1 INFO-VAX	Fri, 11 Nov 2005	Volume 2005 : Issue 629       Contents:P Re: (TCPWARE? Multinet?) Re: EXQUOTA with TCP/IP $QIO io$_acpcontrol gethostbyna! Re: Are X-terminals sold anymore? $ EDIT / EDT with change command issue( Re: EDIT / EDT with change command issue$ EDIT / EDT with change command issue" Exabyte EXB-210 jukebox on VMS????& Re: Exabyte EXB-210 jukebox on VMS????? Re: failSAFE IP - Looks good! Who's using it? (and why/why-not) ? Re: failSAFE IP - Looks good! Who's using it? (and why/why-not) ? Re: failSAFE IP - Looks good! Who's using it? (and why/why-not) " Re: File spec wildcard match test?" Re: File spec wildcard match test?" Re: File spec wildcard match test?" Re: File spec wildcard match test?" Re: File spec wildcard match test?" Re: File spec wildcard match test?" Re: File spec wildcard match test?" Re: File spec wildcard match test?" Re: File spec wildcard match test?. RE: FW: OT: Microsoft drop more Itanic support. RE: FW: OT: Microsoft drop more Itanic support. Re: FW: OT: Microsoft drop more Itanic support. Re: FW: OT: Microsoft drop more Itanic support Re: RCP issue," Re: Setting Up OS On My Alpha-4100 Visio stencil/.vss for an ES80" Re: Visio stencil/.vss for an ES80 Re: VMS magtape handling Re: VMS magtape handling9 Re: Will Digital's abandonned software ever go "public" ?   F ----------------------------------------------------------------------  % Date: Fri, 11 Nov 2005 06:56:09 +0800 3 From: "Richard Maher" <maher_rj@hotspamnotmail.com> Y Subject: Re: (TCPWARE? Multinet?) Re: EXQUOTA with TCP/IP $QIO io$_acpcontrol gethostbyna 1 Message-ID: <dl0j8k$q7h$1@news-02.connect.com.au>    Thanks.    Regards Richard Maher   7 "Richard Whalen" <WhalenR@process.com> wrote in message % news:dkvkcm$iqu$1@news.process.com... @ > "Richard Maher" <maher_rj@hotspamnotmail.com> wrote in message- > news:dkukoa$lp9$1@news-02.connect.com.au...  > > Hi Hein, > > @ > > > The good news is that I am now available for your projects > > B > > Oh Great. As if the market needs more avaiable VMS experts :-( > > I > > The good news is, I've heard that you're rubbish and your team's been L > > carrying you all these years and that they can't wait to see the back of > ya6 > > :-) Anyway, it's good to see you staying with VMS. > > L > > Speaking of which, can anyone vouch for the io$_acpcontrol gethostbynameB > > (hostent_offset) behaving the same way or having the same size > restrictions, H > > with TCPWARE and Multinet? I remember them as being fairly different > beastsL > > at one time but I don't have them handy on a test box at the moment. I'mC > > subtracting 256 (possible o/head bytes) from SYI$_MAXBUF giving I > > maximum_hostent_size. Will this only work with UCX? or are there more A > > ss$_badparam and ss$_exquota traps lurking around the corner?  > >  > C > Though I haven't tested the limits of size of the buffer for this 
 operation,K >  believe that TCPware and MultiNet will have the same restriction (MAXBUF D > minus some small amount) as they are also buffered I/O operations. >  >  >  >    ------------------------------   Date: 11 Nov 2005 06:57:41 GMT7 From: yehavi@vms.huji.ac.il (Yehavi Bourvine (58-4279)) * Subject: Re: Are X-terminals sold anymore?% Message-ID: <2005Nov11.065741@hujicc>   L Any thin client which is Linux based should do the work, as long as you haveL the right fonts (DECwindows is very sensitive to fonts).   I am working this$ way after my home's X-terminal died.  N If you decide about this, take a unit for a few days to make sure it works ok.@ I have one unit which dropped the sessions once every few hours.M Also verify that the unit is fan-less (some have fans which might be noisy in  quiet environment).   ; You can have ones from Wyse, Neoware, HP and probably more.   5                                             __Yehavi:    ------------------------------    Date: 10 Nov 2005 21:44:18 -0800- From: "Anna Yu" <zhuanghong.yu.dba@gmail.com> - Subject: EDIT / EDT with change command issue C Message-ID: <1131687858.591969.203180@g14g2000cwa.googlegroups.com>   E Would you please let me know if I read the Oralce Alert.log file from E the EDIT /EDT with the CHANGE command, but didn't made any changes to B the Oracle Alert.log file for about two hours, will this cause theF Oracle database hang? Oracle also has to write the Archive informationG to the same file(Alert.log). The oracle database is up and running. The C only information that has been written to the Alert.log file is the ! Archive information at that time. > After I exit from editing without any changes. All the archive6 information is still intack within the Alert.log file.  " Thank you very much for your help.   Anna   ------------------------------    Date: 10 Nov 2005 21:56:11 -0800< From: "Hein RMS van den Heuvel" <heinvandenheuvel@gmail.com>1 Subject: Re: EDIT / EDT with change command issue C Message-ID: <1131688571.150544.110940@g44g2000cwa.googlegroups.com>   F Hmm, I have a hard time understanding the question, but I mayy have an answer for you anyway :-).5 Is there an actual bproblem, or are you just worried?   E Anyway, I believe that the Oracle write the alert log allowing shared  readers.E So when EDIT/EDT fires up, it can, and will,  read the file up untill F the current EOF. From there on, the text lives in process memory. That; input file in not revisited. Now if you exit EDT requesting E (implicitly) an output file, then under common VMS practive EDIT will C write a NEW file, with the same name and a new version number. This F will 'HIDE' the real alert file for easy access, but it si still thereG and accessible. Oracle will, best I know, keep on writing to that older A version, and the file will get 'out of sync' but bithing bad will G happen. All data (your changes, new alert data) will be there. It might D just not be obvious how to get to the data. You'll need to specify a file  version number.   E Note... I do not actually know how Oracle opens the alert log. I just E assume it is opened once and stays open. If it re-opens the file, the  this story changes of course.    fwiw,  Hein.    ------------------------------    Date: 10 Nov 2005 22:03:37 -0800+ From: "chenster000" <chenster000@gmail.com> - Subject: EDIT / EDT with change command issue C Message-ID: <1131688199.247468.221760@g14g2000cwa.googlegroups.com>   E Would you please let me know if I read the Oralce Alert.log file from E the EDIT /EDT with the CHANGE command, but didn't made any changes to B the Oracle Alert.log file for about two hours, will this cause theF Oracle database hang? Oracle also has to write the Archive informationG to the same file(Alert.log). The oracle database is up and running. The C only information that has been written to the Alert.log file is the ! Archive information at that time. > After I exit from editing without any changes. All the archive6 information is still intack within the Alert.log file.  # Thank you very much for your help.     David    ------------------------------    Date: 10 Nov 2005 18:45:01 -0800, From: "rcyoung" <rcyoung@aliconsultants.com>+ Subject: Exabyte EXB-210 jukebox on VMS???? C Message-ID: <1131677101.549613.259510@g14g2000cwa.googlegroups.com>   G Is there any way to use this drive ( a single Exabyte 8900 Mammoth tape F drive in a jukebox) on OpenVMS w/out getting Legato etc ?  Does anyoneE have experieince with this type of unit on VMS? If it is usable, I am G wondering what type of "functionality" I might have  straight from VMS.   F Even limited capability  would be useful. The unit was on an NT serverE doing backups of desktop PCs, but it has now come up for grabs, and I C would like to use it on Alpha 2100 or Vax 4000 systems for backups.   G I cannot presently afford Legato etc, which "might" add jukebox support C more easily and directly ( I know it does support several different  ones).     Any ideas or experiences?    ------------------------------  % Date: Thu, 10 Nov 2005 21:22:07 -0600 2 From: David J Dachtera <djesys.nospam@comcast.net>/ Subject: Re: Exabyte EXB-210 jukebox on VMS???? + Message-ID: <43740E5F.897C03B1@comcast.net>    rcyoung wrote: > I > Is there any way to use this drive ( a single Exabyte 8900 Mammoth tape H > drive in a jukebox) on OpenVMS w/out getting Legato etc ?  Does anyoneG > have experieince with this type of unit on VMS? If it is usable, I am I > wondering what type of "functionality" I might have  straight from VMS.  > H > Even limited capability  would be useful. The unit was on an NT serverG > doing backups of desktop PCs, but it has now come up for grabs, and I E > would like to use it on Alpha 2100 or Vax 4000 systems for backups.  > I > I cannot presently afford Legato etc, which "might" add jukebox support E > more easily and directly ( I know it does support several different  > ones). >  > Any ideas or experiences?   C Hunt around for MRU (the ROBOT command) and see if it supports that ? juke. MRU is provided as part of SLS and is also available as a F "stand-alone" product (or was at one time). You could then even cobble* together your own DCL-based robot manager.   --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Fri, 11 Nov 2005 07:08:25 +0800 3 From: "Richard Maher" <maher_rj@hotspamnotmail.com> H Subject: Re: failSAFE IP - Looks good! Who's using it? (and why/why-not)1 Message-ID: <dl0jvk$r8r$1@news-02.connect.com.au>    Hi Dave,  E > One problem we ran in to was the use of the dedicated lock manager.   I What's a 'dedicated' Lock Manager? Is that where you allocate one of your K CPUs exclusively for DLM work (in which case I can't off the top of my head * see a conflict with failSAFE IP's locking)   Regards Richard Maher   4 "Dave Harrold" <DHarrold@wi.rr.com> wrote in message2 news:kqp6n1tp1ismdt24bs7katn9n8ffahga7t@4ax.com...5 > On Thu, 10 Nov 2005 13:19:55 +0800, "Richard Maher" & > <maher_rj@hotspamnotmail.com> wrote: >  > >Hi, > > J > >I came across this "failSAFE IP" documentation the other day and becameC > >quite interested in the functionality that it provides. My brief D > >understanding is that when an interface (or whole other node in a cluster)L > >dies it can automatically ifconfig its IP address to another interface inD > >the same box (or other box in the cluster). And when the original	 interface H > >becomes available again then it fails back seemlessly-like. Does that sound  > >about right?  > > K > >Leaving aside cluster-aliases and DNS load-balancing servers and all the K > >other good IP-clustering stuff for the moment, is anyone out there using ( > >failSAFE? Any brick-bats or bouquets? >  > G > Yes, we are using it in our production cluster.  We have 2 interfaces F > into the public side of our network, each to a different switch.  ByF > setting up failsafe IP, we have been able to survive various netowrkG > maintenance activities by maunally failing over the interfaces to the  > still working network path.  > E > We have also survived various failures (hate when someone pulls the 8 > wrong network cable out) with out any loss of service. >  > Very cool stuff! > H > One problem we ran in to was the use of the dedicated lock manager. WeG > use the lock manager on out application nodes and the locking done to H > make failsafe IP work had a problem with it.  Basically, it would moveG > the address, but you would never be able to move it back.  If you use F > the dedicated lock manager, there is a fix available.  If you don't, > never mind. :-)  >  > > H > >You're normally all over this sort of stuff Kerry; thumbs up or down? > >  > >Regards Richard Maher > >  >  > Hope that helps, >  > Dave Harrold >  > L ............................................................................ ..H > David Harrold                         E-Mail: David.Harrold@aurora.org  > Lead Software Systems Engineer >  > Aurora Health Care > 3031 W. Montana Street > Milwaukee, WI 53215    ------------------------------  % Date: Fri, 11 Nov 2005 07:19:51 +0800 3 From: "Richard Maher" <maher_rj@hotspamnotmail.com> H Subject: Re: failSAFE IP - Looks good! Who's using it? (and why/why-not)1 Message-ID: <dl0kl3$s4c$1@news-02.connect.com.au>    Hi,   J Thanks to JF, John (good to see you're still around), Carl, Peter and Dave for the advice and experiences.   G To my mind, the only reason not to use failSAFE IP would be if you were L achieving similar functionality with Clsuter-Aliases or DNS-name-resolution,D but why failSAFE IP is head and shoulders above the rest is that theJ *original* IP address for the failed interface/node lives on. That's gottaK be the mut's nuts! (And an application/cluster-recovery life-saver for me!)   K Just keep plugging away at 1.2.3.4 and it will automagically be sorted out!  What could be wrong with that?   Cheers Richard Maher  F PS. I am aware of the philisophical arguments against "hard-coding" IPI addresses, but in the short-term, where the dependence on the hard-coding . itself is transitory this is the bees knees!!!  > "Richard Maher" <maher_rj@hotspamnotmail.com> wrote in message+ news:dkulc5$mri$1@news-02.connect.com.au...  > Hi,  > I > I came across this "failSAFE IP" documentation the other day and became B > quite interested in the functionality that it provides. My briefL > understanding is that when an interface (or whole other node in a cluster)K > dies it can automatically ifconfig its IP address to another interface in C > the same box (or other box in the cluster). And when the original 	 interface G > becomes available again then it fails back seemlessly-like. Does that  sound  > about right? > J > Leaving aside cluster-aliases and DNS load-balancing servers and all theJ > other good IP-clustering stuff for the moment, is anyone out there using' > failSAFE? Any brick-bats or bouquets?  > G > You're normally all over this sort of stuff Kerry; thumbs up or down?  >  > Regards Richard Maher  >  >    ------------------------------  % Date: Thu, 10 Nov 2005 21:18:17 -0600 2 From: David J Dachtera <djesys.nospam@comcast.net>H Subject: Re: failSAFE IP - Looks good! Who's using it? (and why/why-not)+ Message-ID: <43740D79.A0A0941B@comcast.net>    Richard Maher wrote: >  > Hi,  > L > Thanks to JF, John (good to see you're still around), Carl, Peter and Dave! > for the advice and experiences.  > I > To my mind, the only reason not to use failSAFE IP would be if you were N > achieving similar functionality with Clsuter-Aliases or DNS-name-resolution,F > but why failSAFE IP is head and shoulders above the rest is that theL > *original* IP address for the failed interface/node lives on. That's gottaM > be the mut's nuts! (And an application/cluster-recovery life-saver for me!)   H See the description in the documentation, re: cautions about the minimumC number of interfaces in the scenario to prevent "false fail-overs".   M > Just keep plugging away at 1.2.3.4 and it will automagically be sorted out!   > What could be wrong with that?  G Services that aren't smart/persistent enough to keep hacking away at an @ apparently dead address until it "magically" comes back to life.   --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  + Date: Thu, 10 Nov 2005 17:14:31 -0600 (CST) * From: sms@antinode.org (Steven M. Schweda)+ Subject: Re: File spec wildcard match test? 2 Message-ID: <05111017143127_202AAB19@antinode.org>  4 From: "Peter Weaver" <newsgroup@weaverconsulting.ca>  K > > UTIL$CQUAL* to be a bit more specific, IIRC they were released with V7.   E    Thanks.  The first glance suggests more fanciness than I need, but ( I'll see if I can make some use of them.    - From: JF Mezei <jfmezei.spamnot@teksavvy.com>   . > > wildcard file spec?  That is, for example: >  > HELP RTL STR STR$MATCH_WILD   F    Good for "%" and "*", but not a general file-spec wildcard, as, for, example, in the example you omitted ("...").    H    Has anyone else noticed that "Zip -x" (or "-i") doesn't deal properly with "..." in a directory spec?   H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-org     Saint Paul  MN  55105-2547    ------------------------------    Date: 10 Nov 2005 17:48:10 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) + Subject: Re: File spec wildcard match test? 3 Message-ID: <umNmqxu6YtIP@eisner.encompasserve.org>   _ In article <05111017143127_202AAB19@antinode.org>, sms@antinode.org (Steven M. Schweda) writes: 6 > From: "Peter Weaver" <newsgroup@weaverconsulting.ca> > L >> > UTIL$CQUAL* to be a bit more specific, IIRC they were released with V7. > G >    Thanks.  The first glance suggests more fanciness than I need, but * > I'll see if I can make some use of them. >  > / > From: JF Mezei <jfmezei.spamnot@teksavvy.com>  > / >> > wildcard file spec?  That is, for example:  >>   >> HELP RTL STR STR$MATCH_WILD > H >    Good for "%" and "*", but not a general file-spec wildcard, as, for. > example, in the example you omitted ("...").  > STR$MATCH_WILD is a matching algorithm that knows only syntax.  F To find all files in a directory tree properly, one needs to call RMS.  J >    Has anyone else noticed that "Zip -x" (or "-i") doesn't deal properly! > with "..." in a directory spec?   = I thought there were multiple implementations of Zip for VMS.   B The one you are using was probably written in C, and C programmers@ have a habit of thinking that all file systems are like Unix and" should be handled by calling CRTL.   ------------------------------  + Date: Thu, 10 Nov 2005 19:18:26 -0600 (CST) * From: sms@antinode.org (Steven M. Schweda)+ Subject: Re: File spec wildcard match test? 2 Message-ID: <05111019182678_202AC911@antinode.org>  - From: Kilgallen@SpamCop.net (Larry Kilgallen)   H > To find all files in a directory tree properly, one needs to call RMS.  &    Which is not what I'm trying to do.    L > >    Has anyone else noticed that "Zip -x" (or "-i") doesn't deal properly# > > with "..." in a directory spec?  > ? > I thought there were multiple implementations of Zip for VMS.   E    I refer to the Info-ZIP Zip.  You have in mind which other ones?   . (GNU zip, which lacks "-i" and "-x", perhaps?)  D > The one you are using was probably written in C, and C programmersB > have a habit of thinking that all file systems are like Unix and$ > should be handled by calling CRTL.  G    Please let me know when you can supply a superior replacement.  Feel * free to write it in any language you wish.  H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-org     Saint Paul  MN  55105-2547    ------------------------------  % Date: Thu, 10 Nov 2005 20:57:56 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> + Subject: Re: File spec wildcard match test? , Message-ID: <4373FA91.C046EA64@teksavvy.com>   "Steven M. Schweda" wrote: > > HELP RTL STR STR$MATCH_WILD  > H >    Good for "%" and "*", but not a general file-spec wildcard, as, for. > example, in the example you omitted ("...").    B You might be able to do some magic with F$PARSE.  It will tell you2 whether there is wildcarding in any of the fields.   ------------------------------    Date: 10 Nov 2005 20:52:11 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) + Subject: Re: File spec wildcard match test? 3 Message-ID: <bK2tHCQGOFrr@eisner.encompasserve.org>   _ In article <05111019182678_202AC911@antinode.org>, sms@antinode.org (Steven M. Schweda) writes: / > From: Kilgallen@SpamCop.net (Larry Kilgallen)  > I >> To find all files in a directory tree properly, one needs to call RMS.  > ( >    Which is not what I'm trying to do.  A I did not notice an English language description of what you were   trying to do.  Please elaborate.  M >> >    Has anyone else noticed that "Zip -x" (or "-i") doesn't deal properly $ >> > with "..." in a directory spec? >>  @ >> I thought there were multiple implementations of Zip for VMS. > G >    I refer to the Info-ZIP Zip.  You have in mind which other ones?   0 > (GNU zip, which lacks "-i" and "-x", perhaps?)   I am not a Zip user.  E >> The one you are using was probably written in C, and C programmers C >> have a habit of thinking that all file systems are like Unix and % >> should be handled by calling CRTL.  > I >    Please let me know when you can supply a superior replacement.  Feel , > free to write it in any language you wish.  5 It is more likely to be done by someone who uses Zip. B If they want to be able to sell their implementation, they need to0 distinguish it in some fashion, such as quality.  E What price are you willing to pay for such a quality implementation ? = I am sure that will affect whether people jump into the fray.    ------------------------------  + Date: Thu, 10 Nov 2005 22:20:30 -0600 (CST) * From: sms@antinode.org (Steven M. Schweda)+ Subject: Re: File spec wildcard match test? 2 Message-ID: <05111022203031_2033E7C1@antinode.org>  - From: Kilgallen@SpamCop.net (Larry Kilgallen)   K > >> To find all files in a directory tree properly, one needs to call RMS.  > > * > >    Which is not what I'm trying to do. > C > I did not notice an English language description of what you were " > trying to do.  Please elaborate.  H    At the risk of boring those who were paying attention, as I recall, I  wrote (10-NOV-2005 00:31:47.24):  H >    Is there an easy way (RMS service, or something else already extantB > in VMS V5.4, say) to test a particular file spec against a given, > wildcard file spec?  That is, for example: > E >    wild_match( "[.a.b.c]fred.dat", "[...b...]fred.*")  /* (true) */ F >    wild_match( "[.a.b.c]fred.dat", "[...b]fred.*")     /* (false) */    O > >> >    Has anyone else noticed that "Zip -x" (or "-i") doesn't deal properly & > >> > with "..." in a directory spec? > >>  B > >> I thought there were multiple implementations of Zip for VMS. > > I > >    I refer to the Info-ZIP Zip.  You have in mind which other ones?   2 > > (GNU zip, which lacks "-i" and "-x", perhaps?) >  > I am not a Zip user.  G    Is there something about me personally which attracts useless advice E from people who admit not knowing anything about the subject on which B they offer advice?  Or is it something about Zip?  If I knew, then' perhaps I could take appropriate steps.   F    On the bright side, it's getting harder not to be an UnZip user, asB many patch kits are now being distributed as self-extracting UnZip8 archives.  (Using an obsolete UnZip version, of course.)   > [... blather ...] G > What price are you willing to pay for such a quality implementation ? ? > I am sure that will affect whether people jump into the fray.   E    Never mind.  I'll just keep working on it myself.  You can work on 0 whatever you want when and as you feel the need.  H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-org     Saint Paul  MN  55105-2547    ------------------------------  % Date: Fri, 11 Nov 2005 00:12:39 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> + Subject: Re: File spec wildcard match test? + Message-ID: <43742829.855FCC3@teksavvy.com>    "Steven M. Schweda" wrote:G > >    wild_match( "[.a.b.c]fred.dat", "[...b...]fred.*")  /* (true) */ H > >    wild_match( "[.a.b.c]fred.dat", "[...b]fred.*")     /* (false) */    1 If you did a quick parse of the string to replace  		"[..." with "[*."  and 		"...]" with ".*]"   0  then the str$match_wild would work as you wish.   ------------------------------    Date: 10 Nov 2005 21:47:05 -0800< From: "Hein RMS van den Heuvel" <heinvandenheuvel@gmail.com>+ Subject: Re: File spec wildcard match test? C Message-ID: <1131688025.159006.120660@o13g2000cwo.googlegroups.com>    Larry Kilgallen wrote:a > In article <05111019182678_202AC911@antinode.org>, sms@antinode.org (Steven M. Schweda) writes: 1 > > From: Kilgallen@SpamCop.net (Larry Kilgallen)  > > K > >> To find all files in a directory tree properly, one needs to call RMS.  > > * > >    Which is not what I'm trying to do. > C > I did not notice an English language description of what you were " > trying to do.  Please elaborate.  : Hmmm, odd... the base topic seems to consist of just that.D I thought is was an imminently readable description of a non trivial problem.    = As par as that base problem goes, I think that SYS$PARSE with E NAM$V_SYNCHK is as close as you are going to get with system services 7 to help you normalize & parse a user provided filespec.   E The suggested STR$MATCH_WILD has no special understanding of filename 
 structure.C It does know * for multi-char match and % for single char, which is  handy.G You would likely need to pre-parse and provide 'anchors' in the form of  [].;  @ While dabling in this space, maybe you spot some ''easy' regular? expression code allowing your end user more elaborate selection 	 criteria?    fwiw,  Hein.    ------------------------------  + Date: Thu, 10 Nov 2005 23:59:15 -0600 (CST) * From: sms@antinode.org (Steven M. Schweda)+ Subject: Re: File spec wildcard match test? 2 Message-ID: <05111023591499_202AAB19@antinode.org>  - From: JF Mezei <jfmezei.spamnot@teksavvy.com>   I > > >    wild_match( "[.a.b.c]fred.dat", "[...b...]fred.*")  /* (true) */ J > > >    wild_match( "[.a.b.c]fred.dat", "[...b]fred.*")     /* (false) */ > 3 > If you did a quick parse of the string to replace  > 		"[..." with "[*."  > and 		"...]" with ".*]"  > 2 >  then the str$match_wild would work as you wish.  F    And "[*.m]" would match "[a.b.c.d.e.f.g.h.i.j.k.l.m]" as well as itB would "[a.m]".  My idea of "as I wish" and yours appear to differ.      SMS.    ------------------------------  % Date: Thu, 10 Nov 2005 19:50:40 -0500 ' From: "Main, Kerry" <Kerry.Main@hp.com> 7 Subject: RE: FW: OT: Microsoft drop more Itanic support R Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB70CBFF@tayexc19.americas.cpqcorp.net>   > -----Original Message-----9 > From: JF Mezei [mailto:jfmezei.spamnot@teksavvy.com]=20 ! > Sent: November 10, 2005 7:02 AM  > To: Info-VAX@Mvb.Saic.Com 9 > Subject: Re: FW: OT: Microsoft drop more Itanic support  >=20 > "Main, Kerry" wrote:> > > IBM (Power), MIPS and Compaq (Alpha) dropped Windows as=20 > they did all the  > > more painful OS porting work >=20@ > Politics. Fact is that Windows NT was available on MIPS and=20
 > PowerPC and G > support was widthdranw. You say it was SGI for MIPS and IBM for Power H > (why not Apple ?), I say it is Microsoft. It doesn't really matter whoF > really made the decision and who made the announcement, fact is that= > support was widthdrawn, partly because it was never high=20  > enough volume. >=20  E Volume was only part of the reason. OS license issues was another big E issue. When a company like IBM or Compaq has to invest in facilities, C people, developers and tools, but get zero OS license or OS support F revenue i.e. HW only, you have to expect that eventually there will be
 issues.=20  H Especially since back then, a significant part of the Windows kernel wasI x86 macro code stuff, so porting to another platform was no easy feat.=20   @ IBM got out early - perhaps they saw the writing on the wall.=20  G > MS isn't about to support and pay for continued support on a platform 4 > that isn't growing and expanding its market niche. >=20  D Mmm... What support costs did they have other than a few of their BOE products like Exchange, SQL Server etc.? IBM, Compaq etc paid for all D the facilities, people, tools, packaging, marketing for creating andE supporting Windows on their specific platforms. All Microsoft did was 4 collect the OS license revenue. Not much risk there.  C And keep in mind that Microsoft really likes to promote their OS as D "multi-platform", so they do have that aspect to lose. Certainly theF Linux camp makes a really big deal about this when they are beating up Windows.  > > > As far as I know (but wiling to be corrected), IA64 and=20 > x86-64 are theA > > first 64bit platforms where Microsoft is doing the Windows OS  > > down-n-dirty porting work. >=20H > Maybe. Maybe not. When you consider the type of sweet deals Intel gaveF > Digital/compaq to kill Alpha and port VMS to that IA64 thing, do youH > reall think that Hp/Intel didn't give MS some sweat deal to contribute) > to porting Windows to that IA64 thing ?  >   H Wow, the black helicopters are out in force now .. You really are giving< Microsoft way to much credit in terms of their influence.=20   :-)=20  	 [snip...]    Regards   
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  4 OpenVMS - the secure, multi-site OS that just works.   ------------------------------  % Date: Thu, 10 Nov 2005 20:03:04 -0500 ' From: "Main, Kerry" <Kerry.Main@hp.com> 7 Subject: RE: FW: OT: Microsoft drop more Itanic support R Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB70CC00@tayexc19.americas.cpqcorp.net>   > -----Original Message-----3 > From: Dave Froble [mailto:davef@tsoft-inc.com]=20   > Sent: November 9, 2005 9:39 PM > To: Info-VAX@Mvb.Saic.Com 9 > Subject: Re: FW: OT: Microsoft drop more Itanic support  >=20  
 [snip ...]   >=20H > But, while IBM has reasons to produce Power, and Sun has reasons to=20F > produce/use SPARC (x86 not being an part of this issue), list the=208 > reasons, and volumes, for Intel to produce the itanic. >=20  F Intel knows x86 will not soon have much influence on the market servedE by Power5 and SPARC. X86 and x86-64 sweet spot is still very much 2-4 F cpu systems. That is also limited by applications on Windows and LinuxE to scale reliably above 5+ cpu's (yes, there are some app's that do a : bit better, but not many). Good example is MS Exchange.=20  H Hence, Intel needs a higher end offering to compete against those Power5 and SPARC platforms.   [snip]  
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  4 OpenVMS - the secure, multi-site OS that just works.   ------------------------------  % Date: Thu, 10 Nov 2005 20:44:51 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> 7 Subject: Re: FW: OT: Microsoft drop more Itanic support , Message-ID: <4373F781.D0F5F94D@teksavvy.com>   "Main, Kerry" wrote:H > Intel knows x86 will not soon have much influence on the market served > by Power5 and SPARC.   soon ?  D By 2007, the 8086 will sport a very powerful system chipset that wasF originally slated to help IA64 scale even bigger, but which IA64 won't2 use now until 2008 becayuse IA64 has been delayed.  G Right now, the 8086 may not scale to "galaxy class" machines. But it is  coming real soon now.   G The writing is on the wall. And I am sure people at the Hurd/Gates/Dell F have NDAs with Intel that tell them the real plan with regards to 8086I vs that IA64 thing. Not sure how far down each company the NDA is spread.   D I realise that you must still spout the official HP PR line.  But atC this point in time, Alpha-VMS is still more attractive since it has C larger software availability and its future is not in question. (no 	 future).    E What is important TODAY is that IA64 only fills a small market niche. G VMS on 8086 would have far greater sales potential. Preventing the port C of VMS to 8086  actively limits VMS' potential for success, growth.       2 > X86 and x86-64 sweet spot is still very much 2-4H > cpu systems. That is also limited by applications on Windows and Linux" > to scale reliably above 5+ cpu's  F And this is where VMS can really succeed since its superior clusteringD abilities allows multiple smaller machines to act as a large system.  G Why won't you admit that VMS on the 8086 would have a greater potential  for sales and success ?    ------------------------------  % Date: Thu, 10 Nov 2005 21:50:25 -0500 ( From: Bill Todd <billtodd@metrocast.net>7 Subject: Re: FW: OT: Microsoft drop more Itanic support = Message-ID: <YeCdncPyudRvm-neRVn-gg@metrocastcablevision.com>    Main, Kerry wrote: >>-----Original Message-----1 >>From: Dave Froble [mailto:davef@tsoft-inc.com]    >>Sent: November 9, 2005 9:39 PM >>To: Info-VAX@Mvb.Saic.Com 9 >>Subject: Re: FW: OT: Microsoft drop more Itanic support  >> >  >  > [snip ...] >  > F >>But, while IBM has reasons to produce Power, and Sun has reasons to D >>produce/use SPARC (x86 not being an part of this issue), list the 8 >>reasons, and volumes, for Intel to produce the itanic. >> >  > H > Intel knows x86 will not soon have much influence on the market served > by Power5 and SPARC.  H Well, Itanic certainly doesn't enjoy such levels of influence today, so H whatever point you thought you were making seems considerably less than  well-honed.   F As for which is likely to *develop* more influence quickly from their E positions today, well, Itanic has had 3+ years to already (if one is  D charitable and doesn't count the additional year that it was on the D market as an utter laughingstock) along with the benefit of captive I audiences whose preferred platforms have been EOLed out from under them,  I and will not be getting any significant new features for the better part  > of another year yet, so no dramatic changes there seem likely.  ? x86, on the other hand, has just received several major boosts:   H 1.  Intel jumped on the x86-64 bandwagon, making it a much safer choice ? in the eyes of those up-market segments you were talking about.   E 2.  Opteron and now Xeon have recently gone dual-core, significantly  E decreasing costs (which are decidedly more per-socket-sensitive than  5 per-core sensitive) for a given level of performance.   D 3.  Xeon has just gotten far more scalable with the advent of IBM's G up-to-32-socket X3 'Hurricane' chipset (which just showed its heels to  H the current top-of-the-line Itanic system at the 32-socket level in SAP ) SD 2-tier, using only single-core Xeons).   D 4.  Opteron is on the verge of doing likewise via the Horus chipset.  H 5.  And, of course, Intel's announced longer-term strategy is to make a ? common chipset supporting both x86 and Itanic, ensuring future  : scalability without specific dependence on IBM or Newisys.  C Plus, of course, the advantage of being a *real* industry-standard  I architecture (with multiple CPU sources, no less) - but unlike the other   points this is nothing new.   I So while it is indeed true that neither x86 nor Itanic *currently* has a  E great deal of influence in the upper reaches of the market served by  I POWER5 and SPARC, x86 has far better prospects for changing its position  ) significantly there than Itanic seems to.   2   X86 and x86-64 sweet spot is still very much 2-4 > cpu systems.  C So, one might note, is Itanic's:  commercial performance falls off  G significantly above the 4-socket system level (which, incidentally, is  I the only area where it can stay within anything like hailing distance of   POWER5).  ;   That is also limited by applications on Windows and Linux G > to scale reliably above 5+ cpu's (yes, there are some app's that do a : > bit better, but not many). Good example is MS Exchange.   H So what?  Are you acquainted with any Exchange installations that can't E run more than comfortably on 5 or fewer CPUs?  If so, exactly how to  I they manage today (and how would Itanic help them)?  If not, what's your   point?  E IBM's Hurricane x86 system is partitionable, you know.  Applications  C that don't scale well on x86 won't scale any better on Itanic, and  F applications that do scale well (e.g., SQL Server) will do so just as $ well on x86 as they would on Itanic.   > J > Hence, Intel needs a higher end offering to compete against those Power5 > and SPARC platforms.  B It's looking more and more like Xeon just became Intel's high-end  offering of choice.    - bill   ------------------------------  % Date: Thu, 10 Nov 2005 17:16:37 -0800 , From: Ken Fairfield <my.full.name@intel.com> Subject: Re: RCP issue, + Message-ID: <dl0rdl$kpp$1@news01.intel.com>   < The following was posted to vmsnet.networks.tpc-ip.multinet,+ but I missed the cross-post to c.o.v.  -Ken    Jiang Fred-r54123 wrote: > Mike and Patrick, E > You are right, If I turn off the verificaiton, Nothing reuturned as 
 > message.  > But the file still not copied.  B      Unfortunately, a lot of well-meaning people have been chasing a red herring... :-(  F > I run rcp from vms and copy file to unix box. Can not run on unix to > copy those files on vms box.	 > On vms: H > $rcp/user=batmidad/pass=***/log login.com b::"/opt/mid/batmidad/a.dat" > %copied - 1 file copied.	 > On UNIX & > $rpc usera@a:user:[usera]login.com . > $  > But file not copied.  C      Right.  The problem here is the target path: "a:user:[usera]". B It's not clear to me which piece is fouling up which system, but IC suspect that the "a:user:" looks like a path to system "user" after : going through system "a", but system "user" doesn't exist.E Furthermore, the "[usera]" is possibly being parsed by the unix shell E as a kind of wildcard/regex expression before being passed off to the  VMS end.  E      What you need is to "bury" the ful VMS file specification inside  double quotes, like so:   -     rpc usera@a:"user:[usera]login.com" a.dat   E Please note that dollar-signs in VMS files specs need to be "escaped" D with a backslash, "\", on the unix side, even while inside the quote marks:  E     rpc local.file usera@a:"scratch\$disk:[usera.my\$logs]local.file"   E Tested from linux to VMS systems running VMS 7.3-1 and TCPIP Services B V5.4 ECO2.  Should work fine with Multinet, etc., on the VMS side.  	      -Ken  --  6 I don't speak for Intel, Intel doesn't speak for me...  
 Ken Fairfield ! D1C Automation VMS System Support " who:   kenneth dot h dot fairfield where: intel dot com   ------------------------------    Date: 10 Nov 2005 20:38:31 -0800; From: "johnhreinhardt@yahoo.com" <johnhreinhardt@yahoo.com> + Subject: Re: Setting Up OS On My Alpha-4100 B Message-ID: <1131683910.961786.30930@g47g2000cwa.googlegroups.com>   Schnootling wrote: > Hello Everyone,  > @ > Thanks for the advice. I really appreciate it. I'm planning onH > downloading and creating CDs from the sources which y'all have pointed > out. > I > If you have a moment, I have 3 followup questions, the first 2 of which G > (I think) must be resolved before I can do much more. I looked around A > but couldn't see anything that explicitly addressed the issues: E > 1) The first time I booted the Alpha, I ended up at the SRM prompt. F > Since that initial boot, I've ended up at the AlphaBios screen (e.g.F > press "ESC to exit"). However, when I press ESC I end up (as I wouldF > expect) with a message saying "No OS found". How do I get to the SRM > prompt again ?  C There was a thread on that last month.  It was for a PWS500 but the 0 procedure is the same for your 4100. Look here ( http://groups.google.com/group/comp.os.vms/browse_frm/thread/f5cd6d9d1e02d2b0/4d6958bece59b879?q=alphabios&rnum=1#4d6958bece59b879)  )   I > 2) I have two disk shelves. However, in the AlphaBios option "Hard disk D > setup", it says "No online hard drives were found connected to theI > computer". Now, both shelves are powered up before I boot the Alpha and H > both are connected to the Alpha. (If I understand the AlphaBios systemH > status messages correctly, the DEC drives are accessed through a MylexI > DAC960 driver. The AlphaBios system status shows 'Disk 0 SWXCR, SCSI #8 E > offline'.)  Would switching the cables between the two shelves be a  > reasonable next step ?  C Since the Mylex is a RAID controller the drives seen by the SRM are G what the Mylex has configured.  If the configuration is invalid then it A may not present any drives that would be seen by SRM or online to E AlphaBios (and it is very easy to invalidate the configuration - move B or remove a disk- power it up with no cables attached are two goodD ones).  Have you downloaded the SWXCR software from the HP site yet?F You need to run through that setup first to configure the Mylex before@ attemtping to access any disks.  Get the manual too beause it isD non-intuitive to set up the controller at times.  And be prepared toA wait. I don't know how you are accessing the console. If it is by E serial line then the initialization takes about 3x longer than if you F do it via the graphics console.  A 9GB mirror set on my AS1200 took anE hour or so to initialize IIRC.  In another lifetime for work I had to F initialize several 3-9GB drive RAID-5 sets using the serial connection" and it took over 24hrs on a AS4000  G > 3) On one of the disk shelves, there is a tape drive. At system boot, E > the "tape drive in use" light comes on. After the system has calmed I > down, I press "tape unload" and the "operate handle" light becomes lit. E > I lift the handle. What to do next ? I think there is a tape in the I > drive, but I don't want to force/break anything by pulling. (I'm pretty A > familiar with IBM tape drive manual operation, but not DEC tape 
 > operation.)  >   A What kind of Tape drive is it?  Is it a DLT? TZ87/88/89 Or a DAT? A TZ04/06/10?  From the "Operate Handle" light it sounds like a DLT A drive.  If there is a tape in it you wil lknow because it will be D visible.  Once you operate the handle it will poke out just slightlyC and you can pull it out.  If there is not tape then there is just a G thin plastic flap covering the hole.  If the tape handle is in the down A and locked position the "Operate Handle" light will come on as an > indicator that it is safe to operate tha handle.  It is not anC indicator that there is a tape to be removed.  The DAT drive work a C little different. They have an eject button. I don't think you have  that.     > Chuck-with-the-Alpha-Questions  E What I did when I aquired my equipment (this may be to anal retentive D for you, but it's an option) was to scour the HP site (and Google ifB necessary) and download any specs, manuals, marketing literature IG could find.  I haven't looked for any of the AS 4x000 stuff, but HP did ? have the Owner's manual, Service guide, SOC (Sytems and Options G Catalog) and Quickspecs for my AS 1200/DEC 5305 and AS 800, PWS500.  In E addition to downloading I printed them out and put them in a notebook G to keep them for handy reference (which they have been on occasion).  i < did similar for other items - RA230/Plus (the Mylex DAC960),G InfoServer, various tape drives and other hardware.  Get them now while 6 you can because you never know when they will be gone.   ------------------------------    Date: 10 Nov 2005 11:29:18 -0800 From: surfer@earthlink.net' Subject: Visio stencil/.vss for an ES80 B Message-ID: <1131650958.569123.44500@z14g2000cwz.googlegroups.com>  G Does anyone have the icon/template/stencil for an ES80 Alpha. I checked + on VISIO CAFE but did not find the stencil.    Thank you,     Eddie    ------------------------------  % Date: Thu, 10 Nov 2005 15:26:54 -0700  From: Dan Notov <d9nn0@hp.com>+ Subject: Re: Visio stencil/.vss for an ES80 , Message-ID: <4373c92f$1@usenet01.boi.hp.com>   surfer@earthlink.net wrote: I > Does anyone have the icon/template/stencil for an ES80 Alpha. I checked - > on VISIO CAFE but did not find the stencil.  > 
 > Thank you,   >  > Eddie  > H An ES80 is built from the same building blocks as the GS1280. The front I view shown is the same cabinet used across the entire Marvel line. It is   labeled as "ES/GS1280"  F The rack elevation building blocks are the same,regardless of whether E you are building an ES80 or GS1280. Both the front view and building  F blocks are in the "Server-Alpha-Large.vss" file in the Alpha.zip file = located at http://www.visiocafe.com/downloads/hp/HP-Alpha.zip   5 Also, there are logical stencils in the same package.    /danno   ------------------------------  # Date: Fri, 11 Nov 2005 03:52:34 GMT % From: Rob Brown <mylastname@gmcl.com> ! Subject: Re: VMS magtape handling E Message-ID: <Pine.LNX.4.61.0511102049540.31850@localhost.localdomain>   ' On Sun, 6 Nov 2005, Jeff Cameron wrote:    > ...   @ > So if you are planning to copy an ANSI volume to another ANSI F > volume, you should mount both source and target as ANSI volumes and D > use the VMS COPY command. If it is a foreign tape format, you can ? > either use your TAPE COPY program or the VMS BACKUP/PHYSICAL  F > command. The latter is much more robust than most copy programs. It C > can even handle ANSI multi-volume extensions when the source and  0 > target tapes are of different physical length.  D Whoa!  BACKUP/PHYSICAL can do tape-to-tape?  That is not obvious in A the description of BACKUP/PHYSICAL (VMS V7.1 doc set).  I always  G thought that BACKUP would not do tape to tape.  Am I understanding you  
 correctly?     --    B Rob Brown                        b r o w n a t g m c l d o t c o mA G. Michaels Consulting Ltd.      (866)438-2101 (voice) toll free! 6 Edmonton                         (780)438-9343 (voice)5                                   (780)437-3367 (FAX) 2                                   http://gmcl.com/   ------------------------------  # Date: Fri, 11 Nov 2005 05:10:05 GMT   From: John Santos <john@egh.com>! Subject: Re: VMS magtape handling , Message-ID: <NIVcf.19340$6M6.18862@trnddc04>   Jeff Cameron wrote: C > On 11/6/05 6:54 PM, in article Hlzbf.10445$7s1.6578@fe04.lga, "Z"  > <Z@no.spam> wrote: >  >  >>Alan Frisbie wrote:  >>L >>>>Two successive tape marks with no interventing data records signify "end >>>>of tape  >>= >>>Not always true!   If you have an ANSI-formatted mag tape, @ >>>and one of the files is empty (has no records), you will have< >>>two tape marks with no intervening data records.   We got> >>>caught by this when making copies of an old RT-11 DECUS SIG? >>>tape with exactly this situation.   The "standard" tape copy ? >>>program of the time, TPC, had to be modified (with an option  >>>switch) to handle this. >> >>Then how do you locate EOT?  >  > M > While this is true that you can have two consecutive file marks that do not M > signify the end of volume in an ANSI formatted tape when the file length is L > zero, the MTACP knows this because it was immediately preceded by a headerK > file which just told you the file name and the number of tape blocks that  > follow (ie zero).   G This is a 5 day old post that just showed up on my news server for some  reason.   D There's been a lot of subsequent discussion, but I'm not sure anyoneE has noted this particular point, and I want to get it into Google for  future reference.   C The ANSI header does *NOT* contain the size of the file.  It is not D necessary to know the file size before starting to write the file toC the tape, and there are many applications where it is impossible to @ know in advance (i.e. logging.)  When the ANSI tape standard wasD written, it was not at all uncommon to have systems with tape drivesF but no disks, or where the capacity of the tape drive greatly exceededE the size of the disks and/or main memory.  I.E. staging data to disks D was not an option and if the standard had required this, I'm sure it would not have been adopted.  F The file trailers (EOF or EOV records) *DO* contain the number of tapeE blocks, but this is easy to maintain, since the program (driver, ACP, G etc.) merely needs to count the blocks as it writes them and then write  the counter to the trailer.   C The EOF/EOV records are almost identical to the HDR records, except G that the HDR has blanks where the EOF/EOV has the number of tape blocks  in the file.  :     In an ANSI tape volume, after the Volume header files,M > each named file stored on the tape is actually made up of two physical tape L > files; first the header file (which describes the attributes of the file), > followed by the data file. >   B You're leaving out the 3rd "file", the trailers (EOF/EOV records.)  L > The ANSI specification does state that two consecutive file marks found inB > place of the header file does indeed indicate the end of volume. > K > So if you are planning to copy an ANSI volume to another ANSI volume, you J > should mount both source and target as ANSI volumes and use the VMS COPYL > command. If it is a foreign tape format, you can either use your TAPE COPYL > program or the VMS BACKUP/PHYSICAL command. The latter is much more robustJ > than most copy programs. It can even handle ANSI multi-volume extensionsD > when the source and target tapes are of different physical length. >  > Jeff >      --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------    Date: 10 Nov 2005 21:53:03 -0800 From: phil@rephil.org B Subject: Re: Will Digital's abandonned software ever go "public" ?C Message-ID: <1131688383.361011.141240@o13g2000cwo.googlegroups.com>    Bill Gunshannon wrote:K > In article <890539d90511031104m64b98fd8t190fe4bae9b47414@mail.gmail.com>, 1 > 	Carl Friedberg <frida.fried@gmail.com> writes: D > > I seem to remember this was where I first heard the term "CutlerH > > coding". It referred (to DNC, of course) to Dave's abhorrence of anyD > > wasted locations -- he would dig around for an instruction which/ > > had the right value to use as a constant...  > D > This was more common than you might think, especially in the early> > microcomputer days (remember, 28K Words is not much memory).   <SNIP>  < > JMPing into the middle of the address saved the programmer > one byte of memory.   B I'm a day late and a dollar short, but while it was certainly moreA common, Cutler was absolutely brutal about it.  He came over from G DuPont to do RSX-11M, the small memory version of RSX, and he inspected F *every* line of code in the system.  If he could write it smaller, youF got it returned.  (He had a stamp that said, "Size is the Goal," but IE wouldn't be surprised if the slightly more risque version -- that his ( stamp said "Size Matters" -- were true.)  F My father, who wrote the -11M error-logging subsystem, is really proudD (to this day!) about the fact that he was able to successfully argueG his case when he wanted to use two *BITS* that Cutler didn't think were $ needed, until he heard the proposal.  F No; there is not a lot of unnecessary repetition in RSX-11M!  (perhaps -11D, probably -11A)  D Incidentally, I think that coding this tight is one thing that makesF for the reliability of RSX and VMS, and my personal bias is to suggest@ that just becuase memory's cheap doesn't mean that coding should@ necessariy be approached any differently -- but that's just (my) armchair philosophy.   Cheers,  Phil Mendelsohn    ------------------------------   End of INFO-VAX 2005.629 ************************