0.208 here:  -   http://www.columbia.edu/kermit/ckdaily.html   ? I haven't been able to build it on the same wide variety of old ? and new VMS systems as before because some of the old ones have @ been dropping offline.  Most of the new binaries are Alpha only,@ VMS 6.2 and later.  If anyone can make VAX binaries (any version> except 7.1, which I have) or VMS binaries (VAX or Alpha) prior6 to VMS 6.2, I'd be grateful if you could send them in.  @ The major new feature of this version is its ability to be built> with SSL/TLS security.  Of course we can't put secure binaries= online for FTP due to USA export law, so I'm only looking for * regular binaries, built in the normal way.   Thanks!    - Frank   B P.S. This version also has the session-log text/binary distinction discussed here last week.    ------------------------------  % Date: Mon, 03 Mar 2003 19:03:34 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca> Subject: Re: C-Kermit 8.0.208 / Message-ID: <3E63DF26.AA4C2261@vl.videotron.ca>    Frank da Cruz wrote:@ > with SSL/TLS security.  Of course we can't put secure binaries? > online for FTP due to USA export law, so I'm only looking for , > regular binaries, built in the normal way.  M Didn't the USA export restrictions get lifted sometime towards the end of the  Clinton administration ?   ------------------------------   Date: 3 Mar 2003 11:49:46 -0800 = From: robert.fairfield@verizonwireless.com (Robert Fairfield)  Subject: Re: CMSVE04041 ? ; Message-ID: <4b4254e.0303031149.21774f8@posting.google.com>   Z p_sture@elias.decus.ch (Paul Sture) wrote in message news:<L8SjEq4xIwXB@elias.decus.ch>...p > In article <i0r8a.197028$Rb4.2379703@news.chello.at>, peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) writes:K > > I just noted on the australian ECO site, that there is a CMSAE04041 and M > > CMSVE04041 since last October but I so far haven't seen them on the usual @ > > ECO sites downloadable. Has anyone managed to see/get them ? > >  >  > I can see the README at M > http://ftp1.support.compaq.com/patches/public/Readmes/vms/cmsae04041.README   @ I went to this web site, downloaded the .EXE file, installed it,E validated it against my 3.9 ECO code, fix the update errors (not real = errors, just CMS version fixes), and it's all working so far.    > C > That's the Alpha version, and contains links to the actual files.  > D > > No, I don't really need them (but I'd be pleased if I get them).& > > And, no, I've no support contract.L > > And, no, I don't think they are only on restricted distribution, becauseD > > 	a) public distribution is growing while restricted is shrinking' > > 	b) ECO 1 is on public distribution  > > B > > It is mere a question of how well is the care of the ECO sitesD > > and who to turn to with questions/corrections for the ECO site ? > >  > > Many TIA > >  > > --   > > Peter "EPLAN" LANGSTOEGER ) > > Network and OpenVMS system specialist   > > E-mail  peter@langstoeger.atJ > > A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist > --   ------------------------------  # Date: Mon, 03 Mar 2003 23:56:53 GMT 6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) Subject: Re: CMSVE04041 ? 3 Message-ID: <9ZR8a.22511$AV5.283139@news.chello.at>   { In article <4b4254e.0303031149.21774f8@posting.google.com>, robert.fairfield@verizonwireless.com (Robert Fairfield) writes: [ >p_sture@elias.decus.ch (Paul Sture) wrote in message news:<L8SjEq4xIwXB@elias.decus.ch>... q >> In article <i0r8a.197028$Rb4.2379703@news.chello.at>, peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) writes: L >> > I just noted on the australian ECO site, that there is a CMSAE04041 andN >> > CMSVE04041 since last October but I so far haven't seen them on the usualA >> > ECO sites downloadable. Has anyone managed to see/get them ?  >>   >> I can see the README at  N >> http://ftp1.support.compaq.com/patches/public/Readmes/vms/cmsae04041.README > A >I went to this web site, downloaded the .EXE file, installed it, F >validated it against my 3.9 ECO code, fix the update errors (not real> >errors, just CMS version fixes), and it's all working so far.  H This URL leads to a page with username/password for entitled patches, soJ the answer is, the ECO is no longer on public distribution. Don't like it, but I've an answer now.   G Then I went on, created an account and tried to download the ECOs again   K 550 /patches/entitled/vms/vax/v7.3/cms/4.1/cmsve04041.a-dcx_vaxexe: No such ' file or directory	for all four files...     Now I wonder how did you do it ?   --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  % Date: Mon, 03 Mar 2003 13:47:11 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>? Subject: Re: Editor wars - Was: Re: Upcoming VMS improvements ? . Message-ID: <3E63951A.3FB2145@vl.videotron.ca>   "Alan E. Feldman" wrote:F > Bonk! I'm not talking about DECwindows. Not everyone has DECwindows.  I But my editor is better than yours because it has both interfaces ... :-)     F > does not make SET NOTRUNCATE moot. Without SET NOTRUNCATE, you can'tB > keep an eye on the data in two distant fields as you scroll down  N Then, you need the TSO editor on 3270 screens. That one had "vertical windows"L where you could show one part of the records on the left and another part on
 the right.  M Note that WPS PLUS does have the NOTRUNCATE capability. You just set a margin I at 80 characters and the records will flow. When finished, you remove the > right margin and the records become one line per record again.    D > If you don't have a need for SET NOTRUNCATE, fine. But stop saying9 > that I or anyone else doesn't. Who are you to say that?    Because we are at war....  :-)   C > So? I have no problem with people using TECO. It is *you* who get < > upset by people using other editors. What is your problem?  J We are the TPU, resistance is futile, you will be assimilated and your EDT; customisations will be adapted to serve the TPU collective.     C > Second, my 262-line EDT init file boils down to 84 lines when you  > remove the comments   M So converting your customisations wouldn't be so bad after all since it isn't / as big as you had lead us to believe initially.   4 > (which was stated in my post). To do the same withH > TPU would probably take thousands of lines and hour and hours of time.  K Out of curiosity, if you posted your customisation file [ok, I read further G along, by all means, do email it to me, just remove the spamnot from my L address], perhaps we could comment on how easy/difficult it would be to portK to TPU. We already know that NOTRUNCATE would be missing but would still be 8 interesting to see how much else might be tough to port.  J This is a bit of a selfish reason. By posting your customisation, it wouldL force me to check out TPU and learn more about it to see how to generate theK same behaviour as your customisation file. Seriously, I don't expect you to . migrate, but the jihad is just too much fun !)    D > Fourth, the question is not which editor is better, but why shouldD > someone who is happy with EDT should bother to switch to an editor4 > (for normal daily use) that he finds to be a pain?  I OK, here is a serious answer:  Over my life, I have always made the wrong J technological choices. I chose Apple, Apple didn't become the prevalent. IN dumped IBM mainframe and chose VMS, VMS went down the drain. I chose ALL-IN-1,M ALL-IN-1 went down the drain, I wrote many gateways to Message Router, Palmer J killed it.  I chose PSION organisers, they went down the drain.  All theseM choices were the best solution technologically and served well. But they were   the wrong "political" decisions.  J I have become extremely conscious of staying attached to "mature" productsI because I know that eventually, i'll have to switch. Getting a DECwindows E workstation has allowed me to move a bit closer to the "real world".    M EDT is far more mature than TPU.  And Even TPU seems dead and I know now that L logically, I should be dumping TPU and going for EMACS which seems to be the+ industry standard editor in the unix world.   I Similarly, I feel strong pressure to move to Perl and stop using ALL-IN-1 L which I am far mroe familiar with and which I know works on VMS with all theL attachements to VMS (SYSUAF, indexed files, queue system, logical,s symbols, right out of the box).  L It becomes a constant battle between using the tools you know and which giveG you better productivity NOW, against the knowledge that those tools are , worthless in the long term for your carreer.  L Continuing to grow roots into a mature product will only make its inevitableF trasition to a newer product more painful. I could have the exact sameI comments/attitude as you when talking about moving from ALL-IN-1 to Perl. T Except that I accept the fact that I must one day get off my dependance on ALL-IN-1.    H > I have tried vi. I didn't like it, but no one could give me a tutorial1 > or a manual or a secret code chart at the time.   M I tried vi too, and I think it seems to be similar to TECO. (the sole user of L TECO has me killfiled so I don't have to worry about him being offended if I comperate vi to TECO).   > > NO EDITOR HAS EVERYTHING.  > & > Sounds like you're arguing my ponit!  K No ! You are the EDT ennemy. I cannot be seen to agree with you. And unless N you are willing to raise the white flag and surrended to the TPU camp, the warN MUST continue :-) :-)  (which event flag would be considered "white" in VMS ?)  D > it does too many things that annoy me. Found in reverse direction?@ > What the hell is the purpose of that? What possible use is it?  N Say you're are the middle of a file. You are looking for "chocolate". It isn'tM found between current position and end of file, but it was found near the top 2 of the file. TPU gives you the option to go there.  : > Well, thank you! How kind of you to allow me to use EDT.  K The president of the United Routines of TPU is running out of patience. You K must comply NOW. You must destroy your EDT initialisation. And stop playing G hide and seek by constantly renaming it to hide it from the inspectors.     . P.S. please don't take the jihad so seriously.   ------------------------------  % Date: Mon, 03 Mar 2003 19:51:46 +0000 ' From: Elliott Roper <elliott@yrl.co.uk> ? Subject: Re: Editor wars - Was: Re: Upcoming VMS improvements ? 2 Message-ID: <030320031951463269%elliott@yrl.co.uk>  7 In article <3E63951A.3FB2145@vl.videotron.ca>, JF Mezei ( <jfmezei.spamnot@vl.videotron.ca> wrote:  O > I tried vi too, and I think it seems to be similar to TECO. (the sole user of N > TECO has me killfiled so I don't have to worry about him being offended if I > comperate vi to TECO).  G The *other* user of teco rather enjoys your posts, but *is* offended by F this comparison. vi shares some less than charming idiosyncrasies withG EVE, like special behavior of searches succeeding/failing at the end of  lines. EVE=evil vi=VILE  TPU=totally puked upF EDT ... words fail me. I still feel the pain of those wilderness years when there was no teco on VMS.   ------------------------------  % Date: Mon, 03 Mar 2003 21:07:55 +0100 9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> ? Subject: Re: Editor wars - Was: Re: Upcoming VMS improvements ? ' Message-ID: <3E63B61B.ABA2915B@aaa.com>   : Talking about editors and writing "code", what about LSE ?> I havn't used it much, but what I'v seen of it, it looks good.   Jan-Erik Sdereholm    ------------------------------  % Date: Mon, 03 Mar 2003 17:31:37 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>? Subject: Re: Editor wars - Was: Re: Upcoming VMS improvements ? / Message-ID: <3E63C99F.EF135E50@vl.videotron.ca>    Elliott Roper wrote:J >> The *other* user of teco rather enjoys your posts, but *is* offended byH > this comparison. vi shares some less than charming idiosyncrasies with > EVE,  J Didn't mean to insult you., My image of TECO is some strange set of singleM keystrokes that produce marvelous results when you know what you're doing, or M totally random results when you don't, and this remonded me of vi which isn't N really a full screen editor since you constantly have to tell it when you wantA to change text. (unless vi changed since I last tried to use it).    ------------------------------  $ Date: Mon, 3 Mar 2003 14:43:27 -0800$ From: Shane Smith <ssmith@icius.com>? Subject: RE: Editor wars - Was: Re: Upcoming VMS improvements ? 0 Message-ID: <01C2E193.40E2C240@sulfer.icius.com>  E VI was still a mess last time I was forced to use it (and I /do/ mean  forced).  F I am a TPU user who respects EDT, and even uses it occasionally. But IF still have a soft spot for something called TED on a PDP-11 in the midE 80s. It was a pure line editor, and could only go one way through the E file, down. If you went past the line you wanted, you had to exit and E start again. You /couldn't/ forget to save occasionally. :-) See? You & can find something good in any editor.   Shane    -----Original Message-----7 From: JF Mezei [mailto:jfmezei.spamnot@vl.videotron.ca] $ Sent: Monday, March 03, 2003 1:32 PM To: Info-VAX@Mvb.Saic.Com ? Subject: Re: Editor wars - Was: Re: Upcoming VMS improvements ?      Elliott Roper wrote:J >> The *other* user of teco rather enjoys your posts, but *is* offended byH > this comparison. vi shares some less than charming idiosyncrasies with > EVE,  C Didn't mean to insult you., My image of TECO is some strange set of  singleC keystrokes that produce marvelous results when you know what you're 	 doing, or G totally random results when you don't, and this remonded me of vi which  isn't E really a full screen editor since you constantly have to tell it when  you wantA to change text. (unless vi changed since I last tried to use it).    ------------------------------  % Date: Mon, 03 Mar 2003 23:39:15 +0000 ' From: Elliott Roper <elliott@yrl.co.uk> ? Subject: Re: Editor wars - Was: Re: Upcoming VMS improvements ? 2 Message-ID: <030320032339152457%elliott@yrl.co.uk>  8 In article <3E63C99F.EF135E50@vl.videotron.ca>, JF Mezei( <jfmezei.spamnot@vl.videotron.ca> wrote:   > Elliott Roper wrote:L > >> The *other* user of teco rather enjoys your posts, but *is* offended byJ > > this comparison. vi shares some less than charming idiosyncrasies with > > EVE, > L > Didn't mean to insult you., My image of TECO is some strange set of singleO > keystrokes that produce marvelous results when you know what you're doing, or ) > totally random results when you don't,  E even worse are the slightly-not-so-marvelous results when you *think* < you know what you are doing...which is part of teco's charm.( > and this remonded me of vi which isn'tP > really a full screen editor since you constantly have to tell it when you wantC > to change text. (unless vi changed since I last tried to use it). F vi and teco have that in common, yet to one whose teco knowledge is inG the fingers, vi is utterly alien. I suspect regular vi users would feel  the same about teco.  A Full screen editors consume so many keystrokes, and so many mouse C operations, that they are not really all that efficient compared to F teco once those single keystroke commands are printed onto your spinalD column. The real strength lies somewhere between search and type andE macros that are too simple to write perl for. You just invent them on G the fly, using a different part of your brain that the part you use for B the managing content of what you are editing. In a funny way it isF quite satisfying. The other satisfying part is other peoples' reactionF after watching you at work. (fingers twirled near ears features a lot)   ------------------------------   Date: 3 Mar 2003 18:41:30 -0800 . From: spamsink2001@yahoo.com (Alan E. Feldman)? Subject: Re: Editor wars - Was: Re: Upcoming VMS improvements ? < Message-ID: <b096a4ee.0303031841.ec2860e@posting.google.com>  f JF Mezei <jfmezei.spamnot@vl.videotron.ca> wrote in message news:<3E63951A.3FB2145@vl.videotron.ca>... > "Alan E. Feldman" wrote:H > > Bonk! I'm not talking about DECwindows. Not everyone has DECwindows. > K > But my editor is better than yours because it has both interfaces ... :-)  >  > H > > does not make SET NOTRUNCATE moot. Without SET NOTRUNCATE, you can'tD > > keep an eye on the data in two distant fields as you scroll down > P > Then, you need the TSO editor on 3270 screens. That one had "vertical windows"N > where you could show one part of the records on the left and another part on > the right. > O > Note that WPS PLUS does have the NOTRUNCATE capability. You just set a margin K > at 80 characters and the records will flow. When finished, you remove the @ > right margin and the records become one line per record again.  + Thanks for the tips on the other editors.    [...]   E > > Second, my 262-line EDT init file boils down to 84 lines when you  > > remove the comments  > O > So converting your customisations wouldn't be so bad after all since it isn't 1 > as big as you had lead us to believe initially.   E Well, let me state it this way. EDT is not so bad that I need 262 EDT E line mode commands. It's really 84 actual EDT line mode commands. But @ the commands themselves were mostly easy to write and are mostlyE useful additions to EDT or just some settings, not multiline programs D to do what should be basically very simple tasks. Yet they are still5 too many to just sit down and convert for use in EVE.   6 > > (which was stated in my post). To do the same withJ > > TPU would probably take thousands of lines and hour and hours of time. > M > Out of curiosity, if you posted your customisation file [ok, I read further I > along, by all means, do email it to me, just remove the spamnot from my N > address], perhaps we could comment on how easy/difficult it would be to portM > to TPU. We already know that NOTRUNCATE would be missing but would still be : > interesting to see how much else might be tough to port.  E Maybe I'll post it and/or e-mail it to you later if I have time. It's F getting late and I really ought to wrap this up right now and go home.  L > This is a bit of a selfish reason. By posting your customisation, it wouldN > force me to check out TPU and learn more about it to see how to generate theM > same behaviour as your customisation file. Seriously, I don't expect you to 0 > migrate, but the jihad is just too much fun !)  @ Well, I certainly would like to have it translated into EVE/TPU!  F > > Fourth, the question is not which editor is better, but why shouldF > > someone who is happy with EDT should bother to switch to an editor6 > > (for normal daily use) that he finds to be a pain? > K > OK, here is a serious answer:  Over my life, I have always made the wrong L > technological choices. I chose Apple, Apple didn't become the prevalent. IP > dumped IBM mainframe and chose VMS, VMS went down the drain. I chose ALL-IN-1,O > ALL-IN-1 went down the drain, I wrote many gateways to Message Router, Palmer L > killed it.  I chose PSION organisers, they went down the drain.  All theseO > choices were the best solution technologically and served well. But they were " > the wrong "political" decisions. > L > I have become extremely conscious of staying attached to "mature" productsK > because I know that eventually, i'll have to switch. Getting a DECwindows G > workstation has allowed me to move a bit closer to the "real world".   > O > EDT is far more mature than TPU.  And Even TPU seems dead and I know now that N > logically, I should be dumping TPU and going for EMACS which seems to be the- > industry standard editor in the unix world.  > K > Similarly, I feel strong pressure to move to Perl and stop using ALL-IN-1 N > which I am far mroe familiar with and which I know works on VMS with all theN > attachements to VMS (SYSUAF, indexed files, queue system, logical,s symbols, > right out of the box). > N > It becomes a constant battle between using the tools you know and which giveI > you better productivity NOW, against the knowledge that those tools are . > worthless in the long term for your carreer. > N > Continuing to grow roots into a mature product will only make its inevitableH > trasition to a newer product more painful. I could have the exact sameK > comments/attitude as you when talking about moving from ALL-IN-1 to Perl. V > Except that I accept the fact that I must one day get off my dependance on ALL-IN-1.  F OK. I see. However, it looks like EDT will continue to work as long asD VMS does. And I am glad that EVE is there for things I can't do with EDT.   [...] F > > it does too many things that annoy me. Found in reverse direction?B > > What the hell is the purpose of that? What possible use is it? > P > Say you're are the middle of a file. You are looking for "chocolate". It isn'tO > found between current position and end of file, but it was found near the top 4 > of the file. TPU gives you the option to go there.  F I'm sorry, but I still see no point to this "found in other direction"@ routine. The case you just mentioned is the only one in which itF almost makes *some* sense, but I'd still prefer that it not do that. I@ *know* that I started in the middle of the file (assuming that ID started in the middle of the file). And I know which way I am going.E If the string is not found at all in the current direction, just tell E me that the string was not found. If I wanted to find all occurrences C of the string, I would have started at the top of the buffer in the = first place! If by some chance I *do* want to try the reverse E direction, I'll do it myself. I don't want to be asked about it every  time.   C Also, this "Found in other direction. Go there? " question *always* D pops up when searching reaches the end of the buffer, even if you'veF already found all occurrences of the string. Even if you've found themD all 13 times. What's the point of that? "Well, yes, sure, take me toF an occurrence of the string you've already found for me before." Nope,F it makes no sense to me. If I want to see all occurrences of a string,B I'll start at the top and see them in order. I never want to startE from the middle, go to the end, and then go back to the other part of @ the file. With "found in other direction", how would I know whenF repetition starts if the file is long and complex? The right way to doE it is to start at the top of the buffer in the first place and making  a single pass.  F Hold on a minute. NEWS FLASH. Beep, de beep beep .. beep, beep. I have= just discovered a way to partially subvert the Found in other D direction nonsense. Just press Find Next again and the evil questionE goes away! I'd still rather it not pop up in the first place, because F it's still an extra keystroke compared to EDT. The problem is that allB these extra keystrokes and delays interrupt my flow of thought and< make editing less efficient. (Uh, yes, I'm tend to be ratherD impatient.) All these faster computers and still we have to wait forE some things. All the tricks to speed things up only to be slowed down C by delayed cursor settelment and found in other direction questions C and forgetting to switch to lower case before searching (okay, that E last one is fixed by a posted TPU code -- thank you to its author! --  so we can drop that one).    [...]    Disclaimer: JMNSHO Alan E. Feldman    ------------------------------  % Date: Mon, 03 Mar 2003 20:18:56 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> ? Subject: Re: Editor wars - Was: Re: Upcoming VMS improvements ? ' Message-ID: <3E640D10.274140E0@fsi.net>    Shane Smith wrote: > G > VI was still a mess last time I was forced to use it (and I /do/ mean 
 > forced).  F True. I found myself on a UNIX box in 1986 when I'd been on IAS (TECO)C for 18 months and RSTS/E and VMS (EDT) for just shy of three years.   D I managed to make vi as EDT-like as I could, but lost the .exrc fileB when I changed jobs (to a VMS/All-in-1 shop). Could probably do it again, I s'pose...   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  $ Date: Mon, 3 Mar 2003 19:12:43 -0800$ From: Shane Smith <ssmith@icius.com>? Subject: RE: Editor wars - Was: Re: Upcoming VMS improvements ? 0 Message-ID: <01C2E1B8.F7A7DA40@sulfer.icius.com>  F NO!!! DON'T DO IT! >>>bullhorn-crackle<<< Mr. Dachtera, step away from the machine!  A If you do do it, don't expect us to clean up the drool or buy the  straightjacket.    Shane    -----Original Message-----6 From: David J. Dachtera [mailto:djesys.nospam@fsi.net]$ Sent: Monday, March 03, 2003 6:19 PM To: Info-VAX@Mvb.Saic.Com ? Subject: Re: Editor wars - Was: Re: Upcoming VMS improvements ?      Shane Smith wrote: > G > VI was still a mess last time I was forced to use it (and I /do/ mean 
 > forced).  F True. I found myself on a UNIX box in 1986 when I'd been on IAS (TECO)C for 18 months and RSTS/E and VMS (EDT) for just shy of three years.   D I managed to make vi as EDT-like as I could, but lost the .exrc fileB when I changed jobs (to a VMS/All-in-1 shop). Could probably do it again, I s'pose...   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  # Date: Tue, 04 Mar 2003 03:15:22 GMT # From: hoff@hp.nospam (Hoff Hoffman) Y Subject: Re: EDT and command tables have invalid format (was: Re: What makes the VMS port - Message-ID: <eTU8a.14$gl.13@news.cpqcorp.net>   w In article <01KSZ13ZT3O29FNHX8@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:    :...I often SPAWN out of  I :EDT while in MAIL (there are several separate reasons I do this, and it  ? :is also handy when one has only one session available).  What  , :specifically annoys me is the error message : : :   command tables have invalid format - see documentation : : :(that's right, no %DCL-E-TBLINV or whatever as a prefix).  H   The error is "%CLI-E-INVTAB, command tables have invalid format - see E   documentation", and this implies some sort of a CLI corruption, and E   some sort of a bug with a call to CLI$DCL_PARSE or other such call.   ' :By the way, in case anyone is curious:  : 8 :DEFINE KEY FUNCTION 29 AS "XLATESPAWN ?*'SPAWN '^ZREF."  E   I'm not particularly familiar with that EDT command syntax, and EDT E   has not seen particular work in eons.  You've defined the DO key to F   some sort of SPAWN, obviously.  (I can't replicate the SPAWN error.)     OpenVMS platform and version?   N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com    ------------------------------  % Date: Mon, 03 Mar 2003 20:03:47 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> ) Subject: Re: Fix for EDT emulation in EVE ' Message-ID: <3E640983.49F8965A@fsi.net>    Rob Brown wrote: >  > > Peter Weaver wrote:  > > >  > > > David J. Dachtera wrote:
 > > > >...L > > > > In EDT, I use GOLD-F to write the current buffer to a file. It wouldH > > > > be nice if EDT could write the current select range to file. I'dK > > > > probably put that on GOLD-S or something that I don't currently use  > > > > for anything else.
 > > > >... > > > ! > > > Does this do what you want?  > > > J > > > define key gold s as "cutsr=writesrbuffer paste=writesrbuffer ext wr8 > > > ?*'Enter the Filename to Write: ' =writesrbuffer." > & > Why not (making this up on the fly): > H >   define key gold s as "ext wr ?*'Enter the Filename to Write: ' sel."  0 Just tried that. Execellent! Thank you so much!   D I would never have suspected that the trailing "SEL" would work thatD way. Is that documented somewhere (cite publication and page number, please)?   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------   Date: 3 Mar 2003 16:57:53 -0800 1 From: keithparris_NOSPAM@yahoo.com (Keith Parris) @ Subject: Re: fork() on VMS, was: Re: Upcoming VMS improvements ?= Message-ID: <cf15391e.0303031657.20e66c98@posting.google.com>   } clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley) wrote in message news:<tHZFEvX9esxD@eisner.encompasserve.org>... s > In article <cf15391e.0302270930.34290dc3@posting.google.com>, keithparris_NOSPAM@yahoo.com (Keith Parris) writes: F > > VMS can meet COE standards without this -- it already has a fork()C > > that works correctly; it's just been slow because of the design 9 > > differences between process creation on VMS and Unix.  > J > This is excellent news; I was unaware of this. Where is it implemented ? > O > Looking at the C RTL reference manual for C V6.5, I cannot find any reference  > to fork, just vfork.  B My apologies.  After looking into this more carefully, it appears:D (1) The old VMS Integrated POSIX (VIP) product had a fork() functionD that could fork a (constrained) VMS process (but that product's been unsupported for some years now) A (2) The initial DII-COE releases (7.2-6Cn) don't have fork() (see 5 http://openvms.compaq.com/solutions/government/coe/).    (1) was what I was thinking of.   2 So this fork() is a future feature on the Roadmap.   ------------------------------  # Date: Mon, 03 Mar 2003 19:33:12 GMT # From: "John Smith" <a@nonymous.com> 0 Subject: Re: HP Reports 1st Quarter 2003 ResultsG Message-ID: <Y5O8a.38994$em1.6051@news04.bloor.is.net.cable.rogers.com>   > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message7 news:cf15391e.0303031053.615b2b0a@posting.google.com... > > keithparris_NOSPAM@yahoo.com (Keith Parris) wrote in message9 news:<cf15391e.0302251421.16842f38@posting.google.com>... ' > > HP Reports 1st Quarter 2003 Results  > 5 > From HP World's e-mail newsletter, Feb. 27 edition:  > --- % > Fiorina defends HP's Q1 performance  > ; > HP CEO Carly Fiorina defended her company's first-quarter  performance before a> > group of investors on Wednesday, calling the period the most profitable sinceE > its merger with Compaq Computer.  Fiorina took direct aim at claims  HPD > reclassified expenses to make its operating unit, including the PC	 division, B > look profitable.  "Nothing could be further from the truth," she said.  But sheD > blamed the large revenue shortfall in the quarter on the company's U.S.F > business.  Fiorina said the quarter's lesson may be that an increase inE > information technology spending is not yet taking place.  She added  that the4 > recovery may be slower than many investors expect./ > http://click.topica.com/maaaSZ6aaWktLa4e8YPe/     > Wait until the Iraq war kicks in and see how much further tech= spending slows down in the US as the costs to the economy and  uncertainty grow.    ------------------------------  $ Date: Mon, 3 Mar 2003 19:26:34 -0500* From: "Bill Todd" <billtodd@metrocast.net>0 Subject: Re: HP Reports 1st Quarter 2003 Results2 Message-ID: <nJacnR0Op60jb_6jXTWc3w@metrocast.net>  > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message7 news:cf15391e.0303031053.615b2b0a@posting.google.com... > > keithparris_NOSPAM@yahoo.com (Keith Parris) wrote in message9 news:<cf15391e.0302251421.16842f38@posting.google.com>... ' > > HP Reports 1st Quarter 2003 Results  > 5 > From HP World's e-mail newsletter, Feb. 27 edition:  > --- % > Fiorina defends HP's Q1 performance  > G > HP CEO Carly Fiorina defended her company's first-quarter performance  before aI > group of investors on Wednesday, calling the period the most profitable  since H > its merger with Compaq Computer.  Fiorina took direct aim at claims HPD > reclassified expenses to make its operating unit, including the PC	 division, H > look profitable.  "Nothing could be further from the truth," she said.  G Wow - given the distance from the truth of some of the whoppers she was I tossing around during the merger battle, that's got to qualify as another  one.   - bill   ------------------------------  % Date: Mon, 03 Mar 2003 20:45:31 +0100  From: Dirk Munk <munk@home.nl>. Subject: I'm going to sleep so well tonight !!2 Message-ID: <b40bcj$9e4$1@news4.tilbu1.nb.home.nl>  M Today I received a beautiful royal blue colored HP fleece blanket, specially   sent to me by Rich Marcello !!  Q Ain't that nice of him ? He tells us that we can rest assured under his blanket,  L   and that we can retain trust in the Alpha. Well, he can rest assured too, 6 because I'm not likely to loose my trust in the Alpha.  M The only problem with the blanket is that it is a bit to short. That means I  < will have cold feet tonight. Now is that symbolic too ?? -:)   ------------------------------  % Date: Mon, 03 Mar 2003 14:18:58 -0600  From: brandon@dalsemi.com 2 Subject: Re: I'm going to sleep so well tonight !!1 Message-ID: <03030314185853@dscis6-0.dalsemi.com>   Q >>> The only problem with the blanket is that it is a bit to short. That means I  @ >>> will have cold feet tonight. Now is that symbolic too ?? -:)  . Hmmmmm.... or do you wear it over your head?         John Brandon VMS Systems Administrator  Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fx    ------------------------------  % Date: Mon, 03 Mar 2003 21:17:43 +0100 9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> 2 Subject: Re: I'm going to sleep so well tonight !!' Message-ID: <3E63B867.4E8E26D2@aaa.com>   / Air intake through the floor in the data room ? ( Maybe your cabs also have "cold feets" ?  	 Jan-Erik.    Dirk Munk wrote: > N > The only problem with the blanket is that it is a bit to short. That means I> > will have cold feet tonight. Now is that symbolic too ?? -:)   ------------------------------  % Date: Tue, 04 Mar 2003 00:02:33 +0100  From: Dirk Munk <munk@home.nl>2 Subject: Re: I'm going to sleep so well tonight !!2 Message-ID: <b40mu1$kv4$1@news4.tilbu1.nb.home.nl>   brandon@dalsemi.com wrote:Q >>>>The only problem with the blanket is that it is a bit to short. That means I  @ >>>>will have cold feet tonight. Now is that symbolic too ?? -:) >  > 0 > Hmmmmm.... or do you wear it over your head?    N No, the blanket is 5 feet long, and I am 6'1". I suppose the blanket was made B with small Asian people in mind. But it's the gesture that counts. >  >  >  > John Brandon > VMS Systems Administrator  > Dallas Semiconductor > john.brandon@dalsemi.com > 972.371.4172 wk  > 972.371.4003 fx    ------------------------------  % Date: Mon, 03 Mar 2003 17:19:46 -0600  From: brandon@dalsemi.com 2 Subject: Re: I'm going to sleep so well tonight !!1 Message-ID: <03030317194675@dscis6-0.dalsemi.com>    > P > No, the blanket is 5 feet long, and I am 6'1". I suppose the blanket was made D > with small Asian people in mind. But it's the gesture that counts. >   _ Nah, I was thinking they just want to have "you" pull the wool over your own eyes...  *snicker*    John Brandon VMS Systems Administrator  Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fx    ------------------------------  % Date: Mon, 03 Mar 2003 20:15:42 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> 2 Subject: Re: I'm going to sleep so well tonight !!' Message-ID: <3E640C4E.8EE43B53@fsi.net>    brandon@dalsemi.com wrote: >  > > Q > > No, the blanket is 5 feet long, and I am 6'1". I suppose the blanket was made F > > with small Asian people in mind. But it's the gesture that counts. > >  > a > Nah, I was thinking they just want to have "you" pull the wool over your own eyes...  *snicker*   * Clev-ER! I bow to the superior wit! Bravo!   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Mon, 03 Mar 2003 16:44:50 -0500 9 From: Hein van den Heuvel <hein_netscape@eps.zko.dec.com> C Subject: Re: Intel Itanium 67% slower than Intel Alpha in SAP tests / Message-ID: <3E63CCD2.9A6E9124@eps.zko.dec.com>    Tsarkon wrote:  8 > Intel Itanium 67% slower than Intel Alpha in SAP tests > H > A REPORT in authoritative German magazine c't said that SAP benchmarks= > demonstrate that an Itanium system is 67% slower than an HP  > AlphaServer GS1280.   8 Cute, for Alpha fans like myself, but highly misleading.  E (fwiw, I actually ran and published that 4500 user benchmark on a 32P   Marvel here in Nashua ZK2 labs).  I The NEC 32P Itanium SYSTEM, running WINDOWS and SQLserver, was 67% slower H than the Alpha SYSTEM running Tru64 Unix and Oracle (9.2). That does not? say much about the the CPU let alone about its potential future  performance.  H That same Itanium CPU in a 4P config rx5670 with the zx1 chipset runningJ hp-ux run 600 sd users. We can not touch that with Alpha today (could comeH close though :^). Albeit a few months earlier, that box running Windows,, it came in 470 users (which Alpha can beat).  E The way I see it, the hp-ux number shows that Itanium is on track for  superior performance. I This hurts me some as an old, diehard, Alpha fan, but it pleases me a lot H as a current hp employee and VMS user! It suggests (but does not prove),C that the Alpha->IA64 transition decision may have been painful, but  correct.  H What the Itanium and it's follow ups needs is a good large SMP SYSTEM toG be put in and flourish and a good OS and DB to get everything out of it G that is potentially in there. Obviously HP is working on such solutions + but I can not say when/where/how-many-CPUs.   > The NEC benchmark is just a stake in the ground. A first step.   Just an opinion! Hein.    ------------------------------  $ Date: Mon, 3 Mar 2003 19:40:25 -0500* From: "Bill Todd" <billtodd@metrocast.net>C Subject: Re: Intel Itanium 67% slower than Intel Alpha in SAP tests 2 Message-ID: <qZ6dnZa-_9VhaP6jXTWcqg@metrocast.net>  F "Hein van den Heuvel" <hein_netscape@eps.zko.dec.com> wrote in message) news:3E63CCD2.9A6E9124@eps.zko.dec.com...  >  >  > Tsarkon wrote: > : > > Intel Itanium 67% slower than Intel Alpha in SAP tests > > J > > A REPORT in authoritative German magazine c't said that SAP benchmarks? > > demonstrate that an Itanium system is 67% slower than an HP  > > AlphaServer GS1280.  > : > Cute, for Alpha fans like myself, but highly misleading.  J In what way?  The quote above specifically compares *systems*, just as you note below.    > G > (fwiw, I actually ran and published that 4500 user benchmark on a 32P " > Marvel here in Nashua ZK2 labs). > K > The NEC 32P Itanium SYSTEM, running WINDOWS and SQLserver, was 67% slower J > than the Alpha SYSTEM running Tru64 Unix and Oracle (9.2). That does notA > say much about the the CPU let alone about its potential future  > performance.  K That's true.  But it says a great deal about scalability, which, of course, H is just as important in the markets Alpha and Itanic are aimed at as raw single-CPU performance.    > J > That same Itanium CPU in a 4P config rx5670 with the zx1 chipset runningL > hp-ux run 600 sd users. We can not touch that with Alpha today (could come > close though :^).   F Indeed:  in small systems Itanic (at least with full feedback-directedB compilation) is quite competitive (as long as you don't take powerG consumption into account, anyway).  But SGI is the only vendor that has L proved able to extend that performance to larger systems, while EV7 *excels*: at doing so (I suspect even better than the SGI platform).  7  Albeit a few months earlier, that box running Windows, . > it came in 470 users (which Alpha can beat). > G > The way I see it, the hp-ux number shows that Itanium is on track for  > superior performance.   L Superior?  Don't forget that a POWER system also beat Itanic2 rather handilyH in those tests (and does so in others as well, though IBM doesn't botherG doing benchmarks for really small - max 4-processor - systems, which of $ course are all HP yet has to offer).  K > This hurts me some as an old, diehard, Alpha fan, but it pleases me a lot J > as a current hp employee and VMS user! It suggests (but does not prove),E > that the Alpha->IA64 transition decision may have been painful, but 
 > correct.  L No, it only suggests that it may not be a total disaster:  there's no reasonF whatsoever to believe that Itanic would ever have become an attractiveE platform for Alpha customers (save possibly at the low end) had Alpha H development continued - even if you ignore the pain of the migration and focus only on utility.   > J > What the Itanium and it's follow ups needs is a good large SMP SYSTEM toI > be put in and flourish and a good OS and DB to get everything out of it I > that is potentially in there. Obviously HP is working on such solutions - > but I can not say when/where/how-many-CPUs.   K While HP may be working on them, its track record does not encourage one to K believe that they'll scale anywhere nearly as well as SGI's systems do (or, I of course, as EV7 does).  And POWER platforms also exhibit better scaling H characteristics than HP's (and Compaq's) large servers have in the past.   - bill   ------------------------------  $ Date: Mon, 3 Mar 2003 19:22:31 -0500* From: "Bill Todd" <billtodd@metrocast.net>8 Subject: Re: Linus trashes Itanium - he's scared of VMS!2 Message-ID: <NDydnSzufaVXbP6jXTWc3Q@metrocast.net>  . "Dan Allen" <dallen@nist.gov> wrote in message4 news:JFEPKAPBPMDFDBOIANGDIEPICNAA.dallen@nist.gov... >  >  > > -----Original Message-----+ > > From: Andrew Harrison SUNUK Consultancy 2 > > [mailto:Andrew_No.Harrison_No@nospamn.sun.com]) > > Sent: Monday, March 03, 2003 10:34 AM  > > To: Info-VAX@Mvb.Saic.Com < > > Subject: Re: Linus trashes Itanium - he's scared of VMS!   ...   < > > Of course part of the cost differential is because Xeons: > > cost Intel a lot less to build than IPF, with a 25% of# > > the die size in a given process   H Your credibility is thin enough in this forum already without the aid ofL such blatantly erroneous statements.  Itanic2's chip area is 421 mm^2 in itsH 180 nm process, while P4's is 140 mm^2 in 130 nm and with only 512 KB ofK cache vs. Itanic2's 3 MB.  That would put P4 over 200 mm^2 in Itanic2's 180 I nm process, and close to 300 mm^2 in that process in a Xeon configuration J with 2 MB cache - in other words, between 50% and 70% of Itanic2's area in the same process.     and earlier access to faster B > > smaller processes this differential is likely to be sustained. > >  > @ > Did I miss the announcement of a 64 bit Xeon or is this simply > comparing apples to zebras?   G No, what you missed is the fact that Xeon very definitely competes with K Itanic in the low-end-to-mid-range (i.e., volume) server market, which more D often than not couldn't care less about 64-bitness.  One of the realB problems Itanic is having getting off the ground in volume is moreI cost-effective (and far better-established) competition from its sibling.    - bill   ------------------------------  $ Date: Mon, 3 Mar 2003 19:28:22 -0000- From: "Steve Spires" <Steve.Spires@torex.com> % Subject: MONITOR SYSTEM screen values E Message-ID: <91947A84607D9D48B8E674A5FAB54DA63CB21E@tahiti.tinuk.com>   F Can anyone tell me how the value scales shown on the output screen forC MONITOR are calculated? More specifically, the value for Direct IO?   G On the node I am looking at the value scale runs from 0 to 60, the user E is seeing values of 200+ and has decided that this is a problem [they F are not technical people as such] so I just need to explain that theseD are just guideline values not 'maximum' values - if that's the case.  C If I could tell them how these values are derived, I think it would H help. It seems as though I should know this, but perhaps I've just never thought about it before!   Thanks   Steve Spires Technical Consultant Torex Health [T] +44 (0) 1295 274200  [F] +44 (0) 1295 275131 
 www.torex.com    ------------------------------  % Date: Mon, 03 Mar 2003 20:59:18 +0100  From: Dirk Munk <munk@home.nl>) Subject: Re: MONITOR SYSTEM screen values 2 Message-ID: <b40c6e$prd$1@news4.tilbu1.nb.home.nl>   Steve Spires wrote: H > Can anyone tell me how the value scales shown on the output screen forE > MONITOR are calculated? More specifically, the value for Direct IO?  > I > On the node I am looking at the value scale runs from 0 to 60, the user G > is seeing values of 200+ and has decided that this is a problem [they H > are not technical people as such] so I just need to explain that theseF > are just guideline values not 'maximum' values - if that's the case. > E > If I could tell them how these values are derived, I think it would J > help. It seems as though I should know this, but perhaps I've just never > thought about it before!  P The Direct IO shows the number of IO's that is going to the disks, or better to L the VMS disk cache. The more you see, the harder your system is working and Q earning money. A value of 200 is not high, I have seen rates of more then 20,000  O on our ES40. A fast SCSI disk can do about 150 random (!!) IO's per second. So  Q it all depends on the kind of system your customer has, the application, and how  O his storage is set up. It can point to a problem if the system and the storage  G are old and slow of course, but without more information that can't be   determined now.    ------------------------------  % Date: Mon, 03 Mar 2003 14:23:31 -0600  From: brandon@dalsemi.com ) Subject: Re: MONITOR SYSTEM screen values 1 Message-ID: <03030314233103@dscis6-0.dalsemi.com>   T >>> The Direct IO shows the number of IO's that is going to the disks, or better to P >>> the VMS disk cache. The more you see, the harder your system is working and U >>> earning money. A value of 200 is not high, I have seen rates of more then 20,000  S >>> on our ES40. A fast SCSI disk can do about 150 random (!!) IO's per second. So  U >>> it all depends on the kind of system your customer has, the application, and how  S >>> his storage is set up. It can point to a problem if the system and the storage  K >>> are old and slow of course, but without more information that can't be   >>> determined now.   0 Are we sure that Direct I/O is disk direct I/O?   w I asked similar questions a year or two ago... I was told that I/O was difficult to quantify; RMS buffering, VMS cache,  controllers, etc...   { I was lead to believe that Dir I/O and Buf I/O were dependent on the device drivers used; therefore, you could have Dir I/O  reported as Buf I/O, etc.     h Is this true of the Disk monitor utility, or only system I/O, or am I total confused? (I can't remember)   John Brandon VMS Systems Administrator  Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fx    ------------------------------  % Date: Mon, 03 Mar 2003 16:08:07 -0500 % From: "Mark D. Jilson" <jilly@hp.com> ) Subject: Re: MONITOR SYSTEM screen values & Message-ID: <3E63C437.E3FD0E86@hp.com>  E The scale for Direct IO is hardcoded and is not adjusted based on the E 'capacity' of the system so don't compare the number reported against 
 the scale.  F Direct IOs are usually mostly disk IO but not completely.  Some TCP/IPH IO is direct.  It is up to the device driver to classify an IO as direct@ or buffered.  For looking at disk IO you should be using MONITOR+ DISK/ITEM=ALL.  Here is what I usually do    MONITOR B DISK/ITEM=ALL/NODISPLAY/SUMMARY=DISK.RPT/INTERVAL=1/END="+0:2:0.0"1 Then look at the AVG column in the DISK.RPT file.   G For a good write up on performanace troubleshooting see the Ask OpenVMS H article titled Cookbook of Performance Slowdown, VAX and Alpha, V6.0 and Up[ http://h18000.www1.hp.com/support/asktima/operating_systems/009BD243-C877FC20-1C0186.html      Dirk Munk wrote: >  > Steve Spires wrote: J > > Can anyone tell me how the value scales shown on the output screen forG > > MONITOR are calculated? More specifically, the value for Direct IO?  > > K > > On the node I am looking at the value scale runs from 0 to 60, the user I > > is seeing values of 200+ and has decided that this is a problem [they J > > are not technical people as such] so I just need to explain that theseH > > are just guideline values not 'maximum' values - if that's the case. > > G > > If I could tell them how these values are derived, I think it would L > > help. It seems as though I should know this, but perhaps I've just never > > thought about it before! > Q > The Direct IO shows the number of IO's that is going to the disks, or better to M > the VMS disk cache. The more you see, the harder your system is working and R > earning money. A value of 200 is not high, I have seen rates of more then 20,000P > on our ES40. A fast SCSI disk can do about 150 random (!!) IO's per second. SoR > it all depends on the kind of system your customer has, the application, and howP > his storage is set up. It can point to a problem if the system and the storageH > are old and slow of course, but without more information that can't be > determined now.    --  C Jilly	- Working from Home in the Chemung River Valley - Waverly, NY 0 	- jilly@clarityconnect.com			- Brett Bodine fan+ 	- Mark.Jilson@hp.com				- since 1975 or so  	- http://www.jilly.baka.com   ------------------------------   Date: 3 Mar 2003 15:54:57 -0800 1 From: keithparris_NOSPAM@yahoo.com (Keith Parris) ) Subject: Re: MONITOR SYSTEM screen values = Message-ID: <cf15391e.0303031554.584a5491@posting.google.com>   z "Steve Spires" <Steve.Spires@torex.com> wrote in message news:<91947A84607D9D48B8E674A5FAB54DA63CB21E@tahiti.tinuk.com>...H > Can anyone tell me how the value scales shown on the output screen forE > MONITOR are calculated? More specifically, the value for Direct IO?  > I > On the node I am looking at the value scale runs from 0 to 60, the user G > is seeing values of 200+ and has decided that this is a problem [they H > are not technical people as such] so I just need to explain that theseF > are just guideline values not 'maximum' values - if that's the case.  D Disk queue lengths ($MONITOR DISK/ITEM=QUEUE_LENGTH) would be a muchF better indicator of a disk performance issue than I/O operation rates.   ------------------------------   Date: 4 Mar 2003 00:27 CDT' From: carl@gerg.tamu.edu (Carl Perkins) ) Subject: Re: MONITOR SYSTEM screen values , Message-ID: <4MAR200300275559@gerg.tamu.edu>  1 "Steve Spires" <Steve.Spires@torex.com> writes... G }Can anyone tell me how the value scales shown on the output screen for D }MONITOR are calculated? More specifically, the value for Direct IO? } H }On the node I am looking at the value scale runs from 0 to 60, the userF }is seeing values of 200+ and has decided that this is a problem [theyG }are not technical people as such] so I just need to explain that these E }are just guideline values not 'maximum' values - if that's the case.  } D }If I could tell them how these values are derived, I think it wouldI }help. It seems as though I should know this, but perhaps I've just never  }thought about it before!  } 
 }Steve Spires   I If you are seeing a direct I/O scale that goes up to 60, you are probably J looking at the MONITOR SYSTEM display (MONITOR PROCESS/TOPDIO only goes upI to 20 and MONITOR IO/CURRENT goes up to 80). The reason you can determine H which monitor command is being talked about is that these values are the same on every system.   D Most of the scales in MONITOR are not calculated. They are fixed. InD fact they have been fixed to the same values pretty much forever. SoB the maximum values you see on your system are, I believe, the sameD maximums that you would have seen on, for example, a VAXstation 2000D many years ago (a VAXstation 2000 is a pitifully slow system, it wasD even slow when it came out). Off hand, they only calculated scales IC can think of are those relating to memory - for these the upper end  can reflect the actual system.  E They are not even guideline values. They are just values that someone H thought were suitable for the display at some point - back in the 1980s,	 I expect.    --- Carl   ------------------------------  % Date: Mon, 03 Mar 2003 16:15:21 -0500 9 From: Hein van den Heuvel <hein_netscape@eps.zko.dec.com> % Subject: Re: oracle benchmarks on VMS / Message-ID: <3E63C5E9.88569D3F@eps.zko.dec.com>    Tim Smith wrote:  + > I see Oracle publishes some benchmarks at E > http://www.oracle.com/apps_benchmark/, but notably there is nothing G > for VMS - is that because Oracle writes to the filesystem, not direct H > to device file files directly i.e. VMS filesystem is a lot slower than > raw devices?  , It has nothing to do with (potential) speed.@ It is just a commercial / marketing decision mostly from Oracle.G They decided there was not enough critical mass to maintain support for G the Oracle Applicaiton suite on VMS. The database itself is and will be I supported at one of the higher tier levels. Details are not up to me, but H it used to be product release on VMS 90 days after first release. PleaseG verify with Oracle. They may also choose to skip 'dot' releases. Dunno, J let's say they had 9.0, skipped 9.1 but released 9.2. Again, this is not aA statement of support, just a line of thinking. Check with Oracle.   H The VMS Filesystem is actually an advantage! The VMS filesystem does NOT buffer data.F On Unix systems the OS tends to waste time and memory buffering OracleH data pages which are better managerd by Oracle in its buffer pool (SGA).H On many Unix implementation, for ultimate Oracle perfromance once has to? deal with hard-to-manage 'Raw Devices' to avoid said buffering. G On VMS you have the comfort of a file system for Alloaction, Naming and G backups yet the speed of a raw device. On HP Tru64 Unix Oracle can (and H will) use the DIRECT IO feature to get the same effect on single systems as well as in clusters.   E > If anyone has older benchmarks that include VMS I would like to see  > them.   H It would be nice to see some VMS / Oracle benchmark, but I will not hold
 my breath.J Benchmarks require major investment which both companies believe is betterB spend on the products itself. VMS will offer comparable (ballpark)) performance as Unix on the same platform. J It will not be 2x slower. It might be a little slower or a a little faster depending on the application. G The performance will be close enough to focus on other, more important, H platform decision factors: Cost-of-ownership, Reliability, Availability,+ Experience, Applications, Installed base,..   A Hope this helps some, but it is just an opinion from a guy in the 
 sidelines.I Be sure to contact officials at Oracle and HP for the official positions.    Cheers, 	     Hein.    ------------------------------  $ Date: Mon, 3 Mar 2003 18:47:03 -0500* From: "Bill Todd" <billtodd@metrocast.net>% Subject: Re: oracle benchmarks on VMS 2 Message-ID: <N4OdnbxOMr7gdP6jXTWcpg@metrocast.net>  F "Hein van den Heuvel" <hein_netscape@eps.zko.dec.com> wrote in message) news:3E63C5E9.88569D3F@eps.zko.dec.com...    ...   H > On Unix systems the OS tends to waste time and memory buffering OracleJ > data pages which are better managerd by Oracle in its buffer pool (SGA).J > On many Unix implementation, for ultimate Oracle perfromance once has toA > deal with hard-to-manage 'Raw Devices' to avoid said buffering.   I Not on the most common Unixes.  Veritas' VxFS file system supports direct K (unbuffered) file I/O, and it's available pretty widely (it may even be the J standard file system on HP-UX - they had their fingers in some part of theI HP-UX data management pie, anyway).  VxFS runs on Solaris, and IIRC Sun's G own file system also supports direct I/O (not too surprising, given how H important Oracle is to them).  Linux has some form of direct I/O in 2.4,I though it may be scheduled for more massaging to eliminate some kludgery. F VxFS was recently ported to AIX (I don't know whether their native JFS supports direct I/O, though).   I (Many Unixes support asynchronous I/O too, but AFAIK none as well as VMS: F they - and the POSIX asynchrony model - never seemed to understand how# completion processing should work.)    - bill   ------------------------------  # Date: Tue, 04 Mar 2003 02:21:47 GMT # From: "John Smith" <a@nonymous.com> % Subject: Re: oracle benchmarks on VMS H Message-ID: <%4U8a.41657$em1.14204@news04.bloor.is.net.cable.rogers.com>  F "Hein van den Heuvel" <hein_netscape@eps.zko.dec.com> wrote in message) news:3E63C5E9.88569D3F@eps.zko.dec.com...  >  >  > Tim Smith wrote: > - > > I see Oracle publishes some benchmarks at ? > > http://www.oracle.com/apps_benchmark/, but notably there is  nothing B > > for VMS - is that because Oracle writes to the filesystem, not directE > > to device file files directly i.e. VMS filesystem is a lot slower  than > > raw devices? > . > It has nothing to do with (potential) speed.B > It is just a commercial / marketing decision mostly from Oracle.E > They decided there was not enough critical mass to maintain support  for & > the Oracle Applicaiton suite on VMS.  $ Sorry Hein, nothing against you.....D Once again HP self-limits the market for VMS by not advertising, and; by doing so, makes it easy for sales to go to Sun/IBM/Dell.     E > It would be nice to see some VMS / Oracle benchmark, but I will not  hold > my breath.E > Benchmarks require major investment which both companies believe is  betterD > spend on the products itself. VMS will offer comparable (ballpark)+ > performance as Unix on the same platform.   F They'll never benchmark (ie. spend the money) on what they consider to be a 'fringe' platform.   A Just thinking about HP's lack of advertising and marketing of VMS ? makes me think of the event horizon of a black hole - VMS keeps B spinning around just microns above the event horizon, with a minorF perturbation in the gravitational flux ready to send it slipping below the event horizon for good.    ------------------------------  # Date: Mon, 03 Mar 2003 23:30:53 GMT ! From: rob.buxton@wcc.spam.govt.nz 5 Subject: Re: OT: America's Cup  [ was Re: I am back ] & Message-ID: <3e63e56b.1225783142@news>  F On 3 Mar 03 16:16:27 +0100, p_sture@elias.decus.ch (Paul Sture) wrote:  i >In article <bRl8a.966$L7.85759@nasal.pacific.net.au>, HARANGOZO CSABA   <spameater@spam.invalid> writes: - >> Paul Sture <p_sture@elias.decus.ch> wrote:  >> [...snip...]  >>  > >> 	Congratulations to Switzerland winning the America's Cup ! >>  A >> 	( I just wonder where they will defend it, being a landlocked  >> 	country... ) >> 						Cheers,  Csaba  > E >Thanks for the thought :-) I didn't realize that it has been so long ! >since a European country won it:  > E >"For the first time since 1851, the America's Cup is thus now on its E >way back to Europe, with Switzerland as the first landlocked country F >to succeed in the battle for the world's oldest and most sought-after >sailing trophy."  >   E Hmmm.... I wonder if Alinghi will keep the kiwi crew that won it from 
 the kiwis? >  >  >--  >Paul Sture    ------------------------------  % Date: Tue, 04 Mar 2003 15:29:16 +1030 % From: Jeremy Begg <jeremy@vsm.com.au> 5 Subject: Re: OT: America's Cup  [ was Re: I am back ] * Message-ID: <3E6432A4.416F87FC@vsm.com.au>   Hello Paul,   D > >       Congratulations to Switzerland winning the America's Cup ! > > G > >       ( I just wonder where they will defend it, being a landlocked  > >       country... )@ > >                                               Cheers,  Csaba > F > Thanks for the thought :-) I didn't realize that it has been so long" > since a European country won it: > F > "For the first time since 1851, the America's Cup is thus now on itsF > way back to Europe, with Switzerland as the first landlocked countryG > to succeed in the battle for the world's oldest and most sought-after  > sailing trophy."  N The quote is slightly misleading.  In fact, this is the first time the Cup hasN been won by a European country, ever.  The competition started in 1851 and wasN named after the first winner, a boat called "America".  Various teams from theL USA successfully defended the trophy for the next 132 years, until 1983 whenL the Australian boat "Australia II" won it, with its (in)famous winged keel. I NZ won it a few years later and retained the Cup until this year's event.   N I was in NZ a couple of weeks ago when the finals of this year's America's CupN were starting.  The locals were *very* upset when their boat fell apart in theK first race!  After the NZ team lost the next two races I don't think anyone B was surprised by the final 5-0 result to "Alinghi" of Switzerland.  N Last I read, the the Swiss team was planning on holding the next America's CupK series in Portugal.  By then they may have recruited the few NZ sailors who  weren't already on the team :-)    Regards,           Jeremy Begg   =   +---------------------------------------------------------+ =   |            VSM Software Services Pty. Ltd.              | =   |                 http://www.vsm.com.au/                  | =   |       "OpenVMS Systems Management & Programming"        | =   |---------------------------------------------------------| =   | P.O.Box 402, Walkerville, |  E-Mail:  jeremy@vsm.com.au | =   | South Australia 5081      |   Phone:  +61 8 8221 5188   | =   |---------------------------|  Mobile:  0414 422 947      | =   |  A.C.N. 068 409 156       |     FAX:  +61 8 8221 7199   | =   +---------------------------------------------------------+    ------------------------------  # Date: Tue, 04 Mar 2003 05:30:06 GMT - From: bdc@world.std.com (Brian 'Jarai' Chase) > Subject: Re: PC developers opinion of 64 bit desktop computing) Message-ID: <71pdj-sso.ln1@pez.jarai.com>   2 In article <cuicnVBSucFr18CjXTWcqg@metrocast.net>,) Bill Todd <billtodd@metrocast.net> wrote:  > : > "Rob Young" <young_r@encompasserve.org> wrote in message/ > news:hpAEzN6RMjUJ@eisner.encompasserve.org... @ > > In article <01C2DD96.8AFC0400@sulfer.icius.com>, Shane Smith > <ssmith@icius.com> writes:; > > > http://slashdot.org/comments.pl?sid=54835&cid=5371889   M > > > Tim Sweeny of Epic (big game developer) about Intel saying 64bit on the + > > > desktop is "end of the decade" stuff:  > > > L > > > "Intel's claims are wholly out of touch with reality. On a daily basisI > > > we're running into the Windows 2GB barrier with our next-generation E > > > content development and preprocessing tools. If cost-effective, K > > > backwards-compatible 64-bit CPU's were available today, we'd buy them H > > > today. We need them today. It looks like we'll get them in April." > > , > > But nothing for the masses to run it on. > N > Exactly what part of 'content development and preprocessing tools' above didN > you fail to understand?  In the above statement Sweeny is talking about whatK > would be useful to developers, not end-users - and since beta versions of L > 64-bit Windows are already running on Hammer, they probably won't have tooA > much difficulty in getting them if they don't develop on Linux.   I Although we're not what you'd call a significant end user market segment, D the feature film vfx and animation industry is certainly hitting theH limits of 32-bit addressing on a daily basis.  Only a week ago, I got toD see a rather large pile of renders fail because they needed a largerI process space than was available under x86 Linux.  It's also not uncommon H for us to have problems on our 32-bit desktops (be they Windows, Linux, 
 or Mac OS X).   J The gaming industry relies on a lot of the same hardware and software thatF we do for their content development, so I'm not surprised that they're having the same problems.   J However, I don't disagree that your average user will need or want 64-bit J addressing.  Well, not until some of the results of those game developers K create something that looks /really/ cool and impressive and that requires  # a larger than 32-bit address space.    -brian.    ------------------------------  % Date: Mon, 03 Mar 2003 17:06:23 -0500 9 From: Hein van den Heuvel <hein_netscape@eps.zko.dec.com> N Subject: Re: Question: why sometimes asctim returns "17-NOV-1858 00:00:00.00"?/ Message-ID: <3E63D1DF.DF41D01C@eps.zko.dec.com>   
 Albert wrote:    > Dear All,  > M > captures two 8-byte VMS_DATE value. They are passed to asctim. But for some J > columns, "17-NOV-1858 00:00:00.00" is returned wrongly, Could you kindly  ; >         /* ------------ Get End Date/Time ------------ */ & >         strncpy(pTmp, BufAdr+90, 8); >         pTmp[8]='\0';    Well Duh....  - You are treating the BINARY time as C String. L strncpy will STOP when it sees a first 0 byte in that 'string' / bitpattern.  8 You chould use memcpy to copy 'opaque' binary structure.  7 Actually... why copy at all? Just point to the source!?   F And while you are at it... just grab a statics string descriptor once.  6 >  printf("Start Time: %s.", AsciiDate.dsc$a_pointer);  M Hmmm... what is garantueeing a terminating 0 on that string? Nothing. Playing 
 with fire!  N Please check out SYS$FAO and friends (LIB$SYS_FAO) as an alternative to printf when playing with all-vms data.     K >         if ( !$VMS_STATUS_SUCCESS(sys$asctim ( &timlen, &AsciiDate, (char  > *)pTmp, 0 )))    lies... lies..   :-).    .Good luck!    ------------------------------   Date: 3 Mar 2003 15:05:16 -0800 1 From: susan_skonetski@hotmail.com (Sue Skonetski) @ Subject: Re: RDB "Tech Day" comes to New York on  March 14, 2003= Message-ID: <857e9e41.0303031505.71c4a665@posting.google.com>   @ From what I understand there are going to be a number of the RDB5 forums, I will try and get the list and post it here.    sue     p "John Smith" <a@nonymous.com> wrote in message news:<OfM8a.37924$em1.32@news04.bloor.is.net.cable.rogers.com>...@ > "Sue Skonetski" <susan_skonetski@hotmail.com> wrote in message9 > news:857e9e41.0303030915.13390934@posting.google.com...  > > Dear Newsgroup, > > > I know the speakers for this event and they are excellent. > > Warm Regards,  > > Sue  > >  > > D > > Subject:      FW: RDB "Tech Day" comes to New York on  March 14, >  2003  > > @ > > Back by Popular Demand!!  Rdb Engineering comes to New York! > > F > > Please forward this invitation to your colleagues with an interest >  in  > > Oracle Rdb.  >  > 
 > Great news.  > " > One suggestion to HP and Oracle:H > Try to announce these sorts of events more than 3 weeks in advance. ItD > makes travel planning and the requisite approvals process for many# > people easier and less expensive.    ------------------------------  % Date: Mon, 03 Mar 2003 20:13:28 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> @ Subject: Re: RDB "Tech Day" comes to New York on  March 14, 2003' Message-ID: <3E640BC8.43075D5B@fsi.net>    Sue Skonetski wrote: > B > From what I understand there are going to be a number of the RDB7 > forums, I will try and get the list and post it here.    Excellent!    D I'll be watching to see if there's a Chicago date. I know, we're theH bastard step-children out here in the heartland, but I keep hoping to be remembered, anyway...    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------   Date: 3 Mar 2003 15:05:45 -0800 1 From: mike_hutcheson@hotmail.com (Mike Hutcheson) 5 Subject: Re: running autogen on alternate system disk = Message-ID: <c7fe9387.0303031505.76ac1e64@posting.google.com>   > My apologies for lack of detail.  The system in question is anC AlphaServer 1200.  No cluster.  I'm just tinkering around a bit.  I F have VMS 7.3 installed on two differend disks: DRA0 is the primary andD DRA5 is the secondary.  DRA5 is just a base install from the CD with< no layered products.  DRA0 is the disk I normally boot from.  B I was wondering if it's possible to boot from DRA5 and run autogenA against DRA0.  Let's assume that we want to use feedback that was  saved on DRA0.  = This is really just an exercise in curiosity.  Are there even B situations in which you would need to run autogen against a system@ disk you didn't boot from?  That's probably the first question I should be asking.   4 And then if there is, how would I go about doing it?   Cheers,    Mike  l brad@.homeportal.2wire.net (Bradford J. Hamilton) wrote in message news:<7M98a.278691$2H6.4795@sccrnsc04>...s > In article <c7fe9387.0303011309.313ccc29@posting.google.com>, mike_hutcheson@hotmail.com (Mike Hutcheson) writes: I > >Hi.  I have a VMS 7.3 system with two system disks.  Is it possible to B > >run autogen against the system disk that I haven't booted from? > 
 > Hi Mike, > < > We might need some more information, in order to help you. > K > Do you mean that you have a single (non-clustered) system with two system P > disks?  Is "the system disk that I haven't booted from" a backup disk?  A disk= > meant to be able to boot the system if the "primary" fails?  >  > >  > >Thanks for your help, > >  > >Mike Hutcheson  > >Systems Manager > >Baylor University > 	 > Thanks,  > C > _________________________________________________________________ 2 > Bradford J. Hamilton			"All opinions are my own"1 > bMradAhamiPltSon@atMtAbi.cPoSm		"Lose the MAPS"    ------------------------------  % Date: Mon, 03 Mar 2003 17:16:51 -0600  From: brandon@dalsemi.com 5 Subject: Re: running autogen on alternate system disk 1 Message-ID: <03030317165187@dscis6-0.dalsemi.com>   D > I was wondering if it's possible to boot from DRA5 and run autogenC > against DRA0.  Let's assume that we want to use feedback that was  > saved on DRA0. > ? > This is really just an exercise in curiosity.  Are there even D > situations in which you would need to run autogen against a systemB > disk you didn't boot from?  That's probably the first question I > should be asking.  > 6 > And then if there is, how would I go about doing it? >   $ I do not know why you would do that.  c If you are interested in getting the SYSGEN parameter setttings to the disk, then do the following:    1) sys$system:modparams.dat; 2) sys$system:alphavmssys.par;  1 If you are building an alternate boot disk, then:   : 3) other system related files like (sysuaf, mail, net,...)  . And if that is the case, why not IMAGE backup?     John Brandon VMS Systems Administrator  Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fx    ------------------------------  % Date: Mon, 03 Mar 2003 17:18:20 -0600  From: brandon@dalsemi.com 5 Subject: Re: running autogen on alternate system disk 1 Message-ID: <03030317182014@dscis6-0.dalsemi.com>   e > If you are interested in getting the SYSGEN parameter setttings to the disk, then do the following:  >  > 1) sys$system:modparams.dat;  > 2) sys$system:alphavmssys.par; >   C Ooooopps, Do the following means COPY from one disk to the other...      John Brandon VMS Systems Administrator  Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fx    ------------------------------  # Date: Tue, 04 Mar 2003 00:52:28 GMT 7 From: brad@.homeportal.2wire.net (Bradford J. Hamilton) 5 Subject: Re: running autogen on alternate system disk . Message-ID: <gNS8a.344821$be.318013@rwcrnsc53>  q In article <c7fe9387.0303031505.76ac1e64@posting.google.com>, mike_hutcheson@hotmail.com (Mike Hutcheson) writes: ? >My apologies for lack of detail.  The system in question is an D >AlphaServer 1200.  No cluster.  I'm just tinkering around a bit.  IG >have VMS 7.3 installed on two differend disks: DRA0 is the primary and E >DRA5 is the secondary.  DRA5 is just a base install from the CD with = >no layered products.  DRA0 is the disk I normally boot from.  >   O I'm just guessing here, but it looks as if you want to have DRA5 around in case I DRA0 goes "belly-up".  If you are familiar with BACKUP, you might want to O consider using BACKUP on a regular basis to get a "truer" copy of DRA0 on DRA5.   K There have been some threads in this NG lately, that have advocated against M doing an online disk-to-disk backup, but I get the feeling that this is not a M "mission-critical" system; if that is true, then regular disk-to-disk backups G will take care of your requirement, if you run autogen, and then do the M disk-to-disk backup.  BACKUP commands, syntax, etc. are located in the System  Manager's Manual.   C >I was wondering if it's possible to boot from DRA5 and run autogen B >against DRA0.  Let's assume that we want to use feedback that was >saved on DRA0.   M Not possible, AFAIK.  If you are concerned about the "viability" of DRA5, you C could take the system down, on occasion, boot standalone, perform a N disk-to-disk backup, and then boot with DRA5.  Then, "reverse" the process theO next time you reboot.  That way, you have two "good" system disks to work from.   O If your system is "mission-critical" (or close to it), then you should consider I other strategies as well: if you have another SCSI bus, and access to the N SHADOWING software and license, consider mounting another disk on another SCSIN bus, and use HBVS (Host-Based Volume Shadowing) to create a two-member "shadowM set", which will keep bothe disks "in sync", and allow your system to survive  the loss of a SCSI bus.   L The System Manager's manual has details on this, and similar configurations.   > > >This is really just an exercise in curiosity.  Are there evenC >situations in which you would need to run autogen against a system A >disk you didn't boot from?  That's probably the first question I  >should be asking.  
 See above.   > 5 >And then if there is, how would I go about doing it?  >  >Cheers, >  >Mike  <snip>  A _________________________________________________________________ 0 Bradford J. Hamilton			"All opinions are my own"/ bMradAhamiPltSon@atMtAbi.cPoSm		"Lose the MAPS"    ------------------------------  # Date: Tue, 04 Mar 2003 02:15:06 GMT 4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk>5 Subject: Re: running autogen on alternate system disk 0 Message-ID: <3E640C01.8CEA9F92@blueyonder.co.uk>   Mike Hutcheson wrote:  > @ > My apologies for lack of detail.  The system in question is anE > AlphaServer 1200.  No cluster.  I'm just tinkering around a bit.  I H > have VMS 7.3 installed on two differend disks: DRA0 is the primary andF > DRA5 is the secondary.  DRA5 is just a base install from the CD with> > no layered products.  DRA0 is the disk I normally boot from. > D > I was wondering if it's possible to boot from DRA5 and run autogenC > against DRA0.  Let's assume that we want to use feedback that was  > saved on DRA0. >   D Its a long time ago but I remember doing something like this. You'll@ need to redefine some  logicals at the process level to point toC the other disk then run autogen in that process. At least you have  F the source for autogen. I'd use search etc to investigate which files C it opens to find out which logicals to redefine if I was attempting )  this again, as I don't remember details.   ? > This is really just an exercise in curiosity.  Are there even D > situations in which you would need to run autogen against a systemB > disk you didn't boot from?  That's probably the first question I > should be asking.   I You might have a very slow cluster satellite (eg MicroVAX II) and want to D run autogen on your fast boot node, to minimize downtime for tuning.    regards, --   tim.llewellyn@blueyonder.co.uk    H * PLEASE NOTE tim.llewellyn@cableinet.co.uk address is NO LONGER VALID *   ------------------------------  % Date: Mon, 03 Mar 2003 20:30:53 +0100  From: Dirk Munk <munk@home.nl> Subject: Re: SCS/DECnet over IP 2 Message-ID: <b40ah6$q3e$1@news4.tilbu1.nb.home.nl>   JF Mezei wrote:  > Dirk Munk wrote: > N >>SCS needs to be fast with very low latency !! Routing overhead would be very >>undesirable. >  > P > Isn't SCS, like LAT, totally unroutable, operating at the raw ethernet level ?  O Exactly, so what I meant is that changing SCS to a routable protocol would not  
 be desirable.    ------------------------------  % Date: Mon, 03 Mar 2003 20:10:22 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>  Subject: Re: SCS/DECnet over IP ' Message-ID: <3E640B0E.C68063C2@fsi.net>    Antony Wardle wrote: >  > SCS/DECnet over IP >  > can we do this yet?  > 2 > I get a hard time about all the bridging/routing > going on, on the network.   G I couldn't help wondering if Cisco can encapulate such things in TCP/IP G and establish, for example, a PP link between two LAN segments (perhaps C "subnets" would be a better term, if more confusing), equivalent to F bridging. Latency would still be an issue, I'm sure, but it might shut
 "them" up.   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Mon, 03 Mar 2003 13:55:16 -0600  From: brandon@dalsemi.com < Subject: Re: SCSI cluster disk thrashing between connections1 Message-ID: <03030313551663@dscis6-0.dalsemi.com>    What is/check the value of ??? QDSKINTERVALV !Beginning with OpenVMS V7.2, the recommended and default value for QSDKINTERVAL is 3. I did some research for one of our remote clusteres and found that there is a recommended value for QDSKINTERVAL - Please NOTE, the P value of 3 is NOT for SCSI.  Please reference WIS (or DSN) for more information.    What are/check the values of ??? MSCP*    I believe that there are weights that can be applied on a particular SYSGEN value that will allow one node to carry the traffic -  it eludes me at this point.   O Also, there is a way to FORCE the disk transport.  Do not remember that either.   + When I remember things I will let you know.    John Brandon VMS Systems Administrator  Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fx    ------------------------------  * Date: Mon, 3 Mar 2003 20:25:14 +0000 (UTC), From: lewis@PROBE.mitre.org (Keith A. Lewis)< Subject: Re: SCSI cluster disk thrashing between connections. Message-ID: <b40dna$918$1@newslocal.mitre.org>  r brandon@dalsemi.com writes in article <03030313551663@dscis6-0.dalsemi.com> dated Mon, 03 Mar 2003 13:55:16 -0600: >What is/check the value of ??? 
 >QDSKINTERVAL   I QDSKINTERVAL is 10 on all nodes.  But I don't think it has anything to do L with anything because I have never set up a quorum disk.  (Though I might if I solve this problem first.)  Q >!Beginning with OpenVMS V7.2, the recommended and default value for QSDKINTERVAL O >is 3. I did some research for one of our remote clusteres and found that there Q >is a recommended value for QDSKINTERVAL - Please NOTE, the value of 3 is NOT for ; >SCSI.  Please reference WIS (or DSN) for more information.  > ! >What are/check the values of ???  >MSCP*  P Parameter Name           Current    Default     Min.      Max.     Unit  DynamicP --------------           -------    -------    -------   -------   ----  -------L MSCP_LOAD                       1          0         0      16384 Coded-valuL MSCP_SERVE_ALL                  2          4         0         15 Bit-EncodeL MSCP_BUFFER                  1024       1024       256         -1 Coded-valuL MSCP_CREDITS                    8          8         2        128 Coded-valuN MSCP_CMD_TMO                    0          0         0 2147483647 Seconds    D     >I believe that there are weights that can be applied on a particular SYSGEN value that will allow one node to carry the traffic - >it eludes me at this point. > P >Also, there is a way to FORCE the disk transport.  Do not remember that either.  G There's a "set device /noenable/path=" command, but that sounds like it H would disable the alternate path completely.  I could have done the sameJ thing in hardware by only connecting the SCSI bus to one of the two hosts, but I want failover.  + --Keith Lewis              klewis$mitre.org > The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Mon, 03 Mar 2003 14:50:34 -0600  From: brandon@dalsemi.com < Subject: Re: SCSI cluster disk thrashing between connections1 Message-ID: <03030314503439@dscis6-0.dalsemi.com>   ! >> What is/check the value of ???  >> QDSKINTERVAL K > QDSKINTERVAL is 10 on all nodes.  But I don't think it has anything to do N > with anything because I have never set up a quorum disk.  (Though I might if > I solve this problem first.)   OK -     Some thoughts here -  Q Why not create a quorum disk on a private SCSI path?   The QDSKINTERVAL set to 2.   d Why not create two or more system disks rather than one common for all nodes?  Especially with SCSI.  Z There is a note about this from HP, specifically dealing with SCSI bus resets during boot.    I > There's a "set device /noenable/path=" command, but that sounds like it J > would disable the alternate path completely.  I could have done the sameL > thing in hardware by only connecting the SCSI bus to one of the two hosts, > but I want failover.  } I was thinking about the Set/Force - however the MACRO was for VAX and I am not sure if it will work on Alpha much less SCSI.     ' What is your VMS cluster interface, NI?   0 Look up pertinenet information about this on WIS  2 Alpha Shared SCSI Cluster Configuration Guidelines   John Brandon VMS Systems Administrator  Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fx    ------------------------------  # Date: Tue, 04 Mar 2003 06:02:03 GMT * From: Damien Cymbal <d_cymbal@hotmail.com>- Subject: simh: ftp access to vms from host os . Message-ID: <vjX8a.305453$iG3.37812@sccrnsc02>  
 Hello all,  J I have TCP/IP setup for my OpenVMS 7.3 system running over simh and it is I mostly working.  I have hit one stumbling block which I assume is a simh  C issue but I was wondering if anyone else has seen it or knows of a   workaround.   J I can't successfully ftp into VMS from the machine that is hosting simh.  H Note that I can telnet into VMS from this machine and also I can ftp in F from a remote machine. From the local machine I can establish the ftp G control connection, but I cannot get the data connection to connect (I  - have tried both PORT and PASV, neither work).   H This is with simh running on a Win2k box, I haven't tried hosting it in  Linux.  J Here is an ftp debug dump from connecting remotely (works) vs. connecting  locally (fails).   $ ftp -d 10.10.10.157 ) Connected to 10.10.10.157 (10.10.10.157). 2 220 simh.attbi.com FTP Server (Version 5.1) Ready." Name (10.10.10.157:XXX): anonymous ---> USER anonymous + 331 Guest login OK, send ident as password. 	 Password:  ---> PASS XXXX. 230 Guest login OK, access restrictions apply.	 ---> SYST , 200 VMS OpenVMS V7.3 on node simh.attbi.com. Remote system type is VMS. ftp> dir, ftp: setsockopt (ignored): Permission denied	 ---> PASV . 227 Entering Passive Mode (10,10,10,157,192,9)	 ---> LIST ? 150 Opening data connection for SYS$SYSDEVICE:[ANONYMOUS]*.*;*   (10.10.10.2,32931)  # Directory SYS$SYSDEVICE:[ANONYMOUS]   D INPUT.DIR;1              1/9           3-MAR-2003 21:45:43  [SYSTEM]    (RWE,RWE,,)< LOGIN.COM;1              1/9           3-MAR-2003 21:45:36   [ANONY,ANONYMOUS]     (RE,RE,,)   Total of 2 files, 2/18 blocks % 226 LIST Directory transfer complete. 	 ftp> quit 	 ---> QUIT  221 Goodbye.   ==============================   $ ftp -d 10.10.10.157  Connected to 10.10.10.157.2 220 simh.attbi.com FTP Server (Version 5.1) Ready." Name (10.10.10.157:XXX): anonymous ---> USER anonymous + 331 Guest login OK, send ident as password. 	 Password:  ---> PASS XXXX. 230 Guest login OK, access restrictions apply.	 ---> SYST , 200 VMS OpenVMS V7.3 on node simh.attbi.com. Remote system type is VMS. ftp> passive Passive mode on. ftp> dir	 ---> PASV / 227 Entering Passive Mode (10,10,10,157,192,10) " ftp: connect: Connection timed out	 ftp> quit 	 ---> QUIT    ------------------------------  % Date: Mon, 03 Mar 2003 20:27:35 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>  Subject: Re: Terminal Servers & Message-ID: <3E640F17.C1DC87C@fsi.net>   David Reed wrote:  > ) > Thanks for the responses to my inquiry. J > I should have been more specific I suppose. I'm running VMS 6.2 and UCX.  C Youch! Any chance you can upgrade to something that would support a $ newer UCX? You'll see why shortly...  K > Using 90TL's with 4 meg upgrade using bootp to load. The terminal servers H > themselves are not at issue, I can connect to them using lat and I canM > telnet to them from a PC or VAX etc. if I define the default protocol for a  > particular port as TELNET.I > The problem is that I have a (very) old piece of software that accesses B > serial devices (bidirectional) through the terminal servers in aE > manufacturing environment. Modifying the software is not an option. M > The software configuration utility requests the name of a terminal "device" J > when starting up. If I enter the name of a LAT device it works fine. ForL > reasons outside the scope of this discussion I would like to eliminate theN > LAT protocol on our network and convert to IP exclusively. What I need to doD > therefore is set up a device on the VAX that my software will findG > acceptable but using IP with the existing hardware. I don't have much J > experience doing this and would appreciate any additional pointers. I'll4 > tinker using the suggestions I've recieved so far.  E Well, check the OpenVMS doc. site: http://www.openvms.compaq.com/doc/ D for the TCP/IP Services (fka "UCX") doc.'s. The newer UCX supports aG /CREATE_SESSION command-line qualifier for TELNET, and a CREATE_SESSION E command from with TELNET. With that, you can create (I believe) "NTY" C devices (like NTY2:) that would equate to LTA devices. When an app. H opens such a device, the RTL (in the background) doe sthe socket connect9 and establishes the VC, then off you go - not unlike LAT.    Cool beans, huh?   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------   Date: 3 Mar 2003 17:44:17 -0800 ' From: eugene@cse.ucsc.edu (Eugene Miya)  Subject: Re: unix & Message-ID: <3e6404f1$1@news.ucsc.edu>  ' In article <3E60E197.C77DC6CF@fsi.net>, 0 David J. Dachtera <djesys.nospam@fsi.net> wrote: >Brian 'Jarai' Chase wrote: + >> In article <3E554E38.7881042@yahoo.com>, . >> Peter Flass  <peter_flass@yahoo.com> wrote:D >> > I'll NEVER understand unix poeople.  Put three of 'em and theirK >> > computers in a room and you'll wind up with at least five more-or-less  >> > incompatible versions.   G >> Consider the role of subtle differences and diversity in the process F >> of natural selection and in evolution in general.  One of the greatF >> strengths of Unix and Unix-like operating systems is their inherentC >> ability to adapt to the needs of a wide variety of target users.  > I >I might buy that if someone could identify those (non-economic) elements C >of VMS that inhibit such adaptation, or those elements of UNIX and H >UN*X-like environments that promote such adaptation, but are lacking inE >VMS and are not found in the POSIX environments provided by/with/for  >current versions.   Post over in comp.sys.convex. D For years, they attempted a DCL/VMS environment to compete and offerH something for the DEC/VMS community.  It was a waste of money in the endH (considering how many other vendors Convex could take people away from).G Instead, they ended up catering to fresh out students who came from SUN I and similar school environments (at the better schools which rejected VMS M in CS Depts. [other depts. like ME, CE, math, etc. didn't count in the end]).   C >The major reasons why I say that revolve around such things as the H >ability to access RMS indexed files from the CLI (DCL) - two of the keyG >differences between VMS and non-VMS environments. A UN*X-land parallel D >might be an extended "shell" that included an ODBC interface to the! >major database engine of choice.   G An interesting hypothesis.  But I doubt that a command line interpreter C (a shell) would be sufficient.  It would be useful to really have a @ comparative OS features list (e.g., file versioning as part of a@ filename or journaling).  And I think what you will find will be@ personal biases (all good OSes will end up converging to Multics9 after all, right? [if we are to believe the Multicians]).    ------------------------------   Date: 3 Mar 2003 17:34:22 -0800 ' From: eugene@cse.ucsc.edu (Eugene Miya)  Subject: Re: unix & Message-ID: <3e64029e$1@news.ucsc.edu>  G In article <b3qhii$nu2$4@bob.news.rcn.net>,  <jmfbahciv@aol.com> wrote: * >In article <rb26j-hcj.ln1@pez.jarai.com>,2 >   bdc@world.std.com (Brian 'Jarai' Chase) wrote:* >>In article <3E554E38.7881042@yahoo.com>,- >>Peter Flass  <peter_flass@yahoo.com> wrote: C >>> I'll NEVER understand unix poeople.  Put three of 'em and their J >>> computers in a room and you'll wind up with at least five more-or-less >>> incompatible versions.   >>F >>Consider the role of subtle differences and diversity in the processE >>of natural selection and in evolution in general.  One of the great E >>strengths of Unix and Unix-like operating systems is their inherent B >>ability to adapt to the needs of a wide variety of target users. > @ >Sorry, you've got it bass-ackwards and this is a very important; >point.  The only reason those adaptations are happening is 9 >because enough users have the sources to implement them.   B Barb's point about source is the critical one, but Brian's comment: is important considering the buggy nature of all software.E It's adaptability was due to a combination of minimalism (the OS just B tries to really be nothing more than a multiplexor) and a level of? partitioning between tools and the OS (a concept foreign to the M big monolithic code types (the old style punch card batch type environments). A The only things you guys left out were the relative accessibility I of the developers (Thanks Dennis for taking a phone call from me in 1977) 4 to unix-wizards when it was an ARPAnet mailing list.    D >No company exists that has both enough manpower and time to it all.A >It was one of TOPS-10's strengths to ship its sources.  Not only > >could customers do customized changes, they could upgrade our@ >stuff and still be able to use their stuff after a <ahem>little? >bit of work.  My point is that it was possible to evolve their < >software as ours evolves.  That can't happen if sources and) >the tools to mush them aren't available.   A Yep.  And she notes the fatal mistake in the next paragraph, too.  As a co-worker once said: ' 	If it ain't source, it ain't software. 1 And the community has to hold develoeprs to that.   ? >Unix could suck eggs but people will still use it if, and only ; >if, they get the sources.  One of DEC's fatal mistakes was ; >to try to keep the "knowledge", a.k.a. sources, to itself; : >it was under the misconception that secrecy would protect= >its investment and intellectual property.  This is one case  " >where the exact opposite is true.   Girl, you got that right. K That was a slightly different time in the history of intellectual property. ! Complexity got the better of DEC.    ------------------------------  $ Date: Tue, 4 Mar 2003 04:09:42 +0100* From: Morten Reistad <mrr@reistad.priv.no> Subject: Re: unix / Message-ID: <md514b.on9.ln@via.reistad.priv.no>   / According to Eugene Miya <eugene@cse.ucsc.edu>: H >In article <b3qhii$nu2$4@bob.news.rcn.net>,  <jmfbahciv@aol.com> wrote:+ >>In article <rb26j-hcj.ln1@pez.jarai.com>, 3 >>   bdc@world.std.com (Brian 'Jarai' Chase) wrote: + >>>In article <3E554E38.7881042@yahoo.com>, . >>>Peter Flass  <peter_flass@yahoo.com> wrote:D >>>> I'll NEVER understand unix poeople.  Put three of 'em and theirK >>>> computers in a room and you'll wind up with at least five more-or-less  >>>> incompatible versions.    >>> G >>>Consider the role of subtle differences and diversity in the process F >>>of natural selection and in evolution in general.  One of the greatF >>>strengths of Unix and Unix-like operating systems is their inherentC >>>ability to adapt to the needs of a wide variety of target users.  >>A >>Sorry, you've got it bass-ackwards and this is a very important < >>point.  The only reason those adaptations are happening is: >>because enough users have the sources to implement them. > C >Barb's point about source is the critical one, but Brian's comment ; >is important considering the buggy nature of all software. F >It's adaptability was due to a combination of minimalism (the OS justC >tries to really be nothing more than a multiplexor) and a level of @ >partitioning between tools and the OS (a concept foreign to theN >big monolithic code types (the old style punch card batch type environments).B >The only things you guys left out were the relative accessibilityJ >of the developers (Thanks Dennis for taking a phone call from me in 1977)5 >to unix-wizards when it was an ARPAnet mailing list.   = Another thing you forget about unix is that the sources were  = in effect submitted to peer review, and they were extensively ; reviewed. Commentaries were written (e.g. Lyons), textbooks 7 referenced them; they were included in lots and lots of = research; and the foremost academic institutions in the world  took them to their hearts.    = That meant they took to adapting and mutating them, and a set = of baselines formed in the form of distributions. All of this 9 was an open, peer reviewed process, in stark contrast to  : corporate development. And, the academic model of r&d held water.  ; This was because the unix developers adhered to the core of : the scientific world-view, namely published, peer reviewed; documents. What was new and revolutionary was that this was ! source code, not human language.    7 For this we have given them far too little recognition.   ; So, to all of you out there writing papers on unix, include 9 a reference to the source of your operating system in the : reference of your paper. (this is how the scientific world tallies importance).  E >>No company exists that has both enough manpower and time to it all. B >>It was one of TOPS-10's strengths to ship its sources.  Not only? >>could customers do customized changes, they could upgrade our A >>stuff and still be able to use their stuff after a <ahem>little @ >>bit of work.  My point is that it was possible to evolve their= >>software as ours evolves.  That can't happen if sources and * >>the tools to mush them aren't available. > B >Yep.  And she notes the fatal mistake in the next paragraph, too. >As a co-worker once said:( >	If it ain't source, it ain't software.2 >And the community has to hold develoeprs to that. > @ >>Unix could suck eggs but people will still use it if, and only< >>if, they get the sources.  One of DEC's fatal mistakes was< >>to try to keep the "knowledge", a.k.a. sources, to itself;; >>it was under the misconception that secrecy would protect > >>its investment and intellectual property.  This is one case # >>where the exact opposite is true.   < From the comments given earlier, these sources from DEC were< not written to the standards of peer review; ref the absence. of comments, personal "ownership" of code etc.  ; These were commercial implementations, not intended for the > public. Publishing code is a whole different thing from making it work.   >Girl, you got that right.L >That was a slightly different time in the history of intellectual property." >Complexity got the better of DEC.  # Yep, DEC was it's own worst enemy.     .. mrr   ------------------------------   Date: 3 Mar 2003 22:34:13 -0800 ' From: eugene@cse.ucsc.edu (Eugene Miya)  Subject: Re: unix & Message-ID: <3e6448e5$1@news.ucsc.edu>  / In article <md514b.on9.ln@via.reistad.priv.no>, , Morten Reistad  <mrr@reistad.priv.no> wrote:> >Another thing you forget about unix is that the sources were > >in effect submitted to peer review, and they were extensively< >reviewed. Commentaries were written (e.g. Lyons), textbooks8 >referenced them; they were included in lots and lots of> >research; and the foremost academic institutions in the world >took them to their hearts.     0 Lyons authored a fine annotation.  I had a copy.F However, I would not say that the best code always came from academia.E Many fine OS features came from industry from both indutrial concerns I (IBM and Tandem) and never saw the light of day, or from places like BBN. A VMS, for instance, had features which created communities of use, E whether Unix fans liked them or not.  Berkeley style networking would D look a lot like DECnet to a pure IBMer familiar with SNA or numerous0 other architectures behind the networking curve.1 Student code, is in many cases, far from perfect.   > >That meant they took to adapting and mutating them, and a set> >of baselines formed in the form of distributions. All of this: >was an open, peer reviewed process, in stark contrast to B >corporate development. And, the academic model of r&d held water.   From /usr/games/fortune:   University, n.: C         Like a software house, except the software's free, and it's F usable, and it works, and if it breaks they'll quickly tell you how to fix it, and ...      That's BSD humor for ya.    < >This was because the unix developers adhered to the core of; >the scientific world-view, namely published, peer reviewed < >documents. What was new and revolutionary was that this was" >source code, not human language.  > 8 >For this we have given them far too little recognition.   Oh sort of. D This view isn't monolithic.  I only need to point to Stephen Wolfram- as an example a less than sharing programmer.   H I prefer to think of a culture of people who realized that it was better/ to share than those who held their cards close.   < >So, to all of you out there writing papers on unix, include: >a reference to the source of your operating system in the; >reference of your paper. (this is how the scientific world  >tallies importance).   G I sort of have to agree that this tallying is commonly done in the face ) of topic ignorance.  He has a point here.  ... = >From the comments given earlier, these sources from DEC were = >not written to the standards of peer review; ref the absence / >of comments, personal "ownership" of code etc.   ' Barb is the best person to answer that.   < >These were commercial implementations, not intended for the? >public. Publishing code is a whole different thing from making 	 >it work.   D It is amazing how much unworkable code was published in pseudo-Algol from the 50s thru the 70s.  D Instead of dividing the world into academic and scientific, considerD looking at different programming communities.  Those sharing numeric> code shared differently than those in business (libraries  andC "packages" [e.g., SPSS, SAS] best well known).  Those who worked in F networking (telecommunication protocols) had to deal with heterogenity9 in ways that others didn't (it takes 2 to share packets). @ Those working in graphics and AI (serious AI [ask to share Cyc]). also having different degrees of code sharing.     >>Girl, you got that right. M >>That was a slightly different time in the history of intellectual property. # >>Complexity got the better of DEC.  > $ >Yep, DEC was it's own worst enemy.    True.  We are all our own worst enemy.    ------------------------------   Date: 4 Mar 2003 06:40:38 GMT - From: djweath@attglobal.net (Dave Weatherall) E Subject: Re: VAXELN vs Posix Threads (Was: DECthreads problem on VAX) 5 Message-ID: <DTiotGxQ0bj6-pn2-nyzyGJ51wCpb@localhost>   = On Sun, 2 Mar 2003 18:09:54 UTC, cwhii_google_spam@yahoo.com   (C.W.Holeman II) wrote:   l > djweath@attglobal.net (Dave Weatherall) wrote in message news:<DTiotGxQ0bj6-pn2-6pHkIZkuND00@localhost>... > D > > One thing I'm sure of, Posix based threading is nowhere near as I > > elegant as VAX/ELN's JOB/Process model and the synchronisation tools   > > it provides. > ( > Do you care to expound/expand on this?  E I don't have time at the minute to writwe a long diatribe as I'm off  E to the UK for the rest of the week. Suffice to say for now that with  B DecThreads (aka Posix) one is obliged to manipulate a lot of data A items that ELN does for you. e..g. (from memory I'm at home). to  C create an ELN process (thread of execution) within an ELN JOB (VMS  E process), I use CREATE_PROCESS with name, argument. In DecThreads, i  E have to create attributes, assign atttributes, and then I can create  - the Thread. Deleting it all again is similar.   C To synchronise with one or more objects, in ELN I call WAIT_ANY or  B WAIT_ALL with a list of things that should be waited for. The ELN E kernel does the work. In Threads I have to manipulate more attribute  C objects and then handle Mutexes and Condition variables myself. It  E works but you can see the Unix as opposed to the RSX/VMS heritage of  0 ELN. I'm a VMS bigot so what can one expect :-).  E Oh yes and then there's the little matter of error/status reporting.  E ELN's got it, Threads expects you to do it with exceptions. Great if  @ you're writing C, there a macros to do it. Not so easy in other  languages. My stuff is Fortran.    Enough, I've got to get to wok.    --   Cheers - Dave.   ------------------------------  $ Date: Mon, 3 Mar 2003 15:25:58 -0500; From: "Brian Tillman" <Brian.Tillman@notnoone.notnohow.com> ! Subject: Re: VMS Backup solutions $ Message-ID: <3e63ba4c$1@news.si.com>  H >Besides the fact that this eliminates a remote boot over the faster NI, >right?    Yep.  Damn!   @ >is it not true that you cannot move the actual SCS traffic backB >over to the faster NI but rather only routing DECnet, LAT, and IP >traffic over the faster NI?  K I don't know if the traffic is actually running over the Nemonix board, all # I know is what ANALYZE/SYSTEM says:    SDA> sho lan LAN Data Structures  -------------------  LSB address  = 93C697C0  Device state = 0003 Inited,Run  <                  -- EWA Unit Summary  3-MAR-2003 15:21:59 --  8 UCB     UCB Addr  Fmt   Value           Client     State; ---     --------  ---   -----           ------  -----------  EWA0    93BB6580I EWA1    93BB6800  Eth   60-07           SCA     0017 Strtn,Len,Uniq,Strtd I EWA4    93BBA180  Eth   60-03           DECNET  0017 Strtn,Len,Uniq,Strtd 4 EWA5    93BBAB80  Eth   60-01           MOPDL   001F Strtn,Len,Uniq,Share,StrtdK EWA6    93BBAE00  Eth   90-00           LOOP    001D Strtn,Uniq,Share,Strtd E EWA7    93BBCC00  Eth   60-04           LAT     0015 Strtn,Uniq,Strtd E EWA9    93BBA680  Eth   08-00           IP      0015 Strtn,Uniq,Strtd E EWA11   93BBB300  Eth   08-06           ARP     0015 Strtn,Uniq,Strtd E EWA12   936A6380  Eth   80-41           LAST    0015 Strtn,Uniq,Strtd   L So, it would seem that cluster traffic is, at least, _able_ to use the path.I Whether or not it actually does is something I don't know how to measure.   I >If you are pushing SCS, are you sure and if  so, how are you doing that?    I'm using the command:   $ mcr config_pci 100f cluster   F The "cluster" parameter purports to enable clustering over the device. --  I Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot com 5 Smiths Aerospace  Addresses modified to prevent SPAM. @ 3290 Patterson Ave. SE, MS Replace "at" with "@", "dot" with "." Grand Rapids, MI 49512-1991 8        This opinion doesn't represent that of my company   ------------------------------  % Date: Mon, 03 Mar 2003 14:22:10 -0700 + From: "Barry Treahy, Jr." <Treahy@MMaz.com> ! Subject: Re: VMS Backup solutions % Message-ID: <3E63C782.90605@MMaz.com>    Brian Tillman wrote:  I >>Besides the fact that this eliminates a remote boot over the faster NI,  >>right? >>     >> >  >Yep.  Damn! >  >    > A >>is it not true that you cannot move the actual SCS traffic back C >>over to the faster NI but rather only routing DECnet, LAT, and IP  >>traffic over the faster NI?  >>     >> > L >I don't know if the traffic is actually running over the Nemonix board, all$ >I know is what ANALYZE/SYSTEM says: > 
 >SDA> sho lan  >LAN Data Structures >------------------- >LSB address  = 93C697C0 >Device state = 0003 Inited,Run  > = >                 -- EWA Unit Summary  3-MAR-2003 15:21:59 --  > 9 >UCB     UCB Addr  Fmt   Value           Client     State < >---     --------  ---   -----           ------  ----------- >EWA0    93BB6580 J >EWA1    93BB6800  Eth   60-07           SCA     0017 Strtn,Len,Uniq,StrtdJ >EWA4    93BBA180  Eth   60-03           DECNET  0017 Strtn,Len,Uniq,Strtd5 >EWA5    93BBAB80  Eth   60-01           MOPDL   001F  >Strtn,Len,Uniq,Share,Strtd L >EWA6    93BBAE00  Eth   90-00           LOOP    001D Strtn,Uniq,Share,StrtdF >EWA7    93BBCC00  Eth   60-04           LAT     0015 Strtn,Uniq,StrtdF >EWA9    93BBA680  Eth   08-00           IP      0015 Strtn,Uniq,StrtdF >EWA11   93BBB300  Eth   08-06           ARP     0015 Strtn,Uniq,StrtdF >EWA12   936A6380  Eth   80-41           LAST    0015 Strtn,Uniq,Strtd > M >So, it would seem that cluster traffic is, at least, _able_ to use the path. J >Whether or not it actually does is something I don't know how to measure. >  >    > J >>If you are pushing SCS, are you sure and if  so, how are you doing that? >>     >> >  >I'm using the command:  >  >$ mcr config_pci 100f cluster > G >The "cluster" parameter purports to enable clustering over the device.  >    > I Perhaps it is not very scientific, but what happens when you copy a very  G large from from one node to the other on the NI?  If you timed it, and  H presuming it were running over the faster link, one would hope that the G simply math of of the file size * 512 / number of seconds might answer   that...    Barry    --    @ Barry Treahy, Jr  *  Midwest Microwave  *  Vice President & CIO   A E-mail: Treahy@mmaz.com * Phone: 480/314-1320 * FAX: 480/661-7028    ------------------------------  % Date: Mon, 03 Mar 2003 16:10:31 -0500 % From: "Mark D. Jilson" <jilly@hp.com> ! Subject: Re: VMS Backup solutions & Message-ID: <3E63C4C7.4DB78743@hp.com>  E Use SDA> SHOW LAN/COUNT and look at the Octets Sent & Received on the 8 EWA1 device to get an idea of how much it is being used.   Brian Tillman wrote: > J > >Besides the fact that this eliminates a remote boot over the faster NI,	 > >right?  > 
 > Yep.  Damn!  > B > >is it not true that you cannot move the actual SCS traffic backD > >over to the faster NI but rather only routing DECnet, LAT, and IP > >traffic over the faster NI? > M > I don't know if the traffic is actually running over the Nemonix board, all % > I know is what ANALYZE/SYSTEM says:  >  > SDA> sho lan > LAN Data Structures  > -------------------  > LSB address  = 93C697C0   > Device state = 0003 Inited,Run > > >                  -- EWA Unit Summary  3-MAR-2003 15:21:59 -- > : > UCB     UCB Addr  Fmt   Value           Client     State= > ---     --------  ---   -----           ------  -----------  > EWA0    93BB6580K > EWA1    93BB6800  Eth   60-07           SCA     0017 Strtn,Len,Uniq,Strtd K > EWA4    93BBA180  Eth   60-03           DECNET  0017 Strtn,Len,Uniq,Strtd 6 > EWA5    93BBAB80  Eth   60-01           MOPDL   001F > Strtn,Len,Uniq,Share,StrtdM > EWA6    93BBAE00  Eth   90-00           LOOP    001D Strtn,Uniq,Share,Strtd G > EWA7    93BBCC00  Eth   60-04           LAT     0015 Strtn,Uniq,Strtd G > EWA9    93BBA680  Eth   08-00           IP      0015 Strtn,Uniq,Strtd G > EWA11   93BBB300  Eth   08-06           ARP     0015 Strtn,Uniq,Strtd G > EWA12   936A6380  Eth   80-41           LAST    0015 Strtn,Uniq,Strtd  > N > So, it would seem that cluster traffic is, at least, _able_ to use the path.K > Whether or not it actually does is something I don't know how to measure.  > K > >If you are pushing SCS, are you sure and if  so, how are you doing that?  >  > I'm using the command: >  > $ mcr config_pci 100f cluster  > H > The "cluster" parameter purports to enable clustering over the device. > --K > Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot com 7 > Smiths Aerospace  Addresses modified to prevent SPAM. B > 3290 Patterson Ave. SE, MS Replace "at" with "@", "dot" with "." > Grand Rapids, MI 49512-1991 : >        This opinion doesn't represent that of my company   --  C Jilly	- Working from Home in the Chemung River Valley - Waverly, NY 0 	- jilly@clarityconnect.com			- Brett Bodine fan+ 	- Mark.Jilson@hp.com				- since 1975 or so  	- http://www.jilly.baka.com   ------------------------------  # Date: Tue, 04 Mar 2003 01:18:52 GMT + From: Jeff Cameron <JCam90502@jcameron.com> ! Subject: Re: VMS Backup solutions 2 Message-ID: <BA893EFB.5536%JCam90502@jcameron.com>  I On 3/3/03 7:09 AM, in article jbr66vctcq43drnl16d6knscl8kv7idipu@4ax.com, ) "jlsue" <jlsuexxxz@screaminet.com> wrote:   = > Actually, I've always wondered if the alias entries for the G > VMS$COMMON.DIR-->[SYSx]SYSCOMMON.DIR are unnecessary, since the SYS$* D > rooted logicals are searchlists anyway.  If they are necessary, itH > could due to the the low-level facilities that are needed during earlyG > boot processes.  But I'm not involved in the product development area / > so I don't get to work on investigating that.   J When backing up the system disk, it is very important to get the directoryK alias structure backed up too, so that it is restored the same way. This is G why. Suppose you back up a VMS system disk without this directory alias J structure. When you restore the disk the common root will be duplicated inH each of the rooted system directories for each node, and each node wouldE boot OK, and each node would find the common files in it's own common L subroot. Now as an example let say a user changes his password, assuming theF UAF is in the common area (as are many cluster common files). Then hisK password change would only be reflected on this one node that he has logged > onto. He must then change the password on all the other roots.  H This is just one example. Any cluster commonality would be lost. This isF true for things like the system rights list, the license database, theI network proxy files, the security auditor, and more importantly you would J not be able to use the queue manager very easily. There are others too! InK essence, you loose most of the functionality of a cluster when you do this.   G Image backups of the system disk are vital to disaster recovery for VMS  systems.   Jeff Cameron   ------------------------------  % Date: Mon, 03 Mar 2003 20:05:16 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> ! Subject: Re: VMS Backup solutions ' Message-ID: <3E6409DC.ED70AA5A@fsi.net>    David Pikcilingis wrote: > L > As part of its OpenVMS emulation tool set, Boston Business Computing sellsH > Vbackup, an OpenVMS BACKUP emulator that can read and restore savesetsK > created by OpenVMS BACKUP on UNIX systems and can also create savesets on . > UNIX that are can be read by OpenVMS BACKUP. > I > Additional information can be found at www.bosbc.com/vbackup.html or by  > contacting info@bosbc.com   ? Y'all got doc.'s for that on-line? I'd like to read up on it...    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Mon, 03 Mar 2003 21:23:54 +0000 - From: Alan Fay <alan.fay@veritas.spamdot.com> 5 Subject: Re: VMS Backup solutions (alias directories) 4 Message-ID: <b40h5b$4bc$1$8302bc10@news.demon.co.uk>  8 Actually the VERITAS NBU OpenVMS client does not backup 8 alias directory data 14 times (within a cluster or not).  2 In fact, it backs up alias directories correctly:-  7  o  This release will backup and restore OpenVMS alias  6     directories. Alias directories do not contain any 8     files, they are simply a link to another directory. 6     A common example is the system disk SYSCOMMON.DIR <     directory which points to the VMS$COMMON.DIR directory. :     Only the files in the real directory are backed up by :     the OpenVMS client the alias directory link is backed '     up as a link to the real directory.   7 There are only two direct system disk restore problems  0 left - and we are working on those. Please see:-  I ftp://ftp.emea.support.veritas.com/pub/support/Products/NetBackup_OpenVMS      nbu_vms.pdf   9 This guide also includes a recommended disaster recovery   procedure (which works now).   Alan    8 > From jlsuexxxz@screaminet.com Mon Mar 03 15:09:38 2003 > Newsgroups: comp.os.vms # > Subject: Re: VMS Backup solutions ( > From: jlsue <jlsuexxxz@screaminet.com>' > Date: Mon, 03 Mar 2003 10:09:38 -0500  > G > On Mon, 03 Mar 2003 10:44:52 +0000, Andrew Harrison SUNUK Consultancy 0 > <Andrew_No.Harrison_No@nospamn.sun.com> wrote: >  >  > > ; > >Why can't you use Veritas I thought it was available for  > >OpenVMS.  > >  > G > That's a very good question.  Actually, there are a few products that * > work with OpenVMS, with some exceptions. > G > The main issue is the OpenVMS system disk.  To support clusters it is D > implemented with file aliases between each system's own, specific,E > system-specific directory, and the cluster-common directory.  These C > are similar (in this context) to links in the Unix File System (I 	 > think).  > D > The problem with most 3rd party backup tools on the OpenVMS system; > disk is that they ignore the alias "links" and backup the F > cluster-common area with each system's own directory.  Thus, in a 13E > node cluster they will backup that info 14 times - utilizes lots of G > additional tape, and will definitely NOT restore to a bootable system  > disk.  > 3 > I think the main issues I've seen are as follows:  > / > 	1.	System disk backups not handled correctly 3 > 	2.	Due to this, restores will not work correctly / > 	3.	No "bare metal" restore capability of the 1 > 		system disk (i.e., booting off of the OpenVMS  > 		distribution CD)/ > 	4.	Often, the extended file semantics of the , > 		VMS file system (protections, ACLs, etc)+ > 		will not be saved or restored correctly  > = > Actually, I've always wondered if the alias entries for the G > VMS$COMMON.DIR-->[SYSx]SYSCOMMON.DIR are unnecessary, since the SYS$* D > rooted logicals are searchlists anyway.  If they are necessary, itH > could due to the the low-level facilities that are needed during earlyG > boot processes.  But I'm not involved in the product development area / > so I don't get to work on investigating that.  > D > I believe that Legato has introduced a new version of their backupE > software for OpenVMS that is supposed to have full OpenVMS support. F > However I have not personally played with it yet so I don't have any$ > direct experience to present here. >  >    ------------------------------  # Date: Mon, 03 Mar 2003 22:55:57 GMT 6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)$ Subject: Re: Volume set across RAID53 Message-ID: <14R8a.21912$AV5.268898@news.chello.at>   a In article <chp66v47gtdrrangik9mo8g1ofp89te4p6@4ax.com>, jlsue <jlsuexxxz@screaminet.com> writes: Z >On Fri, 28 Feb 2003 23:56:37 GMT, peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) wrote:I >>Image an application where uptime is important and so the backup window E >>is minutes if not seconds. With HBVS there is a chance that you use G >>3 member shadowsets and change one of the disks within this couple of I >>minutes (MINICOPY or maybe MINIMERGE) and then do the backup afterwards F >>from the outrotated disk. What is the best way to do this with EVA ? > @ >Well, one way to do this would be to 1) quiesce the applicationF >writes, 2) tell the EVA to take a snapshot (or snapclone) of the LUN,C >3) un-quiesce the application, and then 4) tell the EVA to present E >that new LUN somewhere.  The application is affected only during the F >time it takes for 1-3 to complete.  And #2 is typically very fast, so8 >it depends on how quickly your app responds to #1 & #3.  K That's I see as a solution for part of the whole problem. The backup (time) E problem is solved. Keeps only the a) how to control this from VMS and G b) what happens if EVA controller die(s) = disaster tolerance problems.   2 Don't hold me wrong. I don't mean EVA is useless !H I only need to understand where it fits into the picture (maybe I should" start to read the white papers...)   --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  % Date: Mon, 03 Mar 2003 19:55:00 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> $ Subject: Re: Volume set across RAID5' Message-ID: <3E640774.99B7D89F@fsi.net>    John Santos wrote: > / > On Fri, 28 Feb 2003, David J. Dachtera wrote:  >  > > John Santos wrote: > > >  > > K > > That's *EXACTLY* the part I want to avoid: shadow copies and/or merges.  >  > It's never a merge.   B Never said it was. Merges result from other causes, frequently not within the SysAdmin's control.  , > It's a copy  (or a minicopy, if possible).A > Depends on how long the copy has been out of the shadow set and ! > how much activity has occurred.  > @ > Why do you want to avoid this?  You are suddenly springing new > requirements.   H Negative. Check back thru the thread: it's one of the original criteria.   > [snip]= > But I didn't go into this because I thought you were asking = > how to do the stuff you are doing now for backing up shadow ! > sets by splitting out a member.   B Nope. Looking for how to command an EVA from an OpenVMS batch job,	 remember?    > [snip] > Irrelevent to me.    ...but relevant to me.  . > Once again, you are pulling new requirements > out of the hat.   * Nope. Again, one of the original criteria.  , > Why can you change the underlying hardware< > drastically (go from a just a bunch of disks to EVA, etc.)= > without recertifying, but can't make the slightest software 1 > change?  I think you need to re-certify anyway.   , Does the function of the application change?  A If yes, re-certify the app. Otherwise, so long as the vendor will E support the configuration (meaning THEY have already certified it, or ! some variant of it), you're good.   
 > > *SIGH* > > K > > For my money, full automation remains the way to go, even if that means I > > using some unsupported freeware to make my enterprise backup solution 	 > > work.  > C > You have to certify, but are allowed to use unsupported freeware?   G Doesn't change the function of the app.'s, and the vendor leaves backup H as an afterthought for the site to resolve. So, as I long I don't changeG anything visible to the user, and the vendor is not expected to support 
 it, I'm good.    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------   End of INFO-VAX 2003.123 ************************                                          C+schedule+and+current+stats.">
<NaviLinkData ID=89 url="http://www.nhlcoyotes.com/" sdesc="Phoenix+Coyotes+%28official+site%29" ldesc="The+official+Phoenix+Coyotes+home+page+features+news%2C+stats%2C+game+highlights%2C+player+profiles%2C+Coyotes+history+and+more.">
<NaviLinkData ID=90 url="http://www.canoe.com/HockeyChicago/home.html" sdesc="SLAM%21+Hockey%3A+Chicago+Blackhawks" ldesc="SLAM%27s+Blackhawks+page+features+coverage+from+Canadian+newspapers%2C+detailed+team+stats%2C+commentary+and+more.">
<NaviLinkData ID=91 url="http://www.canoe.com/HockeyAnaheim/home.html" sdesc="SLAM%21+Hockey%3A+Mighty+Ducks+of+Anaheim" ldesc="SLAM%27s+Mighty+Ducks+page+features+coverage+from+Canadian+newspapers%2C+detailed+team+stats%2C+commentary+and+more.">
<NaviLinkData ID=92 url="http://www.canoe.com/HockeyOttawa/home.html" sdesc="SLAM%21+Hockey%3A+Ottawa+Senators" ldesc="SLAM%27s+Senators+page+features+coverage+from+Canadian+newspapers%2C+detailed+team+stats%2C+commentary+and+more.">
<NaviLinkData ID=93 url="http://hockey.dallasnews.com/stars/" sdesc="Dallas+Morning+News%3A+Stars+coverage" ldesc="Stars+hockey+coverage+from+the+Dallas+Morning+News+online+edition.">
<NaviLinkData ID=94 url="http://www.sportsline.com/u/hockey/nhl/teams/VAN/teamrep.htm" sdesc="CBS+SportsLine%3A+Vancouver+Canucks+Report" ldesc="SportsLine%27s+Canucks+report+page+features+an+exclusive+column+with+links+to+Vancouver%27s+roster%2C+depth+chart%2C+schedule+and+other+resources.">
<NaviLinkData ID=95 url="http://www.globe.com/globe/sports/bruins/" sdesc="Boston+Globe%3A+Bruins+coverage" ldesc="Bruins+hockey+coverage+from+the+Boston+Globe+online+edition.">
<NaviLinkData ID=96 url="http://www.sportingnews.com/nhl/teams/predators/" sdesc="The+Sporting+News%3A+Nashville+Predators" ldesc="The+Sporting+News%27+Predators+Team+Report+is+written+by+a+local+correspondent+and+updated+frequently.+The+page+also+links+to+archived+columns+as+well+as+Nashville%27s+roster%2C+schedule+and+current+stats.">
<NaviLinkData ID=97 url="http://www.sportsline.com/u/hockey/nhl/teams/PIT/teamrep.htm" sdesc="CBS+SportsLine%3A+Pittsburgh+Penguins+Report" ldesc="SportsLine%27s+Penguins+report+page+features+an+exclusive+column+with+links+to+Pittsburgh%27s+roster%2C+depth+chart%2C+schedule+and+other+resources.">
<NaviLinkData ID=98 url="http://www.dallasstars.com/" sdesc="Dallas+Stars+%28official+site%29" ldesc="The+official+Dallas+Stars+home+page+features+news%2C+stats%2C+game+highlights%2C+player+profiles%2C+Stars+history+and+more.">
<NaviLinkData ID=99 url="http://www.cano