0 INFO-VAX	Tue, 04 Feb 2003	Volume 2003 : Issue 70      Contents: Re: And the winner is? Re: And the winner is? Re: And the winner is? Re: And the winner is? Re: And the winner is? Re: And the winner is? Re: And the winner is? Re: Batch job log spec Re: Batch job log spec Re: Batch job log spec Re: Batch job log spec Re: Batch job log specA Carly sez "no reason that HP couldn't ship Opteron-based servers"  Re: Columbia Re: Columbia Re: Columbia Re: Columbia Re: Columbia Re: Columbia Re: Columbia Re: DECevent shows errors... Re: DECevent shows errors... Re: DECevent shows errors... DECNet Configuration Help ' Digital Network's Access Server Manager 
 Re: Errors fonts , Re: Frmr VMS eng. (Greg Robert) seeking work, Re: Frmr VMS eng. (Greg Robert) seeking work re: hedder full  Re: Hedder full error = Re: Inspecting VMS SMTP email contents before it gets to user = Re: Inspecting VMS SMTP email contents before it gets to user = Re: Inspecting VMS SMTP email contents before it gets to user = Re: Inspecting VMS SMTP email contents before it gets to user = Re: Inspecting VMS SMTP email contents before it gets to user = Re: Inspecting VMS SMTP email contents before it gets to user = Re: Inspecting VMS SMTP email contents before it gets to user = Re: Inspecting VMS SMTP email contents before it gets to user M Re: Itanium EFI Console (was: Re: OpenVMS Boots on Itanium on Friday Jan  31) P Re: Itanium EFI Console (was: Re: OpenVMS Boots on Itanium on Friday Jan 31) 31) Re: Newbie with dumb questions RE: Nice newsgroup!  Re: Nice newsgroup! # Re: OpenRDA ODBC Driver for OpenVMS  OpenVMS Article on CNet  Re: OpenVMS Article on CNet  Re: OpenVMS Article on CNet  Re: OpenVMS Article on CNet - Re: OpenVMS Boots on Itanium on Friday Jan 31 - Re: OpenVMS Boots on Itanium on Friday Jan 31 - Re: OpenVMS Boots on Itanium on Friday Jan 31 - Re: OpenVMS Boots on Itanium on Friday Jan 31 - Re: OpenVMS Boots on Itanium on Friday Jan 31 - Re: OpenVMS Boots on Itanium on Friday Jan 31 P RE: OpenVMS.org: Marvel article and HP's press release for Marvel and Alpha RetaP Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retai Oracle 9i and ODS-5 ' Re: pathworks client connection problem ' Re: pathworks client connection problem  Re: Question on memory Re: Question on memory2 Re: SKHPC slaps Gartner, naysayers on comp vms ...2 Re: SKHPC slaps Gartner, naysayers on comp vms ...2 Re: SKHPC slaps Gartner, naysayers on comp vms ...' Re: Status of MySQL and Postgres ports? ' Re: Status of MySQL and Postgres ports? ' Re: Status of MySQL and Postgres ports? ' Re: Status of MySQL and Postgres ports? @ Re: TCP/IP Services v5.3 - how to increase the number of "mbufs"/ Type of SCSI disks for Alpha Server 1000A 5/400 # Re: Understanding memory allocation  Update on the Technical Journal # Re: Update on the Technical Journal # Re: Update on the Technical Journal  Re: Very large disks on VMS  Re: Very large disks on VMS  Re: Very large disks on VMS  Re: VMS on Itanium Re: VMS on Itanium - bugcheck? Re: VMS on Itanium - bugcheck? Re: VMS on Itanium - bugcheck? Re: VMS on Itanium - bugcheck? Re: VMS on Itanium - bugcheck?  F ----------------------------------------------------------------------  $ Date: Tue, 4 Feb 2003 09:25:46 -0000* From: "Richard Brodie" <R.Brodie@rl.ac.uk> Subject: Re: And the winner is? , Message-ID: <b1o0us$1a5u@newton.cc.rl.ac.uk>  ] "Alan Frisbie" <Abuse@NelsonUSA.com> wrote in message news:3E3EFA30.72C1E81D@NelsonUSA.com...   @ > I was off (early) by five hours and six minutes, and I haven'tG > seen anyone claim a closer time.   My guess was January 31, at 20:37. H > Now I wonder what the "fabulous OpenVMS gift" is?   Dare we speculate?  J You get to alpha test the code that booted and stayed up long enough to do a DIRECTORY ;)   ------------------------------  % Date: Mon, 03 Feb 2003 15:24:32 -0800 ( From: Alan Frisbie <Abuse@NelsonUSA.com> Subject: Re: And the winner is? - Message-ID: <3E3EFA30.72C1E81D@NelsonUSA.com>    Richard Maher wrote:   > And the winner is?  > I was off (early) by five hours and six minutes, and I haven'tE seen anyone claim a closer time.   My guess was January 31, at 20:37. F Now I wonder what the "fabulous OpenVMS gift" is?   Dare we speculate?  @ I wish I could have been there for the big event (and what I bet@ was one fantastic party that night).   Even more, I wish I could= have actually been working on it.   I really envy those lucky 
 engineers.   Alan   -- Alan E. Frisbie -- Flying Disk Systems, Inc.   ------------------------------  $ Date: Mon, 3 Feb 2003 17:30:00 -0600& From: "Dave Gudewicz" <k9jdk@arrl.net> Subject: Re: And the winner is? / Message-ID: <v3turo6peb2a30@corp.supernews.com>   H In the W.C. Fields movie "The Bank Dick", W. C. was rewarded with a firmL shake of the hand by the Pres. of the bank he fumbled into not being robbed.> Oh and he also got a calendar, I believe of the previous year.  0 Alan and the 4 others, I trust you'll do better.   Dave...   5 "Alan Frisbie" <Abuse@NelsonUSA.com> wrote in message ' news:3E3EFA30.72C1E81D@NelsonUSA.com...  > Richard Maher wrote: >  > > And the winner is? > @ > I was off (early) by five hours and six minutes, and I haven'tG > seen anyone claim a closer time.   My guess was January 31, at 20:37. H > Now I wonder what the "fabulous OpenVMS gift" is?   Dare we speculate? > B > I wish I could have been there for the big event (and what I betB > was one fantastic party that night).   Even more, I wish I could? > have actually been working on it.   I really envy those lucky  > engineers. >  > Alan >  > -- Alan E. Frisbie > -- Flying Disk Systems, Inc.   ------------------------------  % Date: Tue, 04 Feb 2003 06:44:06 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca> Subject: Re: And the winner is? / Message-ID: <3E3F996F.A01D0A9F@vl.videotron.ca>   ! VAXman-, @SendSpamHere.ORG wrote: H > ?????  Of the boot contest?  I know it wasn't me.  I picked the 5th of> > February which was the birthdate of another famous computer.    K I have a vested interest in that date. Which computer was born on the 5th ?    ------------------------------  # Date: Tue, 04 Feb 2003 11:43:51 GMT " From:   VAXman-  @SendSpamHere.ORG Subject: Re: And the winner is? 0 Message-ID: <00A1AFA1.433BCEDE@SendSpamHere.ORG>  ^ In article <AEE%9.37964$IL1.716228@twister.tampabay.rr.com>, "McEagle" <spam@spam.com> writes:& >Which famous computer was born 5-Feb? >  >Mike   * :)  My very own very first VAX, of course.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM              5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------  $ Date: Tue, 4 Feb 2003 18:05:24 +0100$ From: Michael Unger <unger@decus.de> Subject: Re: And the winner is? + Message-ID: <00A1B000.74311210.17@decus.de>   + "Alan Frisbie" <Abuse@NelsonUSA.com> wrote:   @ > I was off (early) by five hours and six minutes, and I haven'tG > seen anyone claim a closer time.   My guess was January 31, at 20:37. H > Now I wonder what the "fabulous OpenVMS gift" is?   Dare we speculate?  E An "OpenVMS golf shirt". And apparently the obligation to participate  in advertising and promotion.   7 It is still mentioned at the "boot contest rules" page:   ? http://h18003.www1.hp.com/hps/ipf-enterprise/contest_rules.html     
 <start quote>    *Official Rules *    [...]   G *6. AWARDS:* [Within ten (10) days after the first "boot" of OpenVMS on G an Itanium? -based system (i.e. OpenVMS commences running and permits a F user to log on and conduct a directory command without error), HP willF determine the five (5) eligible participants who selected the date andG time (U.S. Eastern Standard Time) closest to such first boot of OpenVMS D on an Itanium?-based system, and each of these five (5) participantsF will be awarded one OpenVMS golf shirt, each with an approximate value
 of $25. [...]    [...]    *8. OTHER CONSIDERATIONS* G a. If a prize is returned as undeliverable, the prize will be forfeited = and an alternate winner will be selected in a random drawing. H b. By accepting an award, a participant agrees that HP, and those acting0 under its authority, may use participant's name,G picture/portrait/likeness and/or voice, for advertising and promotional A purposes without further consideration, unless prohibited by law.    [...]   = d. Winners may be required to sign and return an affidavit of E eligibility and liability/publicity release (except where prohibited) D within 21 days. Failure to comply with this condition will result inH forfeiture of prize and an alternate winner will be selected in a randomH drawing. This drawing will take place no earlier than three and one half< months after the unclaimed prize winner had been determined.   <end quote>    Michael    ------------------------------  * Date: Tue, 4 Feb 2003 18:48:38 +0000 (UTC)3 From: "Richard Maher" <maher_rj@hotspamnotmail.com>  Subject: Re: And the winner is? / Message-ID: <b1p1u6$pnm$1@venus.btinternet.com>    Hi,   D For the 2nd time I give notice that (I'm pretty sure?) I was closer!  * And is Golf shirt American for Polo shirt?  4 Regards Richard Maher (Who would much prefer a DS10)  / Michael Unger <unger@decus.de> wrote in message % news:00A1B000.74311210.17@decus.de... - > "Alan Frisbie" <Abuse@NelsonUSA.com> wrote:  > B > > I was off (early) by five hours and six minutes, and I haven'tI > > seen anyone claim a closer time.   My guess was January 31, at 20:37. J > > Now I wonder what the "fabulous OpenVMS gift" is?   Dare we speculate? > G > An "OpenVMS golf shirt". And apparently the obligation to participate  > in advertising and promotion.  > 9 > It is still mentioned at the "boot contest rules" page:  > A > http://h18003.www1.hp.com/hps/ipf-enterprise/contest_rules.html  >  >  > <start quote>  >  > *Official Rules *  >  > [...]  > I > *6. AWARDS:* [Within ten (10) days after the first "boot" of OpenVMS on I > an Itanium? -based system (i.e. OpenVMS commences running and permits a H > user to log on and conduct a directory command without error), HP willH > determine the five (5) eligible participants who selected the date andI > time (U.S. Eastern Standard Time) closest to such first boot of OpenVMS F > on an Itanium?-based system, and each of these five (5) participantsH > will be awarded one OpenVMS golf shirt, each with an approximate value > of $25. [...]  >  > [...]  >  > *8. OTHER CONSIDERATIONS* I > a. If a prize is returned as undeliverable, the prize will be forfeited ? > and an alternate winner will be selected in a random drawing. J > b. By accepting an award, a participant agrees that HP, and those acting2 > under its authority, may use participant's name,I > picture/portrait/likeness and/or voice, for advertising and promotional C > purposes without further consideration, unless prohibited by law.  >  > [...]  > ? > d. Winners may be required to sign and return an affidavit of G > eligibility and liability/publicity release (except where prohibited) F > within 21 days. Failure to comply with this condition will result inJ > forfeiture of prize and an alternate winner will be selected in a randomJ > drawing. This drawing will take place no earlier than three and one half> > months after the unclaimed prize winner had been determined. > 
 > <end quote>  > 	 > Michael    ------------------------------   Date: 3 Feb 2003 15:45:24 -0800 . From: spamsink2001@yahoo.com (Alan E. Feldman) Subject: Re: Batch job log spec = Message-ID: <b096a4ee.0302031545.14089cdc@posting.google.com>   Y briggs@encompasserve.org wrote in message news:<EkQ6LE4o0WlW@eisner.encompasserve.org>... X > In article <1FEB200311465177@gerg.tamu.edu>, carl@gerg.tamu.edu (Carl Perkins) writes:% > > mckinneyj@cpva.saic.com writes... 8 > > }In article <VXgvizKltEip@eisner.encompasserve.org>,> > > }A simpler (DCL only) method might be to get the output ofC > > }SHOW DEVICE/FILES SYS$LOGIN and SEARCH the output for your own A > > }PID and the string "jobname.LOG". You'd only want to do this > > > }after you'd used the F$GETQUI calls to determine that the2 > > }logfile wasn't specified on the command line. > > }  > > }-- 
 > > }- Jim > >  > > = > > This part is pretty easy to do. Here is one way to do it:  > > 
 > > FOO.COM =  > >  > > $ z= f$getjpi("","pid") / > > $ Pipe Show Device/File/NoSys SYS$LOGIN | -  > >     Search Sys$Pipe 'z' | -  > >     @foo2.com  > > $ sh log/job thisfile* > >  > > FOO2.COM = > >  > > $ i = 0 
 > > $LOOPING: 4 > > $ Read/End=ENDLOOP/Err=ENDLOOP Sys$Pipe thisproc0 > > $ thisfile = "[" + F$Element(1,"[",thisproc)
 > > $ i = i+1 & > > $ Define/Job thisfile'i' &thisfile > > $ GoTo LOOPING
 > > $ENDLOOP: 
 > > $ Exit > . > There are a few problems with this approach. > J > 1.  Performance.  You're going to be doing a lot of disk access in orderB > to display all the file names of all the open files on the disk.  E Isn't much of this information cached in physical memory? I mean, you A only need the relevant directory entries and process information, ( right? I don't really know. Just asking.  D > 2.  Audit logs.  If you're running without privileges, you run theD > risk of triggering a bunch of alarms as you try and fail to obtain8 > the file access required to generate those file names.   No comment.   ; > 3.  LOG file not on user's home disk.  You won't find it.  > D > 4.  LOG file not ending in .LOG.  You won't be able to distinguishD > it from your command procedure or any other open file on the disk.  ? It seems to me that Carl's code was intended only as an example E illustrating the basic method. It is not hard to add more DCL code to  take care of 3 and 4.   @ I just ran into this "What is the log file-spec?" problem myself) recently and was rather disappointed that   ;     F$GETQUI("DISPLAY_JOB","LOG_SPECIFICATION",,"THIS_JOB")   D only gets you the information shown by SHOW ENTRY/FULL. I wanted theC fully-qualified actual log file-spec. So I wrote my own DCL (before D this thread appeared) which doesn't suffer from problems 3 and 4. If> anyone wants a copy, let me know (address at bottom, also, theF spamsink2001 in the header is a real address, but I haven't checked it& recently -- I hope it hasn't expired!)  A > 5.  Directory alias.  The file name you get will be obtained by A > traversing the directory backlink chain.  The result may not be  > a correct file specification.   C I don't see why it would be wrong. Unexpected maybe, but wrong? Why  might it be wrong?  F > 6.  Unsupported interface.  You're parsing file names out of a humanG > readable display.  Line wrap, truncation or any change in the display C > format from one release to the next and you could wind up in deep 
 > trouble.  E Yes, I see this as the biggest weakness. But when I wrote my DCL code E I didn't know about your macro! OTOH, I think it's very unlikely that D the format would change enough to break a well-written code (or even9 my code! :-). But yes, you are right, this is a weakness.   E Thanks for posting your macro code. That is definitely the better way B to do this. OTOH, the original poster did ask for DCL. BOTFH, that5 doesn't mean DCL is the best option for this problem.   C And I was also rather confused about THIS_JOB requiring DISPLAY_JOB  instead of DISPLAY_ENTRY!    >  > 	John Briggs   Disclaimer: JMVHO  Alan E. Feldman 
 alan feldman  	 GFIgroup   com    ------------------------------   Date: 4 Feb 03 08:45:30 PST  From: mckinneyj@cpva.saic.com  Subject: Re: Batch job log spec ( Message-ID: <BaVzC$uhx539@cpva.saic.com>  , In article <3FEB200321585022@gerg.tamu.edu>,*  carl@gerg.tamu.edu (Carl Perkins) writes:$ > briggs@encompasserve.org writes...Y > }In article <1FEB200311465177@gerg.tamu.edu>, carl@gerg.tamu.edu (Carl Perkins) writes: & > }> mckinneyj@cpva.saic.com writes...9 > }> }In article <VXgvizKltEip@eisner.encompasserve.org>, ? > }> }A simpler (DCL only) method might be to get the output of D > }> }SHOW DEVICE/FILES SYS$LOGIN and SEARCH the output for your ownB > }> }PID and the string "jobname.LOG". You'd only want to do this? > }> }after you'd used the F$GETQUI calls to determine that the 3 > }> }logfile wasn't specified on the command line.  > }> }  	 > }> }--   > }> }- Jim  > }>   > }>  > > }> This part is pretty easy to do. Here is one way to do it: > }>   > }> FOO.COM = > }>   > }> $ z= f$getjpi("","pid")0 > }> $ Pipe Show Device/File/NoSys SYS$LOGIN | -  > }>     Search Sys$Pipe 'z' | - > }>     @foo2.com > }> $ sh log/job thisfile*  > }>   > }> FOO2.COM =  > }>   > }> $ i = 0 > }> $LOOPING:5 > }> $ Read/End=ENDLOOP/Err=ENDLOOP Sys$Pipe thisproc 1 > }> $ thisfile = "[" + F$Element(1,"[",thisproc)  > }> $ i = i+1' > }> $ Define/Job thisfile'i' &thisfile  > }> $ GoTo LOOPING  > }> $ENDLOOP: > }> $ Exit  > } / > }There are a few problems with this approach.  > } K > }1.  Performance.  You're going to be doing a lot of disk access in order C > }to display all the file names of all the open files on the disk.  > H > If you don't do it very often, performance doesn't normally matter. IfH > you only need it from a few times a month to a few times a day this is > usually irrelevant.  > E > }2.  Audit logs.  If you're running without privileges, you run the E > }risk of triggering a bunch of alarms as you try and fail to obtain 9 > }the file access required to generate those file names.  > I > This could be the case. I suspect that on most systems it won't trigger  > many alarms, if any. > < > }3.  LOG file not on user's home disk.  You won't find it. > L > The above was for the case where the f$getqui request for the log filespecN > fails. You're supposed to try that first, and then this only if that doesn'tK > give you a filespec. When does that fail? When the /LOG qualifier was not O > present on the SUBMIT command used to submit the job. Under these conditions, G > the log file is placed in the sys$login directory. That is why I used ) > sys$login in the show dev/file command.  >   G And for the case where the user may have specified /NOLOG on the SUBMIT  command line there's  5   f$getqui("display_job", "job_log_null",,"this_job")    to catch it.  D If f$getqui tells you that there is a batch log, but that it doesn't< know the name then it is adequate to just PIPE the output ofB SHOW DEVICE/FILES SYS$LOGIN into a SEARCH/MATCH=AND of the PID and "jobname.LOG".  E > }4.  LOG file not ending in .LOG.  You won't be able to distinguish E > }it from your command procedure or any other open file on the disk.  > I > It will end in .LOG when you use this under the intended circumstances.  > B > }5.  Directory alias.  The file name you get will be obtained byB > }traversing the directory backlink chain.  The result may not be  > }a correct file specification. > F > I don't see how this is possible. It may not be the prefered form ofD > the filespec (e.g. with a rooted logical as the base), but it willD > point to the file when prepended with the actual device name, will5 > it not? If not, when will it fail? Give an example.  > G > }6.  Unsupported interface.  You're parsing file names out of a human H > }readable display.  Line wrap, truncation or any change in the displayD > }format from one release to the next and you could wind up in deep > }trouble.  > }  > }	John Briggs  > H > That is true. But it has been the same for a long time and it probablyE > won't change. If it does, it breaks this. If you use extremely long C > filenames or your login directory is very deep in a chain of long E > directory names you might exceed the buffer length and get an error  > instead of a result. Oh well.  > H > The problem is that there may not be a better way to do this from DCL.L > (I havn't seen anyone mention one yet.) Perhaps you could parse the outputM > from an ANA/SYS that was fed the info to do a SHOW PROC/CHAN on the correct I > process. But that has it's own problems, including most (if not all) of 6 > the above problems and probably some others as well. > B > It was just an example of how to do what was suggested. I do useD > something like it in one place myself - I have web servers runningH > on two nodes in a cluster from the same directory tree. The access.logI > files for the two instances are in the same directory. I sometimes want G > to know which of them is the one that is open on a specific node, and F > I find out using a method a lot like the above, but simpler (since IF > know the name and location of the log file, I search the listing forG > the actual desired filespec rather than the process ID: no looping is H > needed to build a list of all opened files to search, it just directly" > provides the desired filespec).  > 
 > --- Carl  D And, I've used schemes like this in the past to provide notification@ of failure in production batch jobs. Whenever a job experience a? failure I'd trap it and mail off the log file to the interested A parties. Depending upon your needs, I think that an approach like B this is reasonable. If the "interface" changes (possible, I guess)? then the code referencing it would have to too. If you use this C code that locates the batch log in multiple instances then bundling D it into a separate command procedure would allow you to only have to fix it in one location.    --   - Jim    ------------------------------   Date: 4 Feb 03 08:56:28 PST  From: mckinneyj@cpva.saic.com  Subject: Re: Batch job log spec ( Message-ID: <ozVcl$20mJlo@cpva.saic.com>  = In article <b096a4ee.0302031545.14089cdc@posting.google.com>, 1  spamsink2001@yahoo.com (Alan E. Feldman) writes: [ > briggs@encompasserve.org wrote in message news:<EkQ6LE4o0WlW@eisner.encompasserve.org>... Y >> In article <1FEB200311465177@gerg.tamu.edu>, carl@gerg.tamu.edu (Carl Perkins) writes: & >> > mckinneyj@cpva.saic.com writes...9 >> > }In article <VXgvizKltEip@eisner.encompasserve.org>, ? >> > }A simpler (DCL only) method might be to get the output of D >> > }SHOW DEVICE/FILES SYS$LOGIN and SEARCH the output for your ownB >> > }PID and the string "jobname.LOG". You'd only want to do this? >> > }after you'd used the F$GETQUI calls to determine that the 3 >> > }logfile wasn't specified on the command line.  >> > }  	 >> > }--   >> > }- Jim  >> >   >> >  > >> > This part is pretty easy to do. Here is one way to do it: >> >   >> > FOO.COM = >> >   >> > $ z= f$getjpi("","pid")0 >> > $ Pipe Show Device/File/NoSys SYS$LOGIN | -  >> >     Search Sys$Pipe 'z' | - >> >     @foo2.com >> > $ sh log/job thisfile*  >> >   >> > FOO2.COM =  >> >   >> > $ i = 0 >> > $LOOPING:5 >> > $ Read/End=ENDLOOP/Err=ENDLOOP Sys$Pipe thisproc 1 >> > $ thisfile = "[" + F$Element(1,"[",thisproc)  >> > $ i = i+1' >> > $ Define/Job thisfile'i' &thisfile  >> > $ GoTo LOOPING  >> > $ENDLOOP: >> > $ Exit  >>  / >> There are a few problems with this approach.  >>  K >> 1.  Performance.  You're going to be doing a lot of disk access in order C >> to display all the file names of all the open files on the disk.  > G > Isn't much of this information cached in physical memory? I mean, you C > only need the relevant directory entries and process information, * > right? I don't really know. Just asking. >   ) Some yes... but, not all. With CMKRNL try    $ SET WATCH/CLASS=ALL FILE! $ SHOW DEVICE/FILES SYS$SYSDEVICE  $ SET WATCH/CLASS=NONE  E and see. It has to be determined if you possess the authority to view  the file specs, etc...   --   - Jim    ------------------------------   Date: 4 Feb 2003 08:57:56 -0800 . From: spamsink2001@yahoo.com (Alan E. Feldman) Subject: Re: Batch job log spec = Message-ID: <b096a4ee.0302040857.20f3bcd6@posting.google.com>   Y briggs@encompasserve.org wrote in message news:<7Hjx6Rr754WN@eisner.encompasserve.org>... E > It has been brought to my attention that the copy of ppf.mar that I C > posted in this thread was corrupt.  In particular, part of a line * > of code around the label 5$ was missing. > G > Probably a data overrun during the cut and paste.  I used the include  > function this time instead.  > ; > Apologies for the duplication and length of this message.  >  > 	John Briggs  @ It works great. But I have a question about one of the comments:   [...] $ > ; FID_FILE_NAME is set as follows: > ;   If nam$l_fid is non zeroA > ;     determine the current file name by using lib$fid_to_name. > > ;     Set FID_FILE_NAME to this value.  Note that this file E > ;     name will be correct if the file was renamed using the RENAME ; > ;     command (but not necessarily if the FID is entered   > ;     multiple times). [...]   A Please explain: What does it mean "if the FID is entered multiple D times". Oh, do you mean via SET FILE/ENTER? In which case, how wouldB it be wrong? You would get the primary name or one of the aliases,A right? So specifying FID_FILE_NAME should be specifying the right - file, even if it's an "unexpected" file-spec.     Thanks, and thanks for the code.   Dislcaimer: JMHO Alan E. Feldman    ------------------------------   Date: 4 Feb 2003 12:10:44 -0600  From: briggs@encompasserve.org Subject: Re: Batch job log spec 3 Message-ID: <SO$w0QKanJVI@eisner.encompasserve.org>   n In article <b096a4ee.0302040857.20f3bcd6@posting.google.com>, spamsink2001@yahoo.com (Alan E. Feldman) writes:B > It works great. But I have a question about one of the comments: >  > [...] % >> ; FID_FILE_NAME is set as follows:  >> ;   If nam$l_fid is non zero B >> ;     determine the current file name by using lib$fid_to_name.? >> ;     Set FID_FILE_NAME to this value.  Note that this file  F >> ;     name will be correct if the file was renamed using the RENAME< >> ;     command (but not necessarily if the FID is entered  >> ;     multiple times).  > [...]  > C > Please explain: What does it mean "if the FID is entered multiple F > times". Oh, do you mean via SET FILE/ENTER? In which case, how wouldD > it be wrong? You would get the primary name or one of the aliases,C > right? So specifying FID_FILE_NAME should be specifying the right / > file, even if it's an "unexpected" file-spec.   < That part is Jon Pinkley's code.  I think you have correctly= understood the point being made.  The name you get back won't @ be the name you expected.  But it be one of the aliases pointing to the right file.  G In some cases the situation gets worse.  e.g. [SYSCOMMON.SYSEXE]SYS.EXE E versus [V4COMMON.SYSEXE]SYS.EXE versus [SYSn.SYSCOMMON.SYSEXE]SYS.EXE    	John Briggs   ------------------------------  # Date: Tue, 04 Feb 2003 16:05:47 GMT % From: Milton <mbhewitt@optonline.net> J Subject: Carly sez "no reason that HP couldn't ship Opteron-based servers"8 Message-ID: <h3pv3vgpdbclns5kq9p88k2eb9en3edhl6@4ax.com>  ; http://www.fortune.com/fortune/print/0,15935,418480,00.html   F "People misunderstand," she says. "We don't have a gun to our head. We. will migrate at the pace our customers want."      Cheers,  Milton   ------------------------------  $ Date: Mon, 3 Feb 2003 16:59:03 -0500* From: "Bill Todd" <billtodd@metrocast.net> Subject: Re: Columbia 2 Message-ID: <INicnXPA7Nyue6OjXTWcqw@metrocast.net>  H "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in message- news:0Xfys8vAIFHm@eisner.encompasserve.org... @ > In article <XdqcnQuLCoKm0KOjXTWcqw@metrocast.net>, "Bill Todd"  <billtodd@metrocast.net> writes: > > J > > It's actually trivial:  you just continually thrust at right-angles to yourH > > current velocity until you've gotten turned around (and time it such thatJ > > you return to the same orbital plane).  But whether this consumes moreE > > energy than that required to decelerate/accelerate is a different 	 question.  > >  > E >    If you could save some of the energy needed by changing the path E >    you took you could use that to build a perpetual motion machine.   L Ignoring any question about the logical merit of that assertion, you seem toH have misread what I said above to suggest that any saving was possible -L when in fact I simply questioned whether *more* energy would be required (my off-hand guess is yes).      The B >    amount of energy needed to change from an easterly orbit to aI >    westerly orbit cannot be less than the amount needed to de-orbit and  >    relaunch.  I You can't prove that without evaluating the amount of energy required and F comparing it with that required for a return (small) plus a subsequentJ Westerly launch (large).  It depends upon how much of the launch energy isE expended fighting gravity and air resistance:  if more than half, for I example, then your statement is incorrect (even for a subsequent Easterly B launch that wouldn't accomplish the same result, for that matter).   - bill   ------------------------------  $ Date: Mon, 3 Feb 2003 17:49:28 -05000 From: "Island Computers" <dbturner@islandco.com> Subject: Re: Columbia / Message-ID: <v3ts9diceu4ra0@news.supernews.com>   " SO what about controlled re-entry?  D It seems to me that our use, in 2003, of late 1920's technology (notJ forgetting that the Chinese had fireworks over 2000 years ago) is somewhat	 outdated.   I The whole concept of pushing a brick upwards with solid boosters is quite F proposterous, and if it was a new concept, would be laughed out of the design meeting.   L Attach some (not enlarged fins as on the Orbiter)wings and a  jet engine forI primary lift until oxygen levels drop too much, and ancilliary rockets to 1 whack it in to orbit seems the logical way to go.  Same goes for re-entry. G Fighting gravity is fuel hungry, but controlling the drop to earth with K retros and again, jet engines on wings providing resistance seems to me the  way to go???  J SCRAMjets are in their infancy, but would be extremely beneficial to those6 who were anticipating coming back to earth in 1 piece.   My 2.53 euros worth    DT      6 "Barry Treahy, Jr." <Treahy@MMaz.com> wrote in message! news:3E3E8CCC.1090300@MMaz.com...  Jan-Erik Sderholm wrote:   = >Well, I don't know, but there must be something here I don't  >understand. > A >7 people got killed in the first shuttle accident in *17 years*. 1 >That is of course terrible for those 7 families.  > > >Now, 40.000 people get killed in car accidents *each year* in >the US. > A >When did the US President hold a specific TV speach about that ?  >  > E Any loss of life is unfortunate; The real shame is the amount of time D the space program will be put on hold, or worse, by space nay-sayersE that will use this accident as a tool to damage or reduce the size of G the space program.  Politics, and how this will be used to spin agendas H in the US Senate and House for power, that is the real shame...  People,E including my own wife, loose site if the innovations and technologies F that have spawned because of the space race that JFK initiated and has since evolved...   Barry    --  ? Barry Treahy, Jr  *  Midwest Microwave  *  Vice President & CIO   A E-mail: Treahy@mmaz.com * Phone: 480/314-1320 * FAX: 480/661-7028    ------------------------------  # Date: Tue, 04 Feb 2003 12:25:39 GMT 0 From: rw@vorpalbunnyeircom.net (Russell Wallace) Subject: Re: Columbia 0 Message-ID: <3e3fb08e.114037061@news.eircom.net>  . On Mon, 3 Feb 2003 19:55:54 -0500, "Bill Todd" <billtodd@metrocast.net> wrote:   L >To say that it would have made no difference assumes that there was nothingJ >that could have been done had the problem (assuming that was the problem)M >been known.  That's roughly equivalent to saying that we should have written A >off the Apollo 13 crew because they were so clearly beyond help.   F This has been thrashed out extensively on sci.astro (complete with the; Apollo 13 analogy); I'll refer you to the discussion there.   L >Valid point in this particular instance, but irrelevant to the larger issueE >of why, when pieces of ET insulation had hit the shuttle on multiple M >previous occasions and on at least one such occasion caused significant tile I >damage, no provision was made earlier to attempt to address the problem.   B Sure, perhaps something could have been done _before_ the ColumbiaD launch. My point is merely that after that, as far as the people whoD actually know something about the issue can tell, the best remainingB option was to attempt re-entry and hope the damage would not prove fatal.   --  3 "Mercy to the guilty is treachery to the innocent." + Remove killer rodent from address to reply. ! http://www.esatclear.ie/~rwallace    ------------------------------   Date: 4 Feb 2003 07:45:44 -0600 ; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)O Subject: Re: Columbiao3 Message-ID: <ljW+lbMeOtb1@eisner.encompasserve.org>n  b In article <3e3ee4e0.61886274@news.eircom.net>, rw@vorpalbunnyeircom.net (Russell Wallace) writes:  D > As for using unmanned rockets to put emergency spares in orbit: ifB > you're going to do that, why not just use the rockets instead ofE > shuttles to carry out the primary missions (or at least the 90% forpD > which human presence isn't needed), thus avoiding the need to risk! > human lives in the first place?b      Two reasons:e  @    1) the science experiments were not all completely automated,F    attempting to make them so severly limits what can be done at great@    cost, you just can't make up for the presence of human beings  @    2) a part of the reason NASA exists is to put people in space   ------------------------------   Date: 4 Feb 2003 07:47:46 -0600o; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)  Subject: Re: Columbia 3 Message-ID: <y6orHIJfXsmR@eisner.encompasserve.org>r  _ In article <INicnXPA7Nyue6OjXTWcqw@metrocast.net>, "Bill Todd" <billtodd@metrocast.net> writes:V >  >   The C >>    amount of energy needed to change from an easterly orbit to anJ >>    westerly orbit cannot be less than the amount needed to de-orbit and >>    relaunch.e > K > You can't prove that without evaluating the amount of energy required andsH > comparing it with that required for a return (small) plus a subsequent > Westerly launch (large).  E    Air resistance is a very small factor.  In fact what I sited aboutEH    the change in paths is a principle of physics.  If you don't want to !    take my expertise, ask Newton.    ------------------------------   Date: 4 Feb 2003 07:52:42 -0600r; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)  Subject: Re: ColumbiaI3 Message-ID: <q5FPrRwO2Bws@eisner.encompasserve.org>m  b In article <v3ts9diceu4ra0@news.supernews.com>, "Island Computers" <dbturner@islandco.com> writes:$ > SO what about controlled re-entry?  N > Attach some (not enlarged fins as on the Orbiter)wings and a  jet engine forK > primary lift until oxygen levels drop too much, and ancilliary rockets toe3 > whack it in to orbit seems the logical way to go.   D    Sounds like the launch of a Pegasus from an L1011.  UnfortunatelyE    Pegasus can only handle fairly small payloads.  Except jet engines D    are not used to provide lift on any current vehicle, only thrust.   > Same goes for re-entry.lI > Fighting gravity is fuel hungry, but controlling the drop to earth withrM > retros and again, jet engines on wings providing resistance seems to me thes > way to go???  E    Who is gong to lift all that fuel into orbit, and why?  OBTW thoser*    jet engines are pretty damn heavy, too.  L > SCRAMjets are in their infancy, but would be extremely beneficial to those8 > who were anticipating coming back to earth in 1 piece.  K    Use of rockets to provide a soft landing for a manned vehichle has been eB    done only in cartoons, or on the moon.  The overall size of the?    original lifting vehicle to get all that stuff into orbit isc    prohibitive.a   ------------------------------   Date: 4 Feb 2003 07:56:09 -0600-; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)- Subject: Re: Columbia 3 Message-ID: <2d3jqmvhgZdj@eisner.encompasserve.org>d  W In article <3E3F597A.4010306@tsoft-inc.com>, David Froble <davef@tsoft-inc.com> writes:w > Q > As taught to student pilots, and mandated/suggested by FAA, a pilot performs a .S > pre-flight inspection before flight.  A mandatory inspection of a shuttle before IN > re-entry would seem to follow this time honored and proven safety procedure.  H    Nope.  I've never, ever seen a pilot perform a pre-landing inspectionD    outside the aircraft prior to landing.  The pre-flight inspectionG    I learned and have done is pre-FLIGHT, once I'm up I will come down.0   ------------------------------  % Date: Tue, 04 Feb 2003 08:34:43 +0000c( From: Nic Clews <sendspamhere@127.0.0.1>% Subject: Re: DECevent shows errors...h) Message-ID: <3E3F7B23.E68DF8C5@127.0.0.1>    Reinhard Eigner wrote: > 8 > I've a Digital Personal Workstation 600au (Miata MX5).: > OpenVMS 7.3-1 runs on it. Yesterday I installed DECeventB > 3.4 to investigate why my errlog.sys is increasing so extremely.& > DECevent shows me following entries: >  > MATRIX:SYSTEM> dia /sinces >  > DECevent V3.4. ... F > This "INCONSTATE, Inconsistent I/O data base" error occurs on nearlyI > every process I work with. For example if I do a telnet on this machinesO > I get an "INCONSTATE, Inconsistent I/O data base" error on the tcpip process.6 > - > Does anyone have any ideas or hints for me?:  G Extended sense means a code sent which is specific to the drive itself, B DIA will probably try to decode it, but it may be meaningless, youB really need manufacturers data to find out what the error means by translating the codes.  E the inconsistent IO database crash is probably NOT unrelated, but I'm4A not going to speculate further. I would check you have the latestbH console firmware, also see if the SCSI drive firmware is up to revision,F and if it is a VMS supported drive, and check the controller revision.* You may of course also have a faulty disk!  G The bugcheck in this case is data protection, the system is telling youtC it is unhappy with what is going on in the IO space, so rather thanwF corrupt data, it bugchecks without causing further damage. If you've aD contract with HP, some analysis on the dump file should reveal  whatG specifically is inconsistent about the IO database, this may or may notk( lead to the source of the inconsistency.   -- i? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences- nclews at csc dot come   ------------------------------  $ Date: Tue, 4 Feb 2003 16:28:01 +0100, From: "Reinhard Eigner" <antispam@garnix.de>% Subject: Re: DECevent shows errors... / Message-ID: <b1om0m$omr$07$1@news.t-online.com>h  G > the inconsistent IO database crash is probably NOT unrelated, but I'mhC > not going to speculate further. I would check you have the latestnJ > console firmware, also see if the SCSI drive firmware is up to revision,H > and if it is a VMS supported drive, and check the controller revision., > You may of course also have a faulty disk!  ? The SCSI-HDD was delivered with the machine. I already bought ac "new" one at ebay.  I > The bugcheck in this case is data protection, the system is telling youmE > it is unhappy with what is going on in the IO space, so rather thanVH > corrupt data, it bugchecks without causing further damage. If you've aF > contract with HP, some analysis on the dump file should reveal  whatI > specifically is inconsistent about the IO database, this may or may not4* > lead to the source of the inconsistency.  F I don't have a contract with HP because this is my private Alpha and IG want to learn OpenVMS. So my next question is: What is the IO database?    Thxa   ------------------------------   Date: 4 Feb 2003 08:57:11 -0700y1 From: nothome@spammers.are.scum (Malcolm Dunnett) % Subject: Re: DECevent shows errors...b- Message-ID: <Vti+ePgA83b9@malvm7.mala.bc.ca.>   > In article <5bfbe1c6.0302032256.569d7c40@posting.google.com>, <     inet-register@planet-eigner.de (Reinhard Eigner) writes:  8 > I've a Digital Personal Workstation 600au (Miata MX5).: > OpenVMS 7.3-1 runs on it. Yesterday I installed DECeventB > 3.4 to investigate why my errlog.sys is increasing so extremely.& > DECevent shows me following entries: >      ....   > 6 > Bugcheck Minor class              2. System Bugcheck > M > Bugcheck Msg                         INCONSTATE, Inconsistent I/O data base     A      IIRC this error was introduced by the VMS731_SYS-V0200 patch.D kit, and seems to only occur on uniprocessor machines. I believe theD suggested workaround was to set the sysgen parameter MULTIPROCESSINGC to a value of 4, so that it always loads the multiprocessing image.$  F      Other than filling the error log with bugchecks it doesn't appear the bug does any damage.   ------------------------------  % Date: Tue, 04 Feb 2003 12:59:44 -0000d+ From: m_thompson@ids.net (Michael Thompson)t" Subject: DECNet Configuration Help/ Message-ID: <v3vea0ctn5v575@corp.supernews.com>n  J I have several VAXen connected with Ethernet. I have TCP/IP installed and N working. I would like to configure DECNet between the machines. I am clueless O about configuring DECNet so I could use some cookbook instructions or pointers  = to tutorial information. The systems are running OpenVMS 7.2.2  I Eventually I would like to configure DECNet on a DPV11 in a VAX 3600 and r? connect it to a DUP11/KMC11 in a PDP-10 KS10 running TOPS 7.03..   Thanks!n   ------------------------------  % Date: Tue, 04 Feb 2003 11:49:36 -0500a+ From: Chuck Aaron <caaron@ceris.purdue.edu>d0 Subject: Digital Network's Access Server Manager/ Message-ID: <3E3FEF20.7010201@ceris.purdue.edu>)   Hello,  D If anyone is using this product can you provide me a recommendation?   Thanks,s Chuck Aaronu CERIS - Purdue Univ.   ------------------------------  % Date: Tue, 04 Feb 2003 11:02:11 +0000I( From: Nic Clews <sendspamhere@127.0.0.1> Subject: Re: Errorse) Message-ID: <3E3F9DB3.8032E160@127.0.0.1>a   > Arindam wrote: > E > Please guide me about the following errors that we often get in our. > VAX 3800 and 3400 systemse >  > PTA0:w > PIA0:  > % > when we give the show error commandi  ; $ ANAL/ERR/OUTPUT=..... (see documentation for full syntax)   F Examine output. Hand to hardware support people, they will tell you if1 there is a problem, providing they can speak DEC.p  F It is fairly common that you see boot up errors, e.g. non-existent IDEH devices on Alphas, other device inquiry type errors. If the error countsD increase over time, then this is the time to want to know more about them.l HTH. --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot comn   ------------------------------  % Date: Tue, 04 Feb 2003 10:37:38 +0000s# From: vervoom <vervoom@hotmail.com>  Subject: fonts- Message-ID: <b1o4rc$ndk$2@cspc1n11.baplc.com>d  E Does anyone here know an internet site where I can get the following w font files?h   menu12.pcf -bigelow & 8 holmes-menu-medium-r-normal--12-120-75-75-p-70-iso8859-1 menu10.pcf -bigelow & 8 holmes-menu-medium-r-normal--10-100-75-75-p-56-iso8859-1   Thanks,    JS.n   ------------------------------  $ Date: Tue, 4 Feb 2003 05:37:21 -0800, From: "James Gessling" <jgessling@yahoo.com>5 Subject: Re: Frmr VMS eng. (Greg Robert) seeking workb5 Message-ID: <b1ofmi$15k6j5$1@ID-46415.news.dfncis.de>d  6 "Swallowbrook" <swallowbrook@aol.com> wrote in message3 news:20030203144004.27916.00000517@mb-ml.aol.com...  > Hi folks,y >e >rD > I'm bored with semi-retirement, unsettled by the stock market, andF > so seek something interesting and challenging to do.  I'm especially  A A search for VMS on Oracle's employment site got three jobs.  OnetI in Nashua and 2 in Callifornia.  So there are jobs out there.  Good luck.e   Jime   ------------------------------   Date: 04 Feb 2003 15:29:06 GMT) From: swallowbrook@aol.com (Swallowbrook)o5 Subject: Re: Frmr VMS eng. (Greg Robert) seeking worka9 Message-ID: <20030204102906.20627.00000136@mb-fb.aol.com>a   re:   N >> Just in case the original poster hasn't been following the newsgroup lately ...d  O No, I haven't been following the group, but I expected and recognize bitterness5A when I see it.  It's not a problem.  Normal human nature and all.:  K What I believe is likely true is that there is a significant number of verygL high quality customers and substantial applications running on [Open]VMS and# that will continue for a long time.e  I That's orthogonal to the [non]-roll of VMS as a "mover and shaker" in the 	 industry.y  O The original promise of VMS, and its consequential handling by DEC, Compaq, and,J HP are what they are.  This is an industry demanding an "adapt, improvise,B overcome" mindset, and the tardy or progress-resistant are doomed.  ! But thanks for the heads-up !  :)o  M re: Oracle jobs (another poster), yes someone else pointed out the Nashua jobtM to me and I immediately applied.  I think the CA jobs were account management-( jobs, which doesn't greatly interest me.   - Greg   ------------------------------  $ Date: Tue, 4 Feb 2003 16:18:17 +0530, From: "Arindam" <arindam-dsp@sail-steel.com> Subject: re: hedder full2 Message-ID: <001101c2cc3b$3c0e44a0$3c03e980@bofpc>  , This is a multi-part message in MIME format.  + ------=_NextPart_000_000A_01C2CC69.069D4E00E Content-Type: text/plain;  	charset="Windows-1252"=+ Content-Transfer-Encoding: quoted-printable+  6 thanks your help has helped me to solve the problem=20     Arindam Paul=20       + ------=_NextPart_000_000A_01C2CC69.069D4E00s Content-Type: text/html; 	charset="Windows-1252" + Content-Transfer-Encoding: quoted-printabled  > <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD><BASE=208 href=3D"file://C:\Program Files\Common Files\Microsoft = Shared\Stationery\">7 <META http-equiv=3DContent-Type content=3D"text/html; =  charset=3Dwindows-1252">9 <META content=3D"MSHTML 5.50.3825.1300" name=3DGENERATOR>nE <META content=3D"C:\PROGRAM FILES\MICROSOFT OFFICE\OFFICE\html.dot" =a name=3DTemplate> <STYLE></STYLE>e </HEAD>o7 <BODY vLink=3D#800080 link=3D#0000ff bgColor=3D#ffffff>s? <DIV>thanks your help has helped me to solve the problem </DIV>n/ <P>&nbsp;</P><B><FONT color=3D#0000ff size=3D2>t1 <P>Arindam Paul <BR></FONT></B></P></BODY></HTML>   - ------=_NextPart_000_000A_01C2CC69.069D4E00--a   ------------------------------   Date: 4 Feb 2003 02:23:32 -0800n( From: nedtrilby@hotmail.com (Ned Trilby) Subject: Re: Hedder full error= Message-ID: <4498b45c.0302040223.10b1ba9d@posting.google.com>d  l Tim Llewellyn <tim.llewellyn@blueyonder.co.uk> wrote in message news:<3E3E9B04.57BE61CC@blueyonder.co.uk>... > > Arindam wrote: > > ; > > I am getting an error in our VAX 3800 system which says- > > "Hedder full"  > >C@ > >  I can log on to the system however cannot create any files. > > Please helpn > ; > This is a pretty standard question for vms system managerm
 > interviews.  > < > Might I suggest your employer stop skimping and hire a VMS > system manager of consultant?  >  ??? Tim,7 Whats the problem with helping someone on the group ???i   ------------------------------  # Date: Tue, 04 Feb 2003 07:12:51 GMTeL From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr")F Subject: Re: Inspecting VMS SMTP email contents before it gets to user6 Message-ID: <00A1AF62.3D3EDCBC@SSRL.SLAC.STANFORD.EDU>  h In article <d7791aa1.0302032014.5b61af9b@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes:h >"Brian Tillman" <tillman_brian@notnoone.notnohow.com> wrote in message news:<3e3ee120$1@news.si.com>...> >> >I need to programmatically intercept email received by VMS@ >> >and inspect it before it gets POPped to its final recipient. >>  ; >> Why not get Sophos' anti-virus program that runs on VMS?  >n@ >that's fine if you want to look for vms viruses (such a thing?); >but to scan smtp bodies you only have one option right nowH7 >and that's pmdf ... lousy but that's the way it is ...l  G PMDF processes incoming email and can have the message body scanned by  F Sophos.  This means that you don't deliver the viral load of infected K messages to PC clients.  Sophos isn't looking for VMS viruses, it's running-M on VMS and looking for PC viruses.  (And seems to do a pretty good job, too.)c  J Are you calling PMDF lousy or saying that it's lousy that there's only one4 option?  I think PMDF is pretty darned good, myself.   -- Alan     O ===============================================================================70  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056EM  Paper mail to: SSRL -- SLAC BIN 99, 2575 Sand Hill Rd, Menlo Park CA   94025eO ===============================================================================r   ------------------------------  * Date: Tue, 4 Feb 2003 11:07:17 +0000 (UTC)+ From: david20@alpha2.mdx.ac.uk (David Webb)iF Subject: Re: Inspecting VMS SMTP email contents before it gets to user+ Message-ID: <b1o6t5$kdr$1@aquila.mdx.ac.uk>   b In article <3e3ee120$1@news.si.com>, "Brian Tillman" <tillman_brian@notnoone.notnohow.com> writes:< >>I need to programmatically intercept email received by VMS> >>and inspect it before it gets POPped to its final recipient. > 9 >Why not get Sophos' anti-virus program that runs on VMS?>  O I don't think he specified that he was just wanting to virus scan the messages.pC He might want to check them for spam or other undesirable contents.i  K Besides which unless Sophos have ported mailmonitor to VMS (which they wereOM supposed to be considering) then you would still need a means of intercepting O the mail message and passing it through Sophos' Vsweep program before delivery.t  O PMDF does this very well as I believe does MX. As far as I am aware none of thet< VMS TCPIP stacks provides such mail interception facilities.  H Alternatively he should be able to do something using the public domain N version of DELIVER. The simplest such solution would be to put a mail.deliveryN file in each individual user's home directory which runs a command file to do P the appropriate scanning and then took the appropriate action eg delivering the J original message, delivering a "censored" message or bouncing the message.    
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  * Date: Tue, 4 Feb 2003 11:48:47 +0000 (UTC)+ From: david20@alpha2.mdx.ac.uk (David Webb)hF Subject: Re: Inspecting VMS SMTP email contents before it gets to user+ Message-ID: <b1o9av$kdr$2@aquila.mdx.ac.uk>   h In article <d7791aa1.0302032014.5b61af9b@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes:h >"Brian Tillman" <tillman_brian@notnoone.notnohow.com> wrote in message news:<3e3ee120$1@news.si.com>...> >> >I need to programmatically intercept email received by VMS@ >> >and inspect it before it gets POPped to its final recipient. >> "; >> Why not get Sophos' anti-virus program that runs on VMS?t >e@ >that's fine if you want to look for vms viruses (such a thing?)  N No Known VMS viruses but Sophos Vsweep is often used on VMS (and Unix systems)J which hold mail or files which can be accessed by vulnerable PC systems to% protect those systems from infection.   ; >but to scan smtp bodies you only have one option right now 7 >and that's pmdf ... lousy but that's the way it is ...n  J MX should also be able to handle this and at the actual delivery point youO could use the public domain version of DELIVER (equivalent to Unix's procmail).   
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  % Date: Tue, 04 Feb 2003 06:53:41 -0400o0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>F Subject: Re: Inspecting VMS SMTP email contents before it gets to user/ Message-ID: <3E3F9BAE.D30BF751@vl.videotron.ca>    Hoff Hoffman wrote:aI >   DELIVER-like mechanism.  If you wish to use an undocumented approach,wH >   please see the MAIL transport image hooks that are shown in a couple< >   of Freeware packages and in the OpenVMS source listings.  J When you mention "MAIL transport image hooks", are you refering to foreignJ mail protocols , or is there another undocumented hook ? Could you provide5 some undocumentation or some pointers on such hooks ?r   ------------------------------   Date: 4 Feb 2003 06:03:27 -0800f( From: bob@instantwhip.com (Bob Ceculski)F Subject: Re: Inspecting VMS SMTP email contents before it gets to user= Message-ID: <d7791aa1.0302040603.5bda25f2@posting.google.com>    winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr") wrote in message news:<00A1AF62.3D3EDCBC@SSRL.SLAC.STANFORD.EDU>...j > In article <d7791aa1.0302032014.5b61af9b@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes:j > >"Brian Tillman" <tillman_brian@notnoone.notnohow.com> wrote in message news:<3e3ee120$1@news.si.com>...@ > >> >I need to programmatically intercept email received by VMSB > >> >and inspect it before it gets POPped to its final recipient. > >>  = > >> Why not get Sophos' anti-virus program that runs on VMS?f > > B > >that's fine if you want to look for vms viruses (such a thing?)= > >but to scan smtp bodies you only have one option right nowe9 > >and that's pmdf ... lousy but that's the way it is ...r > I > PMDF processes incoming email and can have the message body scanned by eH > Sophos.  This means that you don't deliver the viral load of infected M > messages to PC clients.  Sophos isn't looking for VMS viruses, it's runninghO > on VMS and looking for PC viruses.  (And seems to do a pretty good job, too.)e > L > Are you calling PMDF lousy or saying that it's lousy that there's only one6 > option?  I think PMDF is pretty darned good, myself. > 	 > -- Alank >   B there is a sophos vms virus scan the previous poster was referringC to sold separately, but to get smtp virus scans, you need pmdf, andt@ no I didn't call pmdf lousy, but its lousy they will not provide> a smtp virus scan product for other smtp servers (ie tcpware).   ------------------------------   Date: 4 Feb 2003 07:07:15 -0800-# From: sol@adldata.com (Sol Gongola)SF Subject: Re: Inspecting VMS SMTP email contents before it gets to user= Message-ID: <a7337236.0302040707.284f2c35@posting.google.com>0  Q > I don't think he specified that he was just wanting to virus scan the messages.rE > He might want to check them for spam or other undesirable contents.e > M > Besides which unless Sophos have ported mailmonitor to VMS (which they weregO > supposed to be considering) then you would still need a means of intercepting>Q > the mail message and passing it through Sophos' Vsweep program before delivery.m > Q > PMDF does this very well as I believe does MX. As far as I am aware none of theo> > VMS TCPIP stacks provides such mail interception facilities. > J > Alternatively he should be able to do something using the public domain P > version of DELIVER. The simplest such solution would be to put a mail.deliveryP > file in each individual user's home directory which runs a command file to do R > the appropriate scanning and then took the appropriate action eg delivering the L > original message, delivering a "censored" message or bouncing the message. >  >  > David Webb > VMS and Unix team leader > CCSS > Middlesex University  = Sometimes things get in before Antivirus programs know about O them or an update is installed.f  C I was looking to intercept possibly harmful attachments that users oC might just click on without thinking, or could be be processed by a$D misconfigured email client. Renaming an attachment suffix to make it@ not execute right away (and inserting a warning message into the email) would go a long way.    sol gongolao adl data systems inc.    ------------------------------  * Date: Tue, 4 Feb 2003 15:42:39 +0000 (UTC)+ From: david20@alpha2.mdx.ac.uk (David Webb)eF Subject: Re: Inspecting VMS SMTP email contents before it gets to user+ Message-ID: <b1on1f$pkq$1@aquila.mdx.ac.uk>K  h In article <d7791aa1.0302040603.5bda25f2@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes: >winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr") wrote in message news:<00A1AF62.3D3EDCBC@SSRL.SLAC.STANFORD.EDU>... k >> In article <d7791aa1.0302032014.5b61af9b@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes:sk >> >"Brian Tillman" <tillman_brian@notnoone.notnohow.com> wrote in message news:<3e3ee120$1@news.si.com>...sA >> >> >I need to programmatically intercept email received by VMS C >> >> >and inspect it before it gets POPped to its final recipient.  >> >> > >> >> Why not get Sophos' anti-virus program that runs on VMS? >> >C >> >that's fine if you want to look for vms viruses (such a thing?)a> >> >but to scan smtp bodies you only have one option right now: >> >and that's pmdf ... lousy but that's the way it is ... >>  J >> PMDF processes incoming email and can have the message body scanned by I >> Sophos.  This means that you don't deliver the viral load of infected fN >> messages to PC clients.  Sophos isn't looking for VMS viruses, it's runningP >> on VMS and looking for PC viruses.  (And seems to do a pretty good job, too.) >> eM >> Are you calling PMDF lousy or saying that it's lousy that there's only onec7 >> option?  I think PMDF is pretty darned good, myself.. >> .
 >> -- Alan >> a > C >there is a sophos vms virus scan the previous poster was referringuD >to sold separately, but to get smtp virus scans, you need pmdf, andA >no I didn't call pmdf lousy, but its lousy they will not providei? >a smtp virus scan product for other smtp servers (ie tcpware).t  4 The functionality needed isn't in the TCPIP stacks.  You need to be able to :-M  4 1) Split the message up into each of its attachmentsM 2) Mime decode each attachment (or in the case of older mail clients UUDECODE:    attachments).H 3) Scan for the virus and decide what to do with the message/attachment.    ie0      a) Delete the entire messageo    b) Bounce the entire messageEA    c) Replace the infected attachment with a text warning messagenK       and then continue processing other message attachments and eventually;       deliver the message.  P    This requires passing information back to the mailserver from the commandfile    running the virus scanner.t    J 4) Put the message back together taking care that any message replacements-    are correctly mime encoded and structured.l    N It is very easy to disrupt the structure of a Mime message especially if it is4 complicated to start with eg multi-part alternative.  ( PMDF and MX provide this functionality. K (I've never used MX but have been told on previous occasions that it's SITEoK interface provides this functionality. For more information on MX you might " try the vmsnet.mail.mx newsgroup).  eN Sophos were talking about porting their Mailmonitor product to VMS which wouldI also provide this functionality but I don't know whether they have firmlye commited to this.   J I doubt if any of the TCPIP vendors on VMS would be willing to add in thisF functionality since you would be asking them to massively increase theM functionality of the built in SMTP servers they provide. You would almost be -L asking them to produce a complete mail system like Sendmail or Exim on Unix K without the benefit of a large open source community for development or it sJ being a commercial product. Since commercial systems such as PMDF (and theH semi-commercial MX) exist which are better than Sendmail there seems no L real reason for them to spend money on such a development - especially sinceO the vendor of the TCPWARE and Multinet TCIP stacks is also the vendor for PMDF.-    
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  * Date: Tue, 4 Feb 2003 16:00:58 +0000 (UTC)+ From: david20@alpha2.mdx.ac.uk (David Webb)kF Subject: Re: Inspecting VMS SMTP email contents before it gets to user+ Message-ID: <b1oo3p$psq$1@aquila.mdx.ac.uk>e  c In article <a7337236.0302040707.284f2c35@posting.google.com>, sol@adldata.com (Sol Gongola) writes:aR >> I don't think he specified that he was just wanting to virus scan the messages.F >> He might want to check them for spam or other undesirable contents. >> eN >> Besides which unless Sophos have ported mailmonitor to VMS (which they wereP >> supposed to be considering) then you would still need a means of interceptingR >> the mail message and passing it through Sophos' Vsweep program before delivery. >> oR >> PMDF does this very well as I believe does MX. As far as I am aware none of the? >> VMS TCPIP stacks provides such mail interception facilities.  >> dK >> Alternatively he should be able to do something using the public domain iQ >> version of DELIVER. The simplest such solution would be to put a mail.delivery Q >> file in each individual user's home directory which runs a command file to do  S >> the appropriate scanning and then took the appropriate action eg delivering the aM >> original message, delivering a "censored" message or bouncing the message.  >> t >> t
 >> David Webb  >> VMS and Unix team leadere >> CCSSr >> Middlesex Universityp >s> >Sometimes things get in before Antivirus programs know about   >them or an update is installed. >aD >I was looking to intercept possibly harmful attachments that users D >might just click on without thinking, or could be be processed by aE >misconfigured email client. Renaming an attachment suffix to make it3A >not execute right away (and inserting a warning message into thec >email) would go a long way. >e >sol gongola >adl data systems inc.  K PMDF can do the renaming fairly easily and I would expect that MX would be i able to.L I'd think you could probably do it with DELIVER - you would have to edit theC mail message body and replace any filenames on MIME content-type org content-disposition lines.  sL Inserting a warning message explaining what you have done might be trickier.I This is akin to the practise of putting a disclaimer on mail messages andt% suffers from the same problems see :-   : http://www.goldmark.org/jeff/stupid-disclaimers/apply.html       
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  $ Date: Tue, 4 Feb 2003 12:45:02 -0500A From: "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com>cV Subject: Re: Itanium EFI Console (was: Re: OpenVMS Boots on Itanium on Friday Jan  31), Message-ID: <3e3ffc1f_3@hpb10302.boi.hp.com>  = "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in message ) news:3E3F9ADE.9F961B7A@vl.videotron.ca...r > Hoff Hoffman wrote:3K > >   We have various enhancements in mind for the console environment, and,L > >   there are certainly many various and quite useful capabilities already: > >   integrated into the default EFI console environment. > - > If I may make a suggestion out fo the blue.w >0H > If your EFI environment could be used to boot other operating systems, such as F > Linuz especially, it would be a good PR move to make it available to anyone# > and everyone on the VMS web site.a >n  L We are using *the* EFI environment.  So you can imagine a system that startsI up, puts you into an EFI boot screen, with VMS, Linux, HP-UX, and Windowsh< boot disks that you can select with a Down Arrow and Return.  = That *is* the plan.  That *is* what I'd like to clearly show.s  H In *addition* to that, VMS wants to provide a optional VMS-specific bootJ manager (this is loaded from disk by EFI) that would allow system managersJ to set boot options, and view devices in a VMS-friendly way... including a9 "boot" command if they want to start it right from there.o  L > This way, the Linux community would hear about VMS doing good things, theyL > would go to the VMS web site, download the "improved" EFI environment, try itH > on their machine and see how much better it woudl be than the one that cameK > with the machine/linux. They would then have a favourable opinion of VMS.w  J Don't think this really applies well.  One of the main things we are doingG this for is so you don't need to know how to figure out what VMS devicehF (like DKA0) a EFI pathname is.  Or how to set the now familiar boot osG flags.  Or specify a dump device.  None of these things seem useful forb Linux.   ------------------------------  % Date: Tue, 04 Feb 2003 06:50:13 -0400t0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>Y Subject: Re: Itanium EFI Console (was: Re: OpenVMS Boots on Itanium on Friday Jan 31) 31)s/ Message-ID: <3E3F9ADE.9F961B7A@vl.videotron.ca>    Hoff Hoffman wrote: I >   We have various enhancements in mind for the console environment, andtJ >   there are certainly many various and quite useful capabilities already8 >   integrated into the default EFI console environment.  + If I may make a suggestion out fo the blue.a  N If your EFI environment could be used to boot other operating systems, such asK Linuz especially, it would be a good PR move to make it available to anyonee! and everyone on the VMS web site.M  J This way, the Linux community would hear about VMS doing good things, theyM would go to the VMS web site, download the "improved" EFI environment, try it-K on their machine and see how much better it woudl be than the one that cameoI with the machine/linux. They would then have a favourable opinion of VMS.O   ------------------------------   Date: 4 Feb 2003 08:59:08 -0800e+ From: seanobanion@attbi.com (Sean O'Banion) ' Subject: Re: Newbie with dumb questionse= Message-ID: <f883d5a4.0302040859.718f2e0e@posting.google.com>t  C I have a Microsoft Natural Keyboard plugged into my Alpha 2100, and E have had no problems (that come to mind).  And I'm sure I've used EVEnD with the EDT keypad, so that is ussualy a good test.  Make sure it'sD plugged into the serial port for the Keyboard, and try it on another system to make sure it works.,  ; I could also be that the serial port or plug is damaged ...      Sean  ` jasper <jasper@never.tell> wrote in message news:<9hst3vktelps7e7n07g9ld3oraai73cfua@4ax.com>...3 > On Mon, 3 Feb 2003 09:39:57 -0000, "John Travell"- > <john@travell.uk.net> wrote: >  > 1 > >do you have the graphics working under linux ?.N > >You can verify exactly what card you have installed by shutting the machineM > >down and at the console prompt doing a "show config". If you do this, postd > >the results.oE > >We will be able to see from that if you have the graphics hardwaregM > >installed. While it would be a bit bogus to sell a machine as an 'ultimateoN > >workstation' without a graphics card, until we know it is really there, andN > >which variety it is, any advice on how to get the graphics working could be > >a waste of time.) > G > Yes, graphics are fine in linux. When I boot into vms I see a message E > during the boot process that states something to the effect that no D > graphics hardware is detected. Once logged in and I issue the showC > config command I get: %DCL-W-IVEKYW, unrecognized keyboard--check2 > validity and spelling. i > G > When I received this system it did not have a keyboard or mouse, so I-G > got a MS keyboard and a Logitech mouse. I'll go into the system admin0B > manual and see what I can find there for answers to the keyboard: > issue. I have no idea where to obtain a VT200 keyboard.  > D > As an aside, I today returned the Patrick Holmay book as it didn't? > answer many questions I have. Purchased "Getting Started withtF > OpenVMS--a guide for new users." I think it will be beneficial to myG > learning process. Also d/l'ed and printed most of the Open VMS Systemo? > managers manual which I will dig into over the next few days.t >  > @ > -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----C > http://www.newsfeeds.com - The #1 Newsgroup Service in the World!e@ > -----==  Over 80,000 Newsgroups - 16 Different Servers! =-----   ------------------------------  # Date: Tue, 04 Feb 2003 11:41:04 GMTi" From:   VAXman-  @SendSpamHere.ORG Subject: RE: Nice newsgroup!0 Message-ID: <00A1AFA0.DF1D4A0D@SendSpamHere.ORG>  W In article <01C2CB92.579E6FA0@sulfer.icius.com>, Shane Smith <ssmith@icius.com> writes:b >Welcome to the group, T.R.  >eE >What's going on here, there seem to be newbies popping up out of the H >woodwork. Did VMS get an interesting mention somewhere I didn't notice? >a >Shane  C Perhaps it's the requests over in the Micro$haft newsgroups to have-E all of the Micro$haft protrietary formats on the VMS pages translatedo to something useable.  p   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMc            g5   "Well my son, life is like a beanstalk, isn't it?" d   ------------------------------  $ Date: Tue, 4 Feb 2003 09:07:39 -0500+ From: "Mike Kier" <michael.kier@compaq.com>r Subject: Re: Nice newsgroup!/ Message-ID: <pPP%9.398$5L.207@news.cpqcorp.net>t  / "GreyCloud" <cumulus@mist.com> wrote in messaget! news:3E3F0D5F.6236A4F@mist.com...d > Shane Smith wrote: > >r > > Welcome to the group, T.R. > >dH > > What's going on here, there seem to be newbies popping up out of theK > > woodwork. Did VMS get an interesting mention somewhere I didn't notice?  > >0 > 3 > I do my part in spreading the word about OpenVMS.0) > I'm hopeful that OpenVMS will continue.a  C See my sig :-)  (courtesy of one of the lepers at HPETS, St. Louis)3   --	 Mike Kier10 Consultant, HP Consulting & Integration Services Cincinnati, OH, USAk mike.kier@hp.com  ' "Practice random acts of VMS marketing"r   ------------------------------  $ Date: Tue, 4 Feb 2003 10:04:28 -0000 From: "John" <john@dateline.gg>o, Subject: Re: OpenRDA ODBC Driver for OpenVMS( Message-ID: <b1o352$d49$1@ns2.cwgsy.net>  G I have been down this route recently, the prices and features depend onvB budget and personal preferences, but the difference in quality andL availability in tech support was surprising.  Maybe some of those guys don'tK really *do* support to the evaluation version, but you may never need it asm- much.  It made the final choice a lot easier.       G "Craig A. Berry" <craig.berry@SignalTreeSolutions.com> wrote in messagee7 news:7f15589f.0301311205.5c5fda39@posting.google.com...e= > Earl Lakia <Lakia_n_o__s_p_a_m_@ipact.com> wrote in messagec- news:<Z5ydnfx6WZAnCaSjXTWcqw@netnitco.net>...oB > > I was looking a product to allow me to connect my OpenVMS hostD > > to a Microsoft SQL Server database.  Anyway, found the following8 > > product and wondering if anyone has tried it before. >vG > If it's the one I'm thinking of I did try it once and wasn't terriblysH > impressed.  The user guide showed a weak understanding of what logicalB > names are.  Linking a test program showed that the code in theirD > libraries had compiled with warnings (though the user guide said IA > could ignore them).  That was a couple years ago, so you shouldiH > definitely get a trial version and see how well it works for you.  YouA > should also do the same with EasySoft, Attunity, and CONNX, andP8 > compare prices, features, and tech support.  My $0.02.   ------------------------------  $ Date: Tue, 4 Feb 2003 08:32:54 -05005 From: "Sue Skonetski" <susan.skonetski@hp.nospam.com>R  Subject: OpenVMS Article on CNet* Message-ID: <b1ofe9$s6r$1@web1.cup.hp.com>  L http://msnbc-cnet.com.com/2100-1001-983162.html?type=pt&part=msnbc&tag=alert &form=feed&subj=cnetnews   ------------------------------   Date: 4 Feb 2003 09:30:42 -0600e- From: Kilgallen@SpamCop.net (Larry Kilgallen) $ Subject: Re: OpenVMS Article on CNet3 Message-ID: <9qnGnbzHES7D@eisner.encompasserve.org>s  b In article <b1ofe9$s6r$1@web1.cup.hp.com>, "Sue Skonetski" <susan.skonetski@hp.nospam.com> writes:N > http://msnbc-cnet.com.com/2100-1001-983162.html?type=pt&part=msnbc&tag=alert > &form=feed&subj=cnetnews   Good Job !!!   ------------------------------   Date: 4 Feb 2003 12:40:48 -0600w; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)P$ Subject: Re: OpenVMS Article on CNet3 Message-ID: <7JUuJKC3ll5Y@eisner.encompasserve.org>t  b In article <b1ofe9$s6r$1@web1.cup.hp.com>, "Sue Skonetski" <susan.skonetski@hp.nospam.com> writes:N > http://msnbc-cnet.com.com/2100-1001-983162.html?type=pt&part=msnbc&tag=alert > &form=feed&subj=cnetnews  D    Who the hell is Kusnetzky and what makes him think the transition    from VAX to Alpha was hard?   ------------------------------  * Date: Tue, 4 Feb 2003 18:44:41 +0000 (UTC)3 From: "Richard Maher" <maher_rj@hotspamnotmail.com> $ Subject: Re: OpenVMS Article on CNet. Message-ID: <b1p1mp$7d$1@helle.btinternet.com>   Excellent stuff!  > Sue Skonetski <susan.skonetski@hp.nospam.com> wrote in message$ news:b1ofe9$s6r$1@web1.cup.hp.com... >tL http://msnbc-cnet.com.com/2100-1001-983162.html?type=pt&part=msnbc&tag=alert > &form=feed&subj=cnetnews >e >s >s >  >l   ------------------------------  % Date: Mon, 03 Feb 2003 16:48:36 -0400S0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>6 Subject: Re: OpenVMS Boots on Itanium on Friday Jan 31/ Message-ID: <3E3ED59D.216D796C@vl.videotron.ca>Y   warren sander wrote:G > sue will announce the winners when she gets all the stuff finalized..o  L Hope you will strictly enforce the restriction that render ineligible anyoneD who is friends with or has met any of the engineers. :-) :-) :-) ;-)   ------------------------------  # Date: Mon, 03 Feb 2003 23:27:59 GMTg# From: "John Smith" <a@nonymous.com> 6 Subject: Re: OpenVMS Boots on Itanium on Friday Jan 31I Message-ID: <3WC%9.546842$F2h1.2820@news01.bloor.is.net.cable.rogers.com>-  > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message7 news:cf15391e.0302031219.2be9bf68@posting.google.com...k/ > xeio77@hotmail.com (Tsarkon) wrote in message 9 news:<baa86a4a.0302021111.2db94598@posting.google.com>...l0 > > HP Calculators? Not profitable enough - fireC > > all the engineers in the calculator division. Shut the divisionu down.w >d > > HP Calculators. Gone.n >D > Oh, really?  Check this out: >t7 > HP Expands Its Calculator Line with Two New Offerings ; > http://www.hp.com/hpinfo/newsroom/press/2003/030123b.htmle    C A bit OT: http://www.hp.com/hpinfo/newsroom/press/2003/030103a.htmlj  E Marcello & Gorham better make sure that Burson-Marsteller know how to0C spell OpenVMS and Alphaserver really well. I'd suggest for starters B that they write it in 144-point type ten thousand times across theA pages of the Wall Street Journal and lots of technical magazines.    ------------------------------  # Date: Mon, 03 Feb 2003 23:32:43 GMTR# From: "John Smith" <a@nonymous.com>26 Subject: Re: OpenVMS Boots on Itanium on Friday Jan 31K Message-ID: <v_C%9.546850$F2h1.315339@news01.bloor.is.net.cable.rogers.com>   = "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in messageo) news:3E3ED59D.216D796C@vl.videotron.ca...d > warren sander wrote:= > > sue will announce the winners when she gets all the stuff  finalized..  > < > Hope you will strictly enforce the restriction that render ineligible anyoneuF > who is friends with or has met any of the engineers. :-) :-) :-) ;-)    = Everyone knows that an engineer's best friend is their pocketh
 protector.   ------------------------------  % Date: Tue, 04 Feb 2003 06:42:30 -0400a0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>6 Subject: Re: OpenVMS Boots on Itanium on Friday Jan 31/ Message-ID: <3E3F9910.E838C88D@vl.videotron.ca>v   Fred Kleinsorge wrote: >  > Steve is Sue's husband.b  N Does Sue treat all engineers equally, or does Steve get preferential treatmentD when the time comes for Sue to feed the engineers in their habitat ?   :-) :-) :-)r   ------------------------------  $ Date: Tue, 4 Feb 2003 18:08:30 +0100$ From: Michael Unger <unger@decus.de>6 Subject: Re: OpenVMS Boots on Itanium on Friday Jan 31+ Message-ID: <00A1B000.E31A441B.29@decus.de>   B "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com> wrote:   > [...]b >)% > Plus, since they [SUN] are FABless,6M > they will always have a problem getting onto the bleeding edge processes inn > a timely way.0 >0 > [...]7  : So what exactly was the reason Compaq went FABless too ???   Michaelm   ------------------------------  $ Date: Tue, 4 Feb 2003 12:55:46 -0500A From: "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com>o6 Subject: Re: OpenVMS Boots on Itanium on Friday Jan 31. Message-ID: <3e3ffea3$1_3@hpb10302.boi.hp.com>  1 "Michael Unger" <unger@decus.de> wrote in message-% news:00A1B000.E31A441B.29@decus.de...GD > "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com> wrote: >.	 > > [...]r > >e' > > Plus, since they [SUN] are FABless,0L > > they will always have a problem getting onto the bleeding edge processes in > > a timely way.  > >r	 > > [...]  >d< > So what exactly was the reason Compaq went FABless too ??? >c  K Because unless you are building millions chips, keeping up with the processdE race is very expensive.  Not having a FAB seemed to make things a bitnK harder, and I think lengthened the schedules (just my outside observation).nH The chip builders, like Intel and IBM have a real advantage for building4 high performance chips from the process perspective.   ------------------------------  $ Date: Tue, 4 Feb 2003 10:04:58 -0500' From: "Main, Kerry" <Kerry.Main@hp.com>yY Subject: RE: OpenVMS.org: Marvel article and HP's press release for Marvel and Alpha RetayT Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF402660D22@kaoexc01.americas.cpqcorp.net>   Andrew, Andrew ..)  4 There are a number of different types of benchmarks.  F - industry standard benchmarks that use a given workload to simulate a$ theoretical Customer environment.=20= - Customer or ISV application benchmarks that illustrate realb) performance gains with real applications.d  C This is a ISV application benchmark which compared its results to acF previous server they tested on - which I believe was a large GS Series Alpha system.=20  5 Can this result be compared to another vendor system?a   No.c  C Can Customers using WallStreet financial applications use this as a B general indication of performance gains they might expect in their environment?   Absolutely.a  E Can Customers using other applications use this as an indication thatw@ they might expect significant, tbd, performance gains with their applications?=20  B Absolutely. Of course, they need to benchmark their application to8 better understand just how much gains they might expect.   But, you knew this anyway.   Regardsp  
 Kerry Main Senior Consultante Hewlett-Packard (Canada) Co.! Consulting & Integration Services3 Voice: 613-592-4660o Fax   : 613-591-4477 Email: kerryDOTmain@hpDOTcom-     (remove the DOT's and replace with "."'s)  OpenVMS DCL - the original .COMi     -----Original Message-----' From: Andrew Harrison SUNUK Consultancys1 [mailto:Andrew_No.Harrison_No@nospamn.sun.com]=20o Sent: February 3, 2003 9:01 AM To: Info-VAX@Mvb.Saic.Com<C Subject: Re: OpenVMS.org: Marvel article and HP's press release for-E Marvel and Alpha Retain Trust and Alpha Retain Trust and Alpha Retain  Trustb    1 If thats your idea of a comparable benchmark thenr3 you are in a lot more trouble then current evidence. which isn't great suggests.   7 It has no comparable results on other vendors platformsE= and it isn't obvious what can be attributed to changes in the @ application and what can be attributed to the new Alpha servers.  F We don't even know what configuration the benchmark actually ran on on either the old or new systems.  4 Is this the best that OpenVMS marketing can do ?????   Regardsn Andrew Harrison  Main, Kerry wrote:	 > Robert,t >=20E > And as a way to emphasize what you are saying about increases in=20hJ > OpenVMS performance, folks can review this large Customer application=20H > benchmark that was done last Oct (and additional perf optimizations=20% > have been done since then as well):A >=203 > http://www.wallstreetsystems.com/news/hpbench.htmnH > "New York and London, Tuesday, 1st October 2002 - Wall Street Systems,  D > Inc., a leading provider of global treasury and capital markets=20E > solutions and services, today announced the results of benchmark=20aI > testing of The Wall Street System(r) treasury engine performed under=20mI > laboratory conditions at the Hewlett-Packard Company's New Hampshire=20nI > engineering benchmark center. The Wall Street System treasury engine=20tH > exceeded one million FX transactions per day, representing a five-fold  B > improvement from the previously measured peak performance. It=20D > processed a peak load of approximately 45,000 new deals per hour." >=20G > Over a period of several months, the Wall Street Systems Technical=20sE > Design Team, in collaboration with HP's OpenVMS operating system=20vF > developers, redesigned the FX deal processing modules of The Wall=20H > Street System treasury engine, with a view to further increasing FX=20F > transaction processing rates. The FX deal processing modules take=20H > advantage of new operating system features created by HP expressly for   > Wall Street Systems. >=20D > Mark Tirschwell, Chief Technology Officer, Wall Street Systems,=20E > states: "The tremendous performance gains can be attributed to a=20sI > combination of HP AlphaServer systems and software architecture that=20 F > enables our solution to operate resiliently in a high performance=20J > environment. This is particularly important as The Wall Street System=20J > treasury engine must be able to process a high-volume of transactions=201 > for clients 24 hours a day, seven days a week."a >=20E > The global requirements of the customers of Wall Street Systems,=20 F > running an enterprise-wide solution, demand ongoing investment in=20J > benchmarking. In the specific area of FX trading, the need to process=20B > significant numbers of transactions is essential in running a=20G > worldwide FX operation. During the tests, new, advanced techniques=20nI > were developed that enable The Wall Street System treasury engine to=20eF > run many concurrent FX position processing queues and maximize FX=20C > throughput. The new structure allows deals to be submitted and=20 H > completely processed from a front-office perspective in a single step,  G > ensuring users around the world are working with real-time positions.  >=20I > This industry leading performance was achieved running on the latest=20 D > version of HP OpenVMS (Version 7.3-1) running on HP AlphaServer=20B > systems. "I'm delighted to see this performance breakthrough,=20B > resulting from close collaboration between one of our leading=20H > Financial Services ISVs, and the latest technology from OpenVMS and=20J > our AlphaServer engineering teams," said Mark Gorham, vice president,=20 > HP's OpenVMS Software Group."r >=20	 > Regardsa >=20 > Kerry Main > Senior Consultantl > Hewlett-Packard (Canada) Co.# > Consulting & Integration Servicesu > Voice: 613-592-4660  > Fax   : 613-591-4477 > Email: kerryDOTmain@hpDOTcom/ >     (remove the DOT's and replace with "."'s)e! > OpenVMS DCL - the original .COMT >=20 >=20 > -----Original Message-----; > From: Robert Deininger [mailto:rdeininger@mindspring.com]e  > Sent: February 3, 2003 7:33 AM > To: Info-VAX@Mvb.Saic.Com E > Subject: Re: OpenVMS.org: Marvel article and HP's press release fort > Marvel and Alpha Retain Trust: >=20 >=20H > In article <3E3E575E.5090004@nospamn.sun.com>, Andrew.Harrison@Sun.COM > wrote: >=20 >=20F >>>If these were VMS results, you should realize that VMS does not yetI >>>know anything about "near" and "far" memory on Marvel systems.  And=20gH >>>AFAIK, VMS does not provide a way for applications to figure this out >> >=201 >>>and optimize memory allocation within the app.v >>H >>Which is why the results were done as all big Alpha benchmarks have=20J >>been done since the intro of the GS160/320 with Tru64 which does know=20  >>about local and remote memory. >>2 >>The fact that OpenVMS doesn't apparently support1 >>this makes the performance numbers published byi1 >>HP such as they are even more irrelevent to the  >>OpenVMS community. >=20 >=20B > I'll type slowly so you have a better chance of understanding... >=20 >=20H > VMS does not YET support NUMA-awareness on MARVEL systems.  On Marvel,  A > it was clear from the prototype days that NUMA-awareness has=20oH > negligible benefit for Marvel, at least up to 16P.  The extra work was  C > deferred until after first ship. In short, VMS on Marvel isn't=20.+ > NUMA-aware because it doesn't need to be.o >=20G > VMS has supported NUMA-awareness for quite a while on GS320/160/80=20wF > systems.  For many workloads, it improves performance significantly.   ------------------------------  % Date: Tue, 04 Feb 2003 11:05:56 +0000u' From: Andrew Harrison SUNUK ConsultancyeY Subject: Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retaiu. Message-ID: <3E3F9E94.8030002@nospamn.sun.com>   jlsue wrote:G > On Mon, 03 Feb 2003 15:23:21 +0000, Andrew Harrison SUNUK Consultancyr0 > <Andrew_No.Harrison_No@nospamn.sun.com> wrote: >  >  > E >>>VMS has supported NUMA-awareness for quite a while on GS320/160/80.G >>>systems.  For many workloads, it improves performance significantly.U >>D >>But not enough to stop people having to use Oracle Parallel ServerC >>in a box to get decent throughput. This was a recommendation made.G >>to one GS160 customer when they found that it was in fact slower than # >>their GS140 which had fewer CPU'sn >> >  > E > I have heard this claim before, but never found anything of meat totA > it.  If you have a specific instance to which you refer, pleasecF > provide more details.  Who, what was their problem, who advised them0 > to go to multiple instances with OPS in a box? > E > I have looked throught this newsgroup on google and only ever found F > one entry that mentions slower peformance th an their GS140.  It wasG > not clear at all that they'd ever done any kind of performance tuningiC > on the system.  And there was never any further response from theTE > original poster to obtain enough info to draw any real conclusions.s >   : You havn't looked hard enough, there were a number on this4 group. My favourite response being Rob's who pitched6 Marvel as the solution to the problem, this was a year
 or so ago.  8 In the UK there have been a number of customers who have7 replaced GS320/160s with Suns or IBM's because of these 	 problems.   4 If you want more information contact you account rep/ for the UK's largest mobile phone sales outlet.c   Regardse Andrew HarrisonI   ------------------------------  $ Date: Tue, 4 Feb 2003 13:06:00 -0500* From: "Syltrem" <syltremzulu@videotron.ca> Subject: Oracle 9i and ODS-56 Message-ID: <fmT%9.25557$H67.115275@tor-nn1.netcom.ca>  J I saw a note on Metalink to the effect that Oracle 9.2.2 MUST be installed on an ODS-5 disk.s   Can anyone comment on this?4I Is that only for the binaries, or is it also true for the database files?w  L Can I safely convert a drive from ODS-2 to ODS-5, or is is best to buy a new& disk just for installing Oracle on it?   Any advice appreciated.p   Thanks   --   OpenVMS 7.2-1, Oracle 8.1.6.0r   SyltremsI http://pages.infinit.net/syltrem (OpenVMS related web site - en franais) 8 To reply to myself directly, remove zulu from my address   ------------------------------  $ Date: Tue, 4 Feb 2003 11:46:25 +0100  From: "Ralph Henze" <hen@shd.de>0 Subject: Re: pathworks client connection problem& Message-ID: <b1o996$5ln$1@news.shd.de>  K Just change license pwlmxxxca07.02 from group default to group serverbased. 
 thats all   Ralph Henzew Systementwicklungi --J SHD Datentechnik GmbH & Co, Abteilung Systementwicklung, Rennweg 60, 56626	 Andernach 2 Telefon: (0 26 32) 2 95 -0 Fax: (0 26 32) 2 95-615' mailto:hen@shd.de     http://www.shd.den        3 "Manser" <nmanser@progis.de> schrieb im Newsbeitrage7 news:2178d61f.0302031555.61084049@posting.google.com...  > Hi pathworks users,v >i > . > I have the following home LAN configuration: >  > 1 PC (WIN2000)B > 1 AS 800 5/400 (OpenVMS 7.1,DECNET V 7.1,UCX 4.1,Pathworks 6.0B) >m >   > and have the following License >c! > Active licenses on node SERVER:b >e@ > ------- Product ID --------    ---- Rating ----- -- Version --B > Product            Producer    Units Avail Activ Version Release Terminationo > ....L > PWLMXXXCA07.02     DEC           100  0     100    0.0  (none)      (none)L > PWLMXXXFP05.00     DEC           100  0     100    0.0  (none)      (none)L > PWXXWINAT07.00     DEC           100  0     100    0.0  (none)      (none) >e > .... >iH > when trying to open a share (with win explorer) after successful login8 > the following error appears:  (translated from german) >n >i, > in \\SERVER\<share> could not be connected >SG > no connection to the server could be established, because the maximum: number3 > of active connctions for that account was reached  >7 > auf VMS Side >6 > $admin/analyse >c3 >    ================= EVENT #85 ==================> >i: > Event Time:   4-FEB-2003 00:26:13.64       Node:  SERVER > Process Id:   202001CB< > Event:        No server license for client - access denied" > Event Source: LAN Manager Server > Event Class:  WarningP > ! >       Client:   PRIVAT-2R3XWQB6> >  >  > question:s >n > Which License is required ?M >  >> > Any help is appreciated.   ------------------------------  $ Date: Tue, 4 Feb 2003 17:00:29 +0100% From: "Safir" <axica_nopub@yahoo.com>n0 Subject: Re: pathworks client connection problem/ Message-ID: <2tR%9.413$pF.170@news.cpqcorp.net>n  - "Manser" <nmanser@progis.de> wrote in messageh7 news:2178d61f.0302031555.61084049@posting.google.com...d > Hi pathworks users,l >t > . > I have the following home LAN configuration: >  > 1 PC (WIN2000)B > 1 AS 800 5/400 (OpenVMS 7.1,DECNET V 7.1,UCX 4.1,Pathworks 6.0B) >>  5 you should upgrade to V6.0D to correct many problems.t   >e  > and have the following License >F! > Active licenses on node SERVER:> > @ > ------- Product ID --------    ---- Rating ----- -- Version --B > Product            Producer    Units Avail Activ Version Release Terminatione > ....L > PWLMXXXCA07.02     DEC           100  0     100    0.0  (none)      (none)L > PWLMXXXFP05.00     DEC           100  0     100    0.0  (none)      (none)L > PWXXWINAT07.00     DEC           100  0     100    0.0  (none)      (none) >e > .... >   D if you don't need client-based licensing just revert to server-based	 licensinge only (n concurrent accesses)  + for this reconfigure Pathworks ( $pwconfig)w and answer no to the questionc 1 - start the license server  D retsart Pathworks and check you have server-based licenses available by# $ @sys$startup:pwrk$define_commands  $ pwlic    hth.   ------------------------------  % Date: Tue, 04 Feb 2003 09:56:22 +0000 ( From: Nic Clews <sendspamhere@127.0.0.1> Subject: Re: Question on memoryi) Message-ID: <3E3F8E46.20373E96@127.0.0.1>e   Roose Chua wrote:  > C > I was just wondering when I tried to do a show sys on a node, thesD > memory utilization using monit clust doesn't seem to add up when IH > tried to add up all the pages that the sho sys processes are currentlyH > using. My question is, how can I check who has used all the pages thatF > are reported to being In Used when I do a show memor? Does this mean- > that there is a memory leak on this system?c > H > I am also using PSDC, but it seems that I can not totally identify the? > processes that are using a certain timeframe's average memoryr > utilization. Any advise here?o  F I was in the process of replying to your other message when my browserH crashed (hey ho it's only windows, who cares) and I didn't get around to recomposing the reply.  C The effect you are seeing is that of shared memory under OpenVMS. AiG processes total memory allocation comes from private and shared memory, F that shared memory is also counted as the total allocation for another  process, yet it appears 'twice'.   e.g.  D Process A has 10 private pages, and 20 shared pages, total 30 pages.D process B has 20 private pages, and 20 shared pages, total 40 pages.  C Assuming those 20 shared pages are the same for each process above,tG you're talking a total actual memory allocation of 10 plus 20 (private) H plus 20 shared, equal to 50 pages. a SHOW SYSTEM would probably imply 70H pages were used, but that's not quite right. While you could argue it isH incorrect, well no it isn't really, because whether the memory is sharedF or private, it still comes out of the total process addressable range,# so the figures given do have a use.    -- -? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences0 nclews at csc dot comg   ------------------------------   Date: 4 Feb 2003 07:42:47 -0800h) From: jbrankin@ntlworld.com (Jim Brankin)u Subject: Re: Question on memory = Message-ID: <863f19d6.0302040742.5ff0a28c@posting.google.com>e  Y Nic Clews <sendspamhere@127.0.0.1> wrote in message news:<3E3F8E46.20373E96@127.0.0.1>...  >  > F > Process A has 10 private pages, and 20 shared pages, total 30 pages.F > process B has 20 private pages, and 20 shared pages, total 40 pages.  H That's right. You can get the figures using f$getjpi. Use GPGCNT for theG global pages and PPGCNT for private. Add them together and you will geti the figure sho system reports.     - Jim    ------------------------------  * Date: Mon, 3 Feb 2003 17:49:51 -0500 (EST)+ From: Lord Isildur <isildur@andrew.cmu.edu>v; Subject: Re: SKHPC slaps Gartner, naysayers on comp vms ...rO Message-ID: <Pine.GSO.4.44L-027.0302031748380.9864-100000@unix2.andrew.cmu.edu>i  , NetBSD runs on more than those two combined.   Isildurr    ( On Mon, 3 Feb 2003, Brian Tillman wrote:H > It would seem that only Linux now runs on more hardware platforms than
 > OpenVMS.   ------------------------------  % Date: Tue, 04 Feb 2003 11:12:29 +0000b' From: Andrew Harrison SUNUK Consultancya; Subject: Re: SKHPC slaps Gartner, naysayers on comp vms ...a. Message-ID: <3E3FA01D.6020509@nospamn.sun.com>   David Svensson wrote: T > Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote >  > E >>To be fair to Intel they are doing a lower cost Itanium, trouble isoF >>at >200 million transistors and >70 watts power consumption it isn'tH >>destined for the volume desktop, rack optimised server or blade server	 >>market.o >>	 >>Regardsy >>Andrew Harrison  >  > D > As far as I know Deerfield is destined for rack and blade servers.  ' Sure it is but at 70 watts its too hot.i  = The power budget for an entire blade CPU, Memory, Disk etc inn* a 3U 16 blade enclosure is about 75 watts.  9 Its also too expensive, blades should cost arround 1-1.5K & per blade try doing that with Itanium.   Regardsr Andrew Harrison-   ------------------------------   Date: 4 Feb 2003 08:28:32 -0800 ' From: icerq4a@spray.se (David Svensson)-; Subject: Re: SKHPC slaps Gartner, naysayers on comp vms ...r= Message-ID: <734da31c.0302040828.63a34fea@posting.google.com>c   Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<3E3FA01D.6020509@nospamn.sun.com>...  > David Svensson wrote:oV > > Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote >  > > G > >>To be fair to Intel they are doing a lower cost Itanium, trouble isvH > >>at >200 million transistors and >70 watts power consumption it isn'tJ > >>destined for the volume desktop, rack optimised server or blade server > >>market.I > >> > >>Regardsi > >>Andrew Harrison  > > F > > As far as I know Deerfield is destined for rack and blade servers. > ) > Sure it is but at 70 watts its too hot.d > ? > The power budget for an entire blade CPU, Memory, Disk etc ine, > a 3U 16 blade enclosure is about 75 watts. > ; > Its also too expensive, blades should cost arround 1-1.5K ( > per blade try doing that with Itanium. > 	 > Regards  > Andrew Harrison   F With that definition you exclude all but IA-32 CPUs. Are you comparingC Itanium with IA-32 CPUs? Alpha DS10/20L blades are power hungry andtE expensive and still they run very good. I don't see any problems withs a Deerfield in an DS10/20L?s   ------------------------------  % Date: Tue, 04 Feb 2003 21:24:10 +1030./ From: Mark Daniel <Mark.Daniel@wasd.vsm.com.au>t0 Subject: Re: Status of MySQL and Postgres ports?, Message-ID: <3e3f9bc9_3@news.chariot.net.au>   Hey Jean-Franois,  G good to see your script (and the results of your efforts).  Here's one eI (perhaps for Bill Todd - why not ;^).  Why does it appear that Europeans iD have a greater respect and involvement in supporting VMS (in recent C non-official, non-HPQ ports at least) than those elsewhere?  We're eG certainly interested in evidence to the contrary.  Can it be that Bill VC Gates really does represent the epitomy of capitalist, free-market iG enterprise, so often touted as representing the effective mores of the mC free world (and thereby having the covert respect of even the most 0I ardent domestic critic).  In contrast to the Linus Torvolds and - dare I >C say it - Richard Stallmans of this world?  These wish to socialise  3 software (and perhaps other things).  I'm not sure.A  G Perhaps my entire thinking is presently colo(u)red by a recent viewing sF of the the cinema adaptation of Graham Greene's, The Quiet Americian. C Which as a (I know as an off-topic) piece of cinema is a marvelous hA production by Miramax (IMHO; this company produces so many great tE USA-based movies - sorry Bill (whomever), no citations).  It is also p@ compulsive viewing for the serious cinema goer (sorry, again no  citations).  JPPO.  Enjoy.  I Still; perhaps I'm a little (technologically and culturally) confused by  H all this.  Or is it just that people (everywhere) prefer not to pay for 1 the intellectual property they use?  Hmmm. Woops.   F +--------------------------------------------------------------------+E   Mark Daniel                         http://wasd.vsm.com.au/adelaidefF   mailto:Mark.Daniel@wasd.vsm.com.au (Mark.Daniel@dsto.defence.gov.au)F +--------------------------------------------------------------------+     Jean-Franois PIRONNE wrote:f
 > Hi Richard,8 >  > 	 >>Hi all,. >>K >>I noticed some messages toward the end of last year on a few projects forn* >>porting MySQL and PostgreSQL to OpenVMS. >>J >>I'm just wondering what the status of these is and if there are any home >>pages for the projects.e >> >>Thanks for any info. >> >  > N > The port of MySQL 4.0.9 gamma is done, our work will not be include into the> > official distribution until MySQL 4.1 because 4.0 is frozen. > ; > We will provide a better kit during the forthcoming week.i > % > Status of some of our others ports:: >  > IPFilter: canceled > Q > SWISH 2.2.3: port include into the official distribution (http://swish-e.org/),iO > a kit (2.2.1) can be download from http://wasd.vsm.com.au/wasd/, this packagec, > include SWISHesi developed by Mark Daniel. > From the main page: R > SWISH-E is a fast, powerful, flexible, free, and easy to use system for indexing* > collections of Web pages or other files. > R > An example can be found on my site (http://www.pi-net.dyndns.org/search.html) it: > include VMS, Rdb and some others products documentation. >  > Q > SWIG 1.3.16: port include into the official distribution (http://www.swig.org/)  > From the main page: Q > SWIG is a software development tool that connects programs written in C and C++t > with a varietyI > of high-level programming languages. SWIG is primarily used with common  > scripting languagesfM > such as Perl, Python, Tcl/Tk, Ruby, Guile and MzScheme, however the list ofs > supported Q > languages also includes non-scripting languages such as Java and OCAML. SWIG is  > mostO > commonly used to create high-level interpreted programming environments, user  > interfaces, and asM > a tool for testing and prototyping C/C++ software. SWIG may be freely used,  > distributed, and1 > modified for commercial and noncommercial use. n > : > I have used it to build the Python OpenVMS RTL interfaceQ > (documentation at  http://vmspython.dyndns.org/docs/python_vms/python_vms.html)r >  >  >   > Python 2.3: licensing problem,N > The previous maintainer had put on his work an license which is incompatibleM > with the PSF (Python Software Fundation) license, it is not clear if he has-3 > break the PSF license, we are investigating this.cP > It appear that he has absolutely no concern for the VMS (and Python) communityM > and he want to block any distribution of Python on VMS. If I have correctlymQ > understood the problem,  if you want to build a Python for OpenVMS distribution.@ > you have the choice to violate the PSF license or his license.= > So, currently, it seem impossible to have a Python for VMS.t* > If someone has a solution he is welcome. >  >  > O > Samba 2.2.7.a: port done by my colleague Jean-Yves Collot, this is the latesttN > Samba version, which include support for ODS5 and many enhancement. Kits and= > informations on http://www.pi-net.dyndns.org/anonymous/jyc/iJ > If you have any questions about Samba on VMS please use the mailing listC > (samba-vms@lists.samba.org), Jean-Yves is an active member of it.I >  >  >  >  > Jean-Franois    ------------------------------  % Date: Tue, 04 Feb 2003 21:33:35 +1030-/ From: Mark Daniel <Mark.Daniel@wasd.vsm.com.au>m0 Subject: Re: Status of MySQL and Postgres ports?. Message-ID: <3e3f9dfc$1_3@news.chariot.net.au>   Mark Daniel wrote:K > ardent domestic critic).  In contrast to the Linus Torvolds and - dare I i  % Of course, that should be 'Torvalds'.    ------------------------------  % Date: Tue, 04 Feb 2003 21:45:36 +1030 / From: Mark Daniel <Mark.Daniel@wasd.vsm.com.au>i0 Subject: Re: Status of MySQL and Postgres ports?, Message-ID: <3e3fa0ce_2@news.chariot.net.au>   Mark Daniel wrote:H > of the the cinema adaptation of Graham Greene's, The Quiet Americian.   ( And of course that should be 'American'.  @ Cheese, how one inconsequential caning can make you so skittish!  F +--------------------------------------------------------------------+E   Mark Daniel                         http://wasd.vsm.com.au/adelaide0F   mailto:Mark.Daniel@wasd.vsm.com.au (Mark.Daniel@dsto.defence.gov.au)F +--------------------------------------------------------------------+   ------------------------------  $ Date: Tue, 4 Feb 2003 05:45:16 -0800, From: "James Gessling" <jgessling@yahoo.com>0 Subject: Re: Status of MySQL and Postgres ports?5 Message-ID: <b1og5e$15d4uv$1@ID-46415.news.dfncis.de>*  < "Mark Daniel" <Mark.Daniel@wasd.vsm.com.au> wrote in message& news:3e3f9bc9_3@news.chariot.net.au... > Hey Jean-Franois, >gH > good to see your script (and the results of your efforts).  Here's oneJ > (perhaps for Bill Todd - why not ;^).  Why does it appear that EuropeansE > have a greater respect and involvement in supporting VMS (in recentt= > non-official, non-HPQ ports at least) than those elsewhere?e  B It's always seemed to me that the Europeans have been more open toF other alternatives than the current "style".  Even back to the days ofD "no one ever got fired for buying IBM".  I attribute it to a broader? world view in general.  Years and years ago when I was teachingn; school under the British system, I was amazed how much theyo= knew about California geography while I would be hard pressedn to point out Brussels on a map.    Jim   . p.s. Brussels is where the EU is headquartered   ------------------------------   Date: 4 FEB 2003 16:23:45 GMTs4 From: karcher@thuria.waisman.wisc.edu (Carl Karcher)I Subject: Re: TCP/IP Services v5.3 - how to increase the number of "mbufs" 5 Message-ID: <4FEB03.16234537@thuria.waisman.wisc.edu>u  N In a previous article, "Andy Bustamante" <a_c_bustamante@earthlink.net> wrote:  L ->In the meantime, $ TCPIP SET PROTOCOL TCP/NODELAY will provide a temporaryJ ->work around, at least in my environment.  You'll need this in a suitableL ->startup file since it won't survive a reboot or restart of TCPIP services.  8 It seems to survive reboots on my TCP V5.3 ECO-1 system:   $ ucx sho config prot tcpV TCPt!   Delay ACK:             disabledE ..   $ ucx sho prot tcp /param  TCPt!   Delay ACK:             disabled, ..  K This /NODELAY setting is also the Pathworks speedup mentioned by Alan Greigo* back in 2000 (and why I started using it).   --G -- Carl Karcher, Waisman Computing Services, Waisman Center, UW-Madison-6 --               karcher.nomorespam@waisman.wisc.edu     ------------------------------  $ Date: Tue, 4 Feb 2003 14:44:37 +0200$ From: "Thys de Wet" <thys@sun.ac.za>8 Subject: Type of SCSI disks for Alpha Server 1000A 5/4000 Message-ID: <b1ocjo$2bn3$1@news.adamastor.ac.za>   Hi all  : We have an Alphaserver 1000a 5/400 with faulty SCSI disks.K What type of SCSI disks will work in this server or do we need to invest in   DEC/COMPAQ/HP "ordained" drives?  % We have access to some Seagate SCSI'sw   Thank   # T de Wet, Stellenbosch South Africae   ------------------------------  % Date: Tue, 04 Feb 2003 10:17:10 +0000s( From: Nic Clews <sendspamhere@127.0.0.1>, Subject: Re: Understanding memory allocation) Message-ID: <3E3F9326.AB7A2D27@127.0.0.1>k   Roose Chua wrote:sE > We are trying to benchmark one of our nodes, ie. regular cpu/memory.G > utilization, and needs to know first the data in the last line of thet > $ show mem command:s > H > Of the physical pages in use, 12344 pages are permanently allocated to
 > OpenVMS. > H > How do I determine what processes/parameters comprised the 12344 pagesG > here? I know that this is different on every node, but is it always aDG > percentage of the node's memory, ie. OpenVMS always allocate at leasti > 10%? > C > Also, I would like to get some advise regarding one of our nodes,tH > wherein we are experiencing a very high average memory utilization. WeF > would like to lower this memory utilization, this is why, we want toF > know first the "footprint" of a basic VMS system, ie. no other third@ > party product running yet. I have here some data which you canF > reference. This node is an AlphaServer 1000 5/400 with 1GB of memory# > and primarily runs Oracle v7.3.3:P  H OK. Firstly databases like memory, it brings speed. Now, why do you wantE to reduce your memory allocation, so you have a lot of 'free' memory?aH what for, you may as well pull the memory board if you have a large freeG amount of memory. You would only need a free amount of memory to act asiE a buffer for memory demands, so excessive memory reclamation does noth have to take place.   H OpenVMS generally in average situations gets things approximately right.D There are exceptions, and I know some in engineering would [rightly]F disagree. AUTOGEN uses some simple rules to decide the allocation, andC using feedback data can improve things. You can provide input usingnC MODPARAMS. Also, you can set up process quotas in UAF, SYSGEN or at E process creation time within allowable limits. These factors can haveWC significant impact on how the memory is shared out for the system's-G demands. Some process can be favoured, others not, the choice is yours.t  C You need to take a look at the performance management manual in theaC documentation, relevant to your version of the operating system. It.D explains the principles and you need to gain an understanding beforeF assuming that something is right or wrong. The key question you should; answer is, is the overall performance within the businesses B requirements. If it is, then monitor and keep historical data is aH worthwhile task, otherwise observe, learn your system load, and detailedG memory behaviour can be read up from internals manuals. If you then get'A a problem, you'll be far more knowledgeable with dealing with it.-   -- -? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciencese nclews at csc dot comc   ------------------------------  $ Date: Tue, 4 Feb 2003 11:09:02 -05005 From: "Sue Skonetski" <susan.skonetski@hp.nospam.com>r( Subject: Update on the Technical Journal* Message-ID: <b1ooj2$39u$1@web1.cup.hp.com>  L Just letting folks know that the OpenVMS Technical Journal (VTJ) has been on8 the web for 7 days and we have received over 7,300 hits.  ) Many thanks to the Authors and reviewers.:  
 warm regards,i sue    ------------------------------  $ Date: Tue, 4 Feb 2003 18:06:49 +0100$ From: Michael Unger <unger@decus.de>, Subject: Re: Update on the Technical Journal+ Message-ID: <00A1B000.A6ED57D6.21@decus.de>.  6 "Sue Skonetski" <susan.skonetski@hp.nospam.com> wrote:  N > Just letting folks know that the OpenVMS Technical Journal (VTJ) has been on: > the web for 7 days and we have received over 7,300 hits.  F I just noticed there is a new ZIP version including all the single PDF% files which makes downloading easier.   F But what about a "real" journal instead of a collection of individual,C differently formatted articles? Just like "OpenVMS Times" or "Alpha@ System Times" have been.  + > Many thanks to the Authors and reviewers.-   Agreed!B   MichaelB   ------------------------------  % Date: Tue, 04 Feb 2003 12:50:34 -0500c& From: David M Smith <dsmit115@csc.com>, Subject: Re: Update on the Technical Journal8 Message-ID: <a8vv3vs2sfcr05p0qvogjbeqkrsm46oar7@4ax.com>  H On Tue, 4 Feb 2003 18:06:49 +0100, Michael Unger <unger@decus.de> wrote:  G >But what about a "real" journal instead of a collection of individual,dD >differently formatted articles? Just like "OpenVMS Times" or "Alpha >System Times" have been.d  O I would rather have individual PDF files with different formatting than nothingdL at all. It would be nice to include a small "index file" in the ZIP file, toN provide a summary and set of links to the individual files. I put one together@ for my own use by "trimming down" the HTML code on the web page.I -------------------------------------------------------------------------cI David M. Smith 302.391.8533                       dsmit115 at csc dot comAI Computer Sciences Corporation     (Opinions are those of the writer only)nI -------------------------------------------------------------------------t   ------------------------------  % Date: Mon, 03 Feb 2003 14:58:56 -0800 ( From: Alan Frisbie <Abuse@NelsonUSA.com>$ Subject: Re: Very large disks on VMS- Message-ID: <3E3EF430.6EF43686@NelsonUSA.com>o   JF Mezei wrote:S >  > Jim Geier wrote: > > L > > With the evolution towards larger and larger disks, and now with the EVAL > > moreso than its predecessors, it is possible to have VERY large disks on > > a VMs system.  > O > Back in the 1980s, VMS supported HUGE drives. They were so big, they needed 2aK > males to lift, required a 1/3hp electric motor to spin, 15 ams circuit to  > power EACH drive.u  F I would like to see an RP06 lifted by only two people, male or female.? Heck, I know some people who had trouble just lifting the pack!t   Alan   ------------------------------   Date: 4 Feb 2003 08:54:07 -0800o) From: jbrankin@ntlworld.com (Jim Brankin)y$ Subject: Re: Very large disks on VMS= Message-ID: <863f19d6.0302040854.15d2cd26@posting.google.com>a  d "Jim Geier" <jimgeier@jimgeier.com> wrote in message news:<000301c2cbb0$caebaa40$6b3e11ac@jcgt23>...J > With the evolution towards larger and larger disks, and now with the EVAJ > moreso than its predecessors, it is possible to have VERY large disks onG > a VMs system. Are there any problems or issues regarding how well VMSoJ > will handle disks that are 300, 400, 500 GB or even larger? For example,F > how does VMS handle things like the free block list or cache tables?A > Will these things get unmanageable  with extremely large disks?b >  > --jim geier--d  K The file system records free blocks as a bitmap in [00000]BITMAP.SYS. This eM is cached and the cache size is defined by the ACP_MAPCACHE sysgen parameter. L If BITMAP.SYS becomes very big then it seems reasonable to increase the sizeM of the cache. The cache should be able to hold ALL of the biggest bitmap.sys nK on your system and a bit more besides, IMHO. That way when a disk is almostoM full and the file system is scratching around looking for the last few blockse# it will have the bitmap in memory.    I Since the change which allows smaller cluster sizes I have increased the uL ACP_EXTCACHE parameter to 2000. I don't have any scientific reason for doingJ this but it seems to me that if there are more clusters on the disk it is K likely there will be more extents and therefore a bigger cache seems like alH good idea. Increasing the size of the extent cache makes some disks lessK prone to fragmentation and some more so. I cannot figure out what is going t
 on there.      - Jimk   ------------------------------  $ Date: Tue, 4 Feb 2003 12:34:57 -0500A From: "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com>m$ Subject: Re: Very large disks on VMS. Message-ID: <3e3ff9c2$1_2@hpb10302.boi.hp.com>  G We are currently limited to 32-bit LBN's.  We are evaluating a new file2L system to overcome the size and performance limitations of ODS2/5.  When the5 current stuff was designed, the RP06 was cuting edge.r    5 "Bill Todd" <billtodd@metrocast.net> wrote in message0, news:mJ6dnc72H8u4j6KjXTWcoA@metrocast.net... >s6 > "Jim Geier" <jimgeier@jimgeier.com> wrote in message/ > news:000301c2cbb0$caebaa40$6b3e11ac@jcgt23...pL > > With the evolution towards larger and larger disks, and now with the EVAL > > moreso than its predecessors, it is possible to have VERY large disks onI > > a VMs system. Are there any problems or issues regarding how well VMSkL > > will handle disks that are 300, 400, 500 GB or even larger? For example,H > > how does VMS handle things like the free block list or cache tables?C > > Will these things get unmanageable  with extremely large disks?a > K > My impression is that there may still be 32-bit fields in the file systemaH > (and possibly disk driver) internals that limit supported disk size toL > (probably) 2 TB (2^32 512-byte blocks).  Have you checked the FAQ (if it'sD > not there, then perhaps Fred or Hoff will jump in with an answer)? >l   ------------------------------  $ Date: Tue, 4 Feb 2003 18:03:06 +0100$ From: Michael Unger <unger@decus.de> Subject: Re: VMS on Itaniumy+ Message-ID: <00A1B000.21B2446E.10@decus.de>t  + "David Froble" <davef@tsoft-inc.com> wrote:r   > Michael Unger wrote: >  > [...]g >hH > > A "serial interface" -- as used in real "industry standard" devices,I > > not ones claimed to be a "standard" by specific vendors -- usually iseJ > > a RS-232(-C) or RS-422 interface using a DB9 or DB15 connector. RS-4236 > > and MMJ are specific to DEC/Q/HP as far as I know. >q >eQ > A minimal serial interface is 3 connectors, transmit, receive, and ground.  AnyaQ > specific plugs and/or jacks are for ease of use.  Nice, but not mandatory.  :-)   ) Or data, clock and ground as with I2C ...    Michaeli   ------------------------------  % Date: Tue, 04 Feb 2003 08:22:52 +0000i( From: Nic Clews <sendspamhere@127.0.0.1>' Subject: Re: VMS on Itanium - bugcheck?r) Message-ID: <3E3F785C.5EE689CE@127.0.0.1>i   Carl Perkins wrote:u > 6 > Earl Lakia <Lakia_n_o__s_p_a_m_@ipact.com> writes... ...e? > }So does one get the blue screen of death on a VMS bug check?o > }nH > }I hope the system debugger and crash dump analysis will be available." > }Always a pain with Billy boxes. > }u/ > }Just one of the reasons VMS is still around.i > ? > VMS already does the "blue screen of death" on Alpha systems.a@ > When the console is on the graphics card, it is text on a blueB > background. If you crash, the bugcheck (or whatever) informationB > comes up on the blue screen. It is unsettlingly similar to an NT@ > blue screen of death on those very rare occasions that you see" > one (especially the first time).  H Yes, but on a correctly set up system, you get the system context dumped? so you can perform a meaningful and worthwhile post mortem. Bigi difference (as stated).n  C I have been told it is possible to do dump analysis under NT, but I + suspect these people were short of a kebab.l -- o? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciencese nclews at csc dot com    ------------------------------  $ Date: Tue, 4 Feb 2003 12:49:40 -0500A From: "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com>o' Subject: Re: VMS on Itanium - bugcheck?e. Message-ID: <3e3ffd35$1_3@hpb10302.boi.hp.com>  I Hmmm.  I used "blue" so that it would match what the BIOS initializes theoG card in VGA mode to.  Should I add an option to make it something else?'2 Black?  Maroon?  Line green?  Grey?  Chineese Red?      4 "Carl Perkins" <carl@gerg.tamu.edu> wrote in message& news:3FEB200322295197@gerg.tamu.edu...6 > Earl Lakia <Lakia_n_o__s_p_a_m_@ipact.com> writes... > }Dan Allen wrote:=H > }> Just curious: how long after the boot till the crash ;-) Or did you; > }> do an orderly shutdown? Or perhaps it's still running?a > }> > }> Dan > }e? > }So does one get the blue screen of death on a VMS bug check?w > } H > }I hope the system debugger and crash dump analysis will be available." > }Always a pain with Billy boxes. > } / > }Just one of the reasons VMS is still around.e >w? > VMS already does the "blue screen of death" on Alpha systems.o@ > When the console is on the graphics card, it is text on a blueB > background. If you crash, the bugcheck (or whatever) informationB > comes up on the blue screen. It is unsettlingly similar to an NT@ > blue screen of death on those very rare occasions that you see" > one (especially the first time). > 
 > --- Carl   ------------------------------  % Date: Tue, 04 Feb 2003 13:22:00 -0500r! From: Jim Agnew <jpagnew@vcu.edu> ' Subject: Re: VMS on Itanium - bugcheck?n' Message-ID: <3E4004C8.FD4FED0F@vcu.edu>>  F I vote chinese red!!!!!! to match the old manuals... or, white letters on grey to match v5.5..pC or, change it to match the manual colors of the release, so you cant color-code the version!!!o   jimi   Fred Kleinsorge wrote: > K > Hmmm.  I used "blue" so that it would match what the BIOS initializes the I > card in VGA mode to.  Should I add an option to make it something else? 4 > Black?  Maroon?  Line green?  Grey?  Chineese Red? >  > "Cah   ------------------------------  $ Date: Tue, 4 Feb 2003 13:33:10 -0500A From: "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com>o' Subject: Re: VMS on Itanium - bugcheck?s. Message-ID: <3e400767$1_3@hpb10302.boi.hp.com>  ? But the blue matches the DIGITAL in my original business cards.     . "Jim Agnew" <jpagnew@vcu.edu> wrote in message! news:3E4004C8.FD4FED0F@vcu.edu...nH > I vote chinese red!!!!!! to match the old manuals... or, white letters > on grey to match v5.5.. E > or, change it to match the manual colors of the release, so you canu > color-code the version!!!r >  > jime >i > Fred Kleinsorge wrote: > > I > > Hmmm.  I used "blue" so that it would match what the BIOS initializes  thebK > > card in VGA mode to.  Should I add an option to make it something else?c6 > > Black?  Maroon?  Line green?  Grey?  Chineese Red? > >n > > "Caa   ------------------------------   Date: 4 Feb 2003 12:58:22 -0600d- From: Kilgallen@SpamCop.net (Larry Kilgallen)-' Subject: Re: VMS on Itanium - bugcheck?-3 Message-ID: <cyTnlqkjChZA@eisner.encompasserve.org>j  K In article <3E4004C8.FD4FED0F@vcu.edu>, Jim Agnew <jpagnew@vcu.edu> writes:>H > I vote chinese red!!!!!! to match the old manuals... or, white letters > on grey to match v5.5.. E > or, change it to match the manual colors of the release, so you canl > color-code the version!!!s  G That would preclude an Itanium Galaxy implementation capable of runningy5 widely disparate versions of VMS at the same time :-)d   ------------------------------   End of INFO-VAX 2003.070 ************************