1 INFO-VAX	Sat, 28 Dec 2002	Volume 2002 : Issue 717       Contents:7 Re: ";*" vs. ".*" (Was: Re: is VMS really easy to use?) 7 Re: ";*" vs. ".*" (Was: Re: is VMS really easy to use?)  Re: ASCII files & NFS  Re: ASCII files & NFS * Re: C RTL Patch Kit Dependent on SYS Patch( Determining a driver base address on VAX= Re: Dinosaurs on the net? [was Re: vax6k.openecs.org rebirth] ! RE: First Hammer performance test 3 Re: HP/INTEL Migration to Itanium Competence Center 3 Re: HP/INTEL Migration to Itanium Competence Center 3 Re: HP/INTEL Migration to Itanium Competence Center 3 Re: HP/INTEL Migration to Itanium Competence Center 3 Re: HP/INTEL Migration to Itanium Competence Center 3 Re: HP/INTEL Migration to Itanium Competence Center 3 Re: HP/INTEL Migration to Itanium Competence Center  RE: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? RE: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? RE: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: is VMS really easy to use? Re: Link errors.2 Re: Managing increasing number of foreign commands2 Re: Managing increasing number of foreign commands& RE: Merry Christmas to all VMS'ers ...1 Re: Of Galaxy, Infiniband and Distributed Caching  Re: OpenVMS Branding Query Re: OpenVMS Branding Query Re: OpenVMS Branding Query  Re: Pathworks - service Netlogon SCSI Converters  Re: SCSI Converters ? Re: searching for MS640-DA memory and a RF36 drive for 4000-100 " Re: TCPIP FTP fixing up file names RE: vax6k.openecs.org rebirth  Re: vax6k.openecs.org rebirth   F ----------------------------------------------------------------------   Date: 27 Dec 02 20:01:48 +0100) From: p_sture@elias.decus.ch (Paul Sture) @ Subject: Re: ";*" vs. ".*" (Was: Re: is VMS really easy to use?)) Message-ID: <VwE2rQ4shPb2@elias.decus.ch>   T In article <auhj28$9v4$1@helle.btinternet.com>, "matt" <matt987@hotmail.com> writes: > 5 > "Jan-Erik Sderholm" <aaa@aaa.com> wrote in message # > news:3E0C4C77.C32FAB35@aaa.com...  >> matt wrote: >> >. >> > [1] or '.*' - does anyone ever use this ? >>B >> Sure, it's faster to type (single-finger) without shifting (butA >> I'm using the ";*" format in command files just to be sure...)  >>7 >> But this might be a Swedish-keyboard-layout-thing...  >> >> Jan-Erik Sderholm. > L > It must be a slow day today 'cos now I'm intrigued enough to find out when  > and why this syntax got added. > F I'm not sure when the . alternative to ; appeared, or if it was alwaysD there. Particularly useful when typing single handed (phone in otherA hand) on keyboards where ; requires shift. Liie Jan-Erik, I use ;  in procedures.   D > Its a bit like dev:<dir> rather than dev:[dir] which most famouslyI appeared > in STABACKIT.COM. >  > Time for some digging (although it'd be E a lot easier just to ask Hoff :-) >  Also useful for certain keyboard F layouts. Alt is needed for [] on my PC at work. The French VT clones II used many years ago displayed a couple of non-bracket characters instead. H Fortunately, when you DIRECTORY and related commands adjust their outputE accordingly, which circumvents that problem nicely. You do have to be E careful to code procedures which parse filenames for that, of course.    --  
 Paul Sture Switzerland    ------------------------------   Date: 27 Dec 02 21:18:49 +0100) From: p_sture@elias.decus.ch (Paul Sture) @ Subject: Re: ";*" vs. ".*" (Was: Re: is VMS really easy to use?)) Message-ID: <iZ3PoIYd0n28@elias.decus.ch>   U In article <VwE2rQ4shPb2@elias.decus.ch>, p_sture@elias.decus.ch (Paul Sture) writes:   7 Sorry, made a mess of the formatting. Here it is again.   T In article <auhj28$9v4$1@helle.btinternet.com>, "matt" <matt987@hotmail.com> writes: > 5 > "Jan-Erik Sderholm" <aaa@aaa.com> wrote in message # > news:3E0C4C77.C32FAB35@aaa.com...  >> matt wrote: >> >. >> > [1] or '.*' - does anyone ever use this ? >>B >> Sure, it's faster to type (single-finger) without shifting (butA >> I'm using the ";*" format in command files just to be sure...)  >>7 >> But this might be a Swedish-keyboard-layout-thing...  >> >> Jan-Erik Sderholm. > L > It must be a slow day today 'cos now I'm intrigued enough to find out when  > and why this syntax got added.  F I'm not sure when the . alternative to ; appeared, or if it was alwaysD there. Particularly useful when typing single handed (phone in otherA hand) on keyboards where ; requires shift. Liie Jan-Erik, I use ;  in procedures.   D > Its a bit like dev:<dir> rather than dev:[dir] which most famously > appeared in STABACKIT.COM. > ) > Time for some digging (although it'd be # > a lot easier just to ask Hoff :-)   J Also useful for certain keyboard layouts. Alt is needed for [] on my PC atF work. The French VT clones I used many years ago displayed a couple ofB non-bracket characters instead. Fortunately, DIRECTORY and relatedH commands adjust their output accordingly, which circumvents that problemJ nicely. You do have to be careful to code procedures which parse filenames for that, of course.   --  
 Paul Sture Switzerland    ------------------------------  % Date: Fri, 27 Dec 2002 21:17:29 +0100 6 From: Arne =?ISO-8859-1?Q?Vajh=F8j?= <arne@vajhoej.dk> Subject: Re: ASCII files & NFS) Message-ID: <3E0CB559.1050301@vajhoej.dk>    Curtis Justus wrote:  H > I am using Hummingbird's NFS Maestro to transfer text files between anN > OpenVMS system and Windows NT.  My problem is when I create an ASCII file onK > the Windows side and put it on the mounted drive, the record format looks  > like this on the VAX:  > + > Stream_LF, maximum 0 bytes, longest 32787  > K > This is causing a third-party import package to fail.  I have changed the M > format from Stream_LF to Stream and that does not work.  I created the same M > file without the CR, copied it over and had the same problem.  If I use FTP F > to copy the file manually, it works fine (which makes sense with the8 > inherent functionality in FTP to convert ASCII files). > M > Hopefully, it doesn't show, but I have had extensive experience being an NT K > and UNIX administrator, but am very rusty at the VAX.  Would anybody have % > any pointers or things I could try?    Try:  % $ CONVERT/FDL=NLA0: filename filename    Arne   ------------------------------  # Date: Sat, 28 Dec 2002 02:39:04 GMT 4 From: "Curtis Justus" <jjurfvur@uurusr.vurs.yya.com> Subject: Re: ASCII files & NFS: Message-ID: <c98P9.38028$Zt4.467511@twister.rdc-kc.rr.com>   JF,   J Thanks for the insight.  I am using UCX (does that give you enough info?).   I'll give these things a try.   
 Take care, cj= "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in message ) news:3E0CBD7A.164EC190@vl.videotron.ca...  > Curtis Justus wrote:J > > I am using Hummingbird's NFS Maestro to transfer text files between an" > > OpenVMS system and Windows NT. > ) > What TCPIP stack are you using on VMS ?  >  > - > > Stream_LF, maximum 0 bytes, longest 32787  > I > Try DUMP the file to see exactly what is in it (CR LF, LF or whatever).  > J > If you edit the file on VMS, does it look OK ? If so, you could then useB > CONVERT/FDL to create a version of the file in a format your VMS application wants. > L > ( CONVERT/FDL  changes the physical contents but not the logical contents, SET H > FILE/ATTRIB does not change the file but changes how VMS processes the file). > J > You may wish to look at the TCPIP Management reference manual (available onI > the www.openvms.compaq.com web site in the "VMS documentation") for the K > section on NFS, more importantly on the section that discusses ADF files.  IfL > your Windows originated files have a particular file extentions, you might beL > able to have them created with proper attributes by placing an ADF file in SYS$SYSTEM.    ------------------------------  % Date: Fri, 27 Dec 2002 13:27:24 -0800 0 From: Mark Berryman <Mark.Berryman@Mvb.Saic.Com>3 Subject: Re: C RTL Patch Kit Dependent on SYS Patch , Message-ID: <3E0C553B.59B5A356@Mvb.Saic.Com>   Robbie Benton wrote: > F > I downloaded the ACRTL V0100 Patch kit for OpenVMS V7.3-1 yesterday.> > It appears to have a fix for a problem we were experiencing. > E > Howver, it states you must install SYS V0200 Patch kit prior to the E > installation of this kit.  However, SYS V0200 Patch has a note that = > there is a problem with it and should not be installed on a  > uniprocessor system. >  > What is a person to do?  >  > Robbie  E Contact the support center and have them send you the new versions of E the routines IO_ROUTINES.EXE and IO_ROUTINES_MON.EXE.  Install SYS V2 H and then install these two routines on top of that (actually 4 since the( file comes with the .STB files as well).  F I have encountered no problems with this combination (although several$ things broke with SYS V2 by itself).  
 Mark Berryman    ------------------------------  % Date: Fri, 27 Dec 2002 15:16:09 -0500 $ From: Mike Duffy <Duffy@process.com>1 Subject: Determining a driver base address on VAX > Message-ID: <63D30D6E10CFD11190A90000F805FE860492AF2D@LESPAUL>  4 Been playing primarily with Alpha too long and can't4 remember the VAX way.  I need to determine the base 1 address of a device driver from within a program.   7 On Alpha (and VAX for other types of loadable images),  2 I'd rummage through the list of loaded images, do 7 a LDR$REF_INFO, and go from there.  On VAX, the drivers : I'm interested are not included there (LDR$GQ_IMAGE_LIST).   I need the equivalent of:    $ ANALYZE/SYSTEM SDA> EVAL DKDRIVER- Hex = 8091E780	Decimal = -2137921664	DKDRIVER    Thanks, 
 Mike Duffy   ------------------------------    Date: 28 Dec 2002 03:46:00 +0800, From: Paul Repacholi <prep@prep.synonet.com>F Subject: Re: Dinosaurs on the net? [was Re: vax6k.openecs.org rebirth]- Message-ID: <87ptrndz47.fsf@prep.synonet.com>   / bdc@world.std.com (Brian 'Jarai' Chase) writes:   1 > In article <6oseua.f5i.ln@via.reistad.priv.no>, . > Morten Reistad  <mrr@reistad.priv.no> wrote:   B > > I am not too well versed in PDP10s predating the KL and Tops20C > > version 3A; but I believe the KA10 would run versions of Tops10 C > > that supported telnet.  The KA10 is most certainly a transistor  > > (and diode!) construction.   ? > Or if not the KA10, what about DEC's PDP-6 systems?  SAIL was D > originally a PDP-6 when it went online in 1966.  Eventually it wasF > upgraded to a more capable processor, but how long did they run withC > on the PDP-6; was it long enough to have it connected to Arpanet? E > Were there any other PDP-6 systems around that would've been on the   > net sometime during the 1970s?  ? Possibly. UWA had PDP-6 sn 4, the first one sold. In '72 it was G replaced by a Cyber 72... and a few month later, a KA-10 was installed. B There was much joy from many! That machine was one of the first toA use a packet switched net from UWA to WAIT in 74 (I think...I can B never remember if it was 74 or 75) The Arpanet was running to some= degree before that, so the SAIL 6 well could have been on it.      --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Fri, 27 Dec 2002 17:12:09 -0800 $ From: Shane Smith <ssmith@icius.com>* Subject: RE: First Hammer performance test0 Message-ID: <01C2ADCB.32DD3780@sulfer.icius.com>  < Some hammer SPEC figures are now filtering out onto the net.( http://www.theinquirer.org/?article=6940   Shane    ------------------------------  % Date: Sat, 28 Dec 2002 00:30:28 +0100 6 From: Martin Heller <martin.nospam.heller@mheller.org>< Subject: Re: HP/INTEL Migration to Itanium Competence Center* Message-ID: <3E0CE294.2070602@mheller.org>   Greg Cagle schrieb:  [cut] L >> The claim was made in another newsgroup that while no full-fledged serverF >> 64-bit version of Windows existed yet (the closest thing being the  >> 'Limited J >> Edition' of Win2K Advanced Server), a real 64-bit version of WinXP Pro  >> (the J >> workstation version) was released over a year ago.  That surprised me,  >> butE >> no one came forth to dispute it (though that doesn't prove it was  
 >> accurate).  >  > K > If they meant that a "real 64-bit version" of WinXP Pro Workstation *for   > IPF*, K > I'll dispute it. I've seen betas of .NET and XP Pro Workstation for IPF,   > but I > only recently. No one to my knowledge has shipped any official version  M > of Windows for IPF to date. Certainly HP hasn't, and it has the most vested 9 > interest and is shipping the workstations that need it.    Rethoric remark:; All versions of Microsoft software are betas, aren't they ?   B The WinXP 64bit Itanium edition is only available with hardware orC MSDN subscription. Offical shipment date IIRC end of of March 2001. 5 http://www.microsoft.com/windowsxp/64bit/howtobuy.asp   B The "Windows .Net advanced Server Limited Edition 1.2, 64bit" betaC is available with a HP zx6000 - if you can get one of these beasts. ? The Windows license costs a whopping 3852 EUR - calculated from D Linux-box/Windows-box price difference. The price difference between2 Linux/HP-SUX 11.22 is unknown to me at the moment.     Yours, Martin Heller   ------------------------------  % Date: Fri, 27 Dec 2002 17:15:50 -0800 & From: Greg Cagle <gregc@gregcagle.com>< Subject: Re: HP/INTEL Migration to Itanium Competence Center/ Message-ID: <v0pupmp5pegk06@corp.supernews.com>    Martin Heller wrote:   > Rethoric remark:= > All versions of Microsoft software are betas, aren't they ?    No comment!   D > The WinXP 64bit Itanium edition is only available with hardware orE > MSDN subscription. Offical shipment date IIRC end of of March 2001. 7 > http://www.microsoft.com/windowsxp/64bit/howtobuy.asp   D That page is pretty interesting. It was posted (supposedly) in March7 2001, as you note - but seems, well, rather inaccurate.   D > The "Windows .Net advanced Server Limited Edition 1.2, 64bit" betaE > is available with a HP zx6000 - if you can get one of these beasts. A > The Windows license costs a whopping 3852 EUR - calculated from F > Linux-box/Windows-box price difference. The price difference between4 > Linux/HP-SUX 11.22 is unknown to me at the moment.  C It's not actually that much, since IPF Linux from Red Hat is pretty 
 expensive.  C Here's another site that discusses how to get "Microsoft Windows XP B 64-Bit Edition Version 2003 Release Candidate 2" on a workstation:  > http://www.hp.com/workstations/products/itanium/operating.html  
 Note it says:   1 Do NOT use this software for production purposes.    --  
 Greg Cagle gregc at gregcagle dot com   ------------------------------  % Date: Sat, 28 Dec 2002 01:57:00 -0000 ! From: Z  <zarlenga@conan.ids.net> < Subject: Re: HP/INTEL Migration to Itanium Competence Center/ Message-ID: <v0q17cinoa941b@corp.supernews.com>   ' Greg Cagle <gregc@gregcagle.com> wrote: E : It's not actually that much, since IPF Linux from Red Hat is pretty  : expensive.   What's expensive?    ------------------------------  % Date: Fri, 27 Dec 2002 18:09:44 -0800 & From: Greg Cagle <gregc@gregcagle.com>< Subject: Re: HP/INTEL Migration to Itanium Competence Center/ Message-ID: <v0q1uofjs7voe3@corp.supernews.com>    Z wrote:) > Greg Cagle <gregc@gregcagle.com> wrote: G > : It's not actually that much, since IPF Linux from Red Hat is pretty  > : expensive. >  > What's expensive?   F They don't have pricing up on Red Hat's site right now for the ItaniumE version of Advanced Server, but the x86 version is $799. IIRC the 7.2 ) Red Hat for Itanium was about that price.    --  
 Greg Cagle gregc at gregcagle dot com   ------------------------------  % Date: Sat, 28 Dec 2002 02:39:13 -0000 ! From: Z  <zarlenga@conan.ids.net> < Subject: Re: HP/INTEL Migration to Itanium Competence Center/ Message-ID: <v0q3mh25kahr55@corp.supernews.com>   ' Greg Cagle <gregc@gregcagle.com> wrote: H :> : It's not actually that much, since IPF Linux from Red Hat is pretty :> : expensive.    :> What's expensive?  H : They don't have pricing up on Red Hat's site right now for the ItaniumG : version of Advanced Server, but the x86 version is $799. IIRC the 7.2 + : Red Hat for Itanium was about that price.   < Can't you get the CDs for open source s/w for about the cost
 of the media?   + That $799 should be about 99% support fees.    ------------------------------   Date: 28 Dec 2002 02:48:27 GMT, From: bill@gw5.cs.uofs.edu (Bill Gunshannon)< Subject: Re: HP/INTEL Migration to Itanium Competence Center5 Message-ID: <auj3dr$7gi3b$1@ID-135708.news.dfncis.de>   / In article <v0q1uofjs7voe3@corp.supernews.com>, ) 	Greg Cagle <gregc@gregcagle.com> writes: 
 > Z wrote:* >> Greg Cagle <gregc@gregcagle.com> wrote:H >> : It's not actually that much, since IPF Linux from Red Hat is pretty >> : expensive.  >>   >> What's expensive? > H > They don't have pricing up on Red Hat's site right now for the ItaniumG > version of Advanced Server, but the x86 version is $799. IIRC the 7.2 + > Red Hat for Itanium was about that price.  >   @ It looks to me like you can download 7.2 for IA64 from their FTP@ site for nothing.  I didn't find ISO's but there are boot floppy5 images and they would likely allow a network install.    bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  % Date: Fri, 27 Dec 2002 21:07:20 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> < Subject: Re: HP/INTEL Migration to Itanium Competence Center' Message-ID: <3E0D1568.43FFDC0C@fsi.net>    Martin Heller wrote: >  > Greg Cagle schrieb:  > [cut] N > >> The claim was made in another newsgroup that while no full-fledged serverG > >> 64-bit version of Windows existed yet (the closest thing being the 
 > >> 'Limited K > >> Edition' of Win2K Advanced Server), a real 64-bit version of WinXP Pro 	 > >> (the K > >> workstation version) was released over a year ago.  That surprised me,  > >> butF > >> no one came forth to dispute it (though that doesn't prove it was > >> accurate).  > >  > > L > > If they meant that a "real 64-bit version" of WinXP Pro Workstation *for	 > > IPF*, L > > I'll dispute it. I've seen betas of .NET and XP Pro Workstation for IPF, > > but J > > only recently. No one to my knowledge has shipped any official versionO > > of Windows for IPF to date. Certainly HP hasn't, and it has the most vested ; > > interest and is shipping the workstations that need it.  >  > Rethoric remark:= > All versions of Microsoft software are betas, aren't they ?   D I look at it this way: after Beta test comes Gamma test (most people@ call it "general availability" or "general release" or something	 similar).    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Fri, 27 Dec 2002 11:18:59 -0800 $ From: Shane Smith <ssmith@icius.com>' Subject: RE: is VMS really easy to use? 0 Message-ID: <01C2AD99.CC01B260@sulfer.icius.com>  H "man" is short for manual. What Real Programmer would admit reaching for	 a manual?    Shane    -----Original Message-----6 From: David J. Dachtera [mailto:djesys.nospam@fsi.net]( Sent: Friday, December 27, 2002 10:00 AM To: Info-VAX@Mvb.Saic.Com ' Subject: Re: is VMS really easy to use?      Z wrote: > # > VAXman-  @SendSpamHere.ORG wrote: N > : I don't believe that somebody not well versed in unix "crytology" would beF > : able to "interpolate" "rm -r dir" to mean "remove -recursive dir". > > > I believe even a novice can learn to use man in a few hours.  F 1. I find the manpages to be overly verbose at times. IMO, appropriate) for hardcopy, but not for on-line "help".   H 2. IMO, "HELP" is more intuitive than "man", unless you're trying figureG out something and think, "man, what the hell command do I use here??!!"    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------    Date: 27 Dec 2002 11:29:41 -0800. From: spamsink2001@yahoo.com (Alan E. Feldman)' Subject: Re: is VMS really easy to use? = Message-ID: <b096a4ee.0212271129.585c39a4@posting.google.com>   e bill@cs.uofs.edu (Bill Gunshannon) wrote in message news:<auht0l$768v1$1@ID-135708.news.dfncis.de>... ? > In article <b096a4ee.0212270607.1887fb34@posting.google.com>, 3 > 	spamsink2001@yahoo.com (Alan E. Feldman) writes: i > > bill@cs.uofs.edu (Bill Gunshannon) wrote in message news:<aughgr$73t0o$3@ID-135708.news.dfncis.de>... B > >> In article <b096a4ee.0212261156.23786311@posting.google.com>,6 > >> 	spamsink2001@yahoo.com (Alan E. Feldman) writes: > >> >  E > >> > Use DFU if you can. It is the best tool I've seen for deleting D > >> > directory trees. You can get if from the Freeware CD which isK > >> > available at www.openvms.compaq.com. And David can tell you where to # > >> > get the v2.7-1 patch for it.  > >>  N > >> And if we're accepting alternate locally written (as in not a part of theM > >> comercial OS) solutions, I can write a simple script that will mimic the O > >> "rm" command and use much longer but more english-like options in minutes.  > >>  L > >> But, just like writting logicals to mimic Unix on VMS is a bad idea, so > > 9 > > Writing logicals to mimic Unix? What does that mean?   >  > $cd :== set def  > $rm :== delete< > $vi :== "EDIT/EDT/COMMAND=SYS$USERDISK:[BOB237]EDTINI.EDT" >  > And so on..... > B > Surely you saw other people mention this here the last time this > topic made the rounds.  F Those are symbols, not logicals. Also, I, for one, have never tried to Unixize VMS.  L > >> to is trying to make Unix look like something it's not.  Slows down theM > >> learning curve and is non-portable to other people's machines. (hint, if N > >> you came to my site and the only way you knew to delete whole directoriesL > >> full of files was DFU, what would you do when you didn't find DFU on my > >> systems??)  > > : > > 1.) The odds of my coming to your site are very small. > C > True, but some people's jobs require them to frequently visit and ' > work at sites other than their home.    E Fine, so your remarks apply to those people. But I'm not one of them, / nor am I likely to be in the forseeable future.    > But, if you came to my site C > and exepected to find a "cd" or "rm" command on VMS, you would be  > unpleasantly surprised.   C Despite the fact that I use DFU to delete directory trees, I can do C vanilla VMS. Besides, we're talking about deleting directory trees, E not solving partial differential equations with a Green function that E satisfies Neumann boundary conditions! I think I can handle it. And I = never expect, nor do I wish for, cd or rm on systems that are  unfamiliar to me.   H > > 2.) The odds of my forgetting how to delete directory trees are alsoH > > very small. In fact, if I forgot that, I probably also forgot how to > > walk and talk. > I > But you are a long time VMS user and most likely an Admin.  Can you say H > the same thing about all the people who use the machines you service??  E Yes, I am a long time VMS user and an Admin. But the vast majority of D the people who use my machines never use DCL. They run from menus orF use the PC front-end program (if you count front-enders). The very few7 who use DCL have no interest in, nor any need for, DFU.   G > > 3.) Assuming I came to your site and forgot how to delete directory H > > trees (the odds of which are incredibly small), I would learn how toI > > do it! Or, I would simply download and install DFU. How long could it E > > take to learn it, or download it? OK, download and set up takes a 2 > > little while, but you only have to do it once. > F > Say what??  You would take it upon yourself to just install whateverG > you felt like on someone else's system??  Before moving into academia D > I used to work for some major government contractors. I frequentlyI > went to places like the Pentagon, DTSA and other agencies with acronyms C > for names.  Can you even imagine what would have happened if they H > found me trying to install anything other than what I was there for onH > one of their machines??  And, I can assure you that while I don't haveD > legal power they do, if I had someone come in here to do some workJ > and found him trying to install unrelated software on one of my machinesI > he would find himself in the parking lot so fast he would have friction F > burns from the air he passed thru.  And he would find a very unhappyJ > boss waiting for him when he got back to his office as well.  VMS peopleJ > are always talking about security.  How secure is a system if any vendorB > of the street can install any program he wants when he visits??   @ Relax. I said I'd use DEL *or* DFU. I wouldn't install something? without permission or against policy. And you didn't say it was D someone else's system. Maybe you gave me my own test system on whichD to do development. I see no problem installing DFU there. IT DEPENDSE ON THE CIRCUMSTANCES. I never said I'd unconditionally install DFU. I C *did* say that I'd be able to do it without much difficulty with or  without DFU.  < > > Are you saying that the whole Freeware CD is a bad idea? > F > No.  But I know people (like the data center here) who do not freelyD > install anything that is non-commercial.  Somehow, I can't imagineE > them letting outsiders install anything they please when they won't F > install it themselves for the use of their regular users.  And I canE > think of others (like those mentioned above) who would be even more 
 > adamant.  ? OK. I can do vanilla VMS. On those systems, I won't use DFU, or C anything else on the Freeware CD. In a previous job, I was part VMS E Admin, part help desk. And I had to go to users' desks and help them, F even if they customized their PC's different ways and I set up mine myB own way. I did fine with that, too. The bottom line is you need to learn your job.   H > > What am I supposed to do? Do everything the hard way? I should wasteI > > hours of my time deleting directory trees the DEL way? Why? For what?  > D > The VMS engineers apparently didn't see it as so difficult or they- > would have built in an easier way to do it.   B You don't know what they thought. Perhaps they didn't have time toF worry about it. But isn't the author of DFU a VMS engineer? ApparentlyF he saw a need for it. Besides, DFU DEL/DIR/TREE is not only easier, itD is much faster than the corresponding DEL command(s). Sometimes more? than 10 times as fast. Furthermore, it consumes less resources, A leaving said resources more available to do other useful work for 2 other users. And DFU has many other cool features.  A > > And how often does one have to delete directory trees anyway?  > E > I didn't originate the example.  I'm just trying to show how absurd B > the arguments regarding which (VMS vs. Unix) is better or worse.  & OK. I'm in agreement with you on that.  B > And, as most here are bound to remember, I don't think either isC > better or worse, only different.  Two different approaches to the  > same destination.   
 OK. Noted.  I > > Should I walk 4 miles to work every day instead of taking the subway, F > > just in case there's a strike? (And there almost was one about two > > weeks ago!)  > E > But that justifies what I said above.  When they put in the subway, I > they didn't remove all the sidewalks.  And yet, the subway is obviously D > easier.  And you know what, I'll bet there actually are people whoC > prefer to walk even though they have the subway right below them.  > Are they wrong??  D That depends what you mean by wrong. If they want to spend more timeB with their families, they are wrong. If they don't like walking inB rain, wind, snow, sleet, freezing cold, heat and humidity; if theyD don't like listening to drivers leaning on their horns; listening toF ambulance, fire, and police sirens; dodging red light runners; wastingF hours a day; are not in good shape; have bad knees; find walking to be# boring; then they are indeed wrong.   D If there are some people for whom walking works, more power to them.A But most of us prefer the subway; and for many, walking is not an F option. Would you walk from Connecticut to Downtown Manhattan? I thinkF not! (I don't live in Conn., but there are those who do that commute.)  . > > What would I do? I think I'd do all right.   D > Use the system that is best for you and accept the fact that thereE > are alternatives and that they're being different doesn't make them  > wrong, just different.  E I didn't say using DEL was wrong. And I thought you were saying using 4 DFU was wrong. I'm not sure what you're saying here.  B Wrap-up: I was simply recommending what I think is the best way toF delete directory trees. I guess you were more concerned with UnixizingF VMS, but you made a definite point about deleting directory trees, and I was responding to that.    Dislcaimer: JMHO Alan E. Feldman   E Stay tuned! We have some more commercials coming right up! But first,  a few minutes of the program.    ------------------------------   Date: 27 Dec 2002 19:36:26 GMT, From: bill@gw5.cs.uofs.edu (Bill Gunshannon)' Subject: Re: is VMS really easy to use? 5 Message-ID: <auia3q$7gh5c$1@ID-135708.news.dfncis.de>   0 In article <D81P9.2866$oy.3129@news.get2net.dk>,' 	"Dr. Dweeb" <Dweeb@nospam.com> writes: 
 > Clip ... > 6 >> It is much more a reflection on the person than theC >> OS.  Remember, a subtle mistake you don't know you made that has C >> later (and more longterm) effects is probably worse than the one H >> that is obvious from the very start.  Precision is the responsibility >> of the user, ...  > F > Only someone who is not an engineer could say something like that.    H If you can't adequately prove your point, attack the messenger.  we have  now come full circle once again.  L >                                                                    Part ofN > engineering *anything* is to do it in such a way that the inadvertent errors4 > *cannot* occur and major disasters can be averted.  G Misusing any tool can cause a disaster.  If we made all tools incapable G of causing a disaster it would be real hard to build a house becuase we H would have no saws to cut the wood.  Afterall, that same saw can be usedH to cut out the supporting beams.  But we rely on the user to know not to do that.   > J > Whoops, forgot, Unix CLIs are not engineered, they are written by anyoneL > who feels like it - and boy does it show.  There is no overall plan and noF > consistency.  They are powerful and terse, as if these were the onlyM > criteria that mattered.  It is user engineering by nerds (clever nerds, but  > nerds no less).   A I can assure you Dennis Ritchie and Ken Thompson are anything but > "clever nerds". Care to match credentials with either of them?   > J > If we designed user interfaces for power stations, or aircraft like Unix, > CLIs, then we would be in serious trouble.  D I wonder what the ratio of powerstations and aircraft runnng embeded? Unix systems as opposed to running VMS is??  Oh, yeah I seem to D remember a recent investigation into an oil line disaster.  What wasC the controlling system running??  Unix or VMS??  In the hands of an : idiot VMS is no better at averting disasters than Unix is.   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   r   ------------------------------   Date: 27 Dec 2002 19:41:12 GMT, From: bill@gw5.cs.uofs.edu (Bill Gunshannon)' Subject: Re: is VMS really easy to use? 5 Message-ID: <auiaco$7gh5c$2@ID-135708.news.dfncis.de>o  3 In article <oT6La9JZtzxU@eisner.encompasserve.org>,p0 	Kilgallen@SpamCop.net (Larry Kilgallen) writes:b > In article <aui4o1$6tfih$3@ID-135708.news.dfncis.de>, bill@cs.uofs.edu (Bill Gunshannon) writes: > G >> OK, a matter of semantics.  LOGICAL, SYMBOL, ALIAS, NICKNAME, in thee >> end it's all the same.t >  > Not if you care about: > % > 	Using a defined name as a command.e > ? > 	Definitions that are automatically honored in opening files.V > 6 > 	Varying nested definitions within the same process. > - > 	Definitions on a wider-than-process basis.t > = > 	Permanent characteristics of a volume rather than temporals > 	environmental settings. > D > But to the best of my knowledge "nickname" is not a VMS construct,D > so I am not able to say which of Logical Name, DCL Symbol or Alias" > Entry it most closely resembles. > 
 > ======== > G > It is not good to use the terminology loosely, as those who come hereSH > with less VMS experience will dismiss the differences and be confused.  A Obviously.  I didn't know the correct terminology and I've worked @ with VMS off and on (mostly on lately) for 20 years.  I was onlyA pointing out the use of some capability of the CLI to allow users E to assign their own command names to otherwise named system commands.V) And that it is not unique to VMS or UNIX.a   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   n   ------------------------------  % Date: Fri, 27 Dec 2002 14:48:40 -0500n5 From: "Bochnik, William J" <William_Bochnik@acml.com> ' Subject: RE: is VMS really easy to use?CO Message-ID: <2D75787AAF09C64481BDFD89113BE6D5B35FF7@ac2kama0102.ac.lp.acml.com>   J one of VMS's strengths is (or was) long uptime (even better with clusters)L w/o constant babysitting - get the system running properly and you dont haveI to keep an eye on it.  This means fewer sysadmins per boxes ratio. A gooduJ sysadmin can run a couple dozen VMS systems well, but let 'em try watching1 the same NT systems (or to a lesser extent Unix).    -----Original Message-----9 From: bill@gw5.cs.uofs.edu [mailto:bill@gw5.cs.uofs.edu]  ' Sent: Friday, December 27, 2002 2:36 PMw To: Info-VAX@Mvb.Saic.Com0' Subject: Re: is VMS really easy to use?:    0 In article <D81P9.2866$oy.3129@news.get2net.dk>,' 	"Dr. Dweeb" <Dweeb@nospam.com> writes:e
 > Clip ...  J >I wonder what the ratio of powerstations and aircraft runnng embeded UnixF systems as opposed to running VMS is??  Oh, yeah >I seem to remember aI recent investigation into an oil line disaster.  What was the controllingiL system running??  Unix or >>VMS??  In the hands of an idiot VMS is no better# at averting disasters than Unix is.p    I The information contained in this transmission may contain privileged anddJ confidential information and is intended only for the use of the person(s)L named above.  If you are not the intended recipient, or an employee or agentF responsible for delivering this message to the intended recipient, anyK review, dissemination, distribution or duplication of this communication isrJ strictly prohibited. If you are not the intended recipient, please contactD the sender immediately by reply e-mail and destroy all copies of theI original message. Please note that we do not accept account orders and/oraJ instructions by e-mail, and therefore will not be responsible for carrying$ out such orders and/or instructions.   ------------------------------   Date: 27 Dec 2002 21:59:12 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)' Subject: Re: is VMS really easy to use? 5 Message-ID: <auiifg$7em7e$1@ID-135708.news.dfncis.de>:  3 In article <170Hpfry8N24@eisner.encompasserve.org>,e. 	young_r@encompasserve.org (Rob Young) writes:b > In article <aui5re$7etlu$1@ID-135708.news.dfncis.de>, bill@cs.uofs.edu (Bill Gunshannon) writes:6 >> In article <6sWBm5K1PAVC@eisner.encompasserve.org>,1 >> 	young_r@encompasserve.org (Rob Young) writes:8 >>> In article <2D75787AAF09C64481BDFD89113BE6D5B35FE7@ac2kama0102.ac.lp.acml.com>, "Bochnik, William J" <William_Bochnik@acml.com> writes:>L >>>> I think what he was trying to say is setting up a series of SYMBOLS notI >>>> LOGICALS to emulate unix command on VMS is a bad idea as it promotes L >>>> lazyness, as well as a problem if you go to another machine with a diffM >>>> account that doesnt have all your fancy symbols set up. Just think aboutaN >>>> trying to get a machine working at 2am and you can't remember the correct$ >>>> VMS command to do something.    >>>> ; >>> A >>> 	Another fellow here has a  command for everything.  I'm suren* >>> 	many of us have something similar to: >>>  >>> 	$ sq :== show queue >>> A >>> 	We are saving keystrokes, nothing wrong with that.  Unix has C >>> 	you saving keystrokes up front and forcing you to remember the  >>> 	insanity. >> nA >> So, let's see if I have this right.  Unix created really shortuG >> commands because of the high cost (particularly in time) of enteringkG >> these commands and that is now seen as a bad thing.  But today, whenrF >> that cost is minimal to non-existant it's a good thing when someoneD >> does it themself??  You can remember that "sq" means "show queue"E >> but the average Unix user is totally baffled by the idea that "rm"o >> means "remove".  Amazing. >> u > > > 	If you got tired of typing grep, you could create an alias ? > 	"g".  A twisted point you are attempting to make, I suppose.rI > 	Unix commands are crap and you have to know what they really map to.  0G > 	Delete in VMS maps to delete.  If you are tired of typing $ delete  MF > 	... you could create a symbol $ d :== delete.  If you are tired of C > 	typing rm in Unix, you are probably a very tired person I guess.a   Or you could type:   alias delete rmd6 (assuming some version of Unix less than 10 years old)2 and then it would be just like VMS (well, kinda!!)   >  >>> E >>> 	In the "why Unix sucks wars" we end up at the shell or CLI.  TheeH >>> 	reasons are there in a minor sense, but not many.  The real reasonsC >>> 	Unix is inferior are deeper than the CLI and Unix is now less  B >>> 	inferior as Sun almost has a good cluster now it appears.  OrI >>> 	good marketing.  Didn't Sun recently invent long distant clustering?, >> oH >> So you think Unix inferiority is all tied to the lack of clustering?? > * > 	Not only.  That is a good point though. > @ >> Just how many computer users today actually need clusters??   > @ > 	As many that could use high availability with quick recovery." > 	Yes, there are other solutions.  B Not my point.  What percentage of the several million computers inB use today would actually benefit from being a cluster instead of a single machine?o   >  >> I canG >> assure you that in my environment they offer absolutely no advantage E >> and that means the added complexity is a minus rather than a plus.. >> o > ? > 	I would suggest you don't have a good clustering solution int > 	mind.  H No, I don't have a need for higher reliability than I get now that couldH possibly justify the increase in cost and complexity.  And I suggest theF same is true for the majority of systems in use today.  We're not all  running stock exchanges.   >  >>> - >>> http://news.com.com/2100-1001-963615.htmlr >>>  >>> October 28, 2002 >>> O >>> "Sun Microsystems has come up with a way to insulate computer networks fromtS >>> fires, floods and bomb attacks: Split up the machines and put them in different 
 >>> cities. "y >>  J >> I run a collection of servers (about 12 at this point) for the academicJ >> department of a small University in NEPA.  How would being able to haveA >> multiple servers in remote locations make this work better??  w > @ > 	Better isn't the operative word.  More available is.  Is more2 > 	available a good thing?  I would suggest it is.  B See what I said below.  If the building I'm in is destroyed by andB form of disaster, Who is going to need to access the computers?? IG don't offer my services to people outside this area.  If the University-8 is shut down by some disaster then I have no more users.   >  >> If thereBJ >> is a flood (we're on the 4th floor and the building is over 500' higherH >> than the nearest water that is likely to flood), or a fire, or a bombK >> or an earthquake or a [m]ine subsidence (the most likely around here) ite1 >> won't do much good having servers in Denver.  w > ; > 	Denver PA or Denver CO?  Denver CO makes no sense as youSE > 	wouldn't be clustering 1800 miles away.  But if you had a computerYD > 	room fire, having clustered servers elsewhere would make you look > 	good.  A If there was a computer room fire (unlikely as there really isn't B much of anything in there to burn beyond the computers themselves)D no one would be allowed in the area until it was cleaned up and thusB there would be no users.  Also, I can replace all of the computers? in less than 48 hours and replace all the software and files ine@ about another 4 hours.  It would take longer than that to repairD the physical plant (like electricity).  It took me years to convinceC them that a DAT was worth the investment.  Do you really think I am0G likely to convince them to spend the 10's (or is it 100's) of thousands A of dollars needed for equipment and networking to actually have a0@ remote slustered site??  We're not even a for profit business!!!   >  >> we will be shut down andeH >> have no users.  Oh yeah, my uptime performance is much beter than theH >> campus network so there's another reason not to bother worrying about >> clusters. >> o > 
 > 	Too bad.    > @ > 	I get around network problems by having a point to point linkE > 	via Memory Channel with fibre option and it has saved the cluster n > 	on more than one occasion.  E  D And how much does that point to point link cost you??  Porbably moreC than my departments entire annual budget (of which my share is only3 a small fraction).  N >                                   Of course you couldn't get to the cluster . > 	for a time, but jobs were still running :-)  F And, again, of the millions of systems worldwide, how many really need that??   bill   -- <J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   o   ------------------------------   Date: 27 Dec 2002 22:07:03 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)' Subject: Re: is VMS really easy to use? 5 Message-ID: <auiiu7$7em7e$2@ID-135708.news.dfncis.de>   O In article <2D75787AAF09C64481BDFD89113BE6D5B35FF7@ac2kama0102.ac.lp.acml.com>,u8 	"Bochnik, William J" <William_Bochnik@acml.com> writes:L > one of VMS's strengths is (or was) long uptime (even better with clusters)N > w/o constant babysitting - get the system running properly and you dont have > to keep an eye on it.  i  G Hmmm.  I can usually plan on my servers staying up from one campus-widelH power shutdown to the next (annually, on Good Friday).  And I don't have to babysit any of them.   J >                       This means fewer sysadmins per boxes ratio. A goodL > sysadmin can run a couple dozen VMS systems well, but let 'em try watching3 > the same NT systems (or to a lesser extent Unix).   G Actually, the number of sysadmins is more an HR decision than dependant I on the number of servers.  For example, even before they brought in Win2KiM and Unix boxes there were 2-3 admins for the two and later four VMS machines.tL I, on the other hand, have always been the lone Sysadmin for the department.F Originally, for about 50 Sun's and a handful of PC's.  Now for a dozenG or so Unix servers, 40 Win2K boxes (including one server) a small grouptG of QNX boxes and all the networking infrastructure.  Sounds like either>I VMS needs more babysitting or that isn't the rationale behind hiring. :-)w   bill   -- gJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   .   ------------------------------  % Date: Fri, 27 Dec 2002 18:37:17 -0500 5 From: "Bochnik, William J" <William_Bochnik@acml.com> ' Subject: RE: is VMS really easy to use?(O Message-ID: <2D75787AAF09C64481BDFD89113BE6D50BCEEF@ac2kama0102.ac.lp.acml.com>,  K Annually?  Your NT 4.0 and 2k machines stay up an entire year w/o a reboot?eI Dont do much in the way of sw installs huh?  Babysitting is way more than H just reboots you know.  Disk maintenance; backups; errror checking; diskL quotas; email; accounts; etc etc etc.  IMHO our VMS systems ran with minimalL hands on (with the exception of tape swaps, account adds, etc) w/o touching,J reboots, etc.  That's what I meant.  These days the NT side (albeit betterL with 2k but not revolutionary) there are fewer sysadmins per server.  That's what I've seen.      -----Original Message----- r1 From: bill@cs.uofs.edu [mailto:bill@cs.uofs.edu] e Sent: Fri 12/27/2002 5:07 PM e To: Info-VAX@Mvb.Saic.Com  Cc: ' Subject: Re: is VMS really easy to use?       
 In articleE <2D75787AAF09C64481BDFD89113BE6D5B35FF7@ac2kama0102.ac.lp.acml.com>, t@         "Bochnik, William J" <William_Bochnik@acml.com> writes: L > one of VMS's strengths is (or was) long uptime (even better with clusters)  I > w/o constant babysitting - get the system running properly and you dont  have g > to keep an eye on it.     H Hmmm.  I can usually plan on my servers staying up from one campus-wide I power shutdown to the next (annually, on Good Friday).  And I don't have   to babysit any of them.   K >                       This means fewer sysadmins per boxes ratio. A good lL > sysadmin can run a couple dozen VMS systems well, but let 'em try watching  4 > the same NT systems (or to a lesser extent Unix).   H Actually, the number of sysadmins is more an HR decision than dependant J on the number of servers.  For example, even before they brought in Win2K C and Unix boxes there were 2-3 admins for the two and later four VMS 
 machines. L I, on the other hand, have always been the lone Sysadmin for the department.  G Originally, for about 50 Sun's and a handful of PC's.  Now for a dozen 0H or so Unix servers, 40 Win2K boxes (including one server) a small group H of QNX boxes and all the networking infrastructure.  Sounds like either J VMS needs more babysitting or that isn't the rationale behind hiring. :-)    bill p   -- 5K Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolves  E bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. . University of Scranton   | cA Scranton, Pennsylvania   |         #include <std.disclaimer.h>   o      I The information contained in this transmission may contain privileged andnJ confidential information and is intended only for the use of the person(s)L named above.  If you are not the intended recipient, or an employee or agentF responsible for delivering this message to the intended recipient, anyK review, dissemination, distribution or duplication of this communication is0J strictly prohibited. If you are not the intended recipient, please contactD the sender immediately by reply e-mail and destroy all copies of theI original message. Please note that we do not accept account orders and/ormJ instructions by e-mail, and therefore will not be responsible for carrying$ out such orders and/or instructions.   ------------------------------  % Date: Fri, 27 Dec 2002 19:00:35 -0400l0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>' Subject: Re: is VMS really easy to use?s. Message-ID: <3E0CDB81.74AEA76@vl.videotron.ca>   Bill Gunshannon wrote:I > Hmmm.  I can usually plan on my servers staying up from one campus-widepJ > power shutdown to the next (annually, on Good Friday).  And I don't have > to babysit any of them.n  N Just wait till your vaxen become teenagers and start to go out on the internetN and stay out longer than you allowed them to.  And before you know it, they'llM be inviting all sorts of friends over. You'll have to hire a guard (firewall)2E to guard the door to prevent truly undesirable friends from entering.N   :-)r   ------------------------------  % Date: Sat, 28 Dec 2002 00:29:52 -0000e! From: Z  <zarlenga@conan.ids.net>a' Subject: Re: is VMS really easy to use?t/ Message-ID: <v0ps407md2ipb8@corp.supernews.com>   , Rob Young <young_r@encompasserve.org> wrote:+ : http://news.com.com/2100-1001-963615.htmlg : October 28, 2002M : "Sun Microsystems has come up with a way to insulate computer networks from Q : fires, floods and bomb attacks: Split up the machines and put them in different  : cities. "   C My prior employer had been doing that with VMS machines as early as - 1985.  (They switched to AIX in 2001, though)D   ------------------------------  % Date: Sat, 28 Dec 2002 00:30:57 -0000:! From: Z  <zarlenga@conan.ids.net>C' Subject: Re: is VMS really easy to use?3/ Message-ID: <v0ps61l8dnvle7@corp.supernews.com>1  ) Bill Gunshannon <bill@cs.uofs.edu> wrote: ? :> I believe even a novice can learn to use man in a few hours.g  ) : A few hours??  More like a few minutes.t  < I was being conservative.  Some new users have no grasp of a/ shell or redirection of output, pipes and grep.h   ------------------------------  % Date: Sat, 28 Dec 2002 00:36:33 -0000v! From: Z  <zarlenga@conan.ids.net>s' Subject: Re: is VMS really easy to use?u/ Message-ID: <v0psghpe5h7hde@corp.supernews.com>k  # Dr. Dweeb <Dweeb@nospam.com> wrote:e6 :> It is much more a reflection on the person than theC :> OS.  Remember, a subtle mistake you don't know you made that haslC :> later (and more longterm) effects is probably worse than the one H :> that is obvious from the very start.  Precision is the responsibility :> of the user, ...   M : Only someone who is not an engineer could say something like that.  Part oflN : engineering *anything* is to do it in such a way that the inadvertent errors4 : *cannot* occur and major disasters can be averted.  8 Yawn.  That's why many good sysadmins alias rm to rm -i.  % Then users then have to deliberately:f   1 type bin/rm             or 2 (in the C shell) \rm    or 3 unalias rm   to defeat confirmation.h   ------------------------------  % Date: Sat, 28 Dec 2002 00:37:22 -0000o! From: Z  <zarlenga@conan.ids.net>e' Subject: Re: is VMS really easy to use?f/ Message-ID: <v0psi23qcfhm03@corp.supernews.com>   ) Bill Gunshannon <bill@cs.uofs.edu> wrote:iI : Oh and if you are in the "man, what the hell command do I use here??!!"nD : mode, what could be more apropos than the "apropos" command??  :-)   Or "man -k"    ------------------------------    Date: 27 Dec 2002 18:40:35 -0600+ From: young_r@encompasserve.org (Rob Young).' Subject: Re: is VMS really easy to use?t3 Message-ID: <7y9NZxSaTTuX@eisner.encompasserve.org>o  ` In article <auiifg$7em7e$1@ID-135708.news.dfncis.de>, bill@cs.uofs.edu (Bill Gunshannon) writes:   >>  A >>> Just how many computer users today actually need clusters??    >> nA >> 	As many that could use high availability with quick recovery.a# >> 	Yes, there are other solutions.k > D > Not my point.  What percentage of the several million computers inD > use today would actually benefit from being a cluster instead of a > single machine?t >   8 	Every one of them.  How cost effective would that be or< 	how realistic would it be to spend that money?  Not likely.9 	Most of us could use more stable Windows too you know...1     >> 0	 >>> I can H >>> assure you that in my environment they offer absolutely no advantageF >>> and that means the added complexity is a minus rather than a plus. >>>  >>  @ >> 	I would suggest you don't have a good clustering solution in	 >> 	mind.a > J > No, I don't have a need for higher reliability than I get now that couldJ > possibly justify the increase in cost and complexity.  And I suggest theH > same is true for the majority of systems in use today.  We're not all  > running stock exchanges. >   C 	I think you are making a mountain out of a mole hill.  In my case,iG 	I need the CPUs I am using.  It just so happens the CPUs aren't in the F 	same room.  Also, I am sharing my storage with many, I am booting off 	that storage, etc.b   >> t >>>> M. >>>> http://news.com.com/2100-1001-963615.html >>>> " >>>> October 28, 2002  >>>> tP >>>> "Sun Microsystems has come up with a way to insulate computer networks fromT >>>> fires, floods and bomb attacks: Split up the machines and put them in different >>>> cities. " >>> K >>> I run a collection of servers (about 12 at this point) for the academic K >>> department of a small University in NEPA.  How would being able to have B >>> multiple servers in remote locations make this work better??   >>  A >> 	Better isn't the operative word.  More available is.  Is morec3 >> 	available a good thing?  I would suggest it is.e > D > See what I said below.  If the building I'm in is destroyed by andC > form of disaster, Who is going to need to access the computers?? o  + 	The folks on the other side of the campus.o  K > I don't offer my services to people outside this area.  If the Universityf: > is shut down by some disaster then I have no more users.  B 	I would guess they don't all sit in the rooms with the computers.   >>   >>> If thereK >>> is a flood (we're on the 4th floor and the building is over 500' higherrI >>> than the nearest water that is likely to flood), or a fire, or a bombeL >>> or an earthquake or a [m]ine subsidence (the most likely around here) it2 >>> won't do much good having servers in Denver.   >>  < >> 	Denver PA or Denver CO?  Denver CO makes no sense as youF >> 	wouldn't be clustering 1800 miles away.  But if you had a computerE >> 	room fire, having clustered servers elsewhere would make you lookn	 >> 	good.> > C > If there was a computer room fire (unlikely as there really isn'tdD > much of anything in there to burn beyond the computers themselves)F > no one would be allowed in the area until it was cleaned up and thus > there would be no users.     	A fire from below?.  * > Also, I can replace all of the computersA > in less than 48 hours and replace all the software and files in B > about another 4 hours.  It would take longer than that to repairF > the physical plant (like electricity).  It took me years to convinceE > them that a DAT was worth the investment.  Do you really think I am'I > likely to convince them to spend the 10's (or is it 100's) of thousandstC > of dollars needed for equipment and networking to actually have auB > remote slustered site??  We're not even a for profit business!!! >   / 	You are in a segment I'm not in.  That's okay.    >>   >>> we will be shut down andI >>> have no users.  Oh yeah, my uptime performance is much beter than the I >>> campus network so there's another reason not to bother worrying about 
 >>> clusters.b >>>  >> r >> 	Too bad.   >>  A >> 	I get around network problems by having a point to point link"F >> 	via Memory Channel with fibre option and it has saved the cluster   >> 	on more than one occasion.   > 9 > And how much does that point to point link cost you??  b  @ 	Very little.  The fibre was installed long ago.  The cards were 	somewhere 2-3K each.c   > Porbably moreeE > than my departments entire annual budget (of which my share is onlyx > a small fraction). >    	Not sure of that.  O >>                                   Of course you couldn't get to the cluster t/ >> 	for a time, but jobs were still running :-)  > H > And, again, of the millions of systems worldwide, how many really need > that?? >   E 	Not many.  But implicitly we are talking about servers.  If you wanttC 	to shift to desktops, sure.... you got a lot of valid points.  ButgB 	I don't usually think of desktops when discussing VMS/clustering.   				Roba   ------------------------------   Date: 28 Dec 2002 01:26:16 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)' Subject: Re: is VMS really easy to use?n5 Message-ID: <auiujo$7h7pi$1@ID-135708.news.dfncis.de>u  2 In article <3E0CD642.36FB5F0F@firstdbasource.com>,4 	Michael Austin <maustin@firstdbasource.com> writes: > Bill Gunshannon wrote: >> d > <snip>C >> Well, I went over to the Alpha run by the datacenter for generaldB >> academic use to find examples that countered your argument, butE >> it seems that one of my previous arguments has made this one moot.yG >> We used to have "vmstar", "emacs", "kermit" and "GhostScript" (amongmD >> others) available on the system.  I know that there was no "HELP"D >> available for "vmstar", "emacs", or "GhostScript".  But, being asF >> none of these are even available anymore, I guess your right, there@ >> probably is HELP available for everything on the system.  :-) >  > J > They may still be available, you just don't have them installed.  I haveH > used VMSTAR many times over the past 6 months or so... even on VMS7.3  > <snip>  C I didn't mean available in the sense of they exist somewhere in theaC world.  I meant they are no longer being made available to users onsD that machine.  As I said, the current policy seems to be if it isn'tD commercial we don't install it.  So that would pretty much eliminate  everything on the Freeware CD.     bill    -- sJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  % Date: Sat, 28 Dec 2002 01:54:20 -0000t! From: Z  <zarlenga@conan.ids.net>.' Subject: Re: is VMS really easy to use?T. Message-ID: <v0q12cf0phcbd@corp.supernews.com>  , Rob Young <young_r@encompasserve.org> wrote:: : 	Every one of them.  How cost effective would that be or> : 	how realistic would it be to spend that money?  Not likely.; : 	Most of us could use more stable Windows too you know...h   Win 2000 is pretty damn good.   : Not quite Enterprise Critical quality yet, but damn close.   ------------------------------   Date: 28 Dec 2002 02:09:15 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)' Subject: Re: is VMS really easy to use?f5 Message-ID: <auj14a$7pll5$1@ID-135708.news.dfncis.de>m  3 In article <7y9NZxSaTTuX@eisner.encompasserve.org>,s. 	young_r@encompasserve.org (Rob Young) writes:b > In article <auiifg$7em7e$1@ID-135708.news.dfncis.de>, bill@cs.uofs.edu (Bill Gunshannon) writes: > E >> Not my point.  What percentage of the several million computers innE >> use today would actually benefit from being a cluster instead of a  >> single machine? >> r > : > 	Every one of them.  How cost effective would that be or> > 	how realistic would it be to spend that money?  Not likely.; > 	Most of us could use more stable Windows too you know...e  F But that's my point exactly.  What good is it trying to sell ice cubesE to eskimos.  Sure they can put them in their rum & coke, but it won't'D make it any better than using the ice outside.  While VMS clusteringC is a great product, the greatest majority of today's computer usersn probably don't really need it.   > K >> No, I don't have a need for higher reliability than I get now that couldrK >> possibly justify the increase in cost and complexity.  And I suggest the I >> same is true for the majority of systems in use today.  We're not all u >> running stock exchanges.s >> h > E > 	I think you are making a mountain out of a mole hill.  In my case,eI > 	I need the CPUs I am using.  It just so happens the CPUs aren't in theSH > 	same room.  Also, I am sharing my storage with many, I am booting off > 	that storage, etc.r  F Fine, but are you saying that the majority of computer users out thereG have the same needs as you do??  Enough so that it is worth the amazingeH difference in cost getting it involves??  I'm really just saying that ifD it's clustering capability were VMS's real claim to fame, then it is doomed.  Luckily, it isn't.    >>  E >> See what I said below.  If the building I'm in is destroyed by andyD >> form of disaster, Who is going to need to access the computers??  > - > 	The folks on the other side of the campus.l  H Well, they don't have accounts on my servers so who cares.  I serve only my department.   > L >> I don't offer my services to people outside this area.  If the University; >> is shut down by some disaster then I have no more users.a > D > 	I would guess they don't all sit in the rooms with the computers.  D True, but if the department is shut down because of a disaster, thenB they no longer have a need to use the servers.  Remember, you usedA the example of a disaster like a flood or fire.  If it's really aiC disaster, my servers are pretty low on the list of things that needeF to be fixed in order to get back into operation.  And, they constitute: the easiest (and likely the cheapest) piece of the puzzle.   >> kD >> If there was a computer room fire (unlikely as there really isn'tE >> much of anything in there to burn beyond the computers themselves)dG >> no one would be allowed in the area until it was cleaned up and thusn >> there would be no users.    >  > 	A fire from below?h  @ If there is a fire down below capable of burning through the 12"A of concrete floor under neath my raised floor, I doubt that theres@ will be much concern about when I am going to get back up.  MoreB likely would be where are we going to find a new building to houseD all the academic departments and classrooms that would be destroyed.@ If we find ourselves unable to offer courses, then there will beD no assignments for the students to do and thus, no need for servers.   > + >> Also, I can replace all of the computersrB >> in less than 48 hours and replace all the software and files inC >> about another 4 hours.  It would take longer than that to repaircG >> the physical plant (like electricity).  It took me years to convinceeF >> them that a DAT was worth the investment.  Do you really think I amJ >> likely to convince them to spend the 10's (or is it 100's) of thousandsD >> of dollars needed for equipment and networking to actually have aC >> remote slustered site??  We're not even a for profit business!!!  >> l > 1 > 	You are in a segment I'm not in.  That's okay.   G Myself and I would expect a very large percentage of the computer usersiH in the world.  Even a lot who think they are important probably wouldn't8 really be missed if they disappeared for a a week or so.   >> o: >> And how much does that point to point link cost you??   > B > 	Very little.  The fibre was installed long ago.  The cards were > 	somewhere 2-3K each.   G You have a fibre link to where??  We can't even run fibre (or cable) tosF buildings accross the street.  We own no right-of-ways and the PENNDOTF is not about to allow us to trench US Rte. 11 to get them. (I know, weF tried 14 years ago when I was putting in the original Campus network.)G The best I could hope for would be leased lines from the phone company.tD Any guess what that would cost to go any real distance??  2-3K for aC card!!  I build whole servers for that price. (the one advantage ofiF using commodity hardware.) Forgetting the idea that I would need boxesF with a much higher price tag and that I would have to double my server@ count the cost of the cards alone would be 48-72K.  You can restD assured that I don't have anywhere near that amount in my budget forA any given year.  Would be nice to have that kind of money to playd with, though. :-)r   >  >> Porbably moreF >> than my departments entire annual budget (of which my share is only >> a small fraction).s >> n >  > 	Not sure of that.   See above.     > P >>>                                   Of course you couldn't get to the cluster 0 >>> 	for a time, but jobs were still running :-) >> cI >> And, again, of the millions of systems worldwide, how many really needt	 >> that??? >> x > G > 	Not many.  But implicitly we are talking about servers.  If you wantlE > 	to shift to desktops, sure.... you got a lot of valid points.  ButnD > 	I don't usually think of desktops when discussing VMS/clustering.  F I am talking servers.  I still don't have that kind of budget.  That'sD why I usually build them myself.  (and actually repair things.  and B continue to use stuff like a VAX when I can.)  I ran a news serverB that actually made it into top 100 in the world that I built on myC shoestring budget.  I make a habit out of wringing everything I cantD out of what hardware (and what money) I get.  I actually had to shutF down the news server and get out of doing USENET News entirely becauseF the network services people thought I was using too much of the campusD INTERNET bandwidth.  I would love to have the budget to build killerF systems with 99.999 reliability, but it can't be justified financiallyC and I would imagine the same is true for most computer sites today.c  B And to get this at least a little closer to on topic, this idea ofD getting something out of a system that is worth the cost is probablyD another of the reasons that VMS is becoming irrelevant in education.A I have set up VAX in the past to support the few classes that can A still benefit from the VMS environment, particularly because of asC need for COBOL.  This past semester I didn't and they used the lastlD remaining VMS machine for academic use in the data center, an Alpha.D The results were disappointing.  Most students ran out of disk quotaC with just the few assignements they had to do.  I know that's not aaC VMS problem, but it does show you where educational use of VMS fitspB into the overall picture.  Next semester we will have to either goF back to a VAX 4000 or possibly take another look at the current statusD of the OpenSource COBOL Project for Unix.  The second would mean theA last nail in the VMS coffin here.  The first is only an option socD long as I am the Sys Admin for the department.  If I leave (and thatB is a possibility) my replacement will likely have no experience orC interest in VMS.  And anyway, even if he did, all the DEC equipment D left here (PDP's, VAX and one Alpha) actually belong to me and wouldG leave when I did.  Nobody else here in the department cares one bit for B VMS.  I can't imagine VMS is faring any better at other schools.     bill   -- rJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   o   ------------------------------   Date: 28 Dec 2002 02:14:50 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)' Subject: Re: is VMS really easy to use? 5 Message-ID: <auj1eq$7pll5$2@ID-135708.news.dfncis.de>o  . In article <3E0CDB81.74AEA76@vl.videotron.ca>,3 	JF Mezei <jfmezei.spamnot@vl.videotron.ca> writes:  > Bill Gunshannon wrote:J >> Hmmm.  I can usually plan on my servers staying up from one campus-wideK >> power shutdown to the next (annually, on Good Friday).  And I don't have8 >> to babysit any of them. > P > Just wait till your vaxen become teenagers and start to go out on the internetP > and stay out longer than you allowed them to.  And before you know it, they'llO > be inviting all sorts of friends over. You'll have to hire a guard (firewall)mG > to guard the door to prevent truly undesirable friends from entering.h >  >:-)  G Long past that point already.  circle.cs.uofs.edu is a PDP11/44 runningoE Ultrix-11 and vaxsrv3.cs.uofs.edu is a VAX (model indeterminate, it'st6 made of too many differnt parts :-) running ULTRIX-32.  F I am already looking at firewalling the department but not for fear ofD the PDP's or VAXen.  It's the Win2K boxes that seem to be the victim/ of hanging out with the wrong kinds of friends.e  9 Oh yeah, are there any VAXen that are not teenagers?? :-)s   bill   -- nJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   e   ------------------------------    Date: 27 Dec 2002 20:18:19 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)g' Subject: Re: is VMS really easy to use? 3 Message-ID: <kf5iXbYjdOhZ@eisner.encompasserve.org>t  ` In article <auiifg$7em7e$1@ID-135708.news.dfncis.de>, bill@cs.uofs.edu (Bill Gunshannon) writes:  D > Not my point.  What percentage of the several million computers inD > use today would actually benefit from being a cluster instead of a > single machine?   G Certainly all of those that are used in multi-architecture development.OF I find it crucial for VAX-Alpha development of the same software. ThisD should be even _more_ crucial where more architectures are supported such as in Unix.   ------------------------------    Date: 27 Dec 2002 20:20:40 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)r' Subject: Re: is VMS really easy to use? 3 Message-ID: <YYVIT9VgTVL1@eisner.encompasserve.org>>  ` In article <auiujo$7h7pi$1@ID-135708.news.dfncis.de>, bill@cs.uofs.edu (Bill Gunshannon) writes:  E > I didn't mean available in the sense of they exist somewhere in theuE > world.  I meant they are no longer being made available to users onnF > that machine.  As I said, the current policy seems to be if it isn'tF > commercial we don't install it.  So that would pretty much eliminate" > everything on the Freeware CD.    ) And most applications on a Linux machine.s   ------------------------------  % Date: Fri, 27 Dec 2002 21:23:08 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>' Subject: Re: is VMS really easy to use? / Message-ID: <3E0CFCF4.9C7684D2@vl.videotron.ca>g   Z wrote:: > Yawn.  That's why many good sysadmins alias rm to rm -i.    J In a VMS context, it is not good de alias DELETE to DELETE/CONFIRM becauseJ delete is used for many things other than deleting files. (DELET/ENTRY for# instance does not support /CONFIRM)   M However, one could extract the .CLD of "DELETE" and make /CONFIRM the defaulttJ for file delete operations. But it would break a lot of things, especially stuff such as VMSINSTAL.  L I think that the requirement of enetering version numbers in delete makes itN harder to issue such a command by mistake. Having said that, I have had to use* the undelete utility on VMS a few times...   ------------------------------  % Date: Sat, 28 Dec 2002 02:29:43 -0000 ! From: Z  <zarlenga@conan.ids.net>n' Subject: Re: is VMS really easy to use?d/ Message-ID: <v0q34n11k2o2fe@corp.supernews.com>k  1 JF Mezei <jfmezei.spamnot@vl.videotron.ca> wrote:i; :> Yawn.  That's why many good sysadmins alias rm to rm -i.   L : In a VMS context, it is not good de alias DELETE to DELETE/CONFIRM becauseL : delete is used for many things other than deleting files. (DELET/ENTRY for% : instance does not support /CONFIRM)v  > Even thee very-common /LOG creates problems, when using DELETE for symbols, for example.    ------------------------------   Date: 28 Dec 2002 02:29:57 GMT, From: bill@gw5.cs.uofs.edu (Bill Gunshannon)' Subject: Re: is VMS really easy to use?a5 Message-ID: <auj2b4$7dgar$1@ID-135708.news.dfncis.de>r  O In article <2D75787AAF09C64481BDFD89113BE6D50BCEEF@ac2kama0102.ac.lp.acml.com>,t8 	"Bochnik, William J" <William_Bochnik@acml.com> writes:M > Annually?  Your NT 4.0 and 2k machines stay up an entire year w/o a reboot?o  H We were taking servers here.  I have only one Win2K server and (luckily)G no more NY servers.  And while it has never crashed I do have to rebootsD it every once in a while just to clear the memory leaks.  Most of myF real servers run Unix and yes, they can easily go a whole year without	 a reboot.f  / > Dont do much in the way of sw installs huh?  o  8 Don't have to reboot a Unix box to install software. :-)  K >                                              Babysitting is way more thand+ > just reboots you know.  Disk maintenance;N  A Disk space is in a personally built Network Appliance.  Disks are C hot swappable RAID arrays and I haven't had a failure since I built @ it.  Next upgrade will likely be to use bigger disks to increase the available space.  6 >                                            backups;   C Backups are automated.  Only user intervention is sticking in a newe? tape every two days.  If I had the budget for a tape changer ite would be completely automatic.  G >                                                     errror checking; e  B What error checking do you mean??  I read the logs.  Don't get anyB errors that require any action on my part.  Well, maybe every once$ in a while on one of my RA81's.  :-)  K >                                                                      disk 
 > quotas;   C I don't run disk quotas.  I keep an eye on usage and try to keep us)B ahead of the usage curve.  Usually, when there is aproblem it is aG student with a bunch of stuff they shouldn't have on department serverstD and a word from the Chair clears that up.  Never try to solve social6 problems with technological solutions. It never works.   >         email; i  C What about email??  I have an email server handles all the incomingSC and outgoing email, POP and SMTP.  Just keeps on running without soi  much as periodic glance from me.  N >                accounts; etc etc etc.  IMHO our VMS systems ran with minimalN > hands on (with the exception of tape swaps, account adds, etc) w/o touching,' > reboots, etc.  That's what I meant.  n  A As do my Unix servers.  That's what I have been saying all along.e  L >                                      These days the NT side (albeit betterN > with 2k but not revolutionary) there are fewer sysadmins per server.  That's > what I've seen.  r  D Again, as I said, number of sysadmins seesm to be an HR decision andD based on a complete lack of understanding of the task.  I have twice? as many servers and a whole bunch of clients as well and yet mycB counterpart on the VMS side has three times the staff.  Go figure.   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   0   ------------------------------  % Date: Fri, 27 Dec 2002 21:30:16 -0400d0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>' Subject: Re: is VMS really easy to use?c/ Message-ID: <3E0CFE9F.D50FCCED@vl.videotron.ca>    Bill Gunshannon wrote:F > that machine.  As I said, the current policy seems to be if it isn'tF > commercial we don't install it.  So that would pretty much eliminate  > everything on the Freeware CD.  M When I worked for a bank, I was able to have them agree to some exceptions to K this. One such exception was kermit which I was able to convince was one ofeK the more serious comms packages around, to which sources were available etcaN etc. It replaced some really funky stuff that had come with their professionalK PC software which had caused so many problems. Kermit worked right from theo= start and stayed problem-free and they were quite impressed. o  L The fact that it was somehow associated with columbia university  instead ofI just some "basement ware from a some script kiddie" did help convince then, powers of the acceptability of that sofware.   ------------------------------  % Date: Fri, 27 Dec 2002 21:36:07 -0400d0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>' Subject: Re: is VMS really easy to use?w/ Message-ID: <3E0CFFFE.A47829FA@vl.videotron.ca>    Bill Gunshannon wrote: > While VMS clusteringE > is a great product, the greatest majority of today's computer usersr  > probably don't really need it.  H Considering the ability to have some 90 workstattions boot off a clusterJ server(s), I'd say that had Palmer not abandonned all the desktop softwareK that was running on VMS, it would have made a pretty darn nice office "LAN"aK system many many years ahead of what microsoft or novell could offer in theeK late 1980s, early 1990s. And the beauty of it is that you could make use of N spare CPU capacity to run batch jobs on user's workstations while they are out
 for lunch.   ------------------------------  % Date: Sat, 28 Dec 2002 02:37:18 -0000i! From: Z  <zarlenga@conan.ids.net> ' Subject: Re: is VMS really easy to use?d/ Message-ID: <v0q3iuh6ib8h15@corp.supernews.com>n  9 HP _could_ market VMS based on security (it's MUCH better ; than Linux, Unix or any flavor of Windows, even the vauntedo2 2000 / NT when hackers try to break in) or uptime.  < ANYONE can idenitfy with those qualities, especially Windows@ users (oh, the server's hung again? Just reboot!) and Linux/Unix@ users bitten by any one of the numerous buffer overflow vulnera- bilities those OSs have.  ; Clustering is powerful, as is the distributed lock manager a0 but few people will buy based on those features.  8 Alas, I suspect HP has no intention spending even a dime on marketing VMS.I  
 What a shame.    ------------------------------   Date: 28 Dec 2002 02:51:14 GMT, From: bill@gw5.cs.uofs.edu (Bill Gunshannon)' Subject: Re: is VMS really easy to use?a5 Message-ID: <auj3j2$7gi3b$2@ID-135708.news.dfncis.de>>  3 In article <YYVIT9VgTVL1@eisner.encompasserve.org>,e0 	Kilgallen@SpamCop.net (Larry Kilgallen) writes:b > In article <auiujo$7h7pi$1@ID-135708.news.dfncis.de>, bill@cs.uofs.edu (Bill Gunshannon) writes: > F >> I didn't mean available in the sense of they exist somewhere in theF >> world.  I meant they are no longer being made available to users onG >> that machine.  As I said, the current policy seems to be if it isn'teG >> commercial we don't install it.  So that would pretty much eliminate # >> everything on the Freeware CD.  u > + > And most applications on a Linux machine.h  ? We were talking about VMS here.  All of Linux is non-commercials# so the argument hardly makes sense.f   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   a   ------------------------------   Date: 28 Dec 2002 02:54:23 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)' Subject: Re: is VMS really easy to use?c5 Message-ID: <auj3ou$7gi3b$3@ID-135708.news.dfncis.de>   . In article <v0q12cf0phcbd@corp.supernews.com>,$ 	Z  <zarlenga@conan.ids.net> writes:. > Rob Young <young_r@encompasserve.org> wrote:; >: 	Every one of them.  How cost effective would that be ory? >: 	how realistic would it be to spend that money?  Not likely.m< >: 	Most of us could use more stable Windows too you know... >  > Win 2000 is pretty damn good.i > < > Not quite Enterprise Critical quality yet, but damn close.  E Your joking, right??  We have been off since 20 DEC and won;t be backfC til 2 JAN 2003.  The first thing I will need to do is reboot all of B the Win2K boxes in the lab as they will all have a message on themD stating they are out of virtual memory. And that is after 2 weeks of/ doing nothing because the building is locked!!!    And Win2K server is no better.   bill   -- aJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   i   ------------------------------   Date: 28 Dec 2002 03:00:14 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)' Subject: Re: is VMS really easy to use?t5 Message-ID: <auj43u$7gi3b$4@ID-135708.news.dfncis.de>o  3 In article <kf5iXbYjdOhZ@eisner.encompasserve.org>, 0 	Kilgallen@SpamCop.net (Larry Kilgallen) writes:b > In article <auiifg$7em7e$1@ID-135708.news.dfncis.de>, bill@cs.uofs.edu (Bill Gunshannon) writes: > E >> Not my point.  What percentage of the several million computers inyE >> use today would actually benefit from being a cluster instead of a  >> single machine? > I > Certainly all of those that are used in multi-architecture development.aD > I find it crucial for VAX-Alpha development of the same software.   F OK, but the question still remains, what percentage of the millions of/ systems in the world today does this comprise??a  G >                                                                  This F > should be even _more_ crucial where more architectures are supported > such as in Unix.  H Why??  We have frequently had multiple architectures (Sun-3, Sun-4, SGI,I VAX, x86 and others not as important) here and the lack of clustering has-G never seemed to get in the way??  The most that was ever needed was NFSlG and that was merely a convenience.  Realize that with much of the stuffcH for Unix coming in source form it almost always means multi-architecture development.   bill   -- nJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   e   ------------------------------    Date: 27 Dec 2002 21:22:16 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)l' Subject: Re: is VMS really easy to use?i3 Message-ID: <Bo8g4Yn90NP7@eisner.encompasserve.org>   d In article <auj3j2$7gi3b$2@ID-135708.news.dfncis.de>, bill@gw5.cs.uofs.edu (Bill Gunshannon) writes:5 > In article <YYVIT9VgTVL1@eisner.encompasserve.org>,b2 > 	Kilgallen@SpamCop.net (Larry Kilgallen) writes:c >> In article <auiujo$7h7pi$1@ID-135708.news.dfncis.de>, bill@cs.uofs.edu (Bill Gunshannon) writes:e >> .G >>> I didn't mean available in the sense of they exist somewhere in the G >>> world.  I meant they are no longer being made available to users onnH >>> that machine.  As I said, the current policy seems to be if it isn'tH >>> commercial we don't install it.  So that would pretty much eliminate$ >>> everything on the Freeware CD.   >> b, >> And most applications on a Linux machine. > A > We were talking about VMS here.  All of Linux is non-commercial % > so the argument hardly makes sense.   < It seems to me the same standards that prohibit VMS Freeware< should prohibit all of Linux.  But I thought your outfit ran Linux.  Or is that just Unix.v   ------------------------------  % Date: Sat, 28 Dec 2002 03:24:18 -0000s! From: Z  <zarlenga@conan.ids.net>-' Subject: Re: is VMS really easy to use?:/ Message-ID: <v0q6b21uemrr7f@corp.supernews.com>_  ) Bill Gunshannon <bill@cs.uofs.edu> wrote:y= :> Not quite Enterprise Critical quality yet, but damn close.5  G : Your joking, right??  We have been off since 20 DEC and won;t be back E : til 2 JAN 2003.  The first thing I will need to do is reboot all ofsD : the Win2K boxes in the lab as they will all have a message on themF : stating they are out of virtual memory. And that is after 2 weeks of1 : doing nothing because the building is locked!!!     : And Win2K server is no better.  0 I said it's close.  I didn't say it's there now.   And yes, I did mean it.o   ------------------------------   Date: 28 Dec 2002 04:34:37 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)' Subject: Re: is VMS really easy to use?e5 Message-ID: <auj9ks$7logb$1@ID-135708.news.dfncis.de>w  3 In article <Bo8g4Yn90NP7@eisner.encompasserve.org>,t0 	Kilgallen@SpamCop.net (Larry Kilgallen) writes:f > In article <auj3j2$7gi3b$2@ID-135708.news.dfncis.de>, bill@gw5.cs.uofs.edu (Bill Gunshannon) writes:6 >> In article <YYVIT9VgTVL1@eisner.encompasserve.org>,3 >> 	Kilgallen@SpamCop.net (Larry Kilgallen) writes:dd >>> In article <auiujo$7h7pi$1@ID-135708.news.dfncis.de>, bill@cs.uofs.edu (Bill Gunshannon) writes: >>> H >>>> I didn't mean available in the sense of they exist somewhere in theH >>>> world.  I meant they are no longer being made available to users onI >>>> that machine.  As I said, the current policy seems to be if it isn'tgI >>>> commercial we don't install it.  So that would pretty much eliminatep% >>>> everything on the Freeware CD.  i >>> - >>> And most applications on a Linux machine.t >> aB >> We were talking about VMS here.  All of Linux is non-commercial& >> so the argument hardly makes sense. > > > It seems to me the same standards that prohibit VMS Freeware> > should prohibit all of Linux.  But I thought your outfit ran > Linux.  Or is that just Unix.   ; It seems you are confusing two different departments withinT; my organization.  The data center people don't use anythingM: that is not commercial.  I on the other hand see my job as? doing whatever it takes to make my users jobs easier. (Anythingv? within reason, that is!!)  I run various flavors of BSD, Win2K,t8 VMS and even (grudgingly) Linux.  I have no problem with> Freeware as long as it comes with source and can be built from those sources.   bill   -- oJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   a   ------------------------------   Date: 28 Dec 2002 04:38:21 GMT, From: bill@gw5.cs.uofs.edu (Bill Gunshannon)' Subject: Re: is VMS really easy to use?r5 Message-ID: <auj9rt$7logb$2@ID-135708.news.dfncis.de>   / In article <v0q6b21uemrr7f@corp.supernews.com>,t$ 	Z  <zarlenga@conan.ids.net> writes:+ > Bill Gunshannon <bill@cs.uofs.edu> wrote:c> >:> Not quite Enterprise Critical quality yet, but damn close. > H >: Your joking, right??  We have been off since 20 DEC and won;t be backF >: til 2 JAN 2003.  The first thing I will need to do is reboot all ofE >: the Win2K boxes in the lab as they will all have a message on themsG >: stating they are out of virtual memory. And that is after 2 weeks ofl2 >: doing nothing because the building is locked!!! > ! >: And Win2K server is no better.  > 2 > I said it's close.  I didn't say it's there now. >  > And yes, I did mean it.a  B I guess we have different concepts of Enterprise Critical quality.A I hardly consider something with memory leaks big enough to driveeE a Mack truck thru to be even close.  I only work with it here because @ I have to, and I spend a lot of time lobbying to replace it withF one of the current Unix flavors now that they have a quite respectable( Office Suite that can even read MS docs.   bill   -- oJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   i   ------------------------------  % Date: Sat, 28 Dec 2002 19:34:45 -0800: From: "AG" <ang@xtra.co.nz>n' Subject: Re: is VMS really easy to use? 4 Message-ID: <QAbP9.19180$kq6.291770@news.xtra.co.nz>  5 "Bill Gunshannon" <bill@cs.uofs.edu> wrote in messageu/ news:aui4o1$6tfih$3@ID-135708.news.dfncis.de...e   > (I also disagree withtF > teaching that n! = "the product of all integers from 1 to n" becauseA > although it is correct in all but one case, it is still wrong.)t  @ That's nothing, a quick search on the net turned up a definitionA that n! is a product of itself and all whole numbers below it ...i9 Now I'm greasing my abacus since calculating 1! obviously.? requires going to minus infinity, not that it's gonna make muchr: difference since zero is also included, or does it? ... :)  = On the other hand, your statement about "all but one case" ist> also not correct it seems (assuming you allow an integer to be
 negative) ...r   ------------------------------  % Date: Sat, 28 Dec 2002 19:47:23 -0800i From: "AG" <ang@xtra.co.nz>n' Subject: Re: is VMS really easy to use?e4 Message-ID: <GMbP9.19202$kq6.291862@news.xtra.co.nz>  5 "Bill Gunshannon" <bill@cs.uofs.edu> wrote in message / news:aui473$6tfih$2@ID-135708.news.dfncis.de...t1 > In article <v0ovkurn7ush97@corp.supernews.com>,n% > Z  <zarlenga@conan.ids.net> writes:o% > > VAXman-  @SendSpamHere.ORG wrote:eL > >: I don't believe that somebody not well versed in unix "crytology" would beG > >: able to "interpolate" "rm -r dir" to mean "remove -recursive dir".  > > @ > > I believe even a novice can learn to use man in a few hours.  : Guys, let's try to stay off pornography in this thread ok?3 I mean, hrrm, "use man", say what? How *could* that 5 be intuitive to anyone who's not a die-hard Unix fun?f   ------------------------------    Date: 27 Dec 2002 17:17:07 -0800: From: craig.berry@SignalTreeSolutions.com (Craig A. Berry) Subject: Re: Link errors.o< Message-ID: <7f15589f.0212271717.89bbd89@posting.google.com>  0 In article <01C2ACC0.DA217800@sulfer.icius.com>,&  Shane Smith <ssmith@icius.com> wrote:  F > Did you compile with the qualifier /PREFIX=ALL? If not, I think that > might fix your problem.   D That's the default, at least with non-ancient versions of DEC/CompaqF C, and we know it's in effect in Lyndon's case because the presence ofC DECC$ at the beginning of the function name means it's already beenoC prefixed.  We also know it won't help him because snprintf() is notoF mentioned in the C RTL documentation for OVMS 7.3-1, nor is it lurking  in undocumented form in the RTL:  2 $ search sys$share:decc$crtl.exe/nooutput snprintf' %SEARCH-I-NOMATCHES, no strings matchedd  @ The "GX" bit right after the dollar sign in the undefined symbolE warning is interesting.  I suspect it means the compiler believes themB function has something to do with floating point data and will use* G_FLOAT for doubles and X_FLOAT for quads.  D Lyndon, there is an implementation of snprintf() available here, and4 the page includes pointers to other implementations:  & <http://www.ijs.si/software/snprintf/>   > -----Original Message-----7 > From: Lyndon Bartels [mailto:lbartels@pressenter.com]l  ? > %LINK-W-USEUNDEF, undefined symbol DECC$GXSNPRINTF referenced    ------------------------------  % Date: Fri, 27 Dec 2002 13:19:24 -0800h0 From: Mark Berryman <Mark.Berryman@Mvb.Saic.Com>; Subject: Re: Managing increasing number of foreign commandsh, Message-ID: <3E0C535C.1B370971@Mvb.Saic.Com>   David J. Dachtera wrote: >  .' .o .s@ > Other posters are indeed correct. There are some MAJOR caveatsA > surrounding the use of DCL$PATH. I use it everyday, but I don't + > recommend it for "general user" accounts.  >   H Elaborate please.  All my systems make use of DCL$PATH for all users and* I have yet to encounter a problem with it.  
 Mark Berrymanb   ------------------------------  % Date: Fri, 27 Dec 2002 18:10:55 -0600a7 From: "Craig A. Berry" <craigberry@nospam.telocity.com> ; Subject: Re: Managing increasing number of foreign commandsuG Message-ID: <craigberry-64F0DA.18105527122002@news.directvinternet.com>t  6 In article <00A190B3.BA44FC1C@SSRL.SLAC.STANFORD.EDU>,H  winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr")   wrote:   ; > In article <3E0BD561.7A1869BF@vl.videotron.ca>, JF Mezei s+ > <jfmezei.spamnot@vl.videotron.ca> writes:iN > >What is the best way to manage an increasing number of foreign commands on  > >VMS ?  < >  You're kind of stuck with PERL because you need logicals  > for PERLSHR and PERL_ROOTi  H The fact that placing Perl in DCL$PATH is awkward doesn't mean you have H to define it as a foreign command.  Placing it in the command tables is  documented here:  2 <http://www.perldoc.com/perl5.8.0/README.vms.html>  > Look for the section labeled "Installing Perl into DCLTABLES."   ------------------------------  % Date: Fri, 27 Dec 2002 11:06:13 -0800i$ From: Shane Smith <ssmith@icius.com>/ Subject: RE: Merry Christmas to all VMS'ers ...>0 Message-ID: <01C2AD98.12F5D450@sulfer.icius.com>  ? The original Star Wars is a fairy tale set in a science fictionbD universe. Princess kidnapped by evil magician, rescued by prince whoB thinks he's a commoner, big battle, prince beats evil magician who" sneaks off so he can be in sequel.  E Give "Enterprise" a try, Brian. It doesn't suck, and the token VulcanG has a very nice arse. ;-)a   Shaneb   -----Original Message-----B From: VAXman- @SendSpamHere.ORG [mailto:VAXman- @SendSpamHere.ORG]' Sent: Friday, December 27, 2002 6:02 AMt To: Info-VAX@Mvb.Saic.Coma/ Subject: Re: Merry Christmas to all VMS'ers ...e    8 In article <3E0B9594.6A9D47BE@vl.videotron.ca>, JF Mezei) <jfmezei.spamnot@vl.videotron.ca> writes:s" >VAXman-, @SendSpamHere.ORG wrote:K >> I've not seen Star Wars either... well, I tried once and found the firsttK >> few minutes *so* stupid that I decided to take a nap instead.  I supposem; >> that makes me some sort of out-of-touch personality too?h >fJ >How can you function in the 21st century if you don't even know about the >origins of R2D2 and C3PO? >QN >I do hope that you are at least up-to-date with Star Trek.  Hard to imagine a& >VMS guy who doesn't follow Star Trek.  F I watched the original series when it was an original series.  I never really (B got into the new series with Patrick Stewart.  I feel that science
 fiction isF losing in story line and plot to the special effect crowd.  ...and, if you'reC considering Star Wars to be science fiction you're sadly mistaken. -   --2 VAXman- OpenVMS APE certification number: AAA-0001 VAXman(at)TMESIS(dot)COM            m5   "Well my son, life is like a beanstalk, isn't it?" D   ------------------------------    Date: 28 Dec 2002 00:41:28 -0600+ From: young_r@encompasserve.org (Rob Young)a: Subject: Re: Of Galaxy, Infiniband and Distributed Caching3 Message-ID: <zUavRscexpo8@eisner.encompasserve.org>u  _ In article <QRmdnTxQPOVmlJCjXTWc2Q@metrocast.net>, "Bill Todd" <billtodd@metrocast.net> writes:c > : > "Rob Young" <young_r@encompasserve.org> wrote in message/ > news:LChUYQzjHNW$@eisner.encompasserve.org...wA >> In article <XbucnYGGuNz05JSjXTWc2Q@metrocast.net>, "Bill Todd"T" > <billtodd@metrocast.net> writes: >  > ...n > E >> > Surely this cannot be the same Rob Young who not so long ago was- > trumpetingJ >> > the inevitable victory of ADMP (an acronym which HP's web site is not > ableM >> > to find, nor is Google in anything but French and German) over all other  >> > mere MP boxes?t >> > >>< >> Well... spelling it correctly would help.  It isn't ADMP. > A > So what is it, then?  I *did* get hits in French and German for C > 'Galaxy/ADMP' in connection with Tru64 (and I think VMS as well).u >    	APMP/  % 	Adaptive Partitioned MultiProcessingu   > ...p > J >> > No matter, since the comment is the same regardless:  Galaxy/ADMP andL >> > Infiniband are complementary technologies with virtually no overlap, so > ADMP> >> > is worth exactly as much, or as little, as it always was. >>< >> That was the point of my posting.  Showing advantages and >> disadvantages.  > F > No, you were attempting to show *compensatory* aspects of IB vs. theM > possible lack of Galaxy support in Itanic systems (at least that's what youoE > said).  Galaxy is about dynamic reallocation of resources within MPrN > environments, whereas IB is about interconnecting resources in cluster-styleJ > environments (as noted, it's not really fast enough for good distributedL > memory):  they complement each other rather than in any way substitute for
 > each other.a >  >>I >> > Something as low-latency and high-bandwidth as Infiniband is, as youd > note,eK >> > extremely attractive for things like distributed caches (or caching atu > theeI >> > storage device), but in no way helps a system dynamically reallocatei >> > resources.p >>A >> Galaxy in its current incarnation doesn't dynamically allocate E >> much at all, other than CPUs.  Sure, you can change things but yousE >> are at a console prompt when you do it.  There were Galaxy futures|D >> that showed memory being dynamically moved around, but that isn't >> here yet. > N > 'Were'?  Are you suggesting that the fast, fine-grained dynamic reassignmentJ > of CPUs and memory is no longer part of the VMS roadmap (such as it is)? >   7 	Find me a roadmap with that on it.  They are old ones.i   >>E >> >  And IB is not fast enough to support anything but distressingly M >> > NUMA distributed shared memory (why you think that 2 - 10 us. latency ishE >> > comparable to EV7's 390 ns. worst-case latency in a 64-processor-' >> > configuration is not clear at all)n >> >>F >> Part of what I included was application.  Certainly for HPTC GalaxyF >> would be a winner, but as Andrew points out that is a small segment5 >> of the market.  HPTC and VMS?  Very small segment.- > N > My impression was that Galaxy was aimed at commercial activity - especially,J > the ability to divert CPU power to whatever activity in a multi-functionL > server most needed it, on an instantaneous basis.  Very important segment,L > in other words - and a very good way to leverage available processor power! > rather than over-equip systems.l >    	Ok.   > ...  > ? >> Again, XFC would use all free memory.  Galaxy can't/won't ineB >> its current incrarnation so you have more memory available as a' >> file cache (mentioned that earlier).  > M > Huh?  XFC is a distributed cache.  Galaxy is a hardware-level mechanism fortL > dynamically partitioning resources.  You could *use* some of Galaxy memoryK > as some kind of central cache (which we've discussed at length long ago), K > but once again you're talking about two *complementary* facilities rather $ > than any kind of either/or choice. >   < 	Okay.  But what I was thinking here is that XFC in Galactic? 	Shared memory from an old roadmap.  Maybe it is best XFC staysn? 	in local memory, maybe not.  If XFC in Galactic Shared memory,t< 	hard to imagine it larger than 1/2 physical memory in size.   >   The latency of Infinibandi@ >> is very tolerable for remote reads of file cache (from what I, >> understand, correct me if I'm off there). > J > I agree:  while several-microsecond latency is not really acceptable forN > main memory these days, it's very acceptable for a cache hit (cache accessesI > being persumably multiple orders of magnitude less frequent than memory K > accesses) - assuming that the cached data is fetched into local memory onnF > the hit so that subsequent use of it proceeds at main-memory speeds. >  >>D >> Greatest advantage though is recoup or protecting CPU investment.A >> Meaning that like CI it will offload IO and messaging to lowere> >> cost cards.  Querying someone about why they were using theE >> much slower CI as an interconnect, I was told Memory Channel would A >> saturate their CPUs and I know of another situation where thatyC >> happened.  I have no experience with Galaxy SMCI but since it islD >> memory to memory and no helper chips, it too has to be pretty CPU
 >> intensive.n >   > I have no idea what SMCI is.    7 	SMCI - Shared Memory Cluster Interconnect, i.e. Galaxy 7 	instances talking to each other through shared memory.   I http://www.itec.suny.edu/scsys/vms/OVMSDOC073/V73/6512/6512pro_index.html   . 15.1 Shared Memory Cluster Interconnect (SMCI)  H The Shared Memory Cluster Interconnect (SMCI) is a System CommunicationsH Services (SCS) port for communications between Galaxy instances. When anM OpenVMS instance is booted as both a Galaxy and as an OpenVMS Cluster member, O the SMCI driver is loaded. This SCS port driver communicates with other cluster1L instances in the same Galaxy through shared memory. This capability providesD one of the major performance benefits of the OpenVMS Galaxy SoftwareN Architecture. The ability to communicate to another clustered instance throughM shared memory provides dramatic performance benefits over traditional clustern interconnects. a    0 > However, there's no reason for inter-partitionI > transfers (though shared memory) within a Galaxy system to be *anywheretF > nearly* as CPU-intensive as message-structured transfers over commonK > interconnects (including FC and, later, IB) are, so if that's what you'rerF > talking about I strongly suspect that you're reasoning from a flawed
 > premise. >   < 	I couldn't find a reference for that, where is a reference?A 	Also, I discuss this further in the last two or three paragraphsy  	as it came up again down there.   >>> >> > By the way, at the moment IB appears to be heading toward' >> > low-volume/high-price niche statusa >>B >> Regarding IB niche status.  There is a lot of mythology out andD >> about here and you appear to have bought into it.  A lot of folks@ >> are scared of new things and you have networking guys saying:= >> "we do IP, Infiniband? ... shiver me timbers, I'm scared."g > G > No, what I'm referring to is first the departure of Intel from the IB M > enclave ("We're still really enthusiastic about the product, but we'll justtN > let other people develop the hardware for it..."), followed more recently byL > IBM IIRC.  No one who appears to know much about IB's current status seemsK > to expect it to become the true commodity (i.e., including desktop-level)cJ > interconnect that was once envisioned, but rather a high-end, low-volumeN > server product - especially since without commodity-level volumes there's noG > reason to expect its price to drop below that of Fibre Channel (whichhL > already supports 2 Gbit/sec with comparable host CPU overheads to those IBF > should offer, and inter-host communication as well as simple storage > linkage).  >   @ 	Ok.  But FC as an innerconnect is a fallback innerconnect.  The@ 	performance is really sucky, from a recent session.  That makesA 	sense, after all.. it is SCSI commands.  SCS over fibre is quite-C 	a feat it appears but a great plum for failover and the like, i.e.lB 	SCS will talk up until you lose the storage innerconnect when SCS? 	goes over FC.  The cleanup/failover/recovery is then "simpler"  	from session statements.I  M > The story of the past year has been one of IB startups dropping like flies.tJ > There may still be enough industry interest to preserve IB as a high-endJ > niche product, but as of now there's clearly nothing more than that (and< > still some possibility that it will simply be still-born). >   > 	I don't think so.  IBM has a great interest.  Perhaps they doF 	a Linuxification and make that the innerconnect across all platforms.! 	Externally, it appears that way.-   >>E >> Seriously, nothing will be able to compete with IB for price untillF >> about 2006 for high-end interconnect.  IB is certainly cheaper thanJ >> 10 Gbit ethernet at an average of $50000 per port for 10 Gbit ethernet: > N > I guess that depends on what you consider 'high-end'.   Gigabit Ethernet hasL > already become commonplace, with server-quality NICs under $100 (some evenK > including intelligence to off-load the central processor at that price) -eL > and 100 MBytes/sec isn't exactly chopped liver.  FC NIC prices continue toM > drop (IIRC they were starting to flirt with the $300 level well over a year F > ago, so the 2 Gbit versions should be getting down there soon if notM > already).  So quoting $50K/port for the barely-shipping 10 Gbit Ethernet is,I > something of a straw man (and of course assuming that its pricing won't-N > fairly quickly drop to something much closer to commodity levels just as allM > other Ethernet variants have is also flying in the face of rather extensive4N > precedent), while citing $1000/port prices for IB when it first appears nextL > year isn't exactly inspiring (especially considering how difficult it will$ > be to find much to connect it to). >   = 	Point taken, but you are stretching this one.  Yes, I am offr? 	on Gbit pricing.  Yes, I am defining high-end at 10 GBit.  TherB 	ComputerWorld articles are legit.  The big whoopty-doowah in that< 	article is a $30000 per port 10 GBit offering.  The analyst@ 	expects $5000 to $6000 per port pricing by 2006.  Your curve in@ 	10 Gbit pricing decline is way out of line with that analyst's.   > ...- > ? >> That fact that Dell is jumping on Infiniband is an indicatorA >> that it is cost effective.a >  > Reference, please? >   > 	Yes, type: "dell infiniband" in Google, the ones towards the	8 	top are stale but at the bottom of the first page, more  	recent ones including this one:  I http://www.serverworldmagazine.com/newsflash2/2002/12/19_infiniband.shtmlA  J Sun Microsystems, IBM and Dell will outline their commitment to InfiniBandN today, detail future InfiniBand products, and discuss how this next generationJ of "smart" high speed server I/O will radically change the way servers are" configured, operated and designed.  N The Yankee Group predicts that by 2005, 42 percent of all servers shipped willO be InfiniBand-enabled. More than 70 companies worldwide have announced plans tot$ bring InfiniBand products to market.   [snip]  O Dell believes low latency, high-bandwidth technologies -- such as InfiniBand --rM will further enable data centers to be built using low-cost, simple-to-deployaN industry standard systems. Dell's next generation PowerEdge modular blades areJ planned to be InfiniBand-ready enabling customers to take advantage of the" technology's performance benefits.   	"built using low-cost", etc.l  > 	It really is a timing issue.  I would grant this.. if 10 GBit> 	Ethernet was $3000 or $4000 per port today, IB would probably? 	be stillborn.  Again, networking fear of anything not IP wouldn< 	make sure money for 10 GBit enet shows up.  Today, the only= 	folks investing in 10 Gbit enet are early adopters with deep 	 	pockets.    > ...  > C >> So for the next 3 years+, IB is definitely a better interconnectu6 >> for CPU offload reasons, latency reasons, and cost! > F > Gigabit Ethernet costs less than 10% of the IB costs you're citing.   > 	Two different ballparks.  Gbit ethernet has 2 to 10 times the? 	latency of IB (at 2 to 10 us for IB), a tenth of the bandwidthaD 	and no offloading capabilities (for the most part, as I acknowledge 	below it is showing up).h    ItsI > NICs are starting to incorporate 'offload engines' that reduce host CPUw$ > involvement to FC and IB levels.    % 	I see that.  Certainly isn't common.e  + > It can offer one-way latencies in the 10s K > of microseconds range, which is adequate for most things that IB might be,L > usable for (even remote caching).  So unless a link requires more than 100K > MBytes/sec of sustained bandwidth, it will remain a *far* more attractiveT	 > option.h >   @ 	Ok.  But surely a bottleneck as channels would frequently burst< 	above 100 Mbytes with cache transfers, right?  Gbit doesn'tB 	seem to have enough pipe for cache transfers for serious lifting.# 	Pretty fluffy nonsense on my part.n  N > 2 Gbit Fiber Channel costs less than 1/3 of the IB costs you're citing, alsoK > offloads host CPUs to about the same extent that IB does, and IIRC offersrK > more like 10 us. latency, so, again, unless a link requires more than 200 N > MBytes/sec sustained bandwidth it will be a much more attractive option (andK > 10 Gbit FC is on the near horizon, though I don't know anything about itsd > projected costs).  >   3 	Okay.  Sounds like a competing low-end technology.0   > ...1 >  >   Infiniband shouldfI >> have a superior cost structure for quite some time, have no idea whereeE >> people (yourself included) get this idea that IB will be at a cost:E >> disadvantage, if nothing else, that certainly isn't the case.  AretD >> you comparing to Gbit Ethernet?  Cost the same, IB is much faster >> (10 Gbit versus Gbit).e > J > See above:  your cost figures for Gbit Ethernet appear to be quite a fewK > years out of date (or perhaps years of DECpaq experience have conditionedw6 > you to paying 10x the market price for such things). >   ; 	Yes, stale data on my end.  That is what I get for wingingy 	it.   >> >> So ... coming back around...  >>? >> What about Infiniband advantages versus Galaxy?  If you lashoF >> together 16 - 4 CPU servers with Infiniband, would that "do" Oracle4 >> better than a Galaxied partitioned 64 CPU server? >  > No:  not nearly as well. >  >   Wouldn't you> >> get more CPU with IB helping, therefore better performance? > F > Not only are you apparently confused about the relative overheads ofJ > inter-partition communication in Galaxy compared with *any* interconnectI > (even an efficient one like IB), and not only are you assuming that theAI > static partitioning into 4-CPU nodes will be as effective as some otheraB > division of labor, and not only are you ignoring the latency andJ > synchronization overheads of using Oracle's data-shipping 'cache fusion'M > mechanisms rather than one or a small number of larger, centralized caches,eJ > but you're also ignoring Galaxy's ability to shuffle CPUs dynamically toK > adjust to instantaneous load variations (assuming, that is, that it's noto7 > optimal to run all 64 EV7s in a single system image).d >   B 	I'm still a "bit" skeptical as to the CPUs not being impacted.  I? 	do know the old crusty CI can be fastpathed, here is reference.+ 	to that earlier mention of CPU saturation:a  m http://groups.google.com/groups?selm=cf15391e.0109260649.40446623%40posting.google.com&oe=UTF-8&output=gplains  F "Given that you have 10-CPU SMP boxes, and 4 CI adapters per box, it'sE quite possible that you are currently using Fast_Path on CI to spreaddD CI interrupt workload across non-primary CPUs.  Be aware that Memory@ Channel does not support Fast_Path (and since no new MC hardware= development is planned, and there's thus no incentive for VMS/D Engineering to dive back into the driver code, it's unlikely it everC will support Fast_Path).  I worked at a site where we tried to moveiC from CI to MC for SCS traffic in a large cluster (16 6-CPU GS-140s).E and our plans failed due to saturation of CPU 0 in interrupt state oniD lock-master nodes for hot application files.  They had to rip the MCD hardware back out and go to 6 CIs instead.  As a test, you might tryE turning off Fast_Path on CI in your current configuration and measure@" peak CPU 0 interrupt-state usage."  8 	So to further beat the horse, if SMCI can't or won't be; 	fastpathed then is IB the answer and we would assume it is7; 	fastpathed from the outset.  Since SMCI has 94 us for lockn> 	latency and MC II 120us, one would hope IB comes in somewhere8 	less than 200us  (i believe from memory Gbit is 230us).  ; 	If SMCI is or can be fastpathed, Galaxy is a much strongers> 	interconnect as you would have less risk of saturating a CPU.B 	Of course, there is a few other angles here... since it is shared: 	memory you certainly don't have to build up and tear down@ 	checksummed message packets do you?  Maybe you can run 10 timesC 	as many lock requests through shared memory on a single CPU.  That D 	and the fact you wouldn't have 16 CPU partitions, maybe 4 is ideal.B 	Perhaps Galaxy is far superior all the way around.  That would be 	cool.    ; 	Disclaimer:  I dawdle.  If you really want or need to know ; 	what it is all about, contact HP.  I'm not an HP employee,- 	just a usenet denizen.0   				RobG   ------------------------------  % Date: Fri, 27 Dec 2002 21:21:57 +0100s9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>i# Subject: Re: OpenVMS Branding Queryo' Message-ID: <3E0CB665.EE65E464@aaa.com>d  > I think I know *a lot* about computers and I always try to say; "OpenVMS" to anyone I know to (or think) *not* knowing much:B about VMS *specificaly*, but on c.o.v it's just a waste of typing.  < When I speak to someone I know know nothing about computers,' it realy don't matter anyway, does it ?   = To someone older then me (43) that is *in* the IT business, Ik? usualy say VMS, because that's what it probably was called lasti time they heard about it...g  @ And, lastly, the more "official" the context is, the more reason3 of course to use the vendors offical name, OpenVMS.e   Jan-Erik Sderholm.      > Kilgallen quotes Vlems:o > >o > > ? > >That is why "Open" is useful, to sort out the inexperienced.e7 > >I have heard of headhunters looking for someone with-1 > >OpenVMS experience rather than VMS experience.t > >u >    ------------------------------    Date: 28 Dec 2002 04:06:52 +0800, From: Paul Repacholi <prep@prep.synonet.com># Subject: Re: OpenVMS Branding Query-- Message-ID: <87lm2bdy5f.fsf@prep.synonet.com>:  0 peter@langstoeger.at (Peter LANGSTOEGER) writes:  D > In article <aueraf$6ma9q$1@ID-143435.news.dfncis.de>, "Hans Vlems" > <hvlems@iae.nl> writes:n  F > >Nobody I know uses the name openVMS. In fact the few times that theB > >name openVMS pops up in conversation it is used by persons that% > >have no knowledge about computers.G  h > I use it (too). 2 > Just to decrease the name similarity with MVS...  @ I would have thought zOS would have fixed that minor irritation!   -- r< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Fri, 27 Dec 2002 23:55:46 +0100o" From: "Hans Vlems" <hvlems@iae.nl># Subject: Re: OpenVMS Branding Queryu5 Message-ID: <auilpm$7lg87$1@ID-143435.news.dfncis.de>A  ? "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> schreef in bericht ) news:3E0B3E1B.6516263A@vl.videotron.ca...  > Hans Vlems wrote:eK > > Right, VMS ownership might change again in the future. But what companyhI > > would desire that? Don't mention IBM, they will kill VMS overnight tot gainG > > marketshare, especially the financial segment like stock exchanges.( >mK > Au contraire. I think that *IF* IBM were to acquire VMS, they would do soo withF > full knowledge of the past mistakes done on VMS and it would buy VMS becausenG > it would see the potential to make lots of money simply by fixing its7 image0  > (and dropping the silly open).  L IBM has no commercial purpose for VMS. IBM is actually quite sure that theirB current line up of S/390, AS/400 and RS/6000 systems addresses theK marketplace just fine. IBM sales reps are actively looking for customers toiJ lure them away from wandering competitors (if HP et al deserve that name).  K > IBM would be aware that customers need a trustable vendor who really doese putnJ > its money where its mouth is and doesn't just do empty promises that VMS will > be taken seriously.h  D IBM takes its products seriously. That's why they have problems withI retiring products. A couple of years I attended a VSE seminar. The VP foriL VSE stated outright: VSE will be around well beyond your retirement age. TheJ announcement came across as 100% sincere and exceeded everything (in termsL of customer relations) than anything I've seen from VMS owners in the past 8 years.  I > Of course, IBM would port VMS to Power which would give VMS a huge edge G > against HP-UX and Tandem which will be stuck on the failing IA64. VMSL couldBK > then also so battel against Tandem. And I suspect that the "VAX-FT" might- beE > resurected on Power. As long as VMS is owned by the same company asn Tandem,i? > VMS won't really get much more "fault tolerance" added to it.v  L IF IBM would buy VMS then would not port VMS to Power. It would perhaps keep; Alpha alive and possibly consider porting VMS to the S/390.eJ Unfortunately IBM has AS/400, a well loved platform by its customers. MoreH important, OS/400 works the way IBM customers think: each function has aJ screen and a well hidden batch job behind it. Now VMS could be modified to> behave like that of course, but why would IBM spend the money?J IBM managed to do what DEC achieved 10 years ago: IBM can now speed up itsL processors at the pace of double the performance at a slightly higher price. Their customers love it.  K And do not forget that the S/390 has a kind of hardware clustering built ineL that appeals to IBM customers: availability solved in hardware not software.  L > Note that I emphasised the *IF*.  IBM is not the type of company that doesK > things halfway. If they decide to bu VMS, it is because they intend to docK > something with it. They would know that buying VMS just for the customershG > would not work because customers would be very angry if the new owner> didn't > take good care of their VMS.  6 IBM is a professional company, I fully agree with you.  E > Right now, IBM might stand to gain more by letting HP stumble since'	 acquiringl? > the VMS customers who would not trust HP would be quite easy.m  C Sure, by not buying VMS they just lure ex-VMS customers. Dead easy.hL I happen to work for a site that used to have an AXP/NT installed base. JustH up to the moment AXP got murdered. They swapped all their NT systems for Netfinity's.  L > It would be strange hower to think of Sue as an IBM employee, and receicve- > marketing material with the IBM logo on it.- >m :-)l >>G > Come to think of it, it would have made sense for Compaq to have beena brokenJ > up. The ex-Digital part (which HP isn't doing much with) would have gone to> > IBM while the Compaq and Tandem parts would have gone to HP.  K Agreed, if IBM would buy VMS then they'd have a commercial reason for doingpG so and VMS would have been safe with a secure future. But I'm afraid itm
 won't happen.    Hans   ------------------------------  # Date: Sat, 28 Dec 2002 00:39:42 GMTs. From: peter@langstoeger.at (Peter LANGSTOEGER)) Subject: Re: Pathworks - service Netlogonb4 Message-ID: <ip6P9.87555$qq5.1191278@news.chello.at>  V In article <1041002257.57508@krakonosovo>, "Jiri Koutnik" <jkoutnik@ebanka.cz> writes:6 >Netlogon doesnt start, cant move forward, any ideas ?   Yes.   1) Describe the environmentf 2) Describe the problem  3) Read the logfiles@ 4) check for LanMan V2 (Win95) Limits (like max.252 groups, ...). 5) upgrade to the newest version of everything; 6) install all available ECOs (and ask for not public ones)u 7) avoid using WINTEL at all   -- r Peter "EPLAN" LANGSTOEGER$% Network and OpenVMS system specialistc E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------    Date: 27 Dec 2002 17:20:55 -0800& From: brianrota@iname.com (Brian Rota) Subject: SCSI Converters= Message-ID: <a534a137.0212271720.48e59cfe@posting.google.com>n  K I have come into some DEC scsi converters I am looking to get rid of these.l< The part numbers I have are (35) dwzza-wa and (35) dwzzb-vw.B If you are interested please let me know I will let them go cheap.- you can get back to me at brianrota@iname.come  	 Thank Your   Brian    ------------------------------  % Date: Fri, 27 Dec 2002 21:25:10 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca> Subject: Re: SCSI Converters/ Message-ID: <3E0CFD6E.505A1C06@vl.videotron.ca>u   Brian Rota wrote:t > M > I have come into some DEC scsi converters I am looking to get rid of these.e> > The part numbers I have are (35) dwzza-wa and (35) dwzzb-vw.D > If you are interested please let me know I will let them go cheap.  L You should specify your physical location.  Which planet ? Which continent ? Which country? which city ?    ------------------------------    Date: 28 Dec 2002 04:20:37 +0800, From: Paul Repacholi <prep@prep.synonet.com>H Subject: Re: searching for MS640-DA memory and a RF36 drive for 4000-100- Message-ID: <87heczdxii.fsf@prep.synonet.com>4  0 "news.pandora.be" <nieuws@bodymetals.be> writes:  F > I'm trying to restore a vax 4000-100 to use it later as a web-serverC > for the hardware i'm looking for 8 or 16 memmory sticks 16Mb type $ > MS640-DA and a harddrive type RF36  @ See if you can get a HSD controller rather than a RF drive. Then= go hunting for a SCSI of choice. Bit easier to find than Rfs!r Even if they are f'ing ugly...   -- r< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.e@                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Fri, 27 Dec 2002 21:51:28 +0100s6 From: Arne =?ISO-8859-1?Q?Vajh=F8j?= <arne@vajhoej.dk>+ Subject: Re: TCPIP FTP fixing up file namese) Message-ID: <3E0CBD50.1030706@vajhoej.dk>a   JF Mezei wrote:R  K > You know, VMS still lacks one function ALL-IN-1 has had for so much time:oN > MAKE_FILE_NAME. You feed it anything, and it turns it into a valid filename,L > and if it is too strange, it just generates a random filename. You'd thinkM > that every operating system would have a similar function since there is sow8 > much exchange bewteen different file systems nowadays.    ? Unfortunatetly there are also a widespread philosophy: if it is"# not Windows or Unix then ignore it.    :-(    Arne   ------------------------------  % Date: Fri, 27 Dec 2002 11:17:48 -0800s$ From: Shane Smith <ssmith@icius.com>& Subject: RE: vax6k.openecs.org rebirth0 Message-ID: <01C2AD99.A262CA20@sulfer.icius.com>  H The '11 wasn't /allowed/ to compete towards the end. I've seen a pictureF of a multi-cpu PDP/11. It wasn't release (I was told) because it wouldE have eaten into the VAX's market share. Sound familiar? Still, it did9E get to the point where a PDP/11 emulator running on a 32 bit Unix box,H gave you a performance hike over the real thing. Time to stick a fork in it, it was done.   Shane    -----Original Message-----( From: Tom Linden [mailto:tom@kednos.com]' Sent: Friday, December 27, 2002 9:48 AM  To: Info-VAX@Mvb.Saic.Comc& Subject: RE: vax6k.openecs.org rebirth    A Well, there's more to than that.  the '11 wasn't able to compete,oD performance-wise with the 32bit minis like Prime.  I don't know whatC you by "impressive" but I thought it ran very well.  It had a greatiD instruction set, making compiler writing a lot easier the risc.  AndC if Intel can make that miserable accumulator architecture (x86) run 7 at 2.7G who knows what you could have gotten out of VAXb   >-----Original Message-----s. >From: Peter da Silva [mailto:peter@abbnm.com]( >Sent: Friday, December 27, 2002 9:21 AM >To: Info-VAX@Mvb.Saic.Com' >Subject: Re: vax6k.openecs.org rebirthy >n >h0 >In article <6oseua.f5i.ln@via.reistad.priv.no>,- >Morten Reistad  <mrr@reistad.priv.no> wrote:mC >> The VAX was DEC's attempt to repeat the success of the '11, withiG >> a 32-bit SuperCISC architecture. Ran a weird operating system calledmA >> VMS, as well as Unix in variuos flavors. Could make impressivee >performance >> for its CISCyness.  >aI >I don't think the VAX ever had "impressive performance". What it had wasbL >a large address space and legacy support for RSX-11 software via a hardwareL >PDP-11 emulator, and a complex OS that you either love or hate (or love AND >hate).  >e >--tA >I've seen things you people can't imagine. Chimneysweeps on fire  >over the roofsT@ >of London. I've watched kite-strings glitter in the sun at Hyde >Park Gate.  All? >these things will be lost in time, like chalk-paintings in thel >rain.   `-_-'L >Time for your nap.  | Peter da Silva | Har du kramat din varg, idag?    'U` >- >---' >Incoming mail is certified Virus Free.0; >Checked by AVG anti-virus system (http://www.grisoft.com). B >Version: 6.0.431 / Virus Database: 242 - Release Date: 12/17/2002 >D ---a& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.431 / Virus Database: 242 - Release Date: 12/17/2002R   ------------------------------  % Date: Fri, 27 Dec 2002 20:59:20 -0800r' From: pzachary <pzachary@sasquatch.com>.& Subject: Re: vax6k.openecs.org rebirth, Message-ID: <3E0D2FA8.DDE7297@sasquatch.com>   Peter da Silva wrote:? > 1 > In article <6oseua.f5i.ln@via.reistad.priv.no>,l. > Morten Reistad  <mrr@reistad.priv.no> wrote:D > > The VAX was DEC's attempt to repeat the success of the '11, withH > > a 32-bit SuperCISC architecture. Ran a weird operating system calledN > > VMS, as well as Unix in variuos flavors. Could make impressive performance > > for its CISCyness. > J > I don't think the VAX ever had "impressive performance". What it had wasM > a large address space and legacy support for RSX-11 software via a hardwareMM > PDP-11 emulator, and a complex OS that you either love or hate (or love AND. > hate).  H I was impressed by the VAX-11/750, running BSD4.2 with a bunch of RA81s,, as a personal computer, it was quite fast :)   Pavl_o >  > --Q > I've seen things you people can't imagine. Chimneysweeps on fire over the roofs Q > of London. I've watched kite-strings glitter in the sun at Hyde Park Gate.  AllGN > these things will be lost in time, like chalk-paintings in the rain.   `-_-'M > Time for your nap.  | Peter da Silva | Har du kramat din varg, idag?    'U`r   ------------------------------   End of INFO-VAX 2002.717 ************************