1 INFO-VAX	Fri, 09 Jun 2000	Volume 2000 : Issue 321       Contents: blocks incorrectly marked free" Re: blocks incorrectly marked free" Re: blocks incorrectly marked free" Re: blocks incorrectly marked free/ Re: C bashing (was Re: VMS File Caching Futures / re: C bashing (was Re: VMS File Caching Futures / re: C bashing (was Re: VMS File Caching Futures / Re: C bashing (was Re: VMS File Caching Futures / Re: C bashing (was Re: VMS File Caching Futures > Re: Canadian Association of Compaq Users Symposium rescheduled> Re: Canadian Association of Compaq Users Symposium rescheduled> Re: Canadian Association of Compaq Users Symposium rescheduled> Re: Canadian Association of Compaq Users Symposium rescheduled> Re: Canadian Association of Compaq Users Symposium rescheduled. Re: Charon VAX on PC emulator - Asynch DECNET? Re: DEC C versions, Re: Detecting memory leak in C with debugger, Re: Detecting memory leak in C with debuggerP DFU hazard ? (was Large amount of files in one directory causes very slow perforP Re: DFU hazard ? (was Large amount of files in one directory causesvery slow perP Re: DFU hazard ? (was Large amount of files in one directory causesvery slow perP Re: DFU hazard ? (was Large amount of files in one directorycausesvery slow perf Fibre Channel and VMS 7.2-1  Re: Fibre Channel and VMS 7.2-1  Re: Fibre Channel and VMS 7.2-1 $ Re: Hobbyist vms 7.2 alpha questions$ Re: Hobbyist vms 7.2 alpha questionsG Re: Large amount of files in one directory causes very slow performance - Re: LN15AC toner cartridge compatible to ...?  Re: MMOV 2.2 and VMS 7.1 Re: Nic-to-Nic failover A Re: OpenVMS commentaries (was Re: Gartner commentary on Wildfire)  Process States Public Shell Comapny% Re: Question on using ASTs for output ( Submit /After delta time format question, RE: Submit /After delta time format question, Re: Submit /After delta time format question, Re: Submit /After delta time format question, RE: Submit /After delta time format question Re: system slow  TAR FOR VMS? Re: VAX on Intel?  Re: VAX on Intel?  Re: VAX on Intel?  Re: vaxstation LED meanings? Re: VMS File Caching Futures7 Which cpu does Sun server use: I386, Mip, Ppc or Alpha? ; Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha? ; Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha? ; Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha? ; Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha?   F ----------------------------------------------------------------------  $ Date: Fri, 9 Jun 2000 18:51:48 +0200> From: "Jean-Franois Marchal" <jean-francois.marchal@x9000.fr>' Subject: blocks incorrectly marked free 3 Message-ID: <8hr7em$203a$1@s2.feed.news.oleane.net>    Bonjour  tous ...  A Running analyze/disk, I got some "blocks incorrectly marked free" C messages. As I already had to repair my disk yesterday for the same A reasons, I'd like to identify what files are affected, suspecting @ Advanced Server (7.2A on OpenVMS Alpha 7.2-1 + current patches).   I found a example in DSNF "Example-MACRO Givent an LBN, how to find the file it is allocated to"  E But when I can't compile with $ MACRO FIND_LBN.MAR (ends with errors)   ( How could I build the executable image ?& Will it run on an ODS5 shadowed disk ?   Cordialement ...   Jean-Franois Marchal  X9000 - LYON   ------------------------------   Date: 9 Jun 2000 19:15:00 +0200 * From: eplan@kapsch.net (Peter LANGSTOEGER)+ Subject: Re: blocks incorrectly marked free * Message-ID: <39412614$1@news.kapsch.co.at>  t In article <8hr7em$203a$1@s2.feed.news.oleane.net>, "Jean-Franois Marchal" <jean-francois.marchal@x9000.fr> writes:B >Running analyze/disk, I got some "blocks incorrectly marked free"
 >messages.  L Bad sign. That's is the first step. And the next step will then be "multiply+ allocated blocks" and BINGO, you lost data.   : >As I already had to repair my disk yesterday for the sameB >reasons, I'd like to identify what files are affected, suspectingA >Advanced Server (7.2A on OpenVMS Alpha 7.2-1 + current patches).   D It could be, that ASOVMS exploits the bug, but I think the bug is in3 the hardware or in VMS. Check the VMS ECOs again...    >I found a example in DSN G >"Example-MACRO Givent an LBN, how to find the file it is allocated to"  > F >But when I can't compile with $ MACRO FIND_LBN.MAR (ends with errors)  I It is a VAX MACRO32 source code. To compile it on Alpha, you need to make D some modifications (hell, why can't DEQ write such programs for bothI platforms from the beginning - ok, original is dated from 1988, there was H no Alpha then, but this article was updated at least 1999 and there wereI a lot of Alphas at this time - or is this really impossible to write such & MACRO32 programs for both platforms ?)  ) >How could I build the executable image ?    I think, VAXman will know it.   ' >Will it run on an ODS5 shadowed disk ?    Don't know.    --  < Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888 < FBFV/Information Services           E-mail  eplan@kapsch.netF <<< KAPSCH AG  Wagenseilgasse 1     PSImail PSI%(0232)281001141::EPLANH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"N "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998   ------------------------------  # Date: Fri, 09 Jun 2000 17:52:36 GMT 7 From: "Lyle W. West" <lyle.west@childrenshc.org.nospam> + Subject: Re: blocks incorrectly marked free 6 Message-ID: <3940E85D.100BF61E@childrenshc.org.nospam>   Peter LANGSTOEGER wrote: > =   J > In article <8hr7em$203a$1@s2.feed.news.oleane.net>, "Jean-Fran=E7ois Ma=/ rchal" <jean-francois.marchal@x9000.fr> writes:    <snip>   > >I found a example in DSN J > >"Example-MACRO Givent an LBN, how to find the file it is allocated to"=   > > H > >But when I can't compile with $ MACRO FIND_LBN.MAR (ends with errors) > =   J > It is a VAX MACRO32 source code. To compile it on Alpha, you need to ma= keF > some modifications (hell, why can't DEQ write such programs for bothJ > platforms from the beginning - ok, original is dated from 1988, there w= asJ > no Alpha then, but this article was updated at least 1999 and there wer= e J > a lot of Alphas at this time - or is this really impossible to write su= ch( > MACRO32 programs for both platforms ?)   <snip>  F Actually, that article contains 2 source examples, one for Vax and oneB for Alpha. The Alpha example is in the second half of the article.C However, I didn't take the time to build the Alpha version so if it % does not work, feel free to blast me.    --  0 My opinions seldom reflect those of my employer.   Lyle W. West   ------------------------------  $ Date: Fri, 9 Jun 2000 19:59:17 +0200> From: "Jean-Franois Marchal" <jean-francois.marchal@x9000.fr>+ Subject: Re: blocks incorrectly marked free 3 Message-ID: <8hrb68$21eb$1@s2.feed.news.oleane.net>   7 "Peter LANGSTOEGER" <eplan@kapsch.net> wrote in message $ news:39412614$1@news.kapsch.co.at...D > In article <8hr7em$203a$1@s2.feed.news.oleane.net>, "Jean-Franois1 Marchal" <jean-francois.marchal@x9000.fr> writes: D > >Running analyze/disk, I got some "blocks incorrectly marked free" > >messages. > D > Bad sign. That's is the first step. And the next step will then be	 "multiply - > allocated blocks" and BINGO, you lost data.     ! Along the analyze messages, I get = - BADHEADER, invalid file headers (like my post of yesterday) *   - BAD-STRUV_LEVEL, STRUCLEV field is bad2   - INVHEADER_BUSY, invalid file headermarked busy. - ALLOCCLR, blocks incorrecly marked allocated) - ALLOCSET, blocs incorrectly marked free   6 I noticed many backup problems on my configuration ...3 on another disk (just replaced by Compaq yesterday) % ANALYZE/MEDIA/EXERCIZE show no errors 2 but backup can raise verify errors when backing up files from disk to disk   8 I will try to swap some parts and cables this week-end .   Any tip will be welcomed   Cordialement
 Jean-Franois    ------------------------------  % Date: Fri, 09 Jun 2000 18:35:10 +1000 4 From: Huw Davies <Huw.Davies@kerberos.davies.net.au>8 Subject: Re: C bashing (was Re: VMS File Caching FuturesD Message-ID: <4.3.1.2.20000609182616.01d18ba0@kerberos.davies.net.au>  + At 15:31 6/06/00 -0400, Dan Sugalski wrote:   K >Now, yes. At the time C was created, no, since Pascal didn't exist. C's a  G >step up from assembly language, and a half-step down from most of the  L >ALGOL derivatives. For what it did, it beat ALGOL, COBOL, or Fortran. (And ( >B, I assume, though I've never used it)  L Well I did a fair bit of programming in BCPL (from which B was derived) and G I'd say C was a backward step but of course this is a personal opinion.   H >*Now*, yes. When C was created there *were* no optimizing compilers to M >speak of. They didn't come along for quite some time, like a decade or more.   I I have no exact date for the development of the first C compiler, but in  L "The C Programming Language" by Kernighan & Ritchie it's mentioned that the K B compiler was written by Thompson in 1970 to run on a PDP-7, so it's fair  J to assume that C came into existence in 1971-72. I can assure you by this G stage optimizing compilers existed (in fact the first FORTRAN compiler  G written at IBM had significant optimizations built in) but by 1972 the  K BLISS-10 compiler had significant optimization techniques although nothing  K as significant as can be found in modern compilers (instruction scheduling   just makes by brain hurt).  L >I'll admit that at this point I've found very few computer languages worth K >much. I'm fond of perl and Forth, and Python looks reasonably nice though  H >at this point if a language makes me cater to the machine I don't much K >care for it. I've got CPU power and memory to burn, and the languages can  1 >damn well cater to *me*, thank you very much. :)   I Well most of the languages I like to program in don't have compilers for  K OpenVMS, although I belive that there is a BCPL compiler for VAX somewhere  D around and maybe even one for Algol-68. Fortunately there's a BLISS , compiler for those real programming jobs....      @ Huw Davies           | e-mail: Huw.Davies@kerberos.davies.net.au?                       | "If God had wanted soccer played in the =                       | air, the sky would be painted green"     ------------------------------  % Date: Fri, 09 Jun 2000 11:21:42 +0100 / From: Nigel Arnot <sysmgr@maxwell.ph.kcl.ac.uk> 8 Subject: re: C bashing (was Re: VMS File Caching Futures6 Message-ID: <009EB58D.5A2D04B0.5@maxwell.ph.kcl.ac.uk>  % system@niuhep.physics.niu.edu writes:   1 > "Dann Corbit" <dcorbit@solutionsiq.com> writes:  > ) > >A good workman never blames his tools.  > * > A good workman chooses his tools wisely. > B > A master sculpter won't get very far on a block of granite using > a pen knife. > G > I have managed to do many things on the cheap, camping, handyman work E > around the house, etc..  But boy, it was so much nicer when I could * > afford good tools and high quality gear. > ; > Sorry, but that particular bromide rubs me the wrong way.    Me too. I'd add:  M A good toolmaker puts sharp edges only where needed, and where practical puts J guards around them to protect the (tired, inexperienced, careless) worker.  H A good workman improves his tools, if there is a practical way to do so.K (Magnetising a screwdriver is a common example ... and yes, there are a few I occasions when an unmagnetised one is better. You want both in your kit).   D There are indeed folks out there improving the C tool. Some build onI it: TCL/TK, Perl, Python. Others learn from the errors and try to replace ? it: C++ (not a total success in my book), Java. And of course,  F everyone who writes general-purpose libraries and makes them availableB to the public, thereby saving all from the umpteenth re-invention 
 of the wheel.   K I'm not vehemently anti-C, but the protestations of those who seem to think K it's perfect baffle me. As for null-terminated strings ... bleugh, but they D aren't compulsory. The only place they feature in C itself is that aJ double-quoted constant string gets a null tacked onto the end without you L having to code it. UNIX is what to blame for the prevalence of these things.   	Yours, 
 		Nigel Arnot - 		NRA@MAXWELL.PH.KCL.AC.UK                      7 		"In the beginning there was nothing, which exploded."    ------------------------------  # Date: Fri, 09 Jun 2000 11:23:29 GMT * From: westprog 2000 <westprog@my-deja.com>8 Subject: re: C bashing (was Re: VMS File Caching Futures) Message-ID: <8hqk37$n0s$1@nnrp1.deja.com>   6 In article <009EB58D.5A2D04B0.5@maxwell.ph.kcl.ac.uk>,2   Nigel Arnot <sysmgr@maxwell.ph.kcl.ac.uk> wrote:' > system@niuhep.physics.niu.edu writes:  > 3 > > "Dann Corbit" <dcorbit@solutionsiq.com> writes:  > > + > > >A good workman never blames his tools.  ... = > > Sorry, but that particular bromide rubs me the wrong way.    > Me too. I'd add:  @ > A good toolmaker puts sharp edges only where needed, and where > practical putsD > guards around them to protect the (tired, inexperienced, careless)	 > worker.   E For example - why should addEntryToDatabase() be a different function F to addEntrytoDatabase()? Who thought that was a good idea? Until C, noG programming language was case sensitive. It was an obviously dumb idea.   F > A good workman improves his tools, if there is a practical way to do > so. G > (Magnetising a screwdriver is a common example ... and yes, there are  > a few E > occasions when an unmagnetised one is better. You want both in your  > kit).   F > There are indeed folks out there improving the C tool. Some build onC > it: TCL/TK, Perl, Python. Others learn from the errors and try to 9 > replace it: C++ (not a total success in my book), Java.   D C++ took a language where a lot of code was unreadable by humans andD allowed us to produce programs that the compilers got confused by. CF was at least a simple language. Java at least had the sense to abandonA backward compatibility and the libraries, and to try to do things D right. A lot of the worst features of C/C++ have been fixed in Java.   > And of course,H > everyone who writes general-purpose libraries and makes them availableC > to the public, thereby saving all from the umpteenth re-invention  > of the wheel.   G The STL is a great thing - but why is it harder to use than it needs to  be?   G > I'm not vehemently anti-C, but the protestations of those who seem to  > think D > it's perfect baffle me. As for null-terminated strings ... bleugh,
 > but theyF > aren't compulsory. The only place they feature in C itself is that aG > double-quoted constant string gets a null tacked onto the end without  > you F > having to code it. UNIX is what to blame for the prevalence of these	 > things.   F It is the tight coupling of C with UNIX and Windows that has done mostG damage. Instead of mapping language features to the OS in a transparent A way, C forces its way into our lives even when we are using other E languages. If I want to make an API call in Delphi, I have to convert  to a null-terminated string.   -- J.    & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------   Date: 9 Jun 2000 14:10:45 GMT ) From: leslie@clio.rice.edu (Jerry Leslie) 8 Subject: Re: C bashing (was Re: VMS File Caching Futures' Message-ID: <8hqtt5$g6j$1@joe.rice.edu>   5 Huw Davies (Huw.Davies@kerberos.davies.net.au) wrote: - : At 15:31 6/06/00 -0400, Dan Sugalski wrote:   J : >*Now*, yes. When C was created there *were* no optimizing compilers to  : >to speak of.   F Univac's FORTRAN V, ca. ~1967,  was an optimizing compiler. It had an F Algol-like concept of internal subprograms, who could share variables I with the program unit being compiled; e.g. the variable "I" was the SAME  D storage location when used in the SUB subroutine as well as in MAIN:         PROGRAM MAIN       INTEGER I        CALL SUB       SUBROUTINE SUB CCC   INTEGER I        I = 0 	       END   A If "I" was declared in the internal subprogram, it was treated as  a local variable.   F IIRC, that feature caused some of the bugs reported to Univac, caused  by errors in the optimization.  E : >They didn't come along for quite some time, like a decade or more.   K : I have no exact date for the development of the first C compiler, but in  N : "The C Programming Language" by Kernighan & Ritchie it's mentioned that the M : B compiler was written by Thompson in 1970 to run on a PDP-7, so it's fair eL : to assume that C came into existence in 1971-72. I can assure you by this I : stage optimizing compilers existed (in fact the first FORTRAN compiler  I : written at IBM had significant optimizations built in) but by 1972 the oM : BLISS-10 compiler had significant optimization techniques although nothing SM : as significant as can be found in modern compilers (instruction scheduling d : just makes by brain hurt).    4 --Jerry Leslie     (my opinions are strictly my own)   ------------------------------  # Date: Fri, 09 Jun 2000 14:56:26 GMTo! From: Ian Parker <parker@gol.com>o8 Subject: Re: C bashing (was Re: VMS File Caching Futures& Message-ID: <kf$AOAAm8PQ5Ewoq@gol.com>  H In article <4.3.1.2.20000609182616.01d18ba0@kerberos.davies.net.au>, Huw1 Davies <Huw.Davies@kerberos.davies.net.au> writeso, >At 15:31 6/06/00 -0400, Dan Sugalski wrote: >?L >>Now, yes. At the time C was created, no, since Pascal didn't exist. C's a H >>step up from assembly language, and a half-step down from most of the M >>ALGOL derivatives. For what it did, it beat ALGOL, COBOL, or Fortran. (And n) >>B, I assume, though I've never used it)e >oM >Well I did a fair bit of programming in BCPL (from which B was derived) and cH >I'd say C was a backward step but of course this is a personal opinion. >.I >>*Now*, yes. When C was created there *were* no optimizing compilers to nN >>speak of. They didn't come along for quite some time, like a decade or more. >pJ >I have no exact date for the development of the first C compiler, but in M >"The C Programming Language" by Kernighan & Ritchie it's mentioned that the  L >B compiler was written by Thompson in 1970 to run on a PDP-7, so it's fair K >to assume that C came into existence in 1971-72. I can assure you by this nH >stage optimizing compilers existed (in fact the first FORTRAN compiler H >written at IBM had significant optimizations built in) but by 1972 the L >BLISS-10 compiler had significant optimization techniques although nothing L >as significant as can be found in modern compilers (instruction scheduling  >just makes by brain hurt).M > M >>I'll admit that at this point I've found very few computer languages worth  L >>much. I'm fond of perl and Forth, and Python looks reasonably nice though I >>at this point if a language makes me cater to the machine I don't much oL >>care for it. I've got CPU power and memory to burn, and the languages can 2 >>damn well cater to *me*, thank you very much. :) >sJ >Well most of the languages I like to program in don't have compilers for L >OpenVMS, although I belive that there is a BCPL compiler for VAX somewhere E >around and maybe even one for Algol-68. Fortunately there's a BLISS r- >compiler for those real programming jobs....h >. >  >oA >Huw Davies           | e-mail: Huw.Davies@kerberos.davies.net.au @ >                      | "If God had wanted soccer played in the> >                      | air, the sky would be painted green"  >e >a  F And if I remember correctly BCPL was another byte-code/virtual machineE interpreted implementation, almost like Python (which does have a VMSe implementation)!   Regards    Ianp -- a
 Ian Parker   ------------------------------  # Date: Fri, 09 Jun 2000 04:42:17 GMTD. From: "Curtis Rempel" <curtis.rempel@home.com>G Subject: Re: Canadian Association of Compaq Users Symposium rescheduledn9 Message-ID: <JA_%4.6694$Xn5.98659@news1.rdc1.ab.home.com>t  A "Dr Alain Legault" <alain.legault4@sympatico.ca> wrote in messages4 news:W2Y%4.44026$uw6.950817@news20.bellglobal.com... >p? > > If you ask me (from Cambridge, Massachusetts) "How is DECUS @ > > Canada doing?" the answer would be "They still haven't fixedB > > their VMS hobbyist problem."  In fact, if you ask me how DECUSC > > chapters around the world are doing, the answer would be "DECUSS7 > > Canada still can't fix their VMS hobbyist problem."d > >o@ > > Designing a conference and then convincing people to want it@ > > is so much less important than delivering the one thing they( > > _do_ want that only you can provide. >h > All they need do3 > is contact Cathy Zigomanis at 416-228-3934, or atd cathy.zigomanis@compaq.com > to get the ball rolling.  J Talk is cheap.  How many times have we heard this?  (answer: "n", pick you your favorite value for "n")  H How many times have we tried to get an answer on the phone or via email? (answer: "n", same value)   < Net result (careful, don't try this at home): n - n = ZILCH.  J The whole idea with a phone number or email address is to actually use it. Ahem.   J I spoke with three people at work just today with the exact same empirical9 data.  I suggested they send their requests to you Alain.    Curtis   ------------------------------  # Date: Fri, 09 Jun 2000 04:42:17 GMT . From: "Curtis Rempel" <curtis.rempel@home.com>G Subject: Re: Canadian Association of Compaq Users Symposium rescheduledS9 Message-ID: <JA_%4.6693$Xn5.98659@news1.rdc1.ab.home.com>   F "Larry Kilgallen" <Kilgallen@eisner.decus.org.nospam> wrote in message% news:x3x5SVFBYdQa@eisner.decus.org...-K > In article <393FA301.41007155@amse.mets.nrcan.gc.ca>, "Dr. Alain Legault" % <alain@amse.mets.nrcan.gc.ca> writes:r >hA > >> > It is my opinion that the only way in which to rectify theg shortcomings of0I > >> > what has become of DECUS Canada in its incarnanation as CANACU, ist for the7= > >> > entire board to resign and for an election to be held." > >nK > > I fail to see what purpose this would serve.  First because an election  is going on  and wei > > have a lotI > > of difficulty finding people who are "ready to step up to the plate".  Second because I don't > > think thatI > > the board "hasn't done anything".  It certainly failed to communicateu the progresses but toe > > assume thate0 > > nothing has been done is at best simplistic. >t? > The one resounding theme we have heard over and over again is A > that of all the countries in the world it is Canada whose DECUS A > organization is singularly unable to get it together on the VMSl > hobbyist license.e  @ more accurately, "singularly unable to get it together - PERIOD"   ------------------------------  # Date: Fri, 09 Jun 2000 11:38:55 GMTl6 From: "Dr Alain Legault" <alain.legault4@sympatico.ca>G Subject: Re: Canadian Association of Compaq Users Symposium rescheduled : Message-ID: <jH405.50343$uw6.986806@news20.bellglobal.com>   > Talk is cheap.  . So is "Monday morning Quaterbacking"... :-))))  L > I spoke with three people at work just today with the exact same empirical; > data.  I suggested they send their requests to you Alain.k >n  C Communication is a two way process and to blame "everything" on then@ "emitter" is simplistic.  I can't help but to detect a strain ofI "entitlement" posturing....which is one of the reasons of the downfall ofeC the relationship between DECUS and DEC worlwide.  This predated the." "acquisition" by a very long time.  
 Alain Legaulto   ------------------------------  $ Date: Fri, 9 Jun 2000 10:56:36 -0400- From: "Peter Weaver" <peter.weaver@stelco.ca>lG Subject: Re: Canadian Association of Compaq Users Symposium rescheduledl. Message-ID: <sk21d8abh5176@corp.supernews.com>  ! Peter Weaver wrote in message ...  > ...-B >for a new logo design, since then I have not received any renewalE >notices, any invitations to anything, any notices about elections...L > ...   E Correction, when I arrived home last night I had an invitation to then5 now rescheduled Symposium waiting for me in the mail.q   ------------------------------  $ Date: Fri, 9 Jun 2000 11:24:43 -0400- From: "Peter Weaver" <peter.weaver@stelco.ca> G Subject: Re: Canadian Association of Compaq Users Symposium rescheduledn. Message-ID: <sk231uv7h5147@corp.supernews.com>  $ Larry Kilgallen wrote in message ... > ...e< >If you ask me (from Cambridge, Massachusetts) "How is DECUS= >Canada doing?" the answer would be "They still haven't fixed ? >their VMS hobbyist problem."  In fact, if you ask me how DECUS0@ >chapters around the world are doing, the answer would be "DECUS4 >Canada still can't fix their VMS hobbyist problem."    - The answer to "How is DECUS Canada doing?" ise       DECUS Canada has passed on!      This user-group is no more!t     It has ceased to be!-     It's expired and gone to meet it's maker!u6     It's a stiff! Bereft of life, it's rests in peace!  -     It's metabolic processes are now 'istory!      It's off the twig!     It's kicked the bucket,s'     It's shuffled off it's mortal coil,WC     run down the curtain and joined the bleedin' choir invisibile!!.     THIS IS AN EX-USER-GROUP!!    C There is a new "Canadian Association of Compaq Users" (or somethinglF like that) and people who had a paid membership in DECUS Canada becameF members of this new group by default. According to some e-mail message@ I received from Ron Catcheside we should have received a packageF inviting us to become charter members in this new group last December.@ If any of the previous DECUS Canada members did not receive thisB (maybe the post office only lost my package) then let Cathy or RonC know. As I said in a private note to Ron a few minutes ago, "PleaseuC send me whatever you can, I have not yet decided if I will throw mytA money away at this new group. I know that the last time I paid myt% membership fee to DECUS was a waste."u     -- Peter Weaver   ------------------------------  $ Date: Fri, 9 Jun 2000 12:40:23 -0400- From: "Peter Weaver" <peter.weaver@stelco.ca> 7 Subject: Re: Charon VAX on PC emulator - Asynch DECNET?p. Message-ID: <sk28dslrh5128@corp.supernews.com>  ? Peter Moreton wrote in message <8hodnd$2ds$1@nnrp1.deja.com>...f6 >In article <Ioe%4.240$ru5.80198@typhoon.aracnet.com>,6 >  "Zane H. Healy" <healyzh@shell1.aracnet.com> wrote:) >> Alan Greig <agreig@my-deja.com> wrote:y> >> > Now to beg/steal/borrow a libretto (the shirt pocket PC). >>C >> Good luck.  I missed a chance to get one about a year and a half: agoD >andB >> have regretted it since.  Not sure how well it would work for a >Pocket VAX,+ >> but it would make a great pocket PDP-11!w >> >> Zanee >> >"B >I just booted OpenVMS 6.2 on a Libretto 110CT, and I can tell you it'sE >the coolest thing in all of christendom, VMS in my pocket! (well, ifb$ >you have A5 sized pockets that is.) >  >--g >Peter Moreton,A >Northamptonshire, >UK. >> > ' >Sent via Deja.com http://www.deja.com/  >Before you buy.  = This product is the greatest thing since sliced bread, I havem@ installed it on my WIN NT workstation in the office, on my WIN98E desktop and on my WIN98 laptop. It works great on all three, even the 1 P150/32MB laptop is faster than my original MVII.s     -- Peter Weaver   ------------------------------  % Date: Fri, 09 Jun 2000 10:49:51 -0400w( From: Ed Vogel <edward.vogel@compaq.com> Subject: Re: DEC C versionsb* Message-ID: <3941040E.3990078B@compaq.com>   Arne Vajhj wrote:  3 > I thougth DEC C went 5.6-5.7-6.0-6.1, but someonee > just mentioned a 6.0A !t >  > ???? >n > Arne  G    There is not V6.0A, but rather a V6.2A (as another reply mentioned).pE    V6.2A is a release which contains the ECO fixes to V6.2.   I don'ts have a nice H     web pointer for the V6.2A release (it's pretty new), but you can get  (     information on the ECO's to V6.2 at:  J  http://ftp.support.compaq.com/patches/public/Readmes/vms/ccae03062.README  3                                            Ed Vogela@                                            Compaq C Engineering.   ------------------------------  " Date: Fri, 9 Jun 2000 13:52:14 GMT9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)t5 Subject: Re: Detecting memory leak in C with debuggere+ Message-ID: <iE2jjsEX$bLG@eisner.decus.org>s  a In article <8hp8os$9rt@gap.cco.caltech.edu>, mathog@seqaxp.bio.caltech.edu (David Mathog) writes:r  G > The last time I checked there was an interesting bug in that program.qK > Apparently there's no wait state in the code that checks the state of the M > pull down menus.  So if I pull one down in the heap analyzer window and sit M > there thinking, while I'm thinking the program goes to 100% of the CPU and iM > all the other windows come to a grinding halt.  Ok to use it, but be quick   > with those menu choices!  0 How long ago did you report this to DEC/Compaq ?  : Did you get a response (if you are a supported customer) ?   ------------------------------   Date: 9 Jun 2000 15:28:01 GMT22 From: mathog@seqaxp.bio.caltech.edu (David Mathog)5 Subject: Re: Detecting memory leak in C with debugger., Message-ID: <8hr2e1$5s3@gap.cco.caltech.edu>  \ In article <393EAF45.A792C503@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:O >I am running some tests on various modules to ensure quality control. In using N >a couple of modules together, I have noticed a memory leak. But when steppingN >through execution with the debugger, nothing obvious appears. (all the "free", >statements seems to execute as they should. >(F >So obviously, I have coded something wrong. (and yes, this is full of$ >pointers, and dynamic structures).   - The way I usually find these sorts of errors:y  J 1.  Make sure that you use the extended decc$free and not the regular one,J and that the wrapper routine around it sets all ptr's passed to it to NULL after they are free'd.    7 int decc$free(void *ptr); /* prototype for decc$free */t void my_free( void *ptr){h #include <lib$routines.h>e #include <ssdef.h>   if(decc$free(ptr) != 0){1      (void) printf("illegal free() operation\n");d#      (void) lib$signal(SS$_ACCVIO);t   }c   ptr=NULL;o }   I 2.   Replace malloc with my_malloc, which will only allocate memory to a d4 NULL pointer.  (Left as an exercise for the reader!)  " 3.  In a header file somewhere do:   #define free   my_free #define malloc my_malloc  A 3.  Put a breakpoint in each routine at the lines branched to on i conditional failure.  I That should catch most things.  The one thing it won't catch though is a wK situation where you have a linked list that just keeps growing and growing,,F with no pointers being created and then never free'd/deleted.  But if I you're making the common mistake of malloc'ing over a pointer that still  7 points to valid data this will usually find it for you.s   Regards,   David Mathog mathog@seqaxp.bio.caltech.edu ? Manager, sequence analysis facility, biology division, Caltech c   ------------------------------  % Date: Fri, 09 Jun 2000 12:50:47 +0100e2 From: Chris Sharman <Chris.Sharman@CCAgroup.co.uk>Y Subject: DFU hazard ? (was Large amount of files in one directory causes very slow perforh. Message-ID: <3940E827.2B5575EB@CCAgroup.co.uk>   John Nixon wrote:sM > Someone else mentioned DFU to defrag the directory.  I am a big fan of DFU,r > but therefM > are situations where DFU can lose data if acting on an active directory, sot	 > be REAL  > careful about that.a >   F Is this true ? I too am a big fan, and use it on oversized directories4 (Goldfax is another app which can suffer from this).  A I'd always thought DFU was carefully written to high standards toi9 avoid this kind of conflict - what situations risk loss ? 3 Presumably the files are safe enough, just 'lost' ?o  
 Chris Sharmanu   ------------------------------  , Date: Fri, 09 Jun 2000 15:42:06 +0200 (CEST): From: "Gotfryd Smolik, VMS lists" <gotfryd@stanpol.com.pl>Y Subject: Re: DFU hazard ? (was Large amount of files in one directory causesvery slow peraK Message-ID: <Pine.LNX.4.10.10006091519460.17674-100000@irys.stanpol.com.pl>a  ( On Fri, 9 Jun 2000, Chris Sharman wrote:   +John Nixon wrote:N +> Someone else mentioned DFU to defrag the directory.  I am a big fan of DFU,C +> but there are situations where DFU can lose data if acting on an 3 +> active directory, so be REAL careful about that.  +  +Is this true ?   A  I am can't get to the precise answer (ah, yes, second message[1]t; today when I am not exactly check for what I say... flamingo9 me is allowed :]) but *if* DFU uses regular RMS operationt> to modify the DIRECTORY - then I remember, that is  somewhere 9 documented, that RMS write sharing is *not* supported forl directory file !:  That means, you *can* OPEN/READ/WRITE/SHARE=WRITE (in DCL< notation), but RMS does *NOT* guarantee consistency beetween/ *your* IO and the one, where RMS does *itself*.i  5  Yes, years ago I play with "direct directory update"1: (for fun, like mixed-case names ;) and for repairing-tool)7 and this time have see the point... I am can't be sure, 0 if RTFM or someone here warns me for the idea ;)   [...]tB +I'd always thought DFU was carefully written to high standards to: +avoid this kind of conflict - what situations risk loss ?  :  If all have written before is true, then the only concept: where may work properly (i think b.ex. for DFU) is locking; the directory file (or operate only on disk mounted/FOREIGN0" what is not a user-friendly idea).  3  If you will, you can simply check if that is done:e/ $ open/read/write/share=write iotest [.dirname]o  -  and do compress for the "dirname" under DFU.8-  *IF* DFU reject a error - then works safely.a2  *IF* DFU works - then that mean, that *also* uses8   /share=write - and that is AFAIR wrong when RMS itself   will operate on the file.s# (yes, remember "$ close iotest !!")r  4 +Presumably the files are safe enough, just 'lost' ?  5  Yes, even if you corrupt the directory structure the,< "file" in meaning of "header" (owner, protection, structure)0 and "data" are safe. You may see it in [SYSLOST] after ANAL/DISK/REPAIR :]   > BTW: IMHO nothing wrong making a COPY of the directory file...<  Yes, you must fiddle to get the copied file be a directory,9 *and* the copy doesn't have the new files created by RMS,d@ *but* making simple DIFFERRENCE, marking the copy as "directory"> and RENAMing the missing position van always reair the file :)  : Also: because I have't check, what way uses DFU to operate7  on directory file: I can imagine direct ("block mode")T3  operation; and *if that* - then if the file is not ;  locked properly serious corruption (of the directory file)u  can apply.s  6  Hm... must find some free time and see/install DFU...   +Chris Sharman    Regards - Gotfryd- [1] The first was on SAMBA forum, not here...  --E ===================================================================== F $ ON F$ERROR("LANGUAGE","ENGLISH","IN_MESSAGE").GT.F$ERROR("NORMAL") - 		THEN EXCUSE/OBJECT=MEa. $!                        GS@stanpol.zabrze.plE =====================================================================l   ------------------------------  % Date: Fri, 09 Jun 2000 15:15:54 +0100l2 From: Chris Sharman <Chris.Sharman@CCAgroup.co.uk>Y Subject: Re: DFU hazard ? (was Large amount of files in one directory causesvery slow per . Message-ID: <39410A2A.5897FA1C@CCAgroup.co.uk>   +John Nixon wrote:F +> Someone else mentioned DFU to defrag the directory.  I am a big fan of DFU,tC +> but there are situations where DFU can lose data if acting on ano3 +> active directory, so be REAL careful about that.i  ( On Fri, 9 Jun 2000, Chris Sharman wrote: > [...] D > +I'd always thought DFU was carefully written to high standards to< > +avoid this kind of conflict - what situations risk loss ?    Gotfryd Smolik, VMS lists wrote:5 >  If you will, you can simply check if that is done:e1 > $ open/read/write/share=write iotest [.dirname]i > / >  and do compress for the "dirname" under DFU.h/ >  *IF* DFU reject a error - then works safely. 4 >  *IF* DFU works - then that mean, that *also* uses: >   /share=write - and that is AFAIR wrong when RMS itself >   will operate on the file.o  @ Thanks Gotfryd. DFU does error (ie doesn't write-share), so that/ isn't a problem - if indeed there is a problem..  > Perhaps John Nixon and/or Ton Dorland could elucidate please ?? DFU can delete whole trees, but I presume John meant unintendedo loss of data ?  7 > Hm... must find some free time and see/install DFU...E  G Do that - marvellous tool. Recovers corrupted directories too, although   I haven't tried that feature ...   ------------------------------  , Date: Fri, 09 Jun 2000 17:12:03 +0200 (CEST): From: "Gotfryd Smolik, VMS lists" <gotfryd@stanpol.com.pl>Y Subject: Re: DFU hazard ? (was Large amount of files in one directorycausesvery slow perf K Message-ID: <Pine.LNX.4.10.10006091652540.17674-100000@irys.stanpol.com.pl>h  ( On Fri, 9 Jun 2000, Chris Sharman wrote:  ! +Gotfryd Smolik, VMS lists wrote: 6 +>  If you will, you can simply check if that is done:2 +> $ open/read/write/share=write iotest [.dirname] [...] A +Thanks Gotfryd. DFU does error (ie doesn't write-share), so that)0 +isn't a problem - if indeed there is a problem.  ?  Ach, yes. Have two time re-read the statement to be sure, thatt9 DFU is properly written :) (yes, my .sig is proper... ;>)h  8 +> Hm... must find some free time and see/install DFU... +eH +Do that - marvellous tool. Recovers corrupted directories too, although! +I haven't tried that feature ...d  (  Mhm... Have occasion; after (my own :[)  BACKUP/DELETE [...] NL:./SAVE4 from improper directory, search for small important 0 files on 2 GB disk -:> I suspect, that from the 3 time have forget some data :], but this time I have 4 been only some miliards time slower than any utility we can think -:<  7  FLORIAN was not a resolution, because on the disk was r2 hundreths of files [purged before from differrent 3 directories, and the directories renamed from time c to time] of the same name...)a5  But some tool (and haven't this time any on-hand :() 1 was required !! (I have do, don't flame me, IO to 2 the disk with DCL and dump the symbols with little fast-written .COM)    Regards - Gotfryd   --E =====================================================================nF $ ON F$ERROR("LANGUAGE","ENGLISH","IN_MESSAGE").GT.F$ERROR("NORMAL") - 		THEN EXCUSE/OBJECT=MEw. $!                        GS@stanpol.zabrze.plE =====================================================================    ------------------------------  # Date: Fri, 09 Jun 2000 14:56:11 GMTd2 From: Lyle West <lyle.west@childrenshc.org.nospam>$ Subject: Fibre Channel and VMS 7.2-15 Message-ID: <3940BF07.BCB4A59@childrenshc.org.nospam>a  F We are soon to be evaluating several methods of using FC as the inter-F connect for our disk storage. There seems to be two frontrunners as toG the methods we will use. The one I trust the most is StorageWorks usingrF HSG80 etc. The other option is EMC. Actually, EMC isn't really seen by; management as an option, as it is already in the next room.a  F The StorageWorks solution we plan to test will use 2 KGPSA for PCI->FCC in each node, two 16 port switches, and dual HSG80 controllers. The0C alphas are two ES40 boxes clustered. Eventually, the HSG's would be C licensed for DRM for full remote replication ("High Availability" -t> Disaster Recovery). (Those DRM license prices are VERY stiff).  G Testing the EMC will require the KGPSA cards, 16-port switches and thenaE directly into the EMC. Previously, testing EMC via direct scsi showedpB the EMC disks on both nodes of the cluster with the banks of disksB appearing to be 2 seperate banks, but were in fact the same disks.F This implies to me to be a open invitation for a disaster, and I have ; no reason to believe I won't see the same thing with fibre.   E Has anyone done anything similar to either of the above scenarios or .H wish to comment on I have been mumbling about above. Since it is Friday, I welcome all responses.   --   ------------------------------  $ Date: Fri, 9 Jun 2000 17:57:12 +0200, From: "Bart Zorn" <B.Zorn@TrueBit.nospam.nl>( Subject: Re: Fibre Channel and VMS 7.2-1* Message-ID: <8hr467$9fe$1@buty.wanadoo.nl>   Hello,  L I just finished a project which involved installing two ES40's (plus a DS20) in a cluster using FC storage.  I It took my client a long time to decide between Storageworks and EMC, but1K fortunately Storageworks was chosen. To me, EMC did not give me the feelingn that they really support VMS.r  I If you want to do "Disaster Tolerance", I believe you are better off withaJ Host Based Volume Shadowing in stead of DRM. OpenVMS has much more controlK over volumesets than DRM can ever have. If you loose the connection between L the sites, DRM will be at a loss. OpenVMS, however, also may lose track. YouL will want to consider using Compaq's Cockpit manager product to take care of that situation!h  1 If you want more information, please let me know!    Regards,  	 Bart Zorn  OpenVMS consultant.s  ? "Lyle West" <lyle.west@childrenshc.org.nospam> wrote in messageo/ news:3940BF07.BCB4A59@childrenshc.org.nospam...rH > We are soon to be evaluating several methods of using FC as the inter-H > connect for our disk storage. There seems to be two frontrunners as toI > the methods we will use. The one I trust the most is StorageWorks usingcH > HSG80 etc. The other option is EMC. Actually, EMC isn't really seen by= > management as an option, as it is already in the next room.e > H > The StorageWorks solution we plan to test will use 2 KGPSA for PCI->FCE > in each node, two 16 port switches, and dual HSG80 controllers. The-E > alphas are two ES40 boxes clustered. Eventually, the HSG's would bepE > licensed for DRM for full remote replication ("High Availability" -c@ > Disaster Recovery). (Those DRM license prices are VERY stiff). >hI > Testing the EMC will require the KGPSA cards, 16-port switches and thenwG > directly into the EMC. Previously, testing EMC via direct scsi showed D > the EMC disks on both nodes of the cluster with the banks of disksD > appearing to be 2 seperate banks, but were in fact the same disks.G > This implies to me to be a open invitation for a disaster, and I have-= > no reason to believe I won't see the same thing with fibre.r >aF > Has anyone done anything similar to either of the above scenarios orJ > wish to comment on I have been mumbling about above. Since it is Friday, > I welcome all responses. >2 > --   ------------------------------  # Date: Fri, 09 Jun 2000 17:24:30 GMT,% From: Uwe Zessin <zessin@my-deja.com>a( Subject: Re: Fibre Channel and VMS 7.2-1) Message-ID: <8hr97o$7r4$1@nnrp1.deja.com>o  * In article <8hr467$9fe$1@buty.wanadoo.nl>,/   "Bart Zorn" <B.Zorn@TrueBit.nospam.nl> wrote:  > Hello, >lF > I just finished a project which involved installing two ES40's (plus( > a DS20) in a cluster using FC storage. >eG > It took my client a long time to decide between Storageworks and EMC,2E > but fortunately Storageworks was chosen. To me, EMC did not give mel+ > the feeling that they really support VMS.b  E I've never seen an EMC on a VMS system, but I've visited them here inhC Germany for a one-day introduction. The presentation was quite good % and I was impressed about their work. 6 However I'd say they're playing in a different league.  F > If you want to do "Disaster Tolerance", I believe you are better offD > with Host Based Volume Shadowing in stead of DRM. OpenVMS has muchC > more control over volumesets than DRM can ever have. If you loose-: > the connection between the sites, DRM will be at a loss.  A Well, I tell customers that this is the classic 'quorum' problem.nD I've noticed that after that comment anyone who has managed clusters+ before, understood and nodded in agreement.   H > OpenVMS, however, also may lose track. You will want to consider usingB > Compaq's Cockpit manager product to take care of that situation!  D On OpenVMS the members on _both_ sides are simultaneously available.@ With DRM you have to do a site_failover (which, IIRC, requires a reboot), first. D Compaq claims they have fixed this for WindowsNT configurations, butD I haven't seen it and the presentor could not explain how it worked.6 (On the other hand, ACS V8.5P supports a log-history)." [ more comments in the text below]  3 > If you want more information, please let me know!, >J
 > Regards, >4 > Bart Zorni > OpenVMS consultant.g >sA > "Lyle West" <lyle.west@childrenshc.org.nospam> wrote in messageh1 > news:3940BF07.BCB4A59@childrenshc.org.nospam...o? > > We are soon to be evaluating several methods of using FC asp? > > the interconnect for our disk storage.There seems to be twoc? > > frontrunners as to the methods we will use. The one I trustt> > > the most is StorageWorks using HSG80 etc. The other option. > > is EMC. Actually, EMC isn't really seen by? > > management as an option, as it is already in the next room.i > >eB > > The StorageWorks solution we plan to test will use 2 KGPSA for> > > PCI->FC in each node, two 16 port switches, and dual HSG80E > > controllers. The alphas are two ES40 boxes clustered. Eventually,iC > > the HSG's would be licensed for DRM for full remote replication-A > > ("High Availability" - Disaster Recovery). (Those DRM licenseg > > prices are VERY stiff).d  C Yes... Remember that you also need dual controllers with 512 MBytesgA mirrored cache on both sites (initiator & target). You might alsoCF think about spare parts on-site -- you cannot change you configuration: when one controller is down (guess how I found that out?).  F > > Testing the EMC will require the KGPSA cards, 16-port switches andG > > then directly into the EMC. Previously, testing EMC via direct scsiiG > > showed the EMC disks on both nodes of the cluster with the banks of.E > > disks appearing to be 2 seperate banks, but were in fact the samel
 > > disks.D > > This implies to me to be a open invitation for a disaster, and ID > > have no reason to believe I won't see the same thing with fibre.  G What was the problem in the SCSI version? Different controller letters?a8 That could have been fixed with port allocation classes.  A Fibre channel-attached disks always start with "$1$DGA". The unitt- number is assigned to the HSG's unit e.g. D5:e   SC1CL1S1H1> Show unit full  ;   LUN                              Uses             Used byoE ---------------------------------------------------------------------lE   D1                                 M1             SC1CL1S1\RCSCL1S1r9      LUN ID:      6000-1FE1-0000-D3xx-0009-9330-7883-0019  ==>   NOIDENTIFIER      Switches:=        RUN                    NOWRITE_PROTECT      READ_CACHEa-        READAHEAD_CACHE        WRITEBACK_CACHEt(        MAXIMUM_CACHED_TRANSFER_SIZE = 32      Access:H        S1CL1A111, S1CL1A112, S1CL1A221, S1CL1A222, S2CL1A111, S2CL1A112,        S2CL1A221, S2CL1A222d      State:i         ONLINE to this controller        Reservedd'        PREFERRED_PATH = THIS_CONTROLLERn      Size: 35555889 blocks*      Geometry (C/H/S): ( 7000 / 20 / 254 )  = (That is from a non-VMS setup, but I think you get the idea.)   H > > Has anyone done anything similar to either of the above scenarios orD > > wish to comment on I have been mumbling about above. Since it is$ > > Friday, I welcome all responses.  F There are many visions. In a multi-OS environment the reality is some- times very frustrating.i   --
 Uwe Zessin    & Sent via Deja.com http://www.deja.com/ Before you buy.>   ------------------------------  % Date: Fri, 09 Jun 2000 12:02:48 +0100y  From: steven.reece@quintiles.com- Subject: Re: Hobbyist vms 7.2 alpha questionsa> Message-ID: <802568F9.003CC4AB.00@qedilc01.qedi.quintiles.com>   Dave Froble commented:G >>>On Alpha, you upgrade and install from the CD, using mostly the sameiE procedures.  At one point in the install, the question is asked aboutsO 'PRESERVING' what's on the disk.  Choose this, and it's an upgrade, else it's aaP fresh install.  Been a while, but that's the general idea.  Works real well, and is quick.<<<  N Works in most cases, but not in all.  As per a DSN article entitled "[OpenVMS]4 PCSI PRESERVE Option Fails With %PCSI-E-UPGRADERR" :  O "The PRESERVE option is not supported when upgrading to 6.2-1H2 (or 6.2-1H1, orsO 6.2-1H3, or 7.1-1H1).  This is documented in the OpenVMS Alpha, Version 6.2-1H2.+ Release Notes and Installation Procedures."t  E For not supported read "doesn't work as the procedure won't let you".e  M For these instances you have to use VMSINSTAL and the saveset that's found intL the MFD on the CD (called something like AXPVMSU1H1071.A and not even that's% guaranteed to work in all cases IIRC.t  L Unless you have the option to blow away the disk you're installing/upgrading0 onto and making it a new installation of course.   Steve.   ------------------------------   Date: 9 Jun 2000 13:43:29 GMTm8 From: hammond@not@peek.ppb.cpqcorp.net (Charlie Hammond)- Subject: Re: Hobbyist vms 7.2 alpha questionsT6 Message-ID: <8hqsa1$8ns$1@mailint03.im.hou.compaq.com>  ? In article <802568F9.003CC4AB.00@qedilc01.qedi.quintiles.com>, u: steven.reece@quintiles.com writes (quoting documentation):  P >"The PRESERVE option is not supported when upgrading to 6.2-1H2 (or 6.2-1H1, orP >6.2-1H3, or 7.1-1H1).  This is documented in the OpenVMS Alpha, Version 6.2-1H2, >Release Notes and Installation Procedures."  H You may expect that this limitation will NOT be found in future releases of OpenVMS Alpha.o  D You may also expct that for SOME future releases of OpenVMS Alpha --I e.g. "hardware" releases -- choosing the PRESERVE option will often offerDH two choices -- a "full" installation that replaces/refreshes ALL OpenVMSK files, and a "partial" option that replaces only changed files.  The "full"iG choice will always be available; the "partial" option may require that dG certain conditions are met.  When it is available, the "partial" optione6 may significantly reduce the time required to upgrade.   --  K     Charlie Hammond -- Compaq Computer Corporation -- Pompano Beach  FL USAeF          (hammond@peek.ppb.cpqcorp.net -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  # Date: Fri, 09 Jun 2000 14:00:35 GMTu From: d.webb@mdx.ac.ukP Subject: Re: Large amount of files in one directory causes very slow performance) Message-ID: <8hqt9n$u5k$1@nnrp1.deja.com>.  6 In article <882568F8.0073A2BE.00@mailhost.corvel.com>,-   Tad Winters <Tad_Winters@corvel.com> wrote: C > If files are deleted, it sounds like installing DFU and using itssE > DIRECTORY/COMPRESS/TRUNCATE command might be a better solution thanl	 trying too! > move files between directories.u >      Or simply :-   set file/truncate dirname.diri    A Things are apparently better with large directories with VMS 7.2.o@ (And will hopefully be improved even further with the filesystem work being done for VMS 7.3).d  = Up until then the only solution if you have the control is tor; use multiple directories instead of just the one directory.   E If you have the source of the program you can alter that. If not then ; (if the program uses a logical name for the directory) then"D one possible solution might be to redefine the logical name to refer to multiple directories=   eg  7 define testing   dev1:[test1],dev1:[test2],dev1:[test3]-  @ New files created in testing:  would be created in dev1:[test1].  & Then periodically redefine the logical   eg  5 define testing dev1:[test2],dev1:[test3],dev1:[test1]l  A Existing files would still be accessed but new files would now beA created in dev1:[test2]h      
 David Webb VMS and Unix team leader CCSS Middlesex University     > To:   Info-VAX@Mvb.Saic.Come( > cc:    (bcc: Tad Winters/Admin/CorVel)	 > Fax to:nG > Subject:  Re: Large amount of files in one directory causes very slowu >       performance  >tG > You didn't mention if files are created AND deleted in that directoryn ... F > if YES, maybe you should recreate a "new" directory and rename filesD > from the old one, and then switch the names ... the .DIR file size  > can be diminished suche a way. >hF > Maybe using a logical name pointing on several directory files couldA > help make creating files quicker ... and retrieval not so long.F >F1 > May need to be tested, but I can't do that now.a >w > Cordialement > Jean-Fran?ois Marchalt > X9000 - LYON >a% > <salvi@freeler.nl> wrote in message 4 > news:8holi1$vuu$1@rubidium.news.lion-access.net...
 > > Hi there,d > >hA > > At this time we have al lot of files, about 25,000 all in onel directory on > aFE > > VMS 6.2-1H3 2node DSSI cluster with Alpha 4100 machines with each  2Gb of > > memory and two CPU's.eG > > Both Alpha's are connected to two HSZ50 controllers with 64Mb cacher and  > host > > based shadowing. > >'F > > The average RMS performance is slowing down as the amount of files > increases1 > > in one directory.rH > > In the coming months we expect that the amount of files will be near	 > 40,000.w2 > > I'm sure that the RMS performance drops again.5 > > Is there any tuning possible on RMS / INDEXF.SYS.7> > > I was told that when the directoryfile exceeds 128 blocks, performance on7 > > RMS operations will drop very rapidly. (our currentc directoryfilesize is > > about 300 blocks). > >iF > > Maybe needless to say but filehighwatermarking is already switched off, > and @ > > the caching size of the HSZ50 controllers is at maximumsize.F > > Of course you can say that we have to decrease the amount of files in one: > > directory, but that is at the moment no option for us. > >.B > > If you have any idea how i can avoid these performanceproblems please* > > respond to this newsgroup or directly.
 > > Thanxx > >t > >d7 > > For more specific information, feel free to ask me.n > >  > > Salvi Schrijen > > Libertel Network > > The Netherlands." > > s.schrijen-nospam-@libertel.nl > >- > >s > >t >h >f    & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------  # Date: Fri, 09 Jun 2000 13:48:53 GMT + From: Paul Anderson <panderson@genicom.com>h6 Subject: Re: LN15AC toner cartridge compatible to ...?C Message-ID: <panderson-3D4CFF.09485809062000@news.bellatlantic.net>m  D In article <200006090000_MC2-A818-5D5D@compuserve.com>, "Richard B. ' Gilbert" <DRAGON@compuserve.com> wrote:'  H > Allowing Compaq/Digital/Genicom to sell the "key" for a stiff premium?  F The "key" is the set of little tabs on the toner cartridge that match F the inside of the printer.  It is very common for different brands of 6 printers using the same engine to have different keys.  I I don't think it's so one company can charge more for their cartridges.  =I Rather, the printer vendor wants to keep the supplies business for their -E own brand printer.  It's like the razor and blade scenario where the lI razor is sold for cost and the profit is made selling replacement blades.g   Paul   -- ."    Paul Anderson, DCPS Engineering"    GENICOM Corporation, Gardner MA   ------------------------------   Date: 9 Jun 2000 16:53:10 +0100rT From: pmoreau@cenaath.cena.dgac.fr (Patrick MOREAU, CENA Athis, Tel: 01.69.57.64.40)! Subject: Re: MMOV 2.2 and VMS 7.1.! Message-ID: <Q8MUzO90UZA7@gaelic>   H In article <393EAAEB.5CDB80F8@home.nl>, Dirk Munk <munk@home.nl> writes: >  > Hoff Hoffman wrote:r >> l{ >> In article <7a1+fuLOFdUc@gaelic>, pmoreau@cenaath.cena.dgac.fr (Patrick MOREAU, CENA Athis, Tel: 01.69.57.64.40) writes:hQ >> :Multimedia Services for OpenVMS is available on the June CD's with support off5 >> :Ensoniq card and DSx0 workstations and servers !!a > D > And will it support the ESS (?) soundcards in the PWS series too ? >    From the MMOV 2.2 SPD:   HARDWARE REQUIREMENTSt  * o  AlphaStation Sound board, supported on:    AlphaStation 400, 500, 600     DPWS 600 au serieso  C o  Embedded AlphaStation Sound board compatible chip, supported on:b    AlphaStation 200, 255  A o  Creative Labs AudioPCI card (Ensoniq 1373 chip), supported on:e,    Compaq AlphaStation XP900, XP1000 systems-    AlphaStation/AlphaServer DS10, DS20e, ES40n    AlphaServer DS10L    C o  FullVideo Supreme or FullVideo Supreme JPEG option supported on:n"    AlphaStation 250, 400, 500, 600   Patrickl -- pO ===============================================================================+O pmoreau@cena.dgac.fr  (CENA)     ______      ___   _           (Patrick MOREAU)d4 moreau_p@decus.fr (DECUS)       / /   /     / /|  /|J CENA/Athis-Mons FRANCE         / /___/     / / | / |   __   __   __   __  N BP 205                        / /         / /  |/  |  |  | |__| |__  |__| |  |N 94542 ORLY AEROGARE CEDEX    / /   ::    / /       |  |__| | \  |__  |  | |__|N http://www.ath.cena.fr/~pmoreau/            http://www.multimania.com/pmoreau/O ===============================================================================s   ------------------------------  " Date: Fri, 9 Jun 2000 14:14:22 GMT1 From: schreiber@eisner.decus.org (Jeff Schreiber)e  Subject: Re: Nic-to-Nic failover+ Message-ID: <Arfz2WWBHM6+@eisner.decus.org>m  _ In article <WXR%4.15$ye2.10210@news.pacbell.net>, "Joseph Hom" <Joseph_Hom@hotmail.com> writes:hL > Has anyone been able to do this. Basically and alpha ds20 running open vmsL > with two nic cards. Is there any way to have a nic-to-nic failover in case0 > the primary fails. Sort of like 3com's driver. >   K     I'm not much into just blindly plugging product, but it's answering then     question being asked.c  G     Process Software has called that failover "Paired Network InterfacehG     Support".  This was shipped in TCPware 5.4, and will be in the Beta *     release of Multinet 4.3 [coming soon].  &     This is the blurp you can find at:1         http://www.process.com/tcpip/datasht.htmlh  " Paired Network Interface Support: 5 TCPware's Paired Network Interface support allows twop6 network interface cards (NIC) with their own unique IP6 addresses to be connected to the same virtual cable in/ order to create network redundancy and optimizea throughput. 5 This unique feature provides NIC failover support ande: creates network redundancy - if one NIC fails, information6 will be transmitted automatically from the second NIC.3 Additionally, multiple NICs can be used to increasei2 throughput if a data communications bottle neck is8 suspected from the server. Paired Network Interface will5 improve connectivity in e-commerce applications wheret2 Oracle is used for frequent database transactions,5 multimedia applications where there is high bandwidth 7 consumption, and any applications where a single serveri) connection is causing delays for clients.o  + <picture>                                  e  6 Example A:                                            8 Dual 10Mb NICs improve throughput by utilizing bandwidth6 on a 100Mb Ethernet segment downstream from the server7 segment. The Paired Network Interface combines failover$: with an ability to use existing network interface cards as: well as increase transmission performance from the server.  + <picture>                                                                     5 Example B:                                           t9 Throughput is increased using two 10Mb NICs connecting tod; a client via one 10 Mb Ethernet cable to a client. In heavy ; traffic, the second NIC can transmit during the time neededb+ for the first to acknowledge packet arrivalu  )                                     -Jeffn   --1 Jeff Schreiber,            Process Software Corp. 1 schreiber@mx.process.com   http://www.process.com +      TCPware & MultiNet: Stronger than Everf      ------------------------------   Date: 9 Jun 2000 15:18:07 GMTi2 From: mathog@seqaxp.bio.caltech.edu (David Mathog)J Subject: Re: OpenVMS commentaries (was Re: Gartner commentary on Wildfire), Message-ID: <8hr1rf$5s3@gap.cco.caltech.edu>  \ In article <393D6B3D.C03E33CC@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: >David Mathog wrote:G >> Right, but there's no way to avoid it.  As somebody else pointed out H >> before (sorry, don't recall who, but I agree) a true Stream file typeC >> without the requirement for an LF every 64k would eliminate mostt >> of these problems >V >tK >SET file foo.bar/ATRRIB=(RFM=UDF,RAT=NONE) is as close to a stream file asoE >you'll get. It should not be converted/translated in any way by RMS.c  I Isn't it also unreadable by a regular C program unless the file is openedi binary?    >nM >If there is a problem in that the C RTL, by default, creates a text file. IflN >you intend to have binary unformatted data, then create the file accordingly.  B The real problem is that I work in a field where we use letters toJ represent the nucleotide and peptide sequences.  There's no protein that IK know of longer than 64k, but many, many DNA sequences are.  So this data is J sort of halfway between text and binary - it's not really text per se, butH it is represented with characters and is always stored in a "text" file,G and is passed between programs as "text".   Most of the formats used toeH represent these sequences are nice and allow long sequences to be split K down into lines of 80 characters (or something reasonable), but programmerstJ working on Unix systems have made life difficult by being lazy and puttingJ whole sequences on one line.  Some of the software they produce expects toB be able to write 1.2Mb of sequence directly into a text file. Ie,   !   fprintf(fd,"%s\n",verylongseq);t  5 So the 64k limit is very frequently a problem for me.s   Regards,   David Mathog mathog@seqaxp.bio.caltech.edum? Manager, sequence analysis facility, biology division, Caltech d   ------------------------------  # Date: Fri, 09 Jun 2000 14:01:09 GMTp From: nleach@gol.com Subject: Process Statesi) Message-ID: <8hqtap$u5u$1@nnrp1.deja.com>s  E I seem to remember that there used to be a file (in sys$system maybe)rF that contained the list of process state acroynms and their full names# (ie RWAST = Resource Wait For AST).d  B Am I halucinating as it doesn't seem to be around anymore (v7.2) ?   Nigelr      & Sent via Deja.com http://www.deja.com/ Before you buy.k   ------------------------------  % Date: Fri, 09 Jun 2000 02:59:21 +0000 + From: Dempsey Mork <dempsey@magellancc.com>a Subject: Public Shell Comapnys1 Message-ID: <B0000132565@mailserv.desertsurf.com>l  3Bulletin Board shell company available for a reverse merger.  The company is in total compliance and current in all its filings.  I can close a merger and deliver free trading stock within days of completion of due diligence.  If interested please email your name and phone number.  Thank you,  Dempsey Mork-   ------------------------------   Date: 9 Jun 2000 07:50 CST' From: carl@gerg.tamu.edu (Carl Perkins)R. Subject: Re: Question on using ASTs for output, Message-ID: <9JUN200007500171@gerg.tamu.edu>  D "Russell E. Owen" <owen@astroNOJNK.washington.edu.invalid> writes...J }I have a package that consists of multiple processes and multiple users. J }Each process can send messages to all users via a message hub. Each user H }has a command interpreter that reads and executes commands and outputs 
 }messages. } " }The present design is as follows:J }- A message hub receives messages in one mailbox and distributes them in % }to a set of mailboxes, one per user. G }- Each command interpreter queues a read of its message mailbox using  E }SYS$QIO and specifying an AST to run when data is received. The AST iC }reads writes the message using FORTRAN WRITE to unit 6 and queues rG }another read. Meanwhile, the interpreter uses a simple loop to read a n1 }command (FORTRAN READ on unit 5) and execute it.iF }- Messaging system errors are reported by writing directly to unit 6 G }(the log file for background processese, the screen for users) rather 1$ }than going through the message hub. }-- Russellu   My 2 cents worth:m  E Fortran's READ and WRITE are not suited for use at AST level - tryingu5 to do one in the middle of another one is a bad plan.u  B My suggestion: You should do both message reception and user inputF asynchronously. The easiest way to do this is to switch the user inputH to using SMG and its "unsolicited input" feature instead of Fortran READH statements (see HELP RTL SMG$ SMG$Enable_Unsolicited_Input). This causesK user input to set off an AST. Have that AST set an event flag. Also, changeeI the QIO on the mailbox to not do any actual reading/writing itself - this L can cause timing problems and errors if using Fortran's native I/O. Instead,L make the mailbox read's AST also set a (different) event flag. Then the mainI loop of the program sits at an Sys$WFLOR call, waiting for one of the two M event flags to be set. When one of them is set, the program will continue and D you can check to see which event flag was set and process that I/O.   K I have a (relatively, as these things go) simple program written in FortranoH which I wrote about 10 years ago to learn how to do this. It is a singleK program that two people can run which connects them for messaging much like.N a limited version of PHONE. I can send you a copy of the source if you'd like.  K This scheme is good for low to moderate volumes of messages where you won't F get more messages than can fit in the mailbox being used while you areF processing one or processing user input. If you can get bursts of moreE messages than that, you can use a scheme where the AST triggered by ahI message read on the mailbox doesn't set an event flag, but actually readsrG the data from the mailbox and puts it in a queue instead, then issues anK Sys$Wake (likewise, the user input AST would set a flag and do a Sys$Wake - F the flag is no longer waited on, but is used to indicate that the userK wants to enter something that you need to read). The "main loop" then pulls.I things off the message queue and processes them until the queue is empty, J probably checking to see if any user input has been detected between each.I When the queue is empty and there is no user input pending, the main loopiI does a Sys$Hiber (which is what the Sys$Wake in the AST routines is for).iL This is a bit more complex, but it can deal with a sudden influx of messages better.   B Other than that, for even more performance you'd have to deal withG subprocesses or threads (which I would imagine is a bit of a pain to dokD in Fortran, although I havn't tried) to process the commands you getD via the mailbox. Since you've already tried this, you know it can beC difficult to manage, but if you need to be responsive to user inputeH while you are processing commands (you don't say how long this can take)' it may be the only good way to do both.$   --- Carl   ------------------------------  $ Date: Fri, 9 Jun 2000 10:37:13 -0500* From: "Mark-Simon Pope" <mpope@bristol.ca>1 Subject: Submit /After delta time format questionF0 Message-ID: <Ja805.4224$Pz6.52313@news1.mts.net>  2 BAL780::L72> submit test.com /after="+1- 04:00:00"E Job TEST (queue L72_BATCH, entry 244) holding until 10-JUN-2000 14:29c BAL780::L72> sh time    9-JUN-2000 10:29:53  I I want to be able to submit the above job for 4am the next day.  I know IvG could easly write a DCL routine to calculate the correct time. I'm justl= wondering if there is a correct delta time format to do this.t   Mark-Simon Popes   ------------------------------  % Date: Fri, 09 Jun 2000 12:53:53 -0300o1 From: "Boyle, Darren" <boyledj@bankofbermuda.com>s5 Subject: RE: Submit /After delta time format questionsK Message-ID: <9F664D538536D411BD3200508B6FF01A02471C@bdant027.bda.bobda.com>e  5 BMDG_Darren1> submit/after="tomorrow+04:00" login.comvI Job LOGIN (queue BMDG_SYS$BATCH, entry 5001027) holding until 10-JUN-20001 04:00h BMDG_Darren1> sh tii    9-JUN-2000 12:52:42   - Darren   > ----------/ > From: 	Mark-Simon Pope[SMTP:mpope@bristol.ca]d' > Sent: 	Friday, June 09, 2000 12:37 PMw > To: 	Info-VAX@Mvb.Saic.Com4 > Subject: 	Submit /After delta time format question > 4 > BAL780::L72> submit test.com /after="+1- 04:00:00"G > Job TEST (queue L72_BATCH, entry 244) holding until 10-JUN-2000 14:29  > BAL780::L72> sh time >    9-JUN-2000 10:29:53 > K > I want to be able to submit the above job for 4am the next day.  I know I I > could easly write a DCL routine to calculate the correct time. I'm justt? > wondering if there is a correct delta time format to do this.e >  > Mark-Simon Popee >  >     F **********************************************************************C This message and any files transmitted with it are confidential anddJ may be privileged and/or subject to the provisions of privacy legislation.M They are intended solely for the use of the individual or entity to whom theyiL are addressed. If the reader of this message is not the intended recipient, B please notify the sender immediately and then delete this message.I You are notified that reliance on, disclosure of, distribution or copyingI of this message is prohibited.   Bank of BermudafF **********************************************************************   ------------------------------  % Date: Fri, 09 Jun 2000 12:04:48 -0400 ' From: "Zeni B. Schleter" <zzb@ornl.gov>d5 Subject: Re: Submit /After delta time format questionr( Message-ID: <394115A0.72C22D62@ornl.gov>  ) Use the /After="Tomorrow+4:00:00" instead0   Mark-Simon Pope wrote:  4 > BAL780::L72> submit test.com /after="+1- 04:00:00"G > Job TEST (queue L72_BATCH, entry 244) holding until 10-JUN-2000 14:29i > BAL780::L72> sh time >    9-JUN-2000 10:29:53 >tK > I want to be able to submit the above job for 4am the next day.  I know I0I > could easly write a DCL routine to calculate the correct time. I'm juste? > wondering if there is a correct delta time format to do this.t >l > Mark-Simon Popez   ------------------------------  " Date: Fri, 9 Jun 2000 18:01:47 GMT* From: young_r@eisner.decus.org (Rob Young)5 Subject: Re: Submit /After delta time format questiond+ Message-ID: <$sB5iga+NVID@eisner.decus.org>   ] In article <Ja805.4224$Pz6.52313@news1.mts.net>, "Mark-Simon Pope" <mpope@bristol.ca> writes: 4 > BAL780::L72> submit test.com /after="+1- 04:00:00"G > Job TEST (queue L72_BATCH, entry 244) holding until 10-JUN-2000 14:29- > BAL780::L72> sh time >    9-JUN-2000 10:29:53 > K > I want to be able to submit the above job for 4am the next day.  I know I-I > could easly write a DCL routine to calculate the correct time. I'm justg? > wondering if there is a correct delta time format to do this.. >    	Will this work for you?  / $ submit/noprint/nolog/after="tomorrow+4" t.comcB Job T (queue SYS$BATCH, entry 743) holding until 10-JUN-2000 04:00 $ show timem    9-JUN-2000 12:01:10   				RobC   ------------------------------  % Date: Fri, 09 Jun 2000 12:58:50 -0400s# From: John Vottero <John@MVPSI.com>c5 Subject: RE: Submit /After delta time format question : Message-ID: <C15945A9D9EFCF11BA8B08002BBF1CCC0CD71C@BERRY>   You can do:e   $ SUBMIT/AFTER="TOMORROW+04:00"t or $ SUBMIT/AFTER="TODAY+1-04:00"   > -----Original Message-----1 > From: Mark-Simon Pope [mailto:mpope@bristol.ca]e& > Sent: Friday, June 09, 2000 11:37 AM > To: Info-VAX@Mvb.Saic.Comn3 > Subject: Submit /After delta time format questioni >  > 4 > BAL780::L72> submit test.com /after="+1- 04:00:00"G > Job TEST (queue L72_BATCH, entry 244) holding until 10-JUN-2000 14:29- > BAL780::L72> sh time >    9-JUN-2000 10:29:53 > = > I want to be able to submit the above job for 4am the next A > day.  I know I; > could easly write a DCL routine to calculate the correct Q > time. I'm just? > wondering if there is a correct delta time format to do this.e >  > Mark-Simon Popef >  >    ------------------------------  % Date: Fri, 09 Jun 2000 08:46:06 +0100g2 From: Ashley Shepherd <ashley.shepherd@virgin.net> Subject: Re: system slow8 Message-ID: <qn41kss4dl1inurpolt9330bc83cv10p74@4ax.com>  @ You mention that you are running 2 versions of the S/W.  I wouldE suggest that this is where you problem is.  Both versions are causingo? contention for the same resources.  Had you asked, I would haveAB suggested that rather than replace a disk, you added an additional disk.D  C Contention can occur in several places, CPU, Memory, Bus and disks.hC If you HAVE to run both versions (I am assuming you are testing theoE upgrade before you move to it permanently), first of all, try closing9B one down, and see if the performance problem goes away.  It may be? only a short term thing that is only an inconvenience for a fewm< days/weeks, which will not necesitate any hardware upgrades.  5 OK.  having stated the obvious, here's where to look.=  C MONITOR DISK/ITEM=(OP,QU)  -  This will show you the disk operationNA rates followed by the Queue length.  (Average Queue length shouldc ideally be less than 0.25)  D MONITOR FCP   -   Look at the disk Read and Write rate.  Also WindowC Turns.  (Window turns should ideally be ZERO, anything else will be| causing you problems)   D MONITOR IO   -    Look at Direct I/O & Buffered I/O.  Important ones5 here are Split Transfer and Log Name Translation Rate   < MONITOR LOCK   -  Examine Deadlocks and ENQ's Forced to Wait  @ MONITOR MODES  -   Anything other than User Mode is bad for yourF system.  In reality though, the system has to do some work for itself,3 look for a User Mode percentage of greater than 50%c  A MONITOR PAGE   -   Examine  Read Rate, Read I/O Rate, Write Rate,aC Write I/O Rate, Demand Zero Rate, Write in Progress Rate and Systemm) Fault.  The last 2 should always be ZERO.b  C MONITOR PROCESSES   -    Look to see if any are outswapped, this isc BAD.  E MONITOR STATES   -   Again, as above, look for outswapped.  Better toaB use this one first to see, and MONITOR PROCESSES to find out which ones.p  9 MONITOR SYSTEM  -   This will give you a sort of summary.a  E All the above commands should be used together, to get an overview ofnE what is going on.  If you run these for 5 minutes at a bad time (theynC can all be run together) and send me the numbers  (Either type them F into an e:mail or cut'n'Paste into notepad etc.), I'll take a look for youh   RegardsN   Ashley    ) >We have an Alpha 200 running openVMS 6.1rH >System running fine until 2 months ago...at that time we replaced a 1gbJ >drive with a 2gb drive (i do not know drive model off hand) new drive was >not system disk.nF >We added the disk space to accomodate a software upgrade which has us7 >running two different version of synergy concurrently.aI >It seems, that since that time, the system runs much slower. At times itdI >will suspend all activity for 20-30 seconds at a time. Processes will bem >suspended.oI >I need some troubleshooting help. What can I monitor to determine if ther: >hesitation is disk or software or something else related.
 >Thank you >Jim K >a  6 >***************************************************** >Ashley Shepherd o >Consultancy Servicesi% >email to  ashley.shepherd@virgin.neti6 >*****************************************************   ------------------------------   Date: 9 Jun 2000 18:37:35 GMT ' From: andywright@psilink.co.uk (Andy w)t Subject: TAR FOR VMS?b. Message-ID: <39412b5f.0@nnrp1.news.uk.psi.net>   Hello,9       I need to transfer files from Openvms V7.2 on AlphaoK to a Tru64 system via tape (DLT), (no network so it has to be via tape!!!).rF Is there a tar utility for vms or something simillar so I can do this.F As they are both Compaq systems you would think there should be one!!!   thanks Andy   ------------------------------  # Date: Fri, 09 Jun 2000 15:39:21 GMTn# From: alphaman@hsv.sungardtrust.comh Subject: Re: VAX on Intel?) Message-ID: <8hr336$2vj$1@nnrp1.deja.com>t  3 In article <elkZ4.61$ru5.1966@typhoon.aracnet.com>,p5   "Zane H. Healy" <healyzh@shell1.aracnet.com> wrote:J >n0 > Me I wonder what Compaq knows that they don't.  E Me, I wonder the opposite.  Here's the text of an email I just postede3 on the AlphaNT mail list regarding this same topic.-   ------    6 > > 	Have you looked at http://www.charon-vax.com yet?4 > > 	A MicroVAX emulator that runs on Windows NT and4 > > 	boots VMS. It's pretty wild. You could actually > > 	run VMS on a laptop. :)    E Yup, looks really cool.  A laptop, huh?  Wow, a reasonably priced VMSj? laptop?  Like what people have been asking for for a long time?g  E It figures, someone would do something like this.  The merging of VMSf8 and Windows had to happen, and this is one way to do it.  E Of course, the other way would have been for DigiPaq to do what usersmD have been SCREAMING for for the past decade plus -- a decent WINDOWSE emulated environment that ran under VMS, so all the MS apps could runb
 under VMS.  D But that's okay, Compaq -- I'm confident that SRI will be willing toG take the money that could have been yours.  Only they'll do it on an NTlF platform with less margin than what you'd have gotten if you'd sold itD on VMS.  And people don't have to buy Compaq hardware to get to thatG point, either.  "Oh Mr. Dell, could you sell me a 1G P-III so I can runs VMS?"   . *sigh* there goes another boat the Q missed...  > I'd like to know what the collective was thinking when they soE stubbornly refused to put a Windoze environment on VMS?  A decent MS-nC Windows WM plus FX!32 ported to VMS would have been a killer combo.aD THIS is the type of thing that would have kept Alpha NT'ers placated? last August (maybe just a little better than what wasn't done).    peeved at inaction and excuses,  Aaron  --< Aaron Sakovich                 SunGard Huntsville Operations< 256 971-5808                   sakovich@hsv.sungardtrust.com< Make April 15th just another day:    http://www.fairtax.org/3 "Try not!  Do, or do not.  There is no try." (Yoda)i          & Sent via Deja.com http://www.deja.com/ Before you buy.d   ------------------------------   Date: 9 Jun 2000 18:20:50 +0200e* From: eplan@kapsch.net (Peter LANGSTOEGER) Subject: Re: VAX on Intel?* Message-ID: <39411962$1@news.kapsch.co.at>  O In article <8hr336$2vj$1@nnrp1.deja.com>, alphaman@hsv.sungardtrust.com writes::4 >In article <elkZ4.61$ru5.1966@typhoon.aracnet.com>,6 >  "Zane H. Healy" <healyzh@shell1.aracnet.com> wrote: >>1 >> Me I wonder what Compaq knows that they don't.t >mF >Me, I wonder the opposite.  Here's the text of an email I just posted4 >on the AlphaNT mail list regarding this same topic.  L Who wants a (emulated) VAX on a PC/notebook just to run VMS ? Surely not me.  J VMS on VAX is dying or did you see any major improvments in OpenVMS VAX inH the last years ? No IEEE FP, no JAVA, no new hardware support (like PCI,K Fast-Ethernet, Gigabit-Ethernet, ATM, Memory Channel, Fibre Channel, USCSI,3L Sound Cards, ...), no MOTIF/CDE, fewer applications (eg. no SETI) and so on.  M All I want is a cheap and small box to run a REASONABLE version of VMS there.gN And this is now only OpenVMS Alpha. And as Alpha Boxes can be cheap and small,N all I need is the will of Q to sell it (and especially the VMS licenses) cheapG to me and everyone else. And if they do, the selled boxes increases the0M installed base of Alphas and Alpha Chip prices can come down again and so on.n  L So, a 600MHz Alpha Notebook with current commodity hardware supported by VMSF is much more preferable than a 1GHz DELL with an MicroVAX-II emulator.  O And I want VMS as the opsys. And not because I have VMS dependant applications. M Just the opposite, I want normal office applications on VMS. If I have a DELLrL PC, then I could run the M$ crap immediately at a cheaper price (and exactly& the M$ beast is what I want to avoid).  : So, this niche product surely wasn't made for me. Sorry...   -- e< Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888i< FBFV/Information Services           E-mail  eplan@kapsch.netF <<< KAPSCH AG  Wagenseilgasse 1     PSImail PSI%(0232)281001141::EPLANH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"N "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998   ------------------------------  % Date: Fri, 09 Jun 2000 13:44:30 -0400 " From: Dan Sugalski <dan@sidhe.org> Subject: Re: VAX on Intel?: Message-ID: <4.3.2.7.0.20000609134111.01c96ab0@24.8.96.48>  2 At 06:20 PM 6/9/00 +0200, Peter LANGSTOEGER wrote:M >So, a 600MHz Alpha Notebook with current commodity hardware supported by VMSsG >is much more preferable than a 1GHz DELL with an MicroVAX-II emulator.s  L I don't know that Alphas run cool enough to put in a laptop, unfortunately.  They run awfully warm...   					Dan  I --------------------------------------"it's like this"-------------------s2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and evena;                                       teddy bears get drunkn   ------------------------------  # Date: Fri, 09 Jun 2000 10:15:30 GMT 0 From: carlini@true.lkg.dec.com (Antonio Carlini)% Subject: Re: vaxstation LED meanings?W* Message-ID: <8hqfg8$5nc@usenet.pa.dec.com>  J Just for the record, I power-cycled a VAXstation 3100 Model 76/SPX with a   VRT19 and these are the results:         0s - powered upt)     15s - monitor flickers for an instanti"     48s - hard drives can be heard-     64s - SPX test patterns flash on monitor r$     68s - KA43-A V1.2 banner appears?             tests F...B run very quickly, B takes quite a whileh'    129s - tests continue from B onwards /    189s - tests complete, boot ssequence startsh  K So booting should start around three minutes after powering up (unless, of  2 course, the box is set to HALT at the >>> prompt).   Antonioa  I Antonio Carlini                            Mail: carlini@true.lkg.dec.come# DECnet-Plus for OpenVMS Engineeringe6 COMPAQ                                     Reading, UK   ------------------------------  % Date: Fri, 09 Jun 2000 11:07:52 +0100o. From: Roger@natron.demon.co.uk (Roger Barnett)% Subject: Re: VMS File Caching FuturesT- Message-ID: <552952355wnr@natron.demon.co.uk>g  Q In article: <4.3.1.2.20000609182616.01d18ba0@kerberos.davies.net.au>  Huw Davies i+ <Huw.Davies@kerberos.davies.net.au> writes:e >  > K > Well most of the languages I like to program in don't have compilers for  M > OpenVMS, although I belive that there is a BCPL compiler for VAX somewhere n* > around and maybe even one for Algol-68.     E Having worked on it for a short while in the 1980s I can confirm that @ there certainly is (at least one) Algol 68 compiler for VAX/VMS.  J I think it was based on a port of an ICL (now Fujitsu) compiler front-end @ from RSRE (now DRA) and a new backend done by SPL (now EDS) and ? NAG (still NAG?) plus one or two people from Oxford University.h  K If you do a scan for "Sean Leitch" (I think I've spelled that right) you'llu7 find that Algol68 compilers are still being developed !m   -- b
 Roger Barnettt   ------------------------------  # Date: Fri, 09 Jun 2000 14:59:30 GMTn From: lookgood2k@my-deja.com@ Subject: Which cpu does Sun server use: I386, Mip, Ppc or Alpha?) Message-ID: <8hr0o7$12m$1@nnrp1.deja.com>2   Any help will be appreciated.i    & Sent via Deja.com http://www.deja.com/ Before you buy.m   ------------------------------  % Date: Fri, 09 Jun 2000 17:46:20 +02007= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com> D Subject: Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha?) Message-ID: <3941114C.D13D27EC@gtech.com>    SPARCo   ------------------------------   Date: 9 Jun 2000 18:33:13 +0200r* From: eplan@kapsch.net (Peter LANGSTOEGER)D Subject: Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha?* Message-ID: <39411c49$1@news.kapsch.co.at>  H In article <8hr0o7$12m$1@nnrp1.deja.com>, lookgood2k@my-deja.com writes:   In earlier days, 68k.r< 		But then again, these were only workstations, not servers.D Now only SPARC (though there is a current version of SOLARIS for the6 		INTEL platform = I386, SUN's servers are SPARC only)K In the future, who knows what SUN really plan with/for the MERCED/ITHANIUM.    -- 2< Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888@< FBFV/Information Services           E-mail  eplan@kapsch.netF <<< KAPSCH AG  Wagenseilgasse 1     PSImail PSI%(0232)281001141::EPLANH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"N "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998   ------------------------------  $ Date: Fri, 9 Jun 2000 12:30:35 -0400# From: "Rob Kas" <rob@paychoice.com> D Subject: Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha?= Message-ID: <MT805.5455$Q7.209720@news-east.usenetserver.com>o     The wrong one :)   ------------------------------  % Date: Fri, 09 Jun 2000 18:23:54 +0100*B From: Andrew Harrison SUNUK Consultancy <andrew.nospam@uk.sun.com>D Subject: Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha?* Message-ID: <3941282A.EB59E64C@uk.sun.com>   Peter LANGSTOEGER wrote:  J > In article <8hr0o7$12m$1@nnrp1.deja.com>, lookgood2k@my-deja.com writes: >e > In earlier days, 68k.pL >                 But then again, these were only workstations, not servers.  F Huum, most people would think that the  3/180, 3/280's and 3/470/480'sE would be to big to qualify as a deskside or desktop workstation. They*@ were designed as servers. All the Sun 3/XXX servers used members of the Motorola 68K family.r   > F > Now only SPARC (though there is a current version of SOLARIS for theF >                 INTEL platform = I386, SUN's servers are SPARC only)M > In the future, who knows what SUN really plan with/for the MERCED/ITHANIUM.n  E I don't think we have deviated from the SPARC line, the IA-64 versionoB of Solaris simply follows on where the x86 versions of Solaris and SunOS left off.s  : One could if one was being cruel make the same point aboutD Compaq, "who knows what Compaq really plan with/for MERCED/ALPHA :-)   Regardso Andrew Harrisone Enterprise IT Architects   ------------------------------   End of INFO-VAX 2000.321 ************************