1 INFO-VAX	Sat, 31 Mar 2001	Volume 2001 : Issue 181       Contents: Re: "cut" utility for VMS  Re: "cut" utility for VMS - Re: BASIC 1.4 install causing a startup error 0 Re: Beware - It's another Christof Brass bashing- Re: Cluster Breakup - VAX/Alpha Hardware Help - Re: Cluster Breakup - VAX/Alpha Hardware Help - Re: CSWS: are FrontPage extensions supported? - Re: CSWS: are FrontPage extensions supported?  Re: ES40 Benchhmark record Re: Free The OpenVMS Seven!  Re: Free The OpenVMS Seven!  Re: Free The OpenVMS Seven! < Re: FYI: NDS Authentication Services is coming to OpenVMS...' Re: Getting % Merged or Copied from DCL  Re: gmake / port lib problem% Re: I need help sorting a file ahhhh!  Re: Java replacement?  modems Mozilla 0.8.1 graphic bug ?  Re: OpenVMS Educational Program  Re: OpenVMS Educational Program  Re: OpenVMS Educational Program  Re: OpenVMS Educational Program  Re: OpenVMS Educational Program 4 Re: OpenVMS Multi-Site Cluster White Paper Available, Re: OpenVMS Technical Update Days - success!, Re: OpenVMS Technical Update Days - success!P Re: OT: Elements of Linguistics (was Re: POSIX Streams, File Permissions) Permis Re: Redirect output of Com File ( Re: Seeking CD-R/CD-RW SCSI INQUIRY data: Re: Seeking CD-R/CD-RW SCSI INQUIRY data / Yamaha CRW4416S: Re: Seeking CD-R/CD-RW SCSI INQUIRY data / Yamaha CRW4416S Re: Spice for OpenVMS Alpha? Re: Spice for OpenVMS Alpha? Re: Spice for OpenVMS Alpha? Re: Spice for OpenVMS Alpha?  Re: THE EMC Chronicles - revised0 Re: VMS: DECNet to TCP/IP migration help wanted.% When might SYS$SCHDWK not wake me up? ) Re: When might SYS$SCHDWK not wake me up? ) Re: When might SYS$SCHDWK not wake me up? ) Re: When might SYS$SCHDWK not wake me up? ) Re: When might SYS$SCHDWK not wake me up? ) Re: When might SYS$SCHDWK not wake me up? * Re: [Q] param to increase for unsuff SPTE?  F ----------------------------------------------------------------------  % Date: Sat, 31 Mar 2001 16:28:35 +0000 ) From: Christof Brass <brass@infopuls.com> " Subject: Re: "cut" utility for VMS, Message-ID: <3AC605B3.FC7E4078@infopuls.com>   Ben Sego wrote:  >  > Jordan Henderson wrote:  > 2 > > In article <3AC3FF54.10643.15D8791@localhost>,- > > Stanley F. Quayle <stan@stanq.com> wrote: J > > >Has anyone ported the Unix "cut" program to VMS?  I need it for a DCLE > > >script.  I pulled down the GNU source, but it uses that horrible ( > > >shell script configuration stuff... > > >  > > G > > Really, using cut like you would in Unix is very un-VMS like.  What G > > you really want is that every VMS command (and all of your programs D > > as well) to support a much superior, very regular /SELECT_COLUMN > > qualifier. > > * > > I suggest a syntax like the following: > > 8 > >         /SELECT_COLUMN=(LIST:<col-expr>,SEPARATOR=c) > > L > > Where the LIST qualifier takes a <col-expr> which is the same as in UnixL > > 'cut' and the optional SEPARATOR arg is a character to break columns on. > > Q > > Hmmmm... maybe I should have waited until Sunday to make this recommendation.  > > > Wait, maybe we should ask Christof Brass what he thinks! :-) >  > B.S.   Thanks for passing over ...   @ Firstly I have to admit that I'm not sure that this "very un-VMS< like" isn't irony (I suppose it is). Secondly I have to make8 clear that functionality of a program per se isn't an OS< dependent topic though the way it used and some OS dependent features are of course.   ; I don't understand why waiting until Sunday could be of any  help.   = Although I think that basically an output modifying qualifier > would be the best way to go the definition proposed isn't what8 shows the important difference: if the user has to count? character columns the case is lost (then having another program ; will do the trick sufficiently). The point is to be able to ? refer to the content generated by the program e.g. the filename @ column of a DIRECTORY output. If the output of a shell script is< to be cut at a certain character position I offer to write a> simple TPU script which can be used to trim the length of each line in the output file.  @ Anyway I wonder why this is necessary and what the circumstances of this task are.    ------------------------------  % Date: Sat, 31 Mar 2001 12:01:46 -0500 * From: "Stanley F. Quayle" <stan@stanq.com>" Subject: Re: "cut" utility for VMS. Message-ID: <3AC5C72A.20238.8526790@localhost>  B > Anyway I wonder why this is necessary and what the circumstances > of this task are.   = Here's what my resulting code looks like (after renaming the  " wonderful EXTRACT program to CUT):    $!  Get the list of CW club ID'sF $   cut /columns=(3:12) CFFE$COMMERCIAL_BACKUP:COMMERCIAL.SENDING -   .         /output=cffe$scratch:comm_upload_1.tmp4 $   search cffe$scratch:comm_upload_1.tmp "927358" -+         /out=cffe$scratch:comm_upload_2.tmp ) $   sort cffe$scratch:comm_upload_2.tmp - &         cffe$scratch:comm_upload_3.tmp) $   uniq cffe$scratch:comm_upload_3.tmp - '         cffe$reports:comm_upload_cw.rpt  $!# $!  Get the list of NACHA club id's  $   cut /columns=(1:5,42:51) -3         CFFE$COMMERCIAL_BACKUP:COMMERCIAL.SENDING - ,         /output=cffe$scratch:comm_upload_1.p3 $   search cffe$scratch:comm_upload_1.tmp "+5200" - +         /out=cffe$scratch:comm_upload_2.tmp 8 $   cut /columns=(6:15) cffe$scratch:comm_upload_2.tmp -+         /out=cffe$scratch:comm_upload_3.tmp ) $   sort cffe$scratch:comm_upload_3.tmp - &         cffe$scratch:comm_upload_4.tmp) $   uniq cffe$scratch:comm_upload_4.tmp - *         cffe$reports:comm_upload_nacha.rpt    E File COMMERCIAL.SENDING is a concatenation of files in two formats.   C The task is to extract a list of identification numbers that start  A with "927358".  Other fields in the files can also contain those   digits.   C The CW format always has the ID in columns 3-12 of all records.  I  2 sort the list by ID and then eliminate duplicates.  D NACHA ID records always have a "+5200" starting in column one.  The B ID number starts in column 42.  I then sort and un-dup the result.  D It's ugly, but it does exactly what I need done.  And I was able to 2 move it into the production environment on Friday.  ! At least I didn't use "PIPE"   ;)      --Stan  
 ----------G Stanley F. Quayle, P.E.   N8SQ   +1 614-868-1363   Fax: +1 614 868-1671 1 8572 North Spring Ct. NW, Pickerington, OH  43147 = Preferred address:  stan@stanq.com       http://www.stanq.com    ------------------------------  # Date: Sat, 31 Mar 2001 07:10:52 GMT + From: rjordan@mars.mcs.net (Richard Jordan) 6 Subject: Re: BASIC 1.4 install causing a startup error4 Message-ID: <0qfx6.4653$JN5.124815@news.goodnet.com>  9 > This is a known problem that engineering has worked on.  > ...    Karl, F      thanks very much for the info.  Glad to know its being worked on.E I'll take a look at the versions I have available, and perhaps take a 1 look at the VMSIMAGES modification you mentioned.    Rich Jordan  rjordan@mcs.net    ------------------------------  % Date: Sat, 31 Mar 2001 15:58:17 +0000 ) From: Christof Brass <brass@infopuls.com> 9 Subject: Re: Beware - It's another Christof Brass bashing , Message-ID: <3AC5FE99.38A56282@infopuls.com>  @ Good chose of subject - not very helpful content of your message
 wrt the goal.    Paul Sture wrote:  > D > In article <3AC52F13.83C6D434@infopuls.com>, Christof Brass wrote: >  > [big snip] > B > > You obviously have a problem in separating your personal pointA > > of view on technical aspects from the importance of my posts.  > / > Naa. Tis tha wot 'as trubble seein wots reet.    Woa baji dao ...   >  > > To@ > > say it with other words: if you don't like my messages don'tA > > read it but don't compare my postings with the ones of AH and B > > some other threads which mostly contain(ed) nothing related to > > VMS (hundreds of messages).  > D > You mean like hundreds of lengthy messages full of "Unix is crap"?  @ If you are that good in counting compute the ratio of this three< words devided by the rest. You'll see that your complaint is based on bad observation.   A > > If you analyse my posts and the reposts you'll hopefully find = > > them full of technical arguments or complaints of missing D > > technical arguments but only a few lines of texts accusing otherB > > people having said something else in previous messages a.s.o.. > B > To be honest, I found them boring. Hint: "Unix is crap" is NOT a > technical argument.   < Do the calculation. Your response is a clear indication that: your observation is biased by your personal opionion: your? aren't obviously able to get the proportions right of technical  parts and labels.    > D > > Unfortunately I severely missed also with you to communicate the9 > > "UNIX is crap" and the "UNIX apps are crap" messages.  >  > See above.   ?      > [snip] >  > > 1 > > What makes you to play the role of the judge?  > >  > C > Because you refused to listen to other well considered arguments?   = What "well considered arguments"? I listened to the arguments ? but unfortunately most of them weren't either "well considered"  or "on topic".  A > Because you invoked the "I'm a member of a minority" speech?***    Why do you care?  5 > Because you persuaded someone to invoke Odel's Law?   7 Don't confuse the sequence and causality of the events.   , > Because you still haven't got the message?  ? *You* didn't get it obviously. Why would you like to interfere?   K > YOU WERE SINGLE-HANDEDLY MAKING THIS NEWSGROUP AN UNPLEASANT PLACE TO BE.   ? You're repeating your opinion instead of answering or providing < more information. Why do you read my posts? Why did you even? respond to a few lines of one of my posts which wasn't directed  to your business in any means?  < > *** Hah! My family name is not exactly English, and indeedE > unpronounceable by most in the UK, where I grew up. Think about it.   ? Very reasonable. But I have a better suggestion for your family  name ...  % > > What about the freedom of speech?  > M > Whoo Hoo! You have already proved that you don't listen to what others say!   : Unfortunately your are missing here something: "freedom of? speech" is as its words imply the freedom to publish something. @ It remains your secret how you relate this to your response. And- BTW I read your post and I'm answering to it.   ? > > Honestly: I don't see any technical substance in your post.  > C > There wasn't supposed to be any technical substance, apart from a  > statistic.  = Got that. But again: how could 1/8 of posts to a very limited ? set of threads make you feel uncomfortable? I took your numbers < of about 100 posts of me and about 800 posts all together. I@ remember that not all of my posts were directed to the topic you don't like.   = > > What makes you uncomfortable if I post something which is > > > garbage from your point of view? Why do you read my posts? > > 9 > When you make this newsgroup an unpleasant place to be.   4 Is there a problem in skipping/deleting my messages?  A > What makes you think you can wander into a newsgroup and try to 3 > dominate it with your own personal point of view?   7 Nothing. A post is only information. Every educated and = reasonable person cannot be damaged by words. Every civilised ? person would react with words or silence. Of course it happened ? that someone felt insulted and killed the one who insulted him. < But this is history and normal behaviour today. If you don't> like my opinion don't read my posts. It has never been so easy> to ignore someone. If I would call you or send you letters the= amount of work to stop me doing so might be a problem. But to ) delete a message - why do you talk about?   G > What makes you think you can get away with insulting folks who give a G > great deal of their time contributing their technical expertise here?   ; This is your interpretation. Did I insult you? In what way?   F > If your intention was to close this newsgroup down, you were doing aA > pretty good job. That is why I read your posts and oppose them.   < Sorry, but really this seems to be pretty ridiculous. Do you@ really think what you posted? I read this NG for more than about= 18 months IIRC. I didn't post the first about six months. The > number of messages was substantially lower on technical topics> before the idea of spoiling VMS with UNIX SHIT came up and the< discussion of the different aspects like the designless UNIX7 SHIT attitude vs the design based VMS approach started.   = I took the time and checked your complaint about me using the : F-word and insulting someone: nothing. I used it once in a@ wellknown and in no way insulting phrase "you are ... right". If? you can't explain your accusation wrt that I will regard you as 5 someone who is not interested in precision and truth.    > ___  > Paul Sture   ------------------------------  % Date: Fri, 30 Mar 2001 22:54:12 +0200 " From: "Hans Vlems" <hvlems@iae.nl>6 Subject: Re: Cluster Breakup - VAX/Alpha Hardware Help( Message-ID: <9a2rtq$akk$1@news.IAEhv.nl>  ' J.G. Peters heeft geschreven in bericht ' <01c0b893$d8baadb0$32341895@jpeters>... E >We are in the process of breaking up an Alpha 2100/VAX 4300 cluster. F >The VAX is being "decomissioned." The Alpha will remain in productionK >for at least another year, at which time we will replace it with a new one  >-K >AlphaServer, that is... :-)  All data from the disks housed within the VAX  >hasK >been migrated to devices within the Alpha. There is a TLZ06 tape drive and H >an LP27 printer that are physically connected to the VAX and need to beG >moved to the Alpha. I'm planning on chaining the TLZ06 to/from a TKZ9E G >8mm tape drive that is already connected to the Alpha. Beyond ensuring K >a unique scsi id, this would not appear to be terribly complicated. Should G >the Alpha "see" the new tape drive on power up or are there additional  >configuration tasks required?  G A reboot would be enough. OTOH you could just connect it to the AXP2100 1 and run SYSMAN with the IO AUTOCONFIGURE command.   *  As for the LP27 - how can we get that guyH >functioning on the Alpha? I believe there are interface issues here...?  # How's the LP27 connected right now?       
 Hans Vlems > F >When these two devices are moved and operational, I'm sure we'll haveI >additional questions concerning the actual, uh, severing of the cluster.  >  >Thanks, >Joe   ------------------------------    Date: 31 Mar 2001 21:01:03 +0800, From: Paul Repacholi <prep@prep.synonet.com>6 Subject: Re: Cluster Breakup - VAX/Alpha Hardware Help- Message-ID: <87ofuirt2o.fsf@prep.synonet.com>   $ "Hans Vlems" <hvlems@iae.nl> writes:  @ > > As for the LP27 - how can we get that guy functioning on the7 > >Alpha? I believe there are interface issues here...?   % > How's the LP27 connected right now?   F This will possibily be the Data Products vs Centronix problem. I guessF the LP27 hangs off a LPV-11 now? So the question is, does the Alpha doB DP mode IO on the Parport. If it can't, you can possibly re-jumper the LP27 to Centronics.    --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  % Date: Sat, 31 Mar 2001 10:48:49 +0100   From: Paul Sture <paul@sture.ch>6 Subject: Re: CSWS: are FrontPage extensions supported?+ Message-ID: <VA.00000338.8a38c63c@sture.ch>   M In article <009F9CC0.84DA8F55@SSRL04.SLAC.STANFORD.EDU>, Alan Winston - SSRL   Admin Cmptg Mgr wrote:N > From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr") > Newsgroups: comp.os.vms 4 > Subject: CSWS: are FrontPage extensions supported?% > Date: Sat, 31 Mar 2001 00:39:57 GMT  > 9 > VMSers, and especially people using or maintaining the  " > Compaq Secure Web Server (CSWS): > D > I'm told that Microsoft has provided the FrontPage Extensions for E > Apache. Assuming this is true, did they get ported with CSWS?  Are  4 > they available for CSWS, or for any VMS webserver? > L > I'm further told that the extensions consist both of simplified publishingA > and a bunch of programmatic support for things like image maps.  > Q Don't know the details, but here's a snippet from www.netcraft.com - What's that 2 site running?:  A The site is running Apache/1.3.14 (Unix) PHP/4.0.3pl1 PHP/3.0.17 t FrontPage/4.0.4.3 on Linux. 	 ^^^^^^^^^s  > From which I deduce that at least a unix/Linux version exists. ___ 
 Paul Sture Switzerland/   ------------------------------    Date: 31 Mar 2001 08:07:36 -07001 From: nothome@spammers.are.scum (Malcolm Dunnett) 6 Subject: Re: CSWS: are FrontPage extensions supported?, Message-ID: <qPcDvMqlMDPw@malvm1.mala.bc.ca>  , In article <VA.00000338.8a38c63c@sture.ch>, &     Paul Sture <paul@sture.ch> writes: >> cE >> I'm told that Microsoft has provided the FrontPage Extensions for eF >> Apache. Assuming this is true, did they get ported with CSWS?  Are 5 >> they available for CSWS, or for any VMS webserver?  >>  M >> I'm further told that the extensions consist both of simplified publishingcB >> and a bunch of programmatic support for things like image maps. >> a > @ > From which I deduce that at least a unix/Linux version exists. >   H    The extensions exist for a number of Unix platforms and web servers, C including Tru64 Unix and Apache - but VMS is not on the list as farPL as I can see (http://msdn.microsoft.com/workshop/languages/fp/2000/exts.htm)  E    It seems the Unix versions of the extensions are actually provided1E by Ready to Run Software (http://www.rtr.com/Ready-to-Run_Software/),cF you could contact them to see if they have any interest in a VMS port.   ------------------------------    Date: 31 Mar 2001 19:10:28 +0800, From: Paul Repacholi <prep@prep.synonet.com># Subject: Re: ES40 Benchhmark recorde- Message-ID: <871yretcrf.fsf@prep.synonet.com>   6 "Terry C. Shannon" <terryshannon@mediaone.net> writes:  4 > "Keith Brown" <kbrown780@isd.net> wrote in message# > news:3AC52F56.D939FB5E@isd.net...t  < > > http://www.compaq.com/newsroom/pr/2001/pr2001021201.html    E > The same system was the first RISC-based quadprocessor to deliver a-E > $/tpmC of under $20. And equipped with popularly-priced third party<( > memory, the $/tpmC would be about $12.  D Perhaps it's time to bring out the old re-badging tradition. Call it# a "Far Q" and redo the benckmark :)e   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.t@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  % Date: Sat, 31 Mar 2001 09:58:09 +0000e) From: Christof Brass <brass@infopuls.com>e$ Subject: Re: Free The OpenVMS Seven!, Message-ID: <3AC5AA31.3AA0E182@infopuls.com>   Larry Kilgallen wrote: > d > In article <FreeThe.OpenVMS_7@nowhere.nil>, Doc.Cypher <Use-Author-Address-Header@[127.1]> writes: >  > > (c) Developer Ready. > >  > > * Give away (gasp!) DECC., > L > No, charge double for DECC and give away a good language, like Compaq Ada.7 > Better yet, upgrade Compaq Ada to the Ada95 standard.i > < > Developing in C makes it too expensive to achieve quality.  6 As most of the readers of this NG will know I strongly< discourage the use of C/C++ because to my knowledge based on7 experience, analysis and statistics these languages ares? CRAP/SHIT. Ada is an excellent language which might be a littlee> bit too fat. To have the current Ada standard available on VMS@ together with the current API language bindings would be a major
 advantage.  ? But I would recommend to give DECC for a very low charge or foro5 free because it can be used as intermediate language.    ------------------------------    Date: 31 Mar 2001 11:20:32 -00004 From: Doc.Cypher <Use-Author-Address-Header@[127.1]>$ Subject: Re: Free The OpenVMS Seven!5 Message-ID: <20010331112032.4080.qmail@nym.alias.net>h  " -----BEGIN PGP SIGNED MESSAGE-----  ? On Sat, 31 Mar 2001, Christof Brass <brass@infopuls.com> wrote:  >Larry Kilgallen wrote:n >> d9 >> In article <FreeThe.OpenVMS_7@nowhere.nil>, Doc.Cypheru. >> <Use-Author-Address-Header@[127.1]> writes: >> t >> > (c) Developer Ready.o >> > >> > * Give away (gasp!) DECC. >> cM >> No, charge double for DECC and give away a good language, like Compaq Ada.w8 >> Better yet, upgrade Compaq Ada to the Ada95 standard. >>  = >> Developing in C makes it too expensive to achieve quality.a >07 >As most of the readers of this NG will know I strongly.= >discourage the use of C/C++ because to my knowledge based on.8 >experience, analysis and statistics these languages are@ >CRAP/SHIT. Ada is an excellent language which might be a little? >bit too fat. To have the current Ada standard available on VMSrA >together with the current API language bindings would be a major  >advantage.s >:@ >But I would recommend to give DECC for a very low charge or for6 >free because it can be used as intermediate language.   Two points Christoff..."  F 1. People are tired of seeing Cr*p/Sh*t in your posts. It does nothing;    for your position. In fact, it is more likely to provoke     disagreement.  F 2. By its very nature a language such as C is open to abuse. That doesB    not make it a "cr*p" language. Ideally all of the compilers forD    OpenVMS would be available at competitive prices. Then people canE    choose that best suited for the job. After all, you should analyse+<    and decide which to use based on the given circumstances.     Doc.   -----BEGIN PGP SIGNATURE-----2 Version: 2.6.2  @ iQEVAwUBOsUP8sriC3SGiziTAQH8LQgAq9kzkAjCxMnwiA/z8UaDXtuz03+pLmq4@ tu3UoWex4nLyQZZnO//+jD00cBso5EN6tFdAJh9GrUkKTY3JJ5iDp3PhRnRCTtD3@ xJkBy3lUnQ1oOsMSpYqGlXe1d8eeXTTDDt+D4NYDz+1xyKVXwjO3MdVFAH60W7sa@ cwJXIcV1liUyfP+BIznQvM5I2Gyj4SzoRChQh8/GZflP8n4hOpoWwBbKl9Gy9j7P@ t360DsDms5A0NVNSTkdC/D6fGn/7GR8KN+pzbgGw7GleGwM+0jRVYvXTmAGPjaDD8 NxvpcRkiSz3fNfukdhRwW7DxcBkXjLKgE0sfIlrm6w90vq0Ip+NEmQ== =RGzz  -----END PGP SIGNATURE-----i   ------------------------------  % Date: Sat, 31 Mar 2001 17:02:43 +0000 ) From: Christof Brass <brass@infopuls.com>p$ Subject: Re: Free The OpenVMS Seven!+ Message-ID: <3AC60DB3.E8E4BA6@infopuls.com>D   "Doc.Cypher" wrote:n > $ > -----BEGIN PGP SIGNED MESSAGE----- > A > On Sat, 31 Mar 2001, Christof Brass <brass@infopuls.com> wrote:  > >Larry Kilgallen wrote:  > >>; > >> In article <FreeThe.OpenVMS_7@nowhere.nil>, Doc.Cypher-0 > >> <Use-Author-Address-Header@[127.1]> writes: > >> > >> > (c) Developer Ready.8 > >> >  > >> > * Give away (gasp!) DECC. > >>O > >> No, charge double for DECC and give away a good language, like Compaq Ada.-: > >> Better yet, upgrade Compaq Ada to the Ada95 standard. > >>? > >> Developing in C makes it too expensive to achieve quality.C > >i9 > >As most of the readers of this NG will know I strongly.? > >discourage the use of C/C++ because to my knowledge based one: > >experience, analysis and statistics these languages areB > >CRAP/SHIT. Ada is an excellent language which might be a littleA > >bit too fat. To have the current Ada standard available on VMS C > >together with the current API language bindings would be a majore
 > >advantage.- > >-B > >But I would recommend to give DECC for a very low charge or for8 > >free because it can be used as intermediate language. >  > Two points Christoff...n > H > 1. People are tired of seeing Cr*p/Sh*t in your posts. It does nothing= >    for your position. In fact, it is more likely to provokeS >    disagreement.  ? I'm tired of always reading of the well known problems of these @ languages and the consequences of using them. If I use these its@ a simple way to express my opionion about them. What terms would= you propose to let me express my opionion? BTW my opionion ise> based upon experience, analysis and statistics. If you hate my posts, don't read them.'  H > 2. By its very nature a language such as C is open to abuse. That doesD >    not make it a "cr*p" language. Ideally all of the compilers forF >    OpenVMS would be available at competitive prices. Then people canG >    choose that best suited for the job. After all, you should analysed> >    and decide which to use based on the given circumstances.  @ One major problem of C is that it offers very dangerous features< without simple means to control them, without necessary high@ level features and in a way which isn't suitable for most of the; programmers to fully understand and correctly use them. Thet@ rating as Crap PL comes from the fact that there are much better= alternatives *now* for each of the areas where C is used (andn7 assumed to be appropriate) and even a few multi-purposet; languages which are a complete and much better replacement.r6 Using C today is like building cars as did Henry Ford.  @ To respond more specifically: abusing a language is easy and not> per se an argument against any language. The point is how many? programmers are unintentionally abusing the language because itp< offers that many pitfalls which aren't obvious to avoid; how@ many solutions are "strange" because the language forces that by% lack of better ways to accomplish it.o   > Doc. >  > -----BEGIN PGP SIGNATURE-----l > Version: 2.6.2 > B > iQEVAwUBOsUP8sriC3SGiziTAQH8LQgAq9kzkAjCxMnwiA/z8UaDXtuz03+pLmq4B > tu3UoWex4nLyQZZnO//+jD00cBso5EN6tFdAJh9GrUkKTY3JJ5iDp3PhRnRCTtD3B > xJkBy3lUnQ1oOsMSpYqGlXe1d8eeXTTDDt+D4NYDz+1xyKVXwjO3MdVFAH60W7saB > cwJXIcV1liUyfP+BIznQvM5I2Gyj4SzoRChQh8/GZflP8n4hOpoWwBbKl9Gy9j7PB > t360DsDms5A0NVNSTkdC/D6fGn/7GR8KN+pzbgGw7GleGwM+0jRVYvXTmAGPjaDD: > NxvpcRkiSz3fNfukdhRwW7DxcBkXjLKgE0sfIlrm6w90vq0Ip+NEmQ== > =RGzzu > -----END PGP SIGNATURE-----    ------------------------------  % Date: Sat, 31 Mar 2001 08:46:56 +0000 ) From: Christof Brass <brass@infopuls.com>aE Subject: Re: FYI: NDS Authentication Services is coming to OpenVMS... , Message-ID: <3AC59980.EB0BF74C@infopuls.com>   Thanks for your post.r   Chuck Chopp wrote: >  > Bob Koehler wrote: > \ > > In article <3ABF723A.497002D8@rtfmcsi.com>, Chuck Chopp <ChuckChopp@rtfmcsi.com> writes: > >eN > > > 3)  Slightly OT here, but after attending some sessions on NetWare64, itH > > > has become very apparent that Novell has completely redesigned theI > > > kernel [now the nano kernel] of NetWare to be hardware architecture I > > > independent.  I made some comments about why was Novell waiting fortI > > > Intel to get IA64 ready for prime-time when they could have used anwM > > > Alpha processor for 64-bit computing any time during the past 10 years.t > > L > > Novell announced plans to port Novell to Alpha back in about '93.  Still > > waiting.  : I asked DEC people in '95 about porting Netware to Alpha -> answer: we are focused on M$. Maybe at that time porting to an= architecture owned by one company and beeing dependent on theu< information supplied by that company for doing a proper port8 wouldn't have been a good idea. At least DEC should have@ communicated to Novell that porting to the Alpha platform (incl.- mainboards, chipsets, machines) were welcome.y  X > That issue came up in the discussions about NetWare64 and the technical feasibility of^ > porting to Alpha.  The assessment of NetWare 3.x/4.x that was done at that time revealed farV > too many Intel x86 dependencies & assumptions in the core NOS to make it feasible toZ > convert.  It looks like Novell is at least taking some good steps in the right direction\ > this time by making sure that the hardware is abstracted from the OS at the lowest levels. > \ > The reason that eDirectory on Alpha [and even NetWare64 on Alpha] would be desirable comes^ > from the approach that Novell is taking with their marketing of eDirectory these days.  They\ > are not necessarily catering an NOS to companies that will have at most just a few tens of[ > thousands of objects in their directory.  Instead, they're going after the electronic B2B ^ > market and the large-scale directory market where you may literally have tens or hundreds of\ > millions of objects in the directory.  For example, whenever you go to CNN's web site, you! > are issued a cookie to identifyuZ > you.  The cookie also results in an object being created in an NDS eDirectory tree which] > gets some portal site customization information attached to it.  Later on, if you return to V > the CNN site, the cookie held by your browser is used to retrieve this customizationY > information from your personal cookie object in the eDirectory tree.  CNN's servers arerW > handling something in excess of 10 to 20 million objects that get frequently updated.c\ > Powering that eDirectory with Alpha-based systems should provide much more performance and> > scalability than using less capable IA32/IA64/Sparc systems. >  > --
 > Chuck Chopp- > : > ChuckChopp@rtfmcsi.com            http://www.rtfmcsi.com2 >                                   ICQ # 22321532B > RTFM Consulting Services Inc.     864 801 2795 voice & voicemail4 > 103 Autumn Hill Road              864 801 2774 fax6 > Greer, SC  29651                  800 774 0718 pager9 >                                   8007740718@skytel.comH  ? Great. You mentioned the compiler technique Novell is using. Doa: they use their own compiler (what language)? Do they offer? different language bindings for their Netware API - I only know1@ C bindings - and compilers for different languages to write NLMs? (if this term is still correct)? Do you know what is (has been)p> the base for this high performance object database? I remember? attending a Netware programming seminar back in 1994 where theyt= only had very low level C APIs. They did all the object stuffs: manually. Very poor then. I would like to hear about major	 progress.v   ------------------------------  % Date: Fri, 30 Mar 2001 22:56:51 +0200t" From: "Hans Vlems" <hvlems@iae.nl>0 Subject: Re: Getting % Merged or Copied from DCL( Message-ID: <9a2s3a$at0$1@news.IAEhv.nl>  J Wouldn't it be possible to use F$ELEMENT to cut off the string just beforeE the % sign and then use F$EXTRACT to isolate the last two characters?uH The resultant two char string holds then either a two digit integer or aB simple integer with a blank prefix. Which is what you want, right?  
 Hans Vlems  , Dave Harrold heeft geschreven in bericht ... >Hi All, >wG >I would like to be able to get the % merged or % copied information iniA >a DCL procedure.   This is the same information provided by SHOWV >DEVICE. >e >$ sh dev dua117 >rF >Device                  Device           Error    Volume         Free
 >Trans MntG > Name                   Status           Count     Label        Blocks-
 >Count CntG >DSA117:                 Mounted              0  PRODI04       88870563. >121   2F >$1$DUA117:    (MINNIE)  ShadowMergeMbr       0  (merging DSA117:  43% >merged)F >$1$DUA217:     (PLUTO)  ShadowMergeMbr       0  (merging DSA117:  43% >merged) >2E > I haven't been able to identify any argument to F$GETDVI to providerG >that information.  Is this possible?  Is it available from the $GETDVI  >system service? >l. >Any information would be greatly appreciated. >s >Thanks, >n
 >Dave Harrold  >? > L >=========================================================================== ===========y> >Dave Harrold                                          E-Mail: David_Harrold@Aurora.orgD >Sr. Software Systems Engineer                         Phone : (414) 647-6204D >Aurora Health Care                                    FAX   : (414) 647-4999J >3031 W. Montana Street                                Milwaukee, WI 53234 > I >"A common mistake people make when trying to design something completelyW foolproof is toe0 >underestimate the ingenuity of complete fools."   ------------------------------   Date: 31 Mar 2001 01:09 PSTN) From: rankin@eql.caltech.edu (Pat Rankin)t% Subject: Re: gmake / port lib problem./ Message-ID: <31MAR200101090879@eql.caltech.edu>r  . In article <87g0fv2fl3.fsf@prep.synonet.com>,\1  Paul Repacholi <prep@prep.synonet.com> writes... B > I was installing the porting library. The compile went fine, and2 > I went to install it into sys$common:[vms_port.] > ? > The error message is that sysommon:[vms_port] is not found...n >oF > Anyone have any idea where this problem is? Is there a fixed version > of gmake somewhere.r  :      I don't use that software, but it's pretty clear that: it is treating `$c' as a request for macro expansion, with6 the resulting expansion being empty because `c' has no< definition.  Try specifying it as sys$$common or sys\$common; instead of just sys$common.  Or define another logical name < without any dollar sign in its name which points to the same2 location, and then use that instead of sys$common.  2                 Pat Rankin, rankin@eql.caltech.edu   ------------------------------  # Date: Sat, 31 Mar 2001 09:09:26 GMTM From: Dirk Munk <munk@home.nl>. Subject: Re: I need help sorting a file ahhhh!' Message-ID: <3AC59E81.528BAA46@home.nl>u  L I had a similar problem a long time ago, and the only way to solve it was toK write a program that read the sequential file similar to what you have now,tF and inserted the records in a empty indexed file that was created with create/fdl.h  . With a bit of luck you can do it with DCL too.  N The /nosort qualifier with convert only applies to the primary key, other keys are always sorted.  C But of course your main problem is that your key specifications aresN insufficient for the sequence you would like to see. You should always be ableM to reorganize a indexed file, and get a usable file in return. You now dependnM on the moment a record was inserted in the file, and thus the keys were addedtK to the indexes. When a new record is inserted with a key value that alreadygK exists in the file, the record pointer for that key is added just after thec" last entry for the same key value.  . So suppose you you have the following entries:   key value        record number 1000             40000 1000             30000 1000             50000  N Now if you add a new record with the same key value, the pointer will be added( to the end of this row, so you would get   key value        record number 1000             40000 1000             30000 1000             50000 1000             20000  N Please keep in mind with this example that in principle there is no reason whyI you can't insert a record with a lower primairy key than already present.a  G If you depend on this sequence,there is no way you can force convert toA? rebuild it. This is because the logic is not in the file but isp time-dependend.    Regards,   Dirk       Bill Ames wrote:  J > I have a file that is indexed and when new records are added they prettyL > much stay in the order that they were entered only sorted on the main key.K > Well someone reindexed the file and alas lost the sort in which they were H > intered.  I can type the following command to get in sorted correctly: >.K > SORT/KEY=(POS=1,SIZ=6)/KEY=(POS=12,SIZ=23)/KEY=(POS=97,SIZ=12) shipto.idx  > shipto.seq >a= > But notice it is now a sequential file.  Well when I do the  > CONVERT/FDL=CNTRL:SHIPTO.FDL >p > I loose the sort again.  >tD > Is there a way to sort an indexed file and retain the indexed file > format????   ------------------------------  % Date: Sat, 31 Mar 2001 10:08:42 +0000a) From: Christof Brass <brass@infopuls.com>o Subject: Re: Java replacement?, Message-ID: <3AC5ACAA.2C5A8CEE@infopuls.com>   "Steven P. Underwood" wrote: > ! > http://tame.virtualsellers.com/e > C > This company was just brought to my attention, and while I do notoF > personally have a use for it, I thought people here would definitelyG > have an opinion about it.  I am not into web development (as might beA% > obvious from some of my questions).  > E > 1.  Does this site work with the VMS version of Netscape? Lynx?  MySF > Hobby system has been displaced by my newborn son, so I can not testF > it.  It does work from a Wintel system with Java(Script) turned off. > H > 2.  Given the worry about security in Java and/or JavaScript, how doesF > this product compare?  Should I worry about accessing pages that are > using this technology? > F > 3.  There were recent discussions here about the availability of theC > latest Java code for VMS.  Is the server something to try and getSD > ported to VMS to get around this problem?  Does something similiar > already exist? > B > 4.  Other questions/points which are sure to be brought up here. > < > Thank you for you input.  I look forward to your opinions. >  > Steve> > Steven P. Underwood,DNRC > Whitinsville,MAc > StevenU@POBoxes.comn  @ Shockwave/Flush is a graphical oriented technique with much less@ capabilities of damaging the local environment than Java (if run? out of the sandbox). Basically at the moment it is not regardedd? as a major security problem. But it isn't intended as a general+ Java replacement.l   ------------------------------  % Date: Sat, 31 Mar 2001 09:55:54 -0600C. From: Lyndon Bartels <lbartels@pressenter.com> Subject: modemsb. Message-ID: <3AC5A9AA.6515F1D6@pressenter.com>  @ Has anyone done any work with modems connected directly to AlphaB workstations? I want to add on to my 500au for dialing in to work,1 either via SLIP, PPP, or to a terminal server....n  G Any suggestions on hardware, and software would be greatly appreciated.    Thanks in advance,   Lyndon --  G My opinions are mine and mine alone. They seldom align with those of myt	 employer.    ------------------------------    Date: 31 Mar 2001 14:14:53 +0100O From: pmoreau@dev.ath.cena.fr (Patrick MOREAU, CENA Athis, Tel: 01.69.57.64.40)e$ Subject: Mozilla 0.8.1 graphic bug ?  Message-ID: <CF4nbqS6UoPT@sable>   Hi all,o  I I've just upgraded Mozilla to version 0.8.1 (Alpha VMS 7.2-1, DS10 ; ElsadH Gloria Synergy) and I notice a strange behaviour: impossible to see the J bookmarks into the left frame (however the bookmark menu is ok). The left L frame seems never updated: when a menu is occulting a part of the frame, the+ bitmap of the menu stays into the frame ...T   Patrick+ --O ===============================================================================jO pmoreau@cena.dgac.fr  (CENA)     ______      ___   _           (Patrick MOREAU)f4 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 ===============================================================================    ------------------------------  % Date: Sat, 31 Mar 2001 09:45:59 +0000<) From: Christof Brass <brass@infopuls.com> ( Subject: Re: OpenVMS Educational Program, Message-ID: <3AC5A757.AB73FAC2@infopuls.com>   andrew harrison wrote: >  > Christof Brass wrote:g > >k > > andrew harrison wrote: > > >g > > > Christof Brass wrote:s > >r
 > > [SNIP] > >vC > > > > Boring. I and others suggested several viable solutions not B > > > > having the mentioned and analysed risks. Unfortunately youH > > > > didn't get them or you didn't understand them. Bad luck for you.E > > > > Don't use other people for supporting your Anti-VMS attitude. H > > > > And another point: you didn't put in any valuable argument. Your > > > > post is an empty suit. > > > C > > > Really so what were they ? perhaps a recap would be helpfull.e > > C > > Yes, it probably would, but I'm tired of doing it. You can findr@ > > it all in my posts of the last three weeks. Some others alsoC > > contributed at least as good ideas as mine or even better ones.*0 > > But: no COE needed, no "UNIX on VMS" needed. > >n > B > No sorry again you are wrong. You have made very few substantive> > responses, the closest to one I can find is your a number ofB > complaints which turn out to be complaints about the C/C++ specs> > which are equally valid on OpenVMS as they are on UNIX since > they are generic.s  ; Sorry is "the closest to one I can find is your a number ofy complaints" correct English?  @ You are wrong. If you are sure that you are right denote a judge< and give him US$ 1000.- (not very much for you I suppose - I? don't want to block this procedure by rising the sum too high).d> I'll do the same then and post my contributions. If the judges7 agree that I'm right I'll get your US$ 1000.- and mine.l; Otherwise you get them. Or propose a different procedure tot> decide it. You have a very bad record for blaming other people= wrongly of not posting arguments/facts etc.. I won't continues' answering upon your stupids complaints.c  s8 > So instead of the hand waving how about some substance7 > what are your reasons for claiming that UNIX is Crap.    Read mine and other's posts.  4 > Respond in as few lines as possible, you will lose5 > marks for regurgitating sections of the UNIX haters . > handbook and using words like C**P and F**K.  ? See above. BTW I don't accept your rules. Read "The UNIX-HATERSr9 Handbook". Show what is wrong of the substantial critics.n  C > > > But in this case the only person showing an Anti-VMS attitudeT@ > > > is you, you may not think so. But do you really think that< > > > repeating sections of the UNIX haters handbook, an old= > > > and even at the time not very accurate set of anti UNIXl. > > > opinion's is going to advance your case. > >lB > > Unfortunately we had this discussion these days in this NG andD > > it came out that age of a fact doesn't necessarily invalidate itA > > and the problems mentioned in that book are still there. UNIXh8 > > hasn't changed that much - it wouldn't be UNIX then. > >3 > ; > So you are claiming that UNIX and for that matter OpenVMSpA > havn't changed much since the UNIX haters handbook was written.j > : > Any book that has a preface including a long rant from aE > Lisp programmer fresh off a Symbolics Machine complaining about the-> > lack of breakpoints and a debugger in UNIX shell programming > is hardly starting off well.  ; Of course there are anachronisms. But you should be able too; abstract and to focus on the substantial, the architectural  aspects.  3 > It makes a good story for a "Terrible things thatr2 > Managers do to staff" book but hardly rates as a > sensible critique of UNIX.  ? Why is it then that the more reliability (quality) is necessaryt  the more VMS systems are in use?8 Do you know of one important reason VMS has been created
 initially?   [SNIP]  9 > As I said earlier show me a technical argument that youa8 > have made and I will respond. Saying something is Crap7 > of boring laced with liberal injections of the F worde > is no a technical argument.u  
 See above.  N > > > This is hardly helpfull to OpenVMS nor is it advancing your case either. > > > > > Thanks that you're stating this. This gives a bit of hope. > >3 > 3 > Do you want OpenVMS to wither on the vine because 2 > you have weird ideas about architectural purity?  7 Read the posts. You even have a clue what the topic is.b  4 > If you are interested in architectural purity then/ > why are you using OpenVMS at all. Its already . > iredeemably damaged according to your rules.   Prove that.n  - > What is your position on any of the OpenVMSr, > IP stacks and layered ultilities, they are% > based on the BSD IP stack/utlities.h  > Read my other post to you on that topic. Why do you repeat the4 same argument in different posts to one of my posts?  0 > How about CDE, X-windows, Motif and any of the2 > multitude of other foreign interfaces introduced > into OpenVMS.l   See the other post.o  3 > How about pathworks, anything that impliments thee/ > NT file service has to be the spawn of satan.   : I'm sorry to give up my courtesy. It seems that you aren't: really able to understand the basics: layering is the key.  2 > The sad thing about this is that you are holding0 > out for an uncontaminated architecturally pure > OpenVMS which does not exist.r  9 This may be true or not. But even if there have been made @ mistakes in the past the VMS style is very much there and can be? recognised. I really don't understand why you aren't able to doa simple abstractions.  A > > > > > So how would you get major ISV's to support OpenVMS ???l > > > >b > > > > See above! > > >r3 > > > No you havn't answered the question, put yourr, > > > points into short lines and send them. > >a > > See above! > >e > ) > Saying see above when you didn't answerh' > the question "above" does not cut it.n  ? See above is referring to the explanation why I don't repeat myi
 arguments.  	 > regardsu > Andrew Harrisono > Enterprise IT Architectn   ------------------------------  % Date: Sat, 31 Mar 2001 09:48:46 +0000N) From: Christof Brass <brass@infopuls.com>x( Subject: Re: OpenVMS Educational Program, Message-ID: <3AC5A7FE.A4581CC7@infopuls.com>   andrew harrison wrote: > # > steven.reece@quintiles.com wrote:d > >c > > Andrew Harrison wrote:7 > > >>>The sad thing about this is that you are holding 2 > > out for an uncontaminated architecturally pure! > > OpenVMS which does not exist.  > > <<<  > > M > > I almost hate to contribute to this but Andrew was right in his post from G > > which the above extract was taken.  Gosh, that's at least two of us B > > agreeing with Andrew again.  Let me fall on my sword!!!   :-)) > >0N > > There is no justification for VMS to be uncontaminated and architecturallyM > > pure.  TCP/IP on OpenVMS may not be my first choice of network transport, 6 > > but in the present day and age it is a Good Thing.L > > Integration with Gates' Bloatware is also a Good Thing in some respects,N > > since you can store all of your files on VMS using something like Samba orL > > Pathworks/Advanced Server and back them up reliably using BACKUP on your > > VMS system.e > >t > F > Exactly, OpenVMS isn't architecturally pure (in Christofs sense) nowC > and if it really was, e.g No IP, No X, No CIFS, No Threads etc it-< > would have died by now because people would not be able to' > use it in their current environments.  > ; > All OS's have to adapt, sometimes in ways their origional2; > designers did not envisage, if they don't adapt they die.E > 	 > Regardsf > Andrew Harrisone > Enterprise IT Architects  > One point is how the adaptation is done. Another point is what= to adapt. Implementing an open protocol isn't a big deal frome  the architectural point of view.   ------------------------------  % Date: Sat, 31 Mar 2001 09:52:11 +0000n) From: Christof Brass <brass@infopuls.com>o( Subject: Re: OpenVMS Educational Program, Message-ID: <3AC5A8CB.8010E688@infopuls.com>  ! steven.reece@quintiles.com wrote:  >  > Andrew Harrison wrote:5 > >>>The sad thing about this is that you are holding 0 > out for an uncontaminated architecturally pure > OpenVMS which does not exist.  > <<<  > K > I almost hate to contribute to this but Andrew was right in his post fromIE > which the above extract was taken.  Gosh, that's at least two of us @ > agreeing with Andrew again.  Let me fall on my sword!!!   :-))  3 No problem. The point is to what you agree and why.   L > There is no justification for VMS to be uncontaminated and architecturallyK > pure.  TCP/IP on OpenVMS may not be my first choice of network transport, 4 > but in the present day and age it is a Good Thing.  ? Why do you relate architecture and implementation of protocols?=  J > Integration with Gates' Bloatware is also a Good Thing in some respects,L > since you can store all of your files on VMS using something like Samba orJ > Pathworks/Advanced Server and back them up reliably using BACKUP on your
 > VMS system.r  3 Do I have to install Pathworks on every VMS system?   G > Now, a minor thought : if capital letters are taken to be shouting inMA > e-mail, how come nobody objects to capital letters in OpenVMS ?t > / > As I said, it's a minor thought.......   :-))d >  > Steve.  > If you are a VMS type person than I have failed to explain the> major goals of my ideas for fighting against the UNIXification/ of VMS. Please email privatly to sort this out..= If you are a UNIX type person then I'm sure discussion is note( fruitful as I pointed out several times.   ------------------------------    Date: 31 Mar 2001 12:20:56 -00004 From: Doc.Cypher <Use-Author-Address-Header@[127.1]>( Subject: Re: OpenVMS Educational Program4 Message-ID: <20010331122056.568.qmail@nym.alias.net>  " -----BEGIN PGP SIGNED MESSAGE-----  ? On Sat, 31 Mar 2001, Christof Brass <brass@infopuls.com> wrote:s >andrew harrison wrote:t  < >> All OS's have to adapt, sometimes in ways their origional< >> designers did not envisage, if they don't adapt they die.  ? >One point is how the adaptation is done. Another point is whatt> >to adapt. Implementing an open protocol isn't a big deal from! >the architectural point of view.u  ( OK, here's one for you then Christoff...  # http://www.faqs.org/rfc/rfc2440.txt   @ BTW, has anyone else noticed that PGPi have dropped VMS support?  3 http://www.pgpi.org/products/pgp/versions/freeware/,     Doc.   -----BEGIN PGP SIGNATURE-----  Version: 2.6.2  @ iQEVAwUBOsUP8sriC3SGiziTAQGjCAf7B885n0yL/hPAzy3RhwCtrqqeoyPUxbfs@ 9jl3Co1eCmh//zwbddx1UkcA6lzFtRObix3i/9kHVdx/qgOjeugz3u8jP5vz1L6J@ Iqlls101X/tTxlunr91ZTLt9WsXUobxXO+VIatXSWvEbVpyAzhQmI2Nn9Mh+ks8k@ ZU0PtlwEQ9lMm3v0uOgYP/eFIz9XGEUlpZW0fvQStvt79P+DG+dBuoW//UovPO3N@ CwihIvXeTk93hNs5oO4oQALiOHRtEeY3RvgnvMiamaRdFO8ucKYHRnvj2IfCfQ0o8 ixatXto0ofZR0Akg3HvspPYwRRvKN/HMrUlSN1TRuHkUUTaDIYNfJA== =FAjF  -----END PGP SIGNATURE-----p   ------------------------------  % Date: Sat, 31 Mar 2001 16:48:44 +0000o) From: Christof Brass <brass@infopuls.com>i( Subject: Re: OpenVMS Educational Program* Message-ID: <3AC60A6C.582102@infopuls.com>   "Doc.Cypher" wrote:e > $ > -----BEGIN PGP SIGNED MESSAGE----- > A > On Sat, 31 Mar 2001, Christof Brass <brass@infopuls.com> wrote:t > >andrew harrison wrote:o > > > >> All OS's have to adapt, sometimes in ways their origional> > >> designers did not envisage, if they don't adapt they die. > A > >One point is how the adaptation is done. Another point is whaty@ > >to adapt. Implementing an open protocol isn't a big deal from# > >the architectural point of view.  > * > OK, here's one for you then Christoff... > % > http://www.faqs.org/rfc/rfc2440.txt, > B > BTW, has anyone else noticed that PGPi have dropped VMS support? > 5 > http://www.pgpi.org/products/pgp/versions/freeware/0 >  > Doc. >  > -----BEGIN PGP SIGNATURE-----  > Version: 2.6.2 > B > iQEVAwUBOsUP8sriC3SGiziTAQGjCAf7B885n0yL/hPAzy3RhwCtrqqeoyPUxbfsB > 9jl3Co1eCmh//zwbddx1UkcA6lzFtRObix3i/9kHVdx/qgOjeugz3u8jP5vz1L6JB > Iqlls101X/tTxlunr91ZTLt9WsXUobxXO+VIatXSWvEbVpyAzhQmI2Nn9Mh+ks8kB > ZU0PtlwEQ9lMm3v0uOgYP/eFIz9XGEUlpZW0fvQStvt79P+DG+dBuoW//UovPO3NB > CwihIvXeTk93hNs5oO4oQALiOHRtEeY3RvgnvMiamaRdFO8ucKYHRnvj2IfCfQ0o: > ixatXto0ofZR0Akg3HvspPYwRRvKN/HMrUlSN1TRuHkUUTaDIYNfJA== > =FAjFr > -----END PGP SIGNATURE-----h  > The RFC 2440 is a lengthy document. Implementing this protocol; is a major task. But I don't see how this is related to thet; question of the architecture of an OS. Specifically I don'tm: understand why the architecture of VMS will be affected by; implementing this protocol in a library which is used by anp@ email program. The architectural question wrt the OS would be if4 this should be provided as a system servic which can@ tranparently used to encrypt the IP communication or some higher$ level protocol. Do I miss something?  > From the site with the ports of PGPi it seems that the work is; done by volunteers. Or is an organisation behind that whichj actively can drop the VMS port?g   ------------------------------  % Date: Sat, 31 Mar 2001 10:48:49 +0100   From: Paul Sture <paul@sture.ch>= Subject: Re: OpenVMS Multi-Site Cluster White Paper Availablee+ Message-ID: <VA.00000337.8a38c4d3@sture.ch>    In article hN <DC4745D1A85CA04180C83CDC706A9D180D94C1@cthexc02.americas.cpqcorp.net>, Kerry  Main wrote:A- > From: "Main, Kerry" <Kerry.Main@COMPAQ.com>  > Newsgroups: comp.os.vmsh; > Subject: OpenVMS Multi-Site Cluster White Paper Availableb' > Date: Fri, 30 Mar 2001 16:58:38 -0600  >  > Hello All, > M > Apologies if this was previously posted, but the web site for OpenVMS / SAN 4 > technology updates was updated on March 26, 2001.  > M > Among the updates is some new patch information and a technical white paper ! > on OpenVMS Multi-Site clusters.e >  > Reference:8 > http://www.openvms.compaq.com/openvms/fibre/index.html >  Beat me to it :-)r  N From my first reading of the white paper, I've already decided that we should 2 allocate some serious education time to our plans.  K I'll also mention that the July and August 2000 PDFs available at that url n are good reading too.   # This is seriously good stuff folks.P ___e
 Paul Sture Switzerlande   ------------------------------  % Date: Fri, 30 Mar 2001 23:02:33 +0200h" From: "Hans Vlems" <hvlems@iae.nl>5 Subject: Re: OpenVMS Technical Update Days - success! ( Message-ID: <9a2sdg$bdh$1@news.IAEhv.nl>  G Well I do hope we'll get a similar result in Lisbon, early May when the " European Decus event will be held.A I noticed Steve Hoffman's name among those that present sessions.C  
 Hans Vlems  * Jim Becker heeft geschreven in bericht ...8 >In article <3AC275FB.EEF4C63E@ui.urban.org>, Jim Becker <jbecker@ui.urban.org> writes: >s@ >> - The speakers themselves: Rich Marcello, Steve Hoffman, BradF >> McCusker, John Andruszkiewicz (AtoZ), Doug Gordon, Rick Barry, BartF >> Lederman, Barry Kierstein, Leo Demers, Clair Grant, Rich Colarusso,C >> Mark Hollinger, Verell Boaen, Brian Allison, Lynn Glickman, Fred I >> Kleinsorge, and Chet Jacobs from Compaq; Bill Johnson from TECsys; and- >> Norm Lastovica from Oraclea > 7 >Correction: Burns Fisher spoke instead of Clair Grant.  >t >Jim BeckerC, >The Urban Institute (http://www.urban.org/)( >Encompass (http://www.encompassus.org/)/ >ESILUG (http://encompasserve.org/lugs/esilug/)d   ------------------------------  # Date: Sat, 31 Mar 2001 13:47:37 GMT 4 From: "Terry C. Shannon" <terryshannon@mediaone.net>5 Subject: Re: OpenVMS Technical Update Days - success! < Message-ID: <Zdlx6.16879$Wz.6604974@typhoon.ne.mediaone.net>  - "Hans Vlems" <hvlems@iae.nl> wrote in message " news:9a2sdg$bdh$1@news.IAEhv.nl...I > Well I do hope we'll get a similar result in Lisbon, early May when the=$ > European Decus event will be held.  K There's also a VMS Technical Update in London on 15-17 May, and an event insJ Brussels on 15 May. Rumour has it that Charlie Matco will be in attendanceK at both events (although not both at the same time). In addition, there may   be an event in Zurich on 21 May.  H And of course, there's the ever-popular DFWDAYS event in Dallas on April! 17-20. http://dfwdays.dfwcug.org/~   terry s    ------------------------------  % Date: Sat, 31 Mar 2001 16:06:48 +0000=) From: Christof Brass <brass@infopuls.com>=Y Subject: Re: OT: Elements of Linguistics (was Re: POSIX Streams, File Permissions) Permisi, Message-ID: <3AC60098.9BECCFD7@infopuls.com>   Ben Sego wrote:l >  > "David J. Dachtera" wrote: >  > > Christof Brass wrote:h  > > > [snip] There are countless > > > homophones in English. > >o" > > F.Y.I., in (American) English: > >o >  > <snip> >  > > 0 > > As always, I could be (probably am) wrong... > N > Well, DJD, this time I have to side with Christof.  (And you _know_ how thatM > makes me feel.) "Homophones" are words that sound alike, but have differento( > meaning.  Here is a homophonic triple: >  > bightA > bite > byte > N > Incidentally, it is not a requirement that the words be spelled differently;M > but rather that they 1.) sound the same, 2.) have separate origins, and 3.)c > have separate meanings.  > $ > Ben "WKU taught English, too" Sego  > I vote for 1.) and 2.) - I don't regard the third condition as> necessary because having condition 2.) fulfilled the words are? in essence different even if one of their meanings is the same.N; I don't know even of such an example exists. But if not the  third condition is redundant.i   ------------------------------  # Date: Sat, 31 Mar 2001 11:09:37 GMTp1 From: CSABA  HARANGOZO   <csabah@zipworld.com.au>C( Subject: Re: Redirect output of Com File9 Message-ID: <RVix6.1086$CN.180466@nostril.pacific.net.au>r  % Don Chandler <dcq@y12.doe.gov> wrote:uJ > How can I direct the output of a '.com' file to BOTH the screen and to aF > file at the same time. After the run of an interactive '.com' file I> > would like to have the option to print  (or save) the screen > interactions.t > Don)   	Another way could be :n   	$@TT:/OUTPUT=<younameit.log>u" 	_$... enter DCL commands here ... 	_$ .... 	_$CTRL/Z ( *EXIT* or F10 )o 	$  1 	Your session now should be in "<younameit.log>".i  < 	( The above is courtesy of this newsgroup, some kind person 	posted this way back... ) 					Cheers,	   Csabau  I    ----------------------------------------------------------------------vE    * Csaba I. Harangozo     |    'To err is human', said the hedgehog E    * csabah@zipworld.com.au |           as he dismounted a wirebrush.iI    ----------------------------------------------------------------------N;    EARTH::AUSTRALIA:[SYDNEY]HARANGOZO.CSABA;1, delete? [N]:    ------------------------------  # Date: Sat, 31 Mar 2001 06:56:17 GMTk( From: Terry Kennedy <terry@gate.tmk.com>1 Subject: Re: Seeking CD-R/CD-RW SCSI INQUIRY datas' Message-ID: <GB1v9u.IHC@spcuna.spc.edu>u  B In comp.os.vms Hoff Hoffman <hoffman@xdelta.zko.dec.nospam> wrote: >     Yamaha CDR100h >     Yamaha CDR400d  F   FWIW, these are both incredibly obsolete widgets. I've owned both ofD 'em, as well as the CDRW4260S. After that, I started buying internal> IDE burners (TDK VeloCD) as they were much faster/cheaper/etc.  C   If this has something to do with supporting burners under VMS, itm1 would be really cool to support the IDE VeloCD's.a  4         Terry Kennedy             http://www.tmk.com5         terry@tmk.com             Jersey City, NJ USAe   ------------------------------  # Date: Sat, 31 Mar 2001 11:30:16 GMTt= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)rC Subject: Re: Seeking CD-R/CD-RW SCSI INQUIRY data / Yamaha CRW4416Su0 Message-ID: <009F9D34.8377E3DF@SendSpamHere.ORG>  B In article <01033021130179@antinode.org>, sms@antinode.org writes:> >From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) >> Yamaha CRW4416S:i > E >   Coincidentally, I just hooked mine back up to my AlpSta 200 4/233 J >(VMS V7.2-1), after consolidating the two old disks onto one newer disk. I >With the shorter SCSI cables, it no longer completely paralyzes the SCSI  >bus.r >  >> $! Issuing INQUIRY QIOW >? >   Mine's slightly different: >o$ >$!   SCSI Inquiry Data (58 bytes) : >$!]> >$!      000:  05 80 02 02 27 00 00 10 59 41 4D 41 48 41 20 20> >$!      010:  43 52 57 34 34 31 36 53 20 20 20 20 20 20 20 20> >$!      020:  31 2E 30 68 30 39 2F 30 39 2F 39 39 00 00 00 00, >$!      030:  00 00 00 00 00 00 00 00 00 00 >[...]% >$!      Product Revision Level: 1.0hi >y@ >   Even with the shorter cables, it still appears at every LUN: >e >alp $ show device dka > Q >Device                  Device           Error    Volume         Free  Trans MntaQ > Name                   Status           Count     Label        Blocks Count Cnt Q >ALP$DKA0:               Mounted              2  VMS072ALP     15133685   545   1t/ >ALP$DKA400:             Online               0n/ >ALP$DKA500:             Online               0r >s/ >Device                  Device           Errore/ > Name                   Status           Countt/ >ALP$DKA501:             Offline              1a/ >ALP$DKA502:             Offline              1d/ >ALP$DKA503:             Offline              1 / >ALP$DKA504:             Offline              1 / >ALP$DKA505:             Offline              1e/ >ALP$DKA506:             Offline              1h/ >ALP$DKA507:             Offline              1e >cE >   Is there an approved way to suppress the extras, or should I stopt, >caring, or am I the only one who gets them?  G You are NOT the only one that gets them.  I also do and I complained toyG Yamaha numerous times.  Sent the unit to and fro several times with theMI Yamaha service organizataion and somewhere in these exchanges and entire- F ly new unit was given to me.  About 4 different firmware updates were F done by Yamaha and the device still has more personalities than Cybil.    I >   Also, running CDRECORD 1.8.1 (slightly modified), I did a "-dummy" ofpF >about a full disk worth and got about 170 errors on PKA0:.  Normal or >not?t >qI >   Is "%MOUNT-F-FORMAT, invalid media format" the best I can expect when E >trying to mount a normal CD-ROM on the thing?  That's with the block  >size jumper installed.  > G >   I haven't tried it lately, but as I recall, booting from it failed,  >too.t > C >   Comments from anyone with relevant experience would be welcome.-  G I use mine ONLY for burning.  It will not mount and read its own CDr/rwh6 or any CDrom.  I get the annoying sub-lun devices too.  F The only other Yamaha branded item I possess is my electric 12 string F acoustic guitar.  My keyboards -- from a Roland SH-1000 to a KR-570 --G are all Rolands.  With the kind of service and response I got from Yam-oG aha with the CDR4416S, I am damned happy I don't own other Yamaha elec-h
 tronics.     i  F Still, despite the brain-damage in the device, it faithfully burns CDRH and CDRW media.  Ne'er made a coaster with the device.  I suppose that's$ one plus amongst all of the minuses. --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMf            MO city, n., 1. a place where trees are cut down and streets are named after them.b   ------------------------------  # Date: Sat, 31 Mar 2001 11:35:03 GMTa= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)sC Subject: Re: Seeking CD-R/CD-RW SCSI INQUIRY data / Yamaha CRW4416Sy0 Message-ID: <009F9D35.2E310057@SendSpamHere.ORG>  B In article <01033022385291@antinode.org>, sms@antinode.org writes:A >   And while I'm whining, if I turn off the power on the (YamahatF >CDR4416S) thing, I start getting more errors on PKA0:.  If the systemI >can't really use the thing, I would see it as more gracious if it didn't 1 >log "Target Timed Out" errors when it goes away.i  G I want back and looked at your last post.  1.0h eeeek!  I have 1.0f.  I2H have 1.0g on the unit and it was so screwed up that I sent the unit backH to Yamaha and demanded the 1.0f be restored.  Assuming 1.0h is a follow-H up to 1.0g incorporarting 1.0g "fixes", I suspect you are running brain- damaged firmware.    See 1.0f is available.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM2            rO city, n., 1. a place where trees are cut down and streets are named after them.m   ------------------------------  % Date: Sat, 31 Mar 2001 09:06:13 +0100m From: Roy Omond <Roy@Omond.net>r% Subject: Re: Spice for OpenVMS Alpha?t) Message-ID: <3AC58FF5.4104A234@Omond.net>i  
 ualski wrote:a   > Paul Repacholi wrote:o > >t' > > Koloth <koloth@tmisnet.com> writes:c > >bF > > > > DECUS says they're for VAX/VMS but I remember just fortran the1 > > > > last time I compiled it, a few years ago.  > >e2 > > That is an OLD SPICE. The current one is in C. >-F > I should have been more specific.  The first one I listed is FORTRAND > but the newer one, and the second listed, is purported to be C.  I* > don't have any experience with that one. >oD > Anyone know what sort of new features or fixes went along with the > switch to C?   Fortran to C ?   New "features" ??e  # Surely a euphemism for new "bugs" !   	 Roy Omondl Blue Bubble Ltd.   ------------------------------    Date: 31 Mar 2001 18:17:10 +0800, From: Paul Repacholi <prep@prep.synonet.com>% Subject: Re: Spice for OpenVMS Alpha?l- Message-ID: <8766gqtf89.fsf@prep.synonet.com>o  % ualski <ualski@earthlink.net> writes:    > Paul Repacholi wrote:t > > ' > > Koloth <koloth@tmisnet.com> writes:u > > F > > > > DECUS says they're for VAX/VMS but I remember just fortran the1 > > > > last time I compiled it, a few years ago.d > > 2 > > That is an OLD SPICE. The current one is in C.  F > I should have been more specific.  The first one I listed is FORTRAND > but the newer one, and the second listed, is purported to be C.  I* > don't have any experience with that one.  C A new modeling method for MOSFETs, improvments to the other models,-
 bug fixes.  D > Anyone know what sort of new features or fixes went along with the > switch to C?  1 More bugs, harder to optimize, unix braindeath...s  D But, hey, they a proper industry stands conforming bugs! Thats a big improvment isn't it...   -- r< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  % Date: Sat, 31 Mar 2001 14:32:31 +0000c) From: Christof Brass <brass@infopuls.com>l% Subject: Re: Spice for OpenVMS Alpha?y, Message-ID: <3AC5EA7F.A5569BCD@infopuls.com>   andrew harrison wrote: >  > Paul Repacholi wrote:i > > ' > > Koloth <koloth@tmisnet.com> writes:r > >dF > > > > DECUS says they're for VAX/VMS but I remember just fortran the1 > > > > last time I compiled it, a few years ago.7 > >02 > > That is an OLD SPICE. The current one is in C. > >R > $ > Don't let Christof find out :):):) > 	 > RegardsL > Andrew Harrisonv > Enterprise IT ArchitectP  
 Too late ..../   ------------------------------  % Date: Sat, 31 Mar 2001 16:17:30 +00003) From: Christof Brass <brass@infopuls.com>f% Subject: Re: Spice for OpenVMS Alpha?Y, Message-ID: <3AC6031A.8E62723E@infopuls.com>   Paul Repacholi wrote:- > ' > ualski <ualski@earthlink.net> writes:0 >  > > Paul Repacholi wrote:@ > > >s) > > > Koloth <koloth@tmisnet.com> writes:o > > >eH > > > > > DECUS says they're for VAX/VMS but I remember just fortran the3 > > > > > last time I compiled it, a few years ago.r > > >C4 > > > That is an OLD SPICE. The current one is in C. > H > > I should have been more specific.  The first one I listed is FORTRANF > > but the newer one, and the second listed, is purported to be C.  I, > > don't have any experience with that one. > E > A new modeling method for MOSFETs, improvments to the other models,a > bug fixes. > F > > Anyone know what sort of new features or fixes went along with the > > switch to C? > 3 > More bugs, harder to optimize, unix braindeath...n > F > But, hey, they a proper industry stands conforming bugs! Thats a big > improvment isn't it... >  > --> > Paul Repacholi                               1 Crescent Rd.,9 > +61 (08) 9257-1001                           Kalamunda.wB >                                              West Australia 60760 > Raw, Cooked or Well-done, it's all half baked.  / You are doing *my* job :-) (not only this time)0   ------------------------------  # Date: Sat, 31 Mar 2001 11:58:52 GMTA7 From: "Lyle W. West" <lyle.west@childrenshc.org.nospam>R) Subject: Re: THE EMC Chronicles - revised=5 Message-ID: <3AC57208.8A4499F@childrenshc.org.nospam>4   Bill Todd wrote: > N > What at least *seems* to be abundantly clear is that someone landed - hard -K > on Lyndon.  The question now is, was that reaction EMC-instigated, or dide2 > someone in c.o.v. do something foolish, or what? >  > - bill  E You are quite perceptive, Bill. While we have thus far been unable tocA communicate directly with the person we feel used the 'Chronical' E thread to push back on their local sales being, since we were touted n> as a reference site, it appears a reader of COV used the above@ referenced thread as a means of showing their management and theE vendor this may not be the ideal solution. It appears there was localnD opposition to the vendor prior to the appearance of the thread, and C the vendor was provided with a copy of the article (they can't read:- news???)) which wound up at our organization.i  D I read COV 7 days a week, and do not recall anyone posting any  infoF regarding fibre implementation of VMS Alpha -> KGPSA -> Brocade switchE  -> storage without use of a HSGxx device. It is my impression we mayeF have broken new ground here, and without CSC and a couple of dedicatedF (and former VMS engineneers) from the vendor, our current status would probably been much delayed.   D Lyndon came to this organization over two years ago with close to 12C years of VMS experience, yet had never worked with Alpha VMS nor CI F technology. Utilizing a reaonable amount of guidence from myself, CSC,F and COV, I would rate him as a 7 on a scale of 1-10. I add this not soA he will like and respect me, we already have a good relationship. C I would suggest me more people with interest in the content of the mF subject of this thread do as Bill Todd did and read between the lines.  B It is my personal opionion I can neither endorse nor denigrate the7 aforementioned product. Only time will give the result.e  A While I must reinterate the fact that this is my opinion and does D not neccesarily reflect that of Lyndon nor my employer, I do feel itC is becoming a attention attracting situation when Freedom of Speech@> is only appropriate when it coincides with the proper opinion. --  0 My opinions seldom reflect those of my employer.   Lyle W. West   ------------------------------    Date: 31 Mar 2001 19:19:18 +0800, From: Paul Repacholi <prep@prep.synonet.com>9 Subject: Re: VMS: DECNet to TCP/IP migration help wanted.r- Message-ID: <87wv96rxs9.fsf@prep.synonet.com>s  7 danco@cx48228-c.escnd1.sdca.home.com (Dan Cook) writes:M  B > Multinet can do it _without_ DECnet/OSI aka DECnet Phase V.  NotF > only can it do "DECnet over IP", it can also do "IP over DECnet" (of( > all the strange things to want to do.)  B Like routing and flow control that work? Yeah, perverted I know...8 Sadly, BGP and friends can still F*** you along the way.   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.g@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  % Date: Sat, 31 Mar 2001 11:28:48 -0400n+ From: Tim Shoppa <shoppa@trailing-edge.com>:. Subject: When might SYS$SCHDWK not wake me up?1 Message-ID: <3AC5BF70.430F9447@trailing-edge.com>.  H We have a control system (VAX 6000-610, VMS5.5-2, similarly ancient UCX)> that has a process that's supposed to wake up every second, do: some rather trivial processing, and then fall back asleep.  F On wake-up, it uses SYS$SCHDWK to schedule a single wake-up for itselfB one second in the future.  It then does its simple processing, and does a SYS$HIBER.   B This works fine 99.9999% of the time, but every few weeks or every> few months the process won't wake up by itself.  If I kick theA sleeping process by referencing it in a SYS$WAKE call run from an/? interactive program, it wakes back up and then runs for another-$ few weeks or months without a hitch.  ? The status codes returned by the SYS$SCHDWK and SYS$HIBER callsE' are checked and apparently all succeed.e  9 So the question: Why is the process not waking itself up?7  > If I were to rewrite this simple process from scratch, I wouldA use the 4th argument of SYS$SCHDWK to schedule a repeated wake-ups= request for every second.  But I'd rather understand what the > reason for it not working as it currently is rather than go in and start screwing around.  = By simple experimentation, it doesn't seem to hurt any if the_= "trivial" processing between the SYS$SCHDWK and the SYS$HIBERo@ takes more than a second.  I'm tempted to point my finger at the@ decrepit old UCX on the system as responsible for disrupting the@ VMS timer queue, but there is no other apparent problem anywhere in the system.   Tim.   ------------------------------  % Date: Sat, 31 Mar 2001 12:14:46 -0500 * From: "Stanley F. Quayle" <stan@stanq.com>2 Subject: Re: When might SYS$SCHDWK not wake me up?. Message-ID: <3AC5CA37.30347.85E5187@localhost>  + On 31 Mar 2001, at 11:28, Tim Shoppa wrote:eH > On wake-up, it uses SYS$SCHDWK to schedule a single wake-up for itselfD > one second in the future.  It then does its simple processing, and > does a SYS$HIBER.- > ; > So the question: Why is the process not waking itself up?O > @ > If I were to rewrite this simple process from scratch, I wouldC > use the 4th argument of SYS$SCHDWK to schedule a repeated wake-up ? > request for every second.  But I'd rather understand what the @ > reason for it not working as it currently is rather than go in > and start screwing around.  A You may find that your process "walks" a little bit each time it aD wakes up.  It may be a few milliseconds, but they eventually add up  to a second.  C Sometimes we have to abandon understanding and do what works.  The RC recurring wakeup will work.  And if it somehow gets missed on that eC 0.0001%, it'll wake up in another second -- so fast you won't even e notice.s   --Stan  
 ----------G Stanley F. Quayle, P.E.   N8SQ   +1 614-868-1363   Fax: +1 614 868-1671-1 8572 North Spring Ct. NW, Pickerington, OH  43147 = Preferred address:  stan@stanq.com       http://www.stanq.comw   ------------------------------  % Date: Sat, 31 Mar 2001 18:45:21 +0100-+ From: "antonio.carlini" <arcarlini@iee.org>-2 Subject: Re: When might SYS$SCHDWK not wake me up?' Message-ID: <3AC617B1.59E87EBA@iee.org>-   Tim Shoppa wrote:a > H > On wake-up, it uses SYS$SCHDWK to schedule a single wake-up for itselfD > one second in the future.  It then does its simple processing, and > does a SYS$HIBER.n  6 If it's as simple a loop as it sounds, it is difficult! to see what might be going wrong.   . If a wakeup happens while the process is *not*0 hibernating, the subsequent $HIBER will complete/ immediately. However, the number of wakeups and 0 $HIBERs are not matched so if you do two wakeups, and *then* two $HIBERs you'll sleep forever.  . I would consider adding some debugging to keep+ track of the total number of wakeups issuedD( and the total number of $HIBERs invoked.. In addition keep track of the *maximum* number1 of "outstanding" wakeups (increment "outstanding"-/ as a wakeup is scheduled, decrement as a $HIBER ) is invoked; keep track of the maximum in e "maximum outstanding").c  3 (You'll need some way to report this to the outsideW/ world ... perhaps logical names might help herer/ if you have no other handy mechanism built in).>  4 You should hopefully never see "maximum outstanding"* as anything other than 1 (or, briefly, 0).  ? > By simple experimentation, it doesn't seem to hurt any if the ? > "trivial" processing between the SYS$SCHDWK and the SYS$HIBERoB > takes more than a second.  I'm tempted to point my finger at the  ? That should be OK ... the $HIBER ought to complete immediately.N  B > decrepit old UCX on the system as responsible for disrupting theB > VMS timer queue, but there is no other apparent problem anywhere > in the system.  . And it would be strange if it only ever picked on *your* TQEs!    Antonio0     -- 3   --------------- - Antonio Carlini             arcarlini@iee.org0   ------------------------------  % Date: Sat, 31 Mar 2001 12:56:22 -0400 + From: Tim Shoppa <shoppa@trailing-edge.com>s2 Subject: Re: When might SYS$SCHDWK not wake me up?0 Message-ID: <3AC5D3F6.2B9F474@trailing-edge.com>   Stanley F. Quayle wrote: > - > On 31 Mar 2001, at 11:28, Tim Shoppa wrote: J > > On wake-up, it uses SYS$SCHDWK to schedule a single wake-up for itselfF > > one second in the future.  It then does its simple processing, and > > does a SYS$HIBER.P > >0= > > So the question: Why is the process not waking itself up?  > >DB > > If I were to rewrite this simple process from scratch, I wouldE > > use the 4th argument of SYS$SCHDWK to schedule a repeated wake-up A > > request for every second.  But I'd rather understand what the B > > reason for it not working as it currently is rather than go in > > and start screwing around. > B > You may find that your process "walks" a little bit each time itE > wakes up.  It may be a few milliseconds, but they eventually add upd > to a second.  D The $SCHDWK is done with a delta-time (sorry that I left that out!).C Another possibly relevant thing I left out: the process in questionp@ is running at "normal" interactive priority, but there are otherA processes in the system running with elevated priority (up to 15,hD nearly a real-time priority.)  Normal CPU utilization is around 10%.A There doesn't appear to be any particularly heavy CPU or I/O loada@ when the scheduled wakeup fails to come, and my manually-induced+ wakeup *always* brings things back to life.p  C The accuracy or precision of the wake-up period isn't vital, but itrA *is* necessary that the process does eventually wake up (a couples@ of seconds would be good, but sometimes it's been hours before I> got to Rail Central to do a manual wake-up and the system will run happily after that.)  D > Sometimes we have to abandon understanding and do what works.  TheD > recurring wakeup will work.  And if it somehow gets missed on thatD > 0.0001%, it'll wake up in another second -- so fast you won't even	 > notice.e  > The thing is, SYS$HIBER and $SCHDWK are used in gobs of placesE all throughout the basic tools of VMS.  I really doubt that something-= so vital to system operation is failing for reasons beyond myt= control (i.e. I'm assuming this is a case of design or codingC, error in the application, not OS flakiness!)  C This particular system is running a US$10Billion dollar peripheral,dE so I'm reluctant to try random things without understanding the causei of the problem :-).p   Tim.   ------------------------------  % Date: Sat, 31 Mar 2001 13:10:49 -0400t+ From: Tim Shoppa <shoppa@trailing-edge.com>)2 Subject: Re: When might SYS$SCHDWK not wake me up?1 Message-ID: <3AC5D759.57609D9D@trailing-edge.com>    antonio.carlini wrote: >  > Tim Shoppa wrote:  > >1J > > On wake-up, it uses SYS$SCHDWK to schedule a single wake-up for itselfF > > one second in the future.  It then does its simple processing, and > > does a SYS$HIBER.E > 8 > If it's as simple a loop as it sounds, it is difficult# > to see what might be going wrong.   < Yeah, I'm tempted to look at the machine code being produced< (it's built with VAX C) and see if there's a "gotcha" to the> loop that's not apparent from the C source code (i.e. some way= to reach the $HIBER without having done the $SCHDWK, or maybee? some sort of stack corruption that could corrupt the delta-time!/ argument on its way to the SYS$SCHDWK request.)w  0 > If a wakeup happens while the process is *not*2 > hibernating, the subsequent $HIBER will complete1 > immediately. However, the number of wakeups and"2 > $HIBERs are not matched so if you do two wakeups. > and *then* two $HIBERs you'll sleep forever.  = It's always: $SCHDWK, process, $HIBER, repeat.  The "process"O step may do some I/O.   0 > I would consider adding some debugging to keep- > track of the total number of wakeups issued-* > and the total number of $HIBERs invoked.  < We already do thorough logging of the status codes resulting= from $SCHDWK and $HIBER, but I will seriously consider adding 9 the counters you've suggested.  Everything I've read saysh= that a process may only put itself into HIB - nobody else cane be HIBernating me, right?   / >And it would be strange if it only ever pickeda >on *your* TQEs!    = I know, I know!  And $SCHDWK/$HIBER are so fundamental to VMSh* that I know the OS cannot be at fault :-).  A Is there a supported method for looking at any outstanding TQE's?p= I sort-of suspect that somehow the delta-time argument to the.= $SCHDWK is corrupted, and if I look and see outstanding TQE'snB for years or millenia in the future this would be the smoking gun.   Tim.   ------------------------------  % Date: Sat, 31 Mar 2001 13:10:11 -0500-* From: "Stanley F. Quayle" <stan@stanq.com>2 Subject: Re: When might SYS$SCHDWK not wake me up?. Message-ID: <3AC5D733.23290.8910D95@localhost>  D > > You may find that your process "walks" a little bit each time itG > > wakes up.  It may be a few milliseconds, but they eventually add upi > > to a second. > ' > The $SCHDWK is done with a delta-timeS  F A delta-time like "::1"?  If so, your process is definitely waking up C a little later each time.  As your system gets busier, the problem   only gets worse.  @ If it takes 10 milliseconds for the process to recover from the 9 wakeup, you'll lose a whole cycle in just 100 iterations.   E Since the recurring wakeup parameter of $SCHDWK is a delta time, you F8 already have all you need to make that work permanently.     --Stan  
 ----------G Stanley F. Quayle, P.E.   N8SQ   +1 614-868-1363   Fax: +1 614 868-1671e1 8572 North Spring Ct. NW, Pickerington, OH  43147y= Preferred address:  stan@stanq.com       http://www.stanq.come   ------------------------------  % Date: Sat, 31 Mar 2001 12:07:58 +0100c  From: Paul Sture <paul@sture.ch>3 Subject: Re: [Q] param to increase for unsuff SPTE? + Message-ID: <VA.0000033a.8a813c30@sture.ch>   M In article <y4vgosrh3m.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>, Jan   Vorbrueggen wrote:I > From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>R > Newsgroups: comp.os.vmsy5 > Subject: Re: [Q] param to increase for unsuff SPTE?w" > Date: 29 Mar 2001 12:42:53 +0200 > $ > John Santos <JOHN@egh.com> writes: > F > > Since there was an error creating a global section, my guess wouldE > > be one of the global section parameters, GBLSECTIONS, GBLPAGES or  > > maybe GBLPAGFIL.   > K > GBLSECTIONS is dynamic now, isn't it? It only limits the number of global L > sections that can be open, not their size. GBLPAGFIL limits the pages thatH > global sections can occupy in the page file (usually writable sectionsO > selecting the page file as backing store because they don't require permanentAO > storage). GBLPAGES is either irrelevant or dynamic on Alpha, I think, because P > the fixed allocation of S0 space that was needed on VAX is not required on the	 > Alpha. o >   M GBLPAGES and GBLPAGFIL are dynamic on Alpha VMS (since 7.1?). GBLSECTIONS is t not.  - None of the three are dynamic on VAX VMS 7.2.p  N > The best explanation I've heard is that S0 space is full. Put the section inF > S2, or reduce the amount of space required by other parts of the OS. >  >  Jan >    ___h
 Paul Sture Switzerlandn   ------------------------------   End of INFO-VAX 2001.181 ************************