1 INFO-VAX	Wed, 28 Feb 2001	Volume 2001 : Issue 118       Contents:; ??== MadGoat PCX, Alphaworkstation and RX26 diskette drive. ? Re: ??== MadGoat PCX, Alphaworkstation and RX26 diskette drive. @ Re: Alpha: game over. Was RE: Future outlook for OpenVMS admins? AXP vs. Sun... Re: Backups problem...2 Compaq to transfer clustering technology to Oracle6 RE: Compaq to transfer clustering technology to Oracle Re: Compaq wins APAC re-bid # Comparison to Absolute date convert  Re: DCL content (long) RE: DEC Alpha's - FREE T-Shirt' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later ' Re: Drop C/C++ better sooner than later  Re: forum.compaq.com Free to a good home :-) & Re: Future outlook for OpenVMS admins?' FW: Drop C/C++ better sooner than later  GTK+ on VMS C Re: How to get Polycenter Performance Advisor Graphs to a web page? P Re: Installing Availability Manager V1.4 on NT 4.0 breaks ReflectionsV8.0.2 V8.0 Invalid mail notification  Re: Invalid mail notification  Re: Invalid mail notification  Re: Invalid mail notification  Issue dcl command from NT  Re: Issue dcl command from NT  Re: Issue dcl command from NT  Re: Janitor fixes 90L  Re: Janitor fixes 90L  Re: Janitor fixes 90L  Re: Janitor fixes 90L  Re: Low cost workstations  Re: Low cost workstations / RE: My name is Dean, and I'm a [Un*x|NT] hater. / Re: My name is Dean, and I'm a [Un*x|NT] hater. / RE: My name is Dean, and I'm a [Un*x|NT] hater. / Re: My name is Dean, and I'm a [Un*x|NT] hater. / Re: My name is Dean, and I'm a [Un*x|NT] hater. / Re: My name is Dean, and I'm a [Un*x|NT] hater. / Re: My name is Dean, and I'm a [Un*x|NT] hater. / Re: My name is Dean, and I'm a [Un*x|NT] hater. / RE: My name is Dean, and I'm a [Un*x|NT] hater.  OpenVMS Technical Update Day's/ Re: PCSI kits - where was a specific kit built? / Re: PCSI kits - where was a specific kit built? / Re: PCSI kits - where was a specific kit built?   Re: Possible security hole in...  RE: Possible security hole in...4 Re: Problem compiling Fortran program with /SEPARATE4 Re: Problem compiling Fortran program with /SEPARATE4 Re: Problem compiling Fortran program with /SEPARATE4 Re: Problem compiling Fortran program with /SEPARATEC Re: Sun Blade 100 with a VMS monitor ? (was: Low cost workstations) 
 Re: TCPIP 5.0 E the haunted vt100 (was Re: a little humor for this so often humorless  Re: UCX/TCPIP 5.1 for HobbyistI Re: US telephone vote on SPAM in comp.os.vms (was: High Search Engine...)  VMS wanted list. Take II Re: VX1 and PC164LX  Re: VX1 and PC164LX ! WKU FILESERV: Updated DSNLINK_NEW + WKU FILESERV: Updated MMK, plus WASD mirror % [info] Telnet/Rlogin and 7.2-1 broken ) Re: [info] Telnet/Rlogin and 7.2-1 broken + [Q] concatenating executable files is OK??? / Re: [Q] concatenating executable files is OK??? / Re: [Q] concatenating executable files is OK??? / Re: [Q] concatenating executable files is OK???  [Request] f$element delimiter ! RE: [Request] f$element delimiter   F ----------------------------------------------------------------------  % Date: Wed, 28 Feb 2001 13:54:14 +0100 , From: aus@vim.uni-wuerzburg.de (Hans M. Aus)D Subject: ??== MadGoat PCX, Alphaworkstation and RX26 diskette drive.D Message-ID: <aus-2802011354140001@wvia72.virologie.uni-wuerzburg.de>  2 Do I understand the MAdGoat PCX utility correctly?  B PCX can read and write standard 1.44MByte PC diskettes on an AlphaD Workstation 433au, with OpenVMS 7.2-1 and a RX26 drive - without any additional software.   --  B Cheers, Hans M. Aus, Wuerzburg, Germany,  aus@vim.uni-wuerzburg.de   ------------------------------  % Date: Wed, 28 Feb 2001 14:54:01 +0100 ' From: Theo Jakobus <jakobus@iaf.fhg.de> H Subject: Re: ??== MadGoat PCX, Alphaworkstation and RX26 diskette drive.* Message-ID: <3A9D1109.2EC47E80@iaf.fhg.de>   Hans M. Aus wrote: > 4 > Do I understand the MAdGoat PCX utility correctly? > D > PCX can read and write standard 1.44MByte PC diskettes on an AlphaF > Workstation 433au, with OpenVMS 7.2-1 and a RX26 drive - without any > additional software. >  > --D > Cheers, Hans M. Aus, Wuerzburg, Germany,  aus@vim.uni-wuerzburg.de    " Yes!!! It's really a fine utility.  D I'm using PCX to transfer data files, downloaded executables to a PC= which isn't connected to the LAN. You have to use the command ' WRITE/BINARY for writing to the floppy.      Regards, --    ; *********************************************************** ; *                                                         * ; *  Theo Jakobus                                           * ; *  Fraunhofer-Institut fuer Angewandte Festkoerperphysik  * ; *  Tullastr. 72                                           * ; *  D-79108 Freiburg                                       * ; *  Germany                                                * ; *  Phone:   +49-(0)761-5159-325                           * ; *  FAX :    +49-(0)761-5159-200                           * ; *  e-mail:  jakobus@iaf.fhg.de                            * ; *  http://www.iaf.fhg.de                                  * ; *                                                         * ; ***********************************************************    ------------------------------    Date: 28 Feb 2001 09:56:23 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> I Subject: Re: Alpha: game over. Was RE: Future outlook for OpenVMS admins? H Message-ID: <y4itlvyymg.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  # Shane.F.Smith@Healthnet.com writes:   L > I was under the impression they got burned with the 486 because they foundI > out too late that they couldn't copyright a number. Anyone could make a  > chip called a 486.  M For an IP department that registers marks by the busload, that one was really L gross incompetence, or gross negligence ("hey, we're intel, nobody will dareN to abuse our marks!"). I suppose some people were roasted to a slow death over. that failure. Or not. Just another Challenger.   	Jan   ------------------------------  % Date: Wed, 28 Feb 2001 12:37:20 -0000 = From: Nicholas Mark de Smith <Nicholas.MarkdeSmith@liffe.com>  Subject: AXP vs. Sun... M Message-ID: <FCB2387D053DD31180B200805FEA4C3F28B666@ntprdex5.admin.liffe.com>    Probably very old hat, but:   1 http://www.theregister.co.uk/content/2/17100.html    Nick   ------------------------------    Date: 28 Feb 2001 10:06:15 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>  Subject: Re: Backups problem... H Message-ID: <y4d7c3yy60.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  N BACKUP usually writes redundant XOR blocks which allow it to revocer from lostM blocks; this is controlled by the /GROUP switch whose default is 10, IIRC. So M you will add an overhead of 10% for that, plus a little for BACKUP's internal H control information. That means that you 36 GB can just exceed 40 GB. InK addition, you have to make sure you use large block sizes and actually keep K the tape streaming, otherwise the realized capacity can be (sometimes mcuh)  lower.   	Jan   ------------------------------  % Date: Wed, 28 Feb 2001 14:39:21 +0000 % From: Alan Greig <a.greig@virgin.net> ; Subject: Compaq to transfer clustering technology to Oracle 8 Message-ID: <rs2q9t43uc8vs6vfadm951bi9hheal3hk3@4ax.com>  A At the recent analyst conference Winkler stated that Compaq would B license its clustering technology to Oracle and would work closely; with both Oracle and Microsoft to implement this on Windows  Datacentre.   ; Well it looks like this has just been announced officially  N http://news.cnet.com/investor/news/newsitem/0-9900-1028-4961213-CPQ.html?tag=l     Note one particular bit   A "-- Technology Development -- Oracle will license selected Compaq D Tru64 UNIX clustering technology to be integrated into Oracle9i RealE Application Clusters to create Oracle's Portable Cluster Layer (PCL). E This PCL, initially available on Tru64 UNIX, will become available on ' a wide variety of computing platforms."   F Note the "wide variety of computing platforms". To me this is code forA Windows Datacentre which was explicitly mentioned by Winkler as a D primary target for this project. It's also the only other  OS CompaqA sell that could take advantage of this deal unless they intend to  resell HP-UX or Solaris.  E So it seems that Winkler really did pre-announce Compaq policy at the D analysts conference. I wonder how many more of his pre-announcements' will come true over the course of time.    -- Alan   ------------------------------  % Date: Wed, 28 Feb 2001 12:35:19 -0600 + From: "Main, Kerry" <Kerry.Main@COMPAQ.com> ? Subject: RE: Compaq to transfer clustering technology to Oracle N Message-ID: <910612C07BCAD1119AF40000F86AF0D805284E8E@kaoexc3.kao.cpqcorp.net>   Alan,   6 >>> Note the "wide variety of computing platforms".<<<  F My albeit limited understanding is that the code will be integrated byL Oracle as part of the Oracle database / applic offerings which means it willF eventually be offered on the other OS platforms that Oracle supports.   E So, if the Cust at some future point wants to take advantage of these H database clustering features on W2K, when Oracle releases the version ofK their software that supports this platform, then that Customer will be able  to do this.   I This is very similar to what the Oracle Parallel Server (OPS) strategy is  today.   Regards,  
 Kerry Main Senior Consultant  Compaq Canada Inc. Professional Services  Voice: 613-592-4660  Fax  :  819-772-7036 Email: Kerry.Main@Compaq.com     -----Original Message-----, From: Alan Greig [mailto:a.greig@virgin.net] Sent: February 28, 2001 9:39 AM  To: Info-VAX@Mvb.Saic.Com ; Subject: Compaq to transfer clustering technology to Oracle       A At the recent analyst conference Winkler stated that Compaq would B license its clustering technology to Oracle and would work closely; with both Oracle and Microsoft to implement this on Windows  Datacentre.   ; Well it looks like this has just been announced officially  L http://news.cnet.com/investor/news/newsitem/0-9900-1028-4961213-CPQ.html?tag =l     Note one particular bit   A "-- Technology Development -- Oracle will license selected Compaq D Tru64 UNIX clustering technology to be integrated into Oracle9i RealE Application Clusters to create Oracle's Portable Cluster Layer (PCL). E This PCL, initially available on Tru64 UNIX, will become available on ' a wide variety of computing platforms."   F Note the "wide variety of computing platforms". To me this is code forA Windows Datacentre which was explicitly mentioned by Winkler as a D primary target for this project. It's also the only other  OS CompaqA sell that could take advantage of this deal unless they intend to  resell HP-UX or Solaris.  E So it seems that Winkler really did pre-announce Compaq policy at the D analysts conference. I wonder how many more of his pre-announcements' will come true over the course of time.    -- Alan   ------------------------------    Date: 28 Feb 2001 23:51:13 +0800, From: Paul Repacholi <prep@prep.synonet.com>$ Subject: Re: Compaq wins APAC re-bid- Message-ID: <87pug2olfy.fsf@prep.synonet.com>   2 andrew harrison <andrew.nospam@uk.sun.com> writes:  F > Your suggestion was that the E10KS had failed their acceptance tests< > because of the eCache issues. That wasn't the case was it.  F > The only interesting thing about the exchange we had was you abilityF > to attribute blame based on zero evidence, a characteristic that you9 > have demonstrated consistently over the last 4-5 years.    Andrew, you're back!  B So, it wasn't performance, and it wasn't, um, 'temperature induced zinc suficiency' either...   Well, from Tues Australian:    Compaq's solution a super deal.  ( by Selina Mitchel )   / ...Sun Microsystems failed to meet requirements  ... K " We've unexpectadly got more for our money, in a shorter period of time. "  Prof. O'Calligan said. ...   D BTW, the Pacific Peso, aka $AU is worth a lot less now than when Sun penned the papers.   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  % Date: Wed, 28 Feb 2001 08:34:54 -0600 1 From: "Mark-Simon Pope" <mpope@NOSPAM@bristol.ca> , Subject: Comparison to Absolute date convert. Message-ID: <Kc8n6.338$tn1.4324@news1.mts.net>  I It bug's me that Absolute_date = f$cvtime("2001-03-01","ABSOLUTE","DATE") L does not change comparison date to absolute.  I have ways around this, usingH several lines of DCL, but was wondering if there were a simple solution.   Mark-S.    ------------------------------  % Date: Wed, 28 Feb 2001 12:42:05 -0500 - From: "Peter Weaver" <peter.weaver@stelco.ca>  Subject: Re: DCL content (long) 4 Message-ID: <kMan6.135224$Z2.1822333@nnrp1.uunet.ca>  2 <paddy.o'brien@zzz.tg.nsw.gov.au> wrote in message/ news:01K0M88KWNCY00ACL5@tgmail.tg.nsw.gov.au... E > Folks, there has been so little VMS content here lately (in the old  days bits ofF > code), that I thought I might throw some in.  Meritorious or not ... what theC > hell.  (And I am as bad as many for jumping in on u$shaft content  and saying "me > too".) > 0 > Just trying to get c.o.v. back to what it was. > ...   C I thought this group was for M$ bashing, C vs Ada, US Elections and  Sun Marketing. :)   D Thanks for trying to get us back to VMS! I don't have the time to goA through all of these to see what they are doing, but maybe I will  sometime soon.   ------------------------------  % Date: Wed, 28 Feb 2001 09:17:11 -0600 * From: WILLIAM WEBB <WWEBB1@email.usps.gov>' Subject: RE: DEC Alpha's - FREE T-Shirt - Message-ID: <0033000017375136000002L062*@MHS>   5 =0AIf you don't shop web pages for StorageWorks parts + then why are you complaining so vehemently?   1 Your complaints are taking up more bandwidth than   a year's worth of david's posts.   WWWebb   > -----Original Message-----1 > From: Info-VAX-Request@Mvb.Saic.Com at INTERNET - > Sent: Wednesday, February 28, 2001 12:05 PM 8 > To: Webb, William W; Info-VAX@Mvb.Saic.Com at INTERNET) > Subject: RE: DEC Alpha's - FREE T-Shirt  >  > 4 > In article <3A9C445D.60107@wi.rr.com>, Scott Vieth > <svieth@wi.rr.com> wrote:  > H > > I don't shop web pages for Storageworks parts.  We're running a ver= y  > > large enterprise on these : > > systems, not a little DEC 3000 in somebody's basement. > 6 > YOU BETTER apologize to both my DEC 3000s RIGHT NOW! >  >  :-) > ' > And besides, I don't have a basement.  > 0 > The snobbery in this NG is getting intense.... >  >  :-) >  > -- > Robert Deininger > rdeininger@mindspring.com  >=   ------------------------------  % Date: Wed, 28 Feb 2001 02:08:01 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca> 0 Subject: Re: Drop C/C++ better sooner than later, Message-ID: <3A9CA3C9.2DD7BF75@videotron.ca>   Rick Cadruvi wrote: M > it because I don't use it.  When I do need to I always type cast or declare  > thingsJ > as (char *).  Anyone who faults C based on the way pointer arithmetic is > doneH > and the excessive type casting it causes will get NO argument from me.  I Whenever I am not sure of what it actually does, I end up writing a small.N "test.c" program that does the actual operation and I then either use debug orJ lots of printfs to verify that the compiler generate the right arithmetic.  A This is the same with bit fields/arithmetic or other constructs. T    L What I don't like about DEC-C is that it sometimes forces me to make strangeM constructs to shut the compiler up even though the "illegal" constructs wouldr be much easier to understand.t  N I despise having to typecast. There are times where I will move a pointer to aN void and then to a char just to make sure I can do real arithmetic without the compiler complaining.a    M When working with dynamically allocated memory structures, pointer arithmetica and handling is a must.e   ------------------------------  % Date: Wed, 28 Feb 2001 02:16:09 -0500C- From: JF Mezei <jfmezei.spamnot@videotron.ca>t0 Subject: Re: Drop C/C++ better sooner than later, Message-ID: <3A9CA5B0.3614CAE7@videotron.ca>   re: switch / case:  M I have had occasions where I did make use of the lack of automatic "break" atp the end of a case statement.   Think of a dishwasher:     switch(wash_type)  {p 	case heavy_duty:r 					fill_tank();r/ 					wash(5 minutes, no_spoap, no_rinse_agent);+ 					empty_tank(); 	case very_dirty:C 					fill_tank();a/ 					wash(5 minutes, no_spoap, no_rinse_agent);  					empty_tank(); 	case regular: 					fill_tank();C- 					wash(10 minutes, spoap, no_rinse_agent);t 					empty_tank(); 					  	case rinse:		 					fill_tank(); - 					wash(10 minutes, no_spoap, rinse_agent);  					empty_tank(); 					dryer(15 minutes);p 					break;    	case rinse_only:e 					fill_tank(); . 					wash(5 minutes, no soap, no rinse agent); 					empty_tank(); 					break;b   	case dish_warmer: 					dryer(5 minutes); 					break;c      N So, in the case of a heavy_duty wash, it will fall down all the way to the end of the rinse and then stop.i   ------------------------------    Date: 28 Feb 2001 09:46:30 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>u0 Subject: Re: Drop C/C++ better sooner than laterH Message-ID: <y4ofvnyz2x.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  ; Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) writes:   K > About the only aspect of VMS device drivers I can think of that could notmK > theoretically be done in a safe language would be a driver for a physicalyL > device on a VAX, to that @!#@$#!%$$#$% wait-for-interrupt-and-keep-channel
 > linkage   N Yes, that was (is) an odd one, eh? On the level of the internal implementationN of the MMG services with a common frame work and service-dependent co-routinesL being call with a fixed register-based interface. I heard one of the reasonsL that was completely re-written for the Alpha port was that no one understood how it really worked 8-).k   	Jan   ------------------------------    Date: 28 Feb 2001 09:52:01 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>i0 Subject: Re: Drop C/C++ better sooner than laterH Message-ID: <y4lmqryytq.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  / JF Mezei <jfmezei.spamnot@videotron.ca> writes:b   > I despise having to typecast./  N It's the only way to tell the compiler "do what I mean", instead of leaving it
 guessing.   O > When working with dynamically allocated memory structures, pointer arithmeticC > and handling is a must.S   Really? SHow us an example..   	Jan   ------------------------------   Date: 28 Feb 2001 12:39:58 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)e0 Subject: Re: Drop C/C++ better sooner than later+ Message-ID: <97iriu$d2d$1@info.cs.uofs.edu>o  : In article <ZLUm6.1046$CP4.192534@paloalto-snr1.gtei.net>,)  "Rick Cadruvi" <rick@rdperf.com> writes:oM |> And I thought ADA was dead.  Silly me.  Not a language I would choose, butiL |> obviously some people must like it.  I find it telling that DOD no longer |> requires  |> it for contracts.  B Just as another data point.  We used Ada for our early programmingD courses until last semester.  It is now being phased out in favor ofA Java (and you think C/C++ is bad!!)  The last course using Ada isaD being taught this semester and starting this summer I will no longer) be installing Ada on lab machines at all.I   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   ,   ------------------------------    Date: 28 Feb 2001 08:33:47 -0500- From: koehler@encompasserve.org (Bob Koehler)k0 Subject: Re: Drop C/C++ better sooner than later3 Message-ID: <HpNtBhh6YB5V@eisner.encompasserve.org>e  g In article <3A9C5DD9.E437E98A@earthlink.net>, "David J. Dachtera" <djesys.nospam@earthlink.net> writes:t   > I've only I > recently discovered how to use change-mode EDT "nokeypad" commands in ao > /COMMAND procedure.p  D Shame on you, obviously you haven't spent the required time editting files on an LA36 console.  8)e  F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences Corporationu= NASA GSFC Flight Software       | Federal Sector, Civil GroupsE                                 | please remove ".aspm" when replying    ------------------------------    Date: 28 Feb 2001 08:39:26 -0500- From: koehler@encompasserve.org (Bob Koehler) 0 Subject: Re: Drop C/C++ better sooner than later3 Message-ID: <9HJMlkWyghTe@eisner.encompasserve.org>   \ In article <3A9CA3C9.2DD7BF75@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: > O > When working with dynamically allocated memory structures, pointer arithmetic  > and handling is a must.   A Really?  Then why is it I've got tons of Java with massive use of H dynamically allocated memory structures working quite well in a language which does not have pointers?   > As the Java white paper discusses, use of pointers is directlyC coincident with increased error rates.  There's no reason structure G element and array element syntax can't be used to handle dynamic memory- access.-  F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences CorporationR= NASA GSFC Flight Software       | Federal Sector, Civil Group E                                 | please remove ".aspm" when replyingd   ------------------------------  % Date: Wed, 28 Feb 2001 13:48:58 -0000t' From: "LJEB" <LJEB@somewhere.out.there>f0 Subject: Re: Drop C/C++ better sooner than later) Message-ID: <97ivjt$96k$1@soap.pipex.net>e  A Just sticking my 2c in on a few things here late in the thread...    >oE > Believe me, I am NOT a defender of C.  There are many things I HATEt > about C.  Here's 4 of them:o >mG >         1. Case Sensitivity.  I am use to it, but I think it is a BIGiI >            problem.  tmp, TMP, Tmp, tMp, tmP, TMp, tMP should NOT refer-! >            to different things.7  F In a programming language I don't find this problem, then again having simple1 coding standards which prohibit doing this helps.B  J On the other hand having a file system that does this (and worse) is hell.   >tJ >         2. SWITCH/CASE doesn't auto-break.  I NEVER depend on being ableI >            to fall through to the next CASE in a SWITCH statement.  ThejI >            default should be BREAK and allow specifying CONTINUE if youl< >            want to drop through.  THIS IS A BIG PROBLEM!!!   I would have to say NO!P  I I use this feature from time to time, and when needed it can very useful,cH and saves unnecessary tests and duplicated code. Apart from that I thinkH that I cannot remember when a missing "break" caused me any problems the last time, it was so long ago.  = Also "continue" has a completely defined and simple function.a   > D >         3. BREAK.  Break should be allowed from ALL Begin/End ({})H >            constructs (including IF ELSE).  You should also be able toG >            BREAK # where # is the number of Begin/End blocks you want  >            to break out of.i  L Sorry but this is about the most horrible suggestion for any language I have
 ever seen.  I Next you'll suggest that a computed goto's like in Fortran would be nice.eI This would be event worse than using goto in you code (goto is completelyaK band by our C/C++ coding standards), this would make maintaining code hell,oJ every little modification to a module a hazard, causing untold problems if+ just one break# was missed and not updated.e  K Simply speaking if your code is that complex try splitting it into separate   functions, and check your logic.   >lG >         4. Use of "==" for compares.  It's WAY too easy to put in "="aH >            instead.  In fact, I usually do a specialized search of ALLJ >            my code before the first compile just to try and catch these. >a  K I used to make that mistake when I first started in C (about 15 years ago),yL but I soon got over it. Yes this occationally stoll happens, but usually theJ effect is quite obviouse within the first hour or so of testing. Also manyK modern compilers do warn of this at compilation time now, so the problem is  not like it was.  K I have been writing in C and more recently some C++ for about 15 years, andnL I will admit that I have seen some awful code written in both languages. TheL problem with bad code, which is difficult to follow, overly complicated, notL commented, and just down right sloppy, has almost nothing to do the language in my view.K  J I have also seen terrible code written in Fortran, Basic, Java and others,G the problem is poorly skilled programmers or programmers who just don't.E care. You can always say that in one language something could be donelJ simpler (usually simpler for the developer not the machine), but that will always be the case.   K Programming languages are like any other tool you use, be it a paint brush,bI screwdriver, saw, or club hammer. In the hands of someone who understandssI how to use them they can create almost anything, in the hands of a novicen they usually make a mess.c       Laurence Blunt.    ------------------------------   Date: 28 Feb 2001 13:56:48 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)t0 Subject: Re: Drop C/C++ better sooner than later+ Message-ID: <97j030$f4c$1@info.cs.uofs.edu>C  0 In article <009F8479.B6EC89DC@SendSpamHere.ORG>,@  system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes: |> |> mK |> You can shoot yourself in the foot with any language you use but, unlike]J |> C and its bastard relatives, Ada doesn't load the gun and help you aim.  D Sure it does.  It just holds the gun behind it's back while it loadsB it.  I have always said that most of the features of C that peopleC complain about all the time are there becuase they are needed to doaB many serious system level task.  Most languages that expect to seeG use as more than application languages that did not have these featuresmE end out having a vendor put them in as non-standard extensions (such oB as Pascal).  Ada may have standardized them, but it still put themB in.  It just didn't make them visible or intuitively obvious.  TheB textbook we used here had a whole chapter on how to shoot yourself in the foot using Ada.   bill   -- nJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------   Date: 28 Feb 2001 15:03:52 GMT( From: per@nospam.mimer.se (Per Schrder)0 Subject: Re: Drop C/C++ better sooner than later7 Message-ID: <Xns905699EBF1261pescatmimer@192.71.97.199>C  ) "Rick Cadruvi" <rick@rdperf.com> wrote inw0 <cCYm6.1296$CP4.299029@paloalto-snr1.gtei.net>: I >        2. SWITCH/CASE doesn't auto-break.  I NEVER depend on being able-H >           to fall through to the next CASE in a SWITCH statement.  TheH >           default should be BREAK and allow specifying CONTINUE if you; >           want to drop through.  THIS IS A BIG PROBLEM!!!u >u   NO!wI This is a feature! Use it to exploit the statement skip feature in C. An t6 example of perfectly valid C code can illustrate this:   #include <stdio.h>   void main()t {g     int n=2;
     char *p; s     switch (n) { a       case 1:          p = "one"; ,         if (0)  
       case 2:e         p = "two"; e         if (0) t       case 3:          p = "three";           printf("%s", p); a         break;       }a }e       ;-)i   /Per   ------------------------------   Date: 28 Feb 2001 14:22:50 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)M0 Subject: Re: Drop C/C++ better sooner than later+ Message-ID: <97j1jq$g00$1@info.cs.uofs.edu>   : In article <cCYm6.1296$CP4.299029@paloalto-snr1.gtei.net>,)  "Rick Cadruvi" <rick@rdperf.com> writes:o |> s |> tF |> Believe me, I am NOT a defender of C.  There are many things I HATE |> about C.  Here's 4 of them: |> 0H |>         1. Case Sensitivity.  I am use to it, but I think it is a BIGJ |>            problem.  tmp, TMP, Tmp, tMp, tmP, TMp, tMP should NOT refer" |>            to different things.  C Of course they should.  because they are all different things. TheyoD are all words spelled differently.  If you don't want variable namesE like that use programming standards.  Don't hamstring the language tom solve a social problem.c   |> 2K |>         2. SWITCH/CASE doesn't auto-break.  I NEVER depend on being able J |>            to fall through to the next CASE in a SWITCH statement.  TheJ |>            default should be BREAK and allow specifying CONTINUE if you= |>            want to drop through.  THIS IS A BIG PROBLEM!!!<  C Even as far back as Fortran I learned as a programmer that implicit B anything was a bad idea.  Never assume, always code explicitly.  IB think the same applies here.  Maybe the language would be safer ifD the behavior changed, but the behavior of the language is documented1 and the programmer should take that into account.    |> hE |>         3. BREAK.  Break should be allowed from ALL Begin/End ({})aI |>            constructs (including IF ELSE).  You should also be able to H |>            BREAK # where # is the number of Begin/End blocks you want |>            to break out of.  F Sounds like you want GOTO back.  I thought Dyjkstra convinced everyone that was a bad idea??  ;-)   |> tH |>         4. Use of "==" for compares.  It's WAY too easy to put in "="I |>            instead.  In fact, I usually do a specialized search of ALL0K |>            my code before the first compile just to try and catch these.A  C Well, you need two symbols for "equals", comparison and assignment. B That is what the "=" means.   Anything else would be non-intuitve.B Question:  What does ":" have to do with equality??  And yes, whenD we used to use Pascal as our beginning language here one of the mostC common coding errors was leaving out those pesky ":".  Using doubleI@ "=" for comparison is very intuitve when you consider that everyB other comparison that involves equality consists of a digraph with( at least one "=". (ie. ">=", "<=", "!=")   |> AK |> My ideal language would probably be mostly a combination of BLISS and C.   F There is no such thing as "the ideal language".  Ada was an attempt toL create (by committee no less!) a language that could be used for everything.H The result is a bloated monster that requires so much in  resources thatI it was never practical for small machines (and even some large ones like iH PDP-11's).  Most languages were created to fill a discovered need at theJ time of their creation.  A big part of Systems Analysis is identifying theK problem accurately and then picking the right tool for the job.  Everything.= is not a nail so you need more in your toolbox than a hammer.-   |> 9J |> One thing I don't want is a language to get in the way.  A lot of newer |> languages areL |> designed to prevent programmers for doing things to keep them from making |> mistakes.M |> I understand this logic, but what you have is an effort to over compensatea
 |> becauseI |> programmers were never taught to program well and they write BAD code.I |> uF |> In order for a language to be useable, they invariably provide some
 |> "backdoor"sN |> that allows you to do the things you really want to do.  As a result, thoseJ |> "backdoors" get abused.  Along comes someone else who says they have toM |> do something new that removes that "backdoor".  As more code is written inn |> theM |> new language, it becomes clear that new "backdoors" have to be implementedmJ |> because the language is preventing complex things from being done or at |> leastK |> is making them so difficult that they take too long.  This was always mya
 |> problem
 |> with ADA. o  H Now I'm confused.  What you say above is in complete agreement with what5 I have said, as well.  Especially in the case of Ada.   -L |>           Time to implement a project is VERY important.  It's one of the |> BAD: |> things with C/C++ in terms of applications development.  J This I don't agree with.  In most cases, the actual coding is the smallestJ part of the project. (at least well designed projects)  It takes no longerI for an experienced C programmer to code a task in C than it would take an N experienced Ada programmer to do it in Ada.  And believe it or not, regardlessK of the language used, programmers of the same level of experience will make G an equivalent number of mistakes.  It isn't the language that makes the		 mistakes.(   |> sO |> Progrmaming Languages are just part of the tool set programmers use to solve	J |> problems.  What we really do is we solve problems.  The code writing is	 |> merelys@ |> a clerical function through which we express those solutions.   True.e  
 All the best.s   bill   -- fJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  % Date: Wed, 28 Feb 2001 09:33:19 -0500F- From: John Reagan <reagan@hiyall.zko.dec.com>j0 Subject: Re: Drop C/C++ better sooner than later2 Message-ID: <3A9CC5DF.4A94AD04@hiyall.zko.dec.com>   Bill Gunshannon wrote: > < > In article <cCYm6.1296$CP4.299029@paloalto-snr1.gtei.net>,+ >  "Rick Cadruvi" <rick@rdperf.com> writes:  > |>G > |>         3. BREAK.  Break should be allowed from ALL Begin/End ({}) K > |>            constructs (including IF ELSE).  You should also be able to J > |>            BREAK # where # is the number of Begin/End blocks you want  > |>            to break out of. > H > Sounds like you want GOTO back.  I thought Dyjkstra convinced everyone > that was a bad idea??  ;-)  H Not quite.  GOTOs can go backwards as well as forwards making a mess outA of the program's flow.  BREAKs can only go forward.  Dyjkstra wasr% commenting on certain abuses of GOTO.o  C However, even with just going forward, I don't like BREAK since the-G reader of the code doesn't know where the BREAK will go.  Sure, you can.C look and 'guess', but only the compiler knows for sure.  And addingsB another level of BEGIN/END makes you go and visit all the enclosed# BREAKs and increment their "level".f  E Thats why I actually like forward GOTOs.  You get to name and SEE the  label in the source code.   F BLISS has a LEAVE statement and named-block.  You give a name to blockH and say LEAVE with that name.  Its always a forward branch and you don't1 have to worry about other BEGIN/ENDs being added.e -- e John Reagan  Compaq Pascal Project Leader   ------------------------------  % Date: Wed, 28 Feb 2001 09:58:09 -0500u2 From: rdeininger@mindspring.com (Robert Deininger)0 Subject: Re: Drop C/C++ better sooner than laterL Message-ID: <rdeininger-2802010958090001@user-2ive673.dialup.mindspring.com>  G In article <97j1jq$g00$1@info.cs.uofs.edu>, bill@cs.scranton.edu wrote:   < > In article <cCYm6.1296$CP4.299029@paloalto-snr1.gtei.net>,+ >  "Rick Cadruvi" <rick@rdperf.com> writes:o  E > Even as far back as Fortran I learned as a programmer that implicitWC > anything was a bad idea.  Never assume, always code explicitly.  t  & Sounds like a bit of Ada advocacy. :-)     > |> hG > |>         3. BREAK.  Break should be allowed from ALL Begin/End ({})eK > |>            constructs (including IF ELSE).  You should also be able toiJ > |>            BREAK # where # is the number of Begin/End blocks you want  > |>            to break out of. > H > Sounds like you want GOTO back.  I thought Dyjkstra convinced everyone > that was a bad idea??  ;-)  * In no way did Dyjkstra convince everybody.  H There are times when GOTO is just what's needed.  The example I see mostJ often is coding a state machine.  For an eloquent defense of GOTO, look inA the comp.lang.ada archives for Robert Dewer's posts on the topic.r  E > Well, you need two symbols for "equals", comparison and assignment.eD > That is what the "=" means.   Anything else would be non-intuitve.D > Question:  What does ":" have to do with equality??  And yes, whenF > we used to use Pascal as our beginning language here one of the mostE > common coding errors was leaving out those pesky ":".  Using double-B > "=" for comparison is very intuitve when you consider that everyD > other comparison that involves equality consists of a digraph with* > at least one "=". (ie. ">=", "<=", "!=")  H Using ":=" for assignment had quite a lot of precident at the time C wasG developed.  Starting a new convention was unnecessary.  Way too late to  change it now.     > |> BM > |> My ideal language would probably be mostly a combination of BLISS and C.J > H > There is no such thing as "the ideal language".  Ada was an attempt toN > create (by committee no less!) a language that could be used for everything.J > The result is a bloated monster that requires so much in  resources thatK > it was never practical for small machines (and even some large ones like eJ > PDP-11's).  Most languages were created to fill a discovered need at the > time of their creation.   J Though you have supported the use of Ada in a classroom, you're convincingF me you hardly know it yourself.  Ada was IN NO WAY intended to fit all@ problem spaces.  Two of the big design demands were for embeddedH programming, and for a language that scales well to very large programs.  A Your remarks about "bloated monster" and "not practical for smallrG machines" are nonsense.  A number of languages are significantly bigger J than Ada.  And it has been implemented on a number of small systems.  Some, of the first compilers _were_ dogs, however.  > Ada _is_ a lot bigger than C, but it does much more.  From theJ programmer's point of view, you don't have to learn parts you don't need. H There is a fairly concise C-like subset (with somewhat fuzzy boundaries)E for example.  Of course the compiler implementor has to deal with themH whole language.  Ada is also more specific than C -- the implementor has2 much less discretion in how a feature must behave.   > |>  L > |> One thing I don't want is a language to get in the way.  A lot of newer > |> languages areN > |> designed to prevent programmers for doing things to keep them from making > |> mistakes.O > |> I understand this logic, but what you have is an effort to over compensatey > |> becauseK > |> programmers were never taught to program well and they write BAD code..  I I don't think this is quite the right description of Ada.  It is designed G to let the compiler catch a lot of mistakes, so they never get into thelI running program.  This may have a side effect of making programmers a bit % more careful, thus reducing mistakes.   H > |> In order for a language to be useable, they invariably provide some > |> "backdoor" P > |> that allows you to do the things you really want to do.  As a result, thoseL > |> "backdoors" get abused.  Along comes someone else who says they have toO > |> do something new that removes that "backdoor".  As more code is written in  > |> theO > |> new language, it becomes clear that new "backdoors" have to be implementedtL > |> because the language is preventing complex things from being done or at
 > |> leastM > |> is making them so difficult that they take too long.  This was always my  > |> problem > |> with ADA.   > J > Now I'm confused.  What you say above is in complete agreement with what7 > I have said, as well.  Especially in the case of Ada.b  K Well, he can't even spell Ada, so maybe his complaints are second hand? :-)+  G The backdoors in Ada aren't obsure or hard to understand.  There was anmB effort to make use of backdoors (unchecked deallocation, uncheckedG conversion ...) very explicit and obvious to the reader.  And you don't E need such things very often in Ada.  Most programs never see them.  I G think they are most often abused by people trying to write C code usingo8 Ada syntax.  That does lead to horrible code in a hurry.  E It's also not fair to say that more "backdoors" keep getting added tol. Ada.  It has been quite static in this regard.  D Some of the complaints above apply much more to Pascal than to Ada. H Pascal really was too constrained for many purposes, and had some subtleE back doors.  Ada is a descendant of Pascal, but it extends Pascal fora& useability, and closes some backdoors.   >  tN > |>           Time to implement a project is VERY important.  It's one of the > |> BAD< > |> things with C/C++ in terms of applications development. > L > This I don't agree with.  In most cases, the actual coding is the smallest: > part of the project. (at least well designed projects)     Agreed.s   > It takes no longerK > for an experienced C programmer to code a task in C than it would take an P > experienced Ada programmer to do it in Ada.  And believe it or not, regardlessM > of the language used, programmers of the same level of experience will makelI > an equivalent number of mistakes.  It isn't the language that makes ther > mistakes.i  J If you spend time reading the Ada literature (or Eiffel, probably others),I you will find that this has been measured, and refuted.  C really is morerD error prone.  Briefly, you are probably right about speed of initial# coding.  But not about error rates.f  D Don't really want to start a language war. I hope it's not too late.   -- - Robert Deininger rdeininger@mindspring.com8   ------------------------------    Date: 28 Feb 2001 16:11:59 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>o0 Subject: Re: Drop C/C++ better sooner than laterH Message-ID: <y4lmqqlu4g.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  3 bill@triangle.cs.uofs.edu (Bill Gunshannon) writes:h  $ >  Most languages that expect to seeI > use as more than application languages that did not have these featuresuG > end out having a vendor put them in as non-standard extensions (such ID > as Pascal).  Ada may have standardized them, but it still put them@ > in.  It just didn't make them visible or intuitively obvious.   N The important difference is it makes you explicitly say so, instead of strangeN things (like a missing break or an if =) happening by default or accident. AdaL doesn't stop a broken development process and management from wasting over aJ billion Euro on Ariane 501, but the programmers documented, in their code,K what they had been thinking of when they wrote it, and a simple code reviewwI would have uncovered that a number of the assumptions behind them were no 
 longer valid.m  J A similar example: VMS's security defaults are "safe out of the box" - youI have to work to open up and make a VMS system unsafe. By contrast, WNT iseL unsafe out of the box, and in fact so many pieces of software (including theJ OS) assume this remains so that it is a very difficult and reinstallation-I prone activity to try to secure it; even then, some structural unsafenessX" remains (DLL hell, for instance).   F This is just bad engineering. In the case of C, a lot is the result ofG ignoring, if you want a grand word, cognitive psychology or, more down- E to-earth, just common sense. See the Ada "elseif" vs "elsif" example   mentioned elsewhere.   	Jan   ------------------------------    Date: 28 Feb 2001 16:19:18 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> 0 Subject: Re: Drop C/C++ better sooner than laterH Message-ID: <y4itlults9.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  / John Reagan <reagan@hiyall.zko.dec.com> writes:t  J > > Sounds like you want GOTO back.  I thought Dyjkstra convinced everyone > > that was a bad idea??  ;-)  K There is a riposte, about two decades later, to that article in the mode of L the graphics wheel of re-incarnation, discussing abuses of if...else...endifK nested to unlimited depth. Consider that Knuth used GOTO conscientiously in:$ his public software (TeX, Metafont).  H > BLISS has a LEAVE statement and named-block.  You give a name to blockJ > and say LEAVE with that name.  Its always a forward branch and you don't3 > have to worry about other BEGIN/ENDs being added.e  K That's what Fortran95 does, repairing an oversight in F90. You can name allrJ control structures, and if you do, the compiler has to check they are used
 consistently.    	Jan   ------------------------------    Date: 28 Feb 2001 16:29:54 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>.0 Subject: Re: Drop C/C++ better sooner than laterH Message-ID: <y4g0gyltal.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  4 rdeininger@mindspring.com (Robert Deininger) writes:  J > Using ":=" for assignment had quite a lot of precident at the time C wasI > developed.  Starting a new convention was unnecessary.  Way too late to  > change it now.  L In mathematics notation, for one. Also, "=" suggests are symmetry assignmentA doesn't have, and which often throws novices - "how can x=x+7?". t   	Jan   ------------------------------   Date: 28 Feb 2001 16:05:22 GMT& From: "Rick Cadruvi" <rick@rdperf.com>0 Subject: Re: Drop C/C++ better sooner than later0 Message-ID: <97j7k2$se2@dispatch.concentric.net>  ? I don't disagree that the concept is useful.  I just think thate7 BREAK should be automatic and you should use a CONTINUEs6 to drop through since you almost ALWAYS want to BREAK.   Rick...   : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3A9CA5B0.3614CAE7@videotron.ca... > re: switch / case: > L > I have had occasions where I did make use of the lack of automatic "break" at > the end of a case statement. >a > Think of a dishwasher: >s >  > switch(wash_type)h > {s > case heavy_duty: > fill_tank();, > wash(5 minutes, no_spoap, no_rinse_agent); > empty_tank();" > case very_dirty: > fill_tank();, > wash(5 minutes, no_spoap, no_rinse_agent); > empty_tank();  > case regular:c > fill_tank();* > wash(10 minutes, spoap, no_rinse_agent); > empty_tank();s >c
 > case rinse:s > fill_tank();* > wash(10 minutes, no_spoap, rinse_agent); > empty_tank();. > dryer(15 minutes); > break; >d > case rinse_only: > fill_tank();+ > wash(5 minutes, no soap, no rinse agent);  > empty_tank();o > break; >d > case dish_warmer:g > dryer(5 minutes);  > break; >I >= >oL > So, in the case of a heavy_duty wash, it will fall down all the way to the endn > of the rinse and then stop.i   ------------------------------   Date: 28 Feb 2001 16:11:36 GMT& From: "Rick Cadruvi" <rick@rdperf.com>0 Subject: Re: Drop C/C++ better sooner than later0 Message-ID: <97j7vo$qom@dispatch.concentric.net>  C I don't think you can blame pointers for errors that are really duee: to lazy or bad programming.  Just because the whoite paperA says it, doesn't make it true.  It is true that languages without / pointers can deal with dynamic memory, however.t   Rick Cadruvi...a  : "Bob Koehler" <koehler@encompasserve.org> wrote in message- news:9HJMlkWyghTe@eisner.encompasserve.org...t7 > In article <3A9CA3C9.2DD7BF75@videotron.ca>, JF Mezeir& <jfmezei.spamnot@videotron.ca> writes: > >nF > > When working with dynamically allocated memory structures, pointer
 arithmetic > > and handling is a must.o >fC > Really?  Then why is it I've got tons of Java with massive use ofgJ > dynamically allocated memory structures working quite well in a language > which does not have pointers?  >r@ > As the Java white paper discusses, use of pointers is directlyE > coincident with increased error rates.  There's no reason structure I > element and array element syntax can't be used to handle dynamic memoryu	 > access.t >gH > ----------------------------------------------------------------------A > Bob Koehler                     | Computer Sciences Corporation ? > NASA GSFC Flight Software       | Federal Sector, Civil GroupeG >                                 | please remove ".aspm" when replyingm   ------------------------------   Date: 28 Feb 2001 16:28:48 GMT& From: "Rick Cadruvi" <rick@rdperf.com>0 Subject: Re: Drop C/C++ better sooner than later0 Message-ID: <97j900$ppc@dispatch.concentric.net>  L > >         2. SWITCH/CASE doesn't auto-break.  I NEVER depend on being ableK > >            to fall through to the next CASE in a SWITCH statement.  ThetK > >            default should be BREAK and allow specifying CONTINUE if youi> > >            want to drop through.  THIS IS A BIG PROBLEM!!! >n > I would have to say NO!i >iK > I use this feature from time to time, and when needed it can very useful,kJ > and saves unnecessary tests and duplicated code. Apart from that I thinkJ > that I cannot remember when a missing "break" caused me any problems the  > last time, it was so long ago. >d? > Also "continue" has a completely defined and simple function.v >nE BREAK could be defined to only deal with loops too, but it is used in(: SWITCH/CASE.  There is no reason CONTINUE couldn't be used@ as well.  My point is only that the NORMAL use is to BREAK afterC a case rather than dropping down.  The default should always be theu+ normal usage, not the exception in my view.n   > >rF > >         3. BREAK.  Break should be allowed from ALL Begin/End ({})J > >            constructs (including IF ELSE).  You should also be able toI > >            BREAK # where # is the number of Begin/End blocks you want  > >            to break out of.  >dI > Sorry but this is about the most horrible suggestion for any language I0 have > ever seen.  G The thing a language should do is provide the tools to make the code ast simpleK as possible.  Many times you get inside of something and you really want toh get E out of multiple levels.  If the point is that the BREAK # has to be a, constantF number rather than a variable, I agree entirely.  However, setting and testingbI flags to exit levels makes for more code and makes bugs more likely.  TheFH real problem is not when you are writing the code, but when someone elseJ comes along to maintain the code and misses the flags.  Here's an example:       for (i = 0, i < 10; i++)       {o       if (condition[i] == TRUE) 
          {           for (j = 0; j < 5; j++)
             {w=              if  (stuff[j] == FALSE) break 3    // Get me outo
             }o
          }       }s  D I know the example is simplistic, but if expanded into more involvedF code, the value of the contruct becomes more clear.  The truth is thatI I would probably have made the inner loop a function call and the problemnE goes away.  What I am saying is that occasionally, I would like to beDB able to leave more than a single BEGIN/END.  Also, I would like toD be able to leave an IF or ELSE BEGIN/END early sometimes to simplify	 the code.h   >aK > Next you'll suggest that a computed goto's like in Fortran would be nice. K > This would be event worse than using goto in you code (goto is completelynG > band by our C/C++ coding standards), this would make maintaining codee hell,RL > every little modification to a module a hazard, causing untold problems if- > just one break# was missed and not updated.i >rD > Simply speaking if your code is that complex try splitting it into separate" > functions, and check your logic.  = Agreed.  Absolutely right.  No I wouldn't want computed GOTOs1   >0 > >fI > >         4. Use of "==" for compares.  It's WAY too easy to put in "="4J > >            instead.  In fact, I usually do a specialized search of ALLL > >            my code before the first compile just to try and catch these. > >t >gG > I used to make that mistake when I first started in C (about 15 years  ago),fJ > but I soon got over it. Yes this occationally stoll happens, but usually therL > effect is quite obviouse within the first hour or so of testing. Also manyJ > modern compilers do warn of this at compilation time now, so the problem is > not like it was.L I still slip once in awhile and I have as many C years under my belt as you.G Maybe my brain just locks up once in awhile.  I have noticed my fingersaJ get dylexsia once in awhile, perhaps my brain does also.  My point is onlyK that it was a BAD choice of operators in my opinion that has the poitentiale of causing bugs. >TI > I have been writing in C and more recently some C++ for about 15 years,c and J > I will admit that I have seen some awful code written in both languages. The J > problem with bad code, which is difficult to follow, overly complicated, notsE > commented, and just down right sloppy, has almost nothing to do the  language
 > in my view.  >eE Absolutely correct.  This is actually the starting point of the whole"
 discussion thread.=   ------------------------------   Date: 28 Feb 2001 16:31:57 GMT& From: "Rick Cadruvi" <rick@rdperf.com>0 Subject: Re: Drop C/C++ better sooner than later0 Message-ID: <97j95t$ppc@dispatch.concentric.net>  A This has been my point all along.  The problem in laguages is NOTnB the features that may allow programmers to get into problems.  TheC problems is that those features in a language get abused and codingoF is carelessly undertaken.  Blame programmers for being lazy or writing* bad code rather than blaming the language.   Rick...s  > "Bill Gunshannon" <bill@triangle.cs.uofs.edu> wrote in message% news:97j030$f4c$1@info.cs.uofs.edu... 2 > In article <009F8479.B6EC89DC@SendSpamHere.ORG>,B >  system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes: > |> > |>F > |> You can shoot yourself in the foot with any language you use but, unlikeL > |> C and its bastard relatives, Ada doesn't load the gun and help you aim. > F > Sure it does.  It just holds the gun behind it's back while it loadsD > it.  I have always said that most of the features of C that peopleE > complain about all the time are there becuase they are needed to doaD > many serious system level task.  Most languages that expect to seeI > use as more than application languages that did not have these features F > end out having a vendor put them in as non-standard extensions (suchD > as Pascal).  Ada may have standardized them, but it still put themD > in.  It just didn't make them visible or intuitively obvious.  TheD > textbook we used here had a whole chapter on how to shoot yourself > in the foot using Ada. >d > bill >  > --L > Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesF > bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. > University of Scranton   |@ > Scranton, Pennsylvania   |         #include <std.disclaimer.h>   ------------------------------   Date: 28 Feb 2001 16:42:11 GMT& From: "Rick Cadruvi" <rick@rdperf.com>0 Subject: Re: Drop C/C++ better sooner than later0 Message-ID: <97j9p3$ppc@dispatch.concentric.net>  E > However, even with just going forward, I don't like BREAK since therI > reader of the code doesn't know where the BREAK will go.  Sure, you canhE > look and 'guess', but only the compiler knows for sure.  And adding D > another level of BEGIN/END makes you go and visit all the enclosed% > BREAKs and increment their "level".   @ True.  Sometimes you create new problems when trying to solve anB existing problem.  However, through careful indentation of my code< BEGIN/ENDs I can make it obvious where the BREAK ends up at.H Obviously a language construct that eliminates this is better.  My pointF is just that I could write simpler and therefore better code with this notion.    >oG > Thats why I actually like forward GOTOs.  You get to name and SEE thet > label in the source code.m >mH > BLISS has a LEAVE statement and named-block.  You give a name to blockJ > and say LEAVE with that name.  Its always a forward branch and you don't3 > have to worry about other BEGIN/ENDs being added.D  G Exactly.  I rarely use this feature in BLISS, but when I do want it, itw workstD well.  However, you still have to locate the correct BEGIN/END as inE the BREAK # case.  I'm not sure I see how this is entirely different.    Rick...r   ------------------------------   Date: 28 Feb 2001 16:50:36 GMT& From: "Rick Cadruvi" <rick@rdperf.com>0 Subject: Re: Drop C/C++ better sooner than later0 Message-ID: <97ja8s$ql1@dispatch.concentric.net>  I I'm not arguing that you shouldn't be able to fall down to the next CASE.yB I'm arguing that allowing CONTINUE for that makes more sense sinceB most SWITCH/CASE statements don't want to fall through, but intend	 to BREAK.a   rick...   5 "Per Schrder" <per@nospam.mimer.se> wrote in messaget1 news:Xns905699EBF1261pescatmimer@192.71.97.199...a+ > "Rick Cadruvi" <rick@rdperf.com> wrote in 1 > <cCYm6.1296$CP4.299029@paloalto-snr1.gtei.net>:,K > >        2. SWITCH/CASE doesn't auto-break.  I NEVER depend on being ablenJ > >           to fall through to the next CASE in a SWITCH statement.  TheJ > >           default should be BREAK and allow specifying CONTINUE if you= > >           want to drop through.  THIS IS A BIG PROBLEM!!!  > >  >s > NO!tJ > This is a feature! Use it to exploit the statement skip feature in C. An8 > example of perfectly valid C code can illustrate this: >  > #include <stdio.h> >o
 > void main()d > {o >     int n=2; >     char *p; >     switch (n)   >       case 1:  >         p = "one"; >         if (0) >       case 2:d >         p = "two"; >         if (0) >       case 3:e >         p = "three"; >         printf("%s", p); >         break; >     }a > }e >a >a >  > ;-)  >t > /Per   ------------------------------  % Date: Wed, 28 Feb 2001 11:55:26 -0500," From: Dan Sugalski <dan@sidhe.org>0 Subject: Re: Drop C/C++ better sooner than later: Message-ID: <5.0.2.1.0.20010228114851.021b7270@24.8.96.48>  0 At 04:31 PM 2/28/2001 +0000, Rick Cadruvi wrote:B >This has been my point all along.  The problem in laguages is NOTC >the features that may allow programmers to get into problems.  The-D >problems is that those features in a language get abused and codingG >is carelessly undertaken.  Blame programmers for being lazy or writing + >bad code rather than blaming the language.l  J Generally speaking it's better to blame people for choosing inappropriate H languages. C is, in my experience, wildly inappropriate for most of the J uses to which it is put. It's a horrible language for writing pretty much  any user-level application.h  D It's not a bad language for writing low-level things such as device L drivers, but for higher-level things it a bad choice. I wouldn't even write K most pieces of an OS or compiler in it if better languages were available, iJ though often there isn't one. (I'd choose something better if I could for D implementing perl, but there's nothing better given the portability  constraints)  ? >"Bill Gunshannon" <bill@triangle.cs.uofs.edu> wrote in messager& >news:97j030$f4c$1@info.cs.uofs.edu...4 > > In article <009F8479.B6EC89DC@SendSpamHere.ORG>,D > >  system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes: > > |> > > |>H > > |> You can shoot yourself in the foot with any language you use but, >unlikeoN > > |> C and its bastard relatives, Ada doesn't load the gun and help you aim. > >sH > > Sure it does.  It just holds the gun behind it's back while it loadsF > > it.  I have always said that most of the features of C that peopleG > > complain about all the time are there becuase they are needed to dotF > > many serious system level task.  Most languages that expect to seeK > > use as more than application languages that did not have these featureswH > > end out having a vendor put them in as non-standard extensions (suchF > > as Pascal).  Ada may have standardized them, but it still put themF > > in.  It just didn't make them visible or intuitively obvious.  TheF > > textbook we used here had a whole chapter on how to shoot yourself > > in the foot using Ada. > >+ > > bill > >  > > --N > > Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesH > > bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. > > University of Scranton   |B > > Scranton, Pennsylvania   |         #include <std.disclaimer.h>     					Dan  I --------------------------------------"it's like this"-------------------i2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and event;                                       teddy bears get drunka   ------------------------------   Date: 28 Feb 2001 16:55:28 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)r0 Subject: Re: Drop C/C++ better sooner than later+ Message-ID: <97jai0$klq$1@info.cs.uofs.edu>9  L In article <rdeininger-2802010958090001@user-2ive673.dialup.mindspring.com>,5  rdeininger@mindspring.com (Robert Deininger) writes:o |> |> y) |> Sounds like a bit of Ada advocacy. :-)-  0 Sorry, you won't find much of that from me.  :-)B I don't particularly like the language and I've been involved with7 peripherally since before the first validated compiler.g     |> s- |> In no way did Dyjkstra convince everybody.   ? Tongue must have been too firmly planted in cheek on the above.: Dijkstra never convinced me.   |> pK |> There are times when GOTO is just what's needed.  The example I see mosteM |> often is coding a state machine.  For an eloquent defense of GOTO, look in D |> the comp.lang.ada archives for Robert Dewer's posts on the topic.  @ This surprises me.  If Dewer is who I think he is, he was one of? Ada's strongest advocates and GOTO is very much contrary to ther) paradigm put forward by it's developers. /   |> tK |> Using ":=" for assignment had quite a lot of precident at the time C wasoJ |> developed.  Starting a new convention was unnecessary.  Way too late to |> change it now.   C Really??  With C and Pascal being pretty much parallel developments F (both starting around 1971 somewhere) I assume oyu must mean somethingB earlier.  Of course, it could have been worse.  PL/1 used the sameE symbol "=" for both assignment and comparison.  Maybe you mean Algol?m@ But the use there wasn't particularly clear or unambiguous, if I remember correctly.    |> h< |> Though you have supported the use of Ada in a classroom,   E Where did I say I supported it's use inthe classroom??  My preferencetE for a first programming language remains Pascal.  After all, that wastE the intended use of it's creator.  We're back to that "the right tool  for the job" idea again.  M |>                                                          you're convincinga# |> me you hardly know it yourself. m  G While I have not done any million line projects with it, I know it wellsJ enough to program in it when required.  And to help debug others Ada code.  rH |>                                 Ada was IN NO WAY intended to fit allC |> problem spaces.  Two of the big design demands were for embeddednK |> programming, and for a language that scales well to very large programs.1  F Well, I was in data processing in the Military when it was first beingF developed and we were being told that all DOD systems would have to beE eventually, rewritten in Ada.  Luckily, that never happened.  DOD haduE every intention of impossing Ada on everyone under their domain.  How,D else do you think they justified the millions (billions?) of dollars wasted on the project??r   |> vD |> Your remarks about "bloated monster" and "not practical for smallJ |> machines" are nonsense.  A number of languages are significantly biggerM |> than Ada.  And it has been implemented on a number of small systems.  Somew/ |> of the first compilers _were_ dogs, however.l  I I guess your concept of a "small system" differs from mine.  The smallest I system I have ever seen a validated Ada compiler (and remember, there are K no subsets of Ada!!) on was a VAX.  At the time I saw my first Ada compilerhH I was doing a lot of work on PDP-11's, Z80 and M68000 microcomputers andI even a Univac 1100.  I know there was never an Ada compiler for the first ) three and I never saw one for the Univac.l   |> y8 |> Ada _is_ a lot bigger than C, but it does much more.   D Actually, it does exactly the same thing.  It just does it differnt.F One of the most common ada compilers is GNAT which is just a differentC parser front end on a C compiler.  I guess the biggest question is, @ "Does the language provide enough advantage to warrant the extra+ resources it requires?"  Some of us say no.e  8@ |>                                                      From theM |> programmer's point of view, you don't have to learn parts you don't need. h   Which is true of any language.  K |> There is a fairly concise C-like subset (with somewhat fuzzy boundaries)t  = No, according to the proponent agency, there are no subsets.    H |> for example.  Of course the compiler implementor has to deal with theK |> whole language.  Ada is also more specific than C -- the implementor hasa5 |> much less discretion in how a feature must behave.i  F Not true, and this is one of my main complaints.  Look up tasking.  ItG is totally implementor dependant.  Two "parallel" tasks need not run in G parallel and may in fact run asynchronously under some implementations. D Which enters into a whole other argument about wether or not taskingG belongs in a language at all, as opposed to being a function of the OS.s   |>  L |> I don't think this is quite the right description of Ada.  It is designedJ |> to let the compiler catch a lot of mistakes, so they never get into theL |> running program.  This may have a side effect of making programmers a bit( |> more careful, thus reducing mistakes.  J Experience has shown that it has quite the opposite effect. Why desk checkI your code when you can rely on the compiler to do it for you.  The result#J is sloppier code that is syntactically correct but with some rather unique
 logic errors.     M |> > Now I'm confused.  What you say above is in complete agreement with whate: |> > I have said, as well.  Especially in the case of Ada. |>  N |> Well, he can't even spell Ada, so maybe his complaints are second hand? :-) |> aJ |> The backdoors in Ada aren't obsure or hard to understand.  There was anE |> effort to make use of backdoors (unchecked deallocation, uncheckedlJ |> conversion ...) very explicit and obvious to the reader.  And you don'tH |> need such things very often in Ada.  Most programs never see them.  IJ |> think they are most often abused by people trying to write C code using; |> Ada syntax.  That does lead to horrible code in a hurry.  |>  H |> It's also not fair to say that more "backdoors" keep getting added to1 |> Ada.  It has been quite static in this regard.t |> ]  < This was true of C as well until ANSI got their hands on it.  G |> Some of the complaints above apply much more to Pascal than to Ada. nK |> Pascal really was too constrained for many purposes, and had some subtlet |> back doors. y  I Yes, this is a prime example of using the wrong tool for the job.  PascalIG was created to "teach" algorithms and data structures. Not to implementtI them in the real world.  It is, of course, interesting to note that afterLH seeing this terrible misuse of the language, wirth went back and createdH a language specifically to put the ideas he had put forward in Pascal inH a practical and usable form.  And thus was Modula born.  And we all knowJ how popular this language became. (he says with tongue planted very firmly in cheek!) :-)  fG |>             Ada is a descendant of Pascal, but it extends Pascal fork) |> useability, and closes some backdoors.y  F This is one of the biggest things in Ada's favor. (IMHO)  Not only didH it keep a lot of this Pascal flavor, it had input from and was developedE by many people who were firmly planted in earlier Pascal development.m    O |> > This I don't agree with.  In most cases, the actual coding is the smallest = |> > part of the project. (at least well designed projects)  h |> s
 |> Agreed.  G I'm glad of this.  One of the things I constantly lament around here isfF how little pre-coding thought goes into most student work and the fact4 that this is not seen as a problem worth addressing.   |> t |> > It takes no longer N |> > for an experienced C programmer to code a task in C than it would take anS |> > experienced Ada programmer to do it in Ada.  And believe it or not, regardlessaP |> > of the language used, programmers of the same level of experience will makeL |> > an equivalent number of mistakes.  It isn't the language that makes the |> > mistakes. |> aM |> If you spend time reading the Ada literature (or Eiffel, probably others),tL |> you will find that this has been measured, and refuted.  C really is moreG |> error prone.  Briefly, you are probably right about speed of initiale& |> coding.  But not about error rates.  F I would dispute this, but as you state in the next line, I think it isF more of a language war thing.  Trust me on this one thing.  In workingH with students, I see really bad programs written in Pascal, Ada, C, C++. Visual Basic and now even Java.i   |> hG |> Don't really want to start a language war. I hope it's not too late.   F No war intended.  While I do most of my work in C (as I work primarilyE on Unix Machines and C is very much at home there.)  I can do programtF in more than a dozen other languages.  I believe in the right languageC for the job and that 90% of the project preceeds even the choice of J language.  The question is, how do you convince enough people to translate this into practice.l  
 All the best.a   bill   -- fJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  % Date: Wed, 28 Feb 2001 12:45:00 -0500>- From: John Reagan <reagan@hiyall.zko.dec.com>a0 Subject: Re: Drop C/C++ better sooner than later1 Message-ID: <3A9CF2CC.33F0C69@hiyall.zko.dec.com>    Rick Cadruvi wrote:  > J > > BLISS has a LEAVE statement and named-block.  You give a name to blockL > > and say LEAVE with that name.  Its always a forward branch and you don't5 > > have to worry about other BEGIN/ENDs being added.r > I > Exactly.  I rarely use this feature in BLISS, but when I do want it, ita > worksnF > well.  However, you still have to locate the correct BEGIN/END as inG > the BREAK # case.  I'm not sure I see how this is entirely different.   G Yes, you still have to find the right BEGIN/END to label to begin with,i8 but the addition of other code doesn't 'move' the label.   --   John Reagan  Compaq Pascal Project Leader   ------------------------------    Date: 28 Feb 2001 12:48:42 -0500- From: koehler@encompasserve.org (Bob Koehler)v0 Subject: Re: Drop C/C++ better sooner than later3 Message-ID: <UD5lzBTgT+T4@eisner.encompasserve.org>|  _ In article <97j1jq$g00$1@info.cs.uofs.edu>, bill@triangle.cs.uofs.edu (Bill Gunshannon) writes:-  E > Well, you need two symbols for "equals", comparison and assignment.> > That is what the "=" means.   D No, "=" does not mean "equals".  Not in any progamming language I'veG ever used.  In C, C++, Java, and Fortran it means the exression on the  B left gets the value on the right.  In some other langauge it means compare for equality.'  @ If "=" meant "equals", the following statement would be illegal:     a = a + 1m  F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences Corporationo= NASA GSFC Flight Software       | Federal Sector, Civil Group E                                 | please remove ".aspm" when replying.   ------------------------------    Date: 28 Feb 2001 12:59:11 -0500- From: koehler@encompasserve.org (Bob Koehler)e0 Subject: Re: Drop C/C++ better sooner than later3 Message-ID: <f9yZLDIRFnH3@eisner.encompasserve.org>m  Y In article <97j7vo$qom@dispatch.concentric.net>, "Rick Cadruvi" <rick@rdperf.com> writes:cE > I don't think you can blame pointers for errors that are really duey< > to lazy or bad programming.  Just because the whoite paperC > says it, doesn't make it true.  It is true that languages withoutI1 > pointers can deal with dynamic memory, however.v >  > Rick Cadruvi...o >   E Yes I can, because it becomes a cost issue.  Why worry about the timehB and effort needed to find and eliminate those bugs when I have theG option of programming in a language that doens't permit those mistakes?.  H Well written code is well written code, but a compiler is a tool and I'm9 not about to give up on using the right tool for the job.o  F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences Corporatione= NASA GSFC Flight Software       | Federal Sector, Civil Group.E                                 | please remove ".aspm" when replying0   ------------------------------    Date: 28 Feb 2001 13:05:18 -0500- From: koehler@encompasserve.org (Bob Koehler)a0 Subject: Re: Drop C/C++ better sooner than later3 Message-ID: <KForX+5U$4bD@eisner.encompasserve.org>   _ In article <97jai0$klq$1@info.cs.uofs.edu>, bill@triangle.cs.uofs.edu (Bill Gunshannon) writes:dK > I guess your concept of a "small system" differs from mine.  The smallestrK > system I have ever seen a validated Ada compiler (and remember, there arei$ > no subsets of Ada!!) on was a VAX.  C The software which controls the space shuttle is written in Ada.  I A don't think DEC ever shipped a VAX as small as the computers that  shuttle software runs on.  )  C By small, I'm taking RAM size, but I think CPU speed is near 11/780a+ in orders of magnitude.  No disks are used.)  F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences Corporation = NASA GSFC Flight Software       | Federal Sector, Civil Group2E                                 | please remove ".aspm" when replyinge   ------------------------------    Date: 28 Feb 2001 13:11:16 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)r0 Subject: Re: Drop C/C++ better sooner than later3 Message-ID: <vgTkciHxgEnV@eisner.encompasserve.org>p   In article <rdeininger-2802010128330001@user-2iveaji.dialup.mindspring.com>, rdeininger@mindspring.com (Robert Deininger) writes:e  K > And if you need a nested if-else in both the if part and the else part, a  > switch isn't simple anymore.   But something like:m   	if WINE = RED 	thenp 		if MOON = FULL 		then 		    ACTION_1;c 		else 		    ACTION_2;o	 		end if;h 	elsif SIZE > 500t 	then. 		if SEASON = SPRING 		then 		    ACTION_3;c 		else 		    ACTION_4;a	 		end if;t 	elsea
 		if OS = VMSo 		then 		    ACTION_5;s 		else 		    ACTION_6;,	 		end if;g 	end if;   Can be deconstructed into:   	if WINE = RED and MOON = FULL 	then  	    ACTION_1; 	elsif WINE = REDw 	theni 	    ACTION_2;% 	elsif SIZE > 500 and SEASON = SPRINGe 	thene 	    ACTION_3; 	elsif SIZE > 500p 	thent 	    ACTION_4; 	elsif OS = VMS  	    ACTION_5; 	elset 	    ACTION_6; 	end if;  9 Remember that at this level it is the job of the compilerc8 and not the programmer to perform optimization.  The two; examples above are quite likely to produce the same machinek3 code (presuming I made no error in deconstruction).u   ------------------------------    Date: 28 Feb 2001 13:22:11 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)e0 Subject: Re: Drop C/C++ better sooner than later3 Message-ID: <rFwVS93L2YC0@eisner.encompasserve.org>    In article <rdeininger-2802010958090001@user-2ive673.dialup.mindspring.com>, rdeininger@mindspring.com (Robert Deininger) writes:e  I > The backdoors in Ada aren't obsure or hard to understand.  There was anoD > effort to make use of backdoors (unchecked deallocation, uncheckedI > conversion ...) very explicit and obvious to the reader.  And you don'tsG > need such things very often in Ada.  Most programs never see them.  I I > think they are most often abused by people trying to write C code using : > Ada syntax.  That does lead to horrible code in a hurry.  C I have found significant uses for UNCHECKED_CONVERSION and also for B UNCHECKED_DEALLOCATION, but I try to follow the "spirit" of Ada in, my instantiations with name choices such as:  * 	function UNCHECKED_LONGWORD_TO_POINTER is! 	    new UNCHECKED_CONVERSION ...c  D so at each _use_ I am reminded that I am using UNCHECKED_CONVERSION.   ------------------------------   Date: 28 Feb 2001 18:26:05 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon) 0 Subject: Re: Drop C/C++ better sooner than later+ Message-ID: <97jfrt$l7u$2@info.cs.uofs.edu>   3 In article <UD5lzBTgT+T4@eisner.encompasserve.org>,-0  koehler@encompasserve.org (Bob Koehler) writes:b |> In article <97j1jq$g00$1@info.cs.uofs.edu>, bill@triangle.cs.uofs.edu (Bill Gunshannon) writes: |> tH |> > Well, you need two symbols for "equals", comparison and assignment.! |> > That is what the "=" means. u |> dG |> No, "=" does not mean "equals".  Not in any progamming language I've  |> ever used.  i  I In the statement above I put it in quotes to try to force the implication H that I was using it n the english and not the technical sense.  When oneH is reading a statement in english and encounters the "=" symbol they say/ equals regardless of which meaning is intended.e  J |>             In C, C++, Java, and Fortran it means the exression on the E |> left gets the value on the right.  In some other langauge it means  |> compare for equality.  8 And in PL/1 it meant both, just like it does in english. Example:         IF START = NULLg         THEN)              START = CURRENT_PRODUCT_PTR;hD (code fragment from "PL/I Structured Programming" by Joan K. Hughes)   |> sC |> If "=" meant "equals", the following statement would be illegal:a |>   |>   a = a + 1  ? If I were reading that statement to you in english I would say,nB " a equals a plus one ".  Funny thing that english.  Many teachersA have this penchant for using the term  "gets" for that usage, butC? that is jargon and is pretty much meaningless outside of a veryi6 small inner circle.  People just don't talk that way.    bill   -- fJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   e   ------------------------------    Date: 28 Feb 2001 13:34:55 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) 0 Subject: Re: Drop C/C++ better sooner than later3 Message-ID: <ZRC$0VIY5C9A@eisner.encompasserve.org>c  _ In article <97jai0$klq$1@info.cs.uofs.edu>, bill@triangle.cs.uofs.edu (Bill Gunshannon) writes: N > In article <rdeininger-2802010958090001@user-2ive673.dialup.mindspring.com>,7 >  rdeininger@mindspring.com (Robert Deininger) writes:o  M > |> There are times when GOTO is just what's needed.  The example I see mostcO > |> often is coding a state machine.  For an eloquent defense of GOTO, look inlF > |> the comp.lang.ada archives for Robert Dewer's posts on the topic. > B > This surprises me.  If Dewer is who I think he is, he was one ofA > Ada's strongest advocates and GOTO is very much contrary to thef+ > paradigm put forward by it's developers. t  E Robert Dewar seems to take particular delight in showing non-kneejerkiE reactions.  As a compiler vendor he hears lots of requirements, above E and beyond his personal experience.  Indeed state machines are one ofe3 his most frequent examples for "never say 'never'".c  K > I guess your concept of a "small system" differs from mine.  The smallestaK > system I have ever seen a validated Ada compiler (and remember, there arerM > no subsets of Ada!!) on was a VAX.  At the time I saw my first Ada compilerwJ > I was doing a lot of work on PDP-11's, Z80 and M68000 microcomputers andK > even a Univac 1100.  I know there was never an Ada compiler for the first + > three and I never saw one for the Univac.   F But those using Ada to _target_ small systems are cross-compiling fromD something larger (like a VAX :-).  But many are also doing so from a Windows machine.  J > |> for example.  Of course the compiler implementor has to deal with theM > |> whole language.  Ada is also more specific than C -- the implementor hasb7 > |> much less discretion in how a feature must behave.a > H > Not true, and this is one of my main complaints.  Look up tasking.  ItI > is totally implementor dependant.  Two "parallel" tasks need not run intI > parallel and may in fact run asynchronously under some implementations.rF > Which enters into a whole other argument about wether or not taskingI > belongs in a language at all, as opposed to being a function of the OS.   G Tasks must conform to synchronization rules of the language, and as youtH say they are free from there on out.  But tasking is _critical_ for code1 that is to be portable between operating systems.p  N > |> I don't think this is quite the right description of Ada.  It is designedL > |> to let the compiler catch a lot of mistakes, so they never get into theN > |> running program.  This may have a side effect of making programmers a bit* > |> more careful, thus reducing mistakes. > L > Experience has shown that it has quite the opposite effect. Why desk checkK > your code when you can rely on the compiler to do it for you.  The result L > is sloppier code that is syntactically correct but with some rather unique > logic errors.i  J That has not been my experience.  I find I am more free to deal with logic3 when the compiler is detecting the spelling errors.r   > |> > It takes no longerdP > |> > for an experienced C programmer to code a task in C than it would take anU > |> > experienced Ada programmer to do it in Ada.  And believe it or not, regardlessuR > |> > of the language used, programmers of the same level of experience will makeN > |> > an equivalent number of mistakes.  It isn't the language that makes the > |> > mistakes.  J I don't think the issue is the number of mistakes made.  I think it is theI number of mistakes made that are not caught at compile-time.  That is the J second-cheapest place to catch errors (the cheapest being design-time, but3 of course the truly incompetent skip that step :-).e  N ==============================================================================N Great Inventors of our time: Al Gore -> Internet; Sun Microsystems -> ClustersN ==============================================================================   ------------------------------   Date: 28 Feb 2001 18:31:25 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)o0 Subject: Re: Drop C/C++ better sooner than later+ Message-ID: <97jg5t$l7u$3@info.cs.uofs.edu>b  3 In article <f9yZLDIRFnH3@eisner.encompasserve.org>,y0  koehler@encompasserve.org (Bob Koehler) writes:\ |> In article <97j7vo$qom@dispatch.concentric.net>, "Rick Cadruvi" <rick@rdperf.com> writes:H |> > I don't think you can blame pointers for errors that are really due? |> > to lazy or bad programming.  Just because the whoite paperrF |> > says it, doesn't make it true.  It is true that languages without4 |> > pointers can deal with dynamic memory, however. |> I/ |> Yes I can, because it becomes a cost issue. c  E There's never time to do it right, there's always time to do it over.r  iG |>                                             Why worry about the timehE |> and effort needed to find and eliminate those bugs when I have the J |> option of programming in a language that doens't permit those mistakes?  E No language eliminates the possibility of writting buggy code.  It isoF the analysis steps that precede the coding that do that.  Language not
 withstanding.c   |>  K |> Well written code is well written code, but a compiler is a tool and I'm < |> not about to give up on using the right tool for the job.  F This is true.  But the crux of this discussion has been that it is theD fault of the C/C++ languages that lead to bad code.  I can write badD code in any language.  There may be an Obfuscated C Contest, but theG same could be done in most languages.  C programmers just have a betterj7 sense of humor and don't mind poking fun at themselves.h   bill9 (Who has never found the time to actually enter the OCC.)    --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------    Date: 28 Feb 2001 13:37:32 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)  Subject: Re: forum.compaq.com 3 Message-ID: <xFEXGh$E5J0C@eisner.encompasserve.org>>  l In article <200102272238_MC2-C714-C4F7@compuserve.com>, "Richard B. Gilbert" <DRAGON@compuserve.com> writes:  L > If a VMS forum is moderated to keep it "on topic", I'd be in favor of it.= >  = > L > The bandwidth that has been wasted here over the last six or eight years = > onL > topics such as "Is VMS dying", "VMS is dead", "No it's not", "the future = > of< > VMS", etc, ad nauseum, does not need an additional outlet!  G Try DECUServe.   Telnet to Eisner.DECUS.org and log in as REGISTRATION.+   ------------------------------   Date: 28 Feb 2001 17:27:21 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)   Subject: Free to a good home :-)+ Message-ID: <97jcdp$l7u$1@info.cs.uofs.edu>i  F I have an SMD disk free to anyone who can use it.  Only requirement isF that you come and pick it up (not literally).  It is a GEAC 8775 whichF appears to be a rebadged CDC SMD disk of some sort.  I do not know theD capacity or anything, but if someone knew where I might look or whatC number might identify it.  It was functional when the GEAC was shut C down, but that was probably two years ago (maybe more).  It has notaG been used since and is currently occupying valuable space on my machinecB room floor.  The down side is it weighs in at sell over 100 pounds and is not easy to transport.-  F I saved it from it's first trip towards the dumpster in hopes of usingF it on a PDP-11.  That seems unlikely because of the age of my Qbus SMD( controllers, but it might work on a VAX.  E Sadly, if I can find no takers, I will have to dismantle it and throwm- it piece by piece into the dumpster out back.    It is located in Scranton, PA.   bill   -- oJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------    Date: 28 Feb 2001 09:19:48 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>i/ Subject: Re: Future outlook for OpenVMS admins?'H Message-ID: <y4wvabz0bf.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  3 "Dave Gudewicz" <david.gudewicz@abbott.com> writes::  4 > I think Tru64, like VMS is Compaq's own product.    N Coming as it does from OSF/1 (which was its (close to) original name), I thinkM the OSF is still collecting a license fee for every copy sold. Motif as well,n
 I presume.   	Jan   ------------------------------  % Date: Wed, 28 Feb 2001 09:30:14 -0500i! From: Dan Allen <dallen@nist.gov>t0 Subject: FW: Drop C/C++ better sooner than later: Message-ID: <NEBBIALHDHJMJINPGMOAMEHNDEAA.dallen@nist.gov>   -----Original Message-----+ From: Rick Cadruvi [mailto:rick@rdperf.com]n( Sent: Tuesday, February 27, 2001 8:36 PM To: Info-VAX@Mvb.Saic.Com 0 Subject: Re: Drop C/C++ better sooner than later   <snip>  D >Believe me, I am NOT a defender of C.  There are many things I HATE >about C.  Here's 4 of them:  
 Me either ;-)   ; >3. BREAK.  Break should be allowed from ALL Begin/End ({})fG >           constructs (including IF ELSE).  You should also be able to F >           BREAK # where # is the number of Begin/End blocks you want >           to break out of.  T The second part of that suggestion is a Bad idea and exemplifies the philosophy thatT typically inculcates the modern day C/C++ programer's thinking. The language alreadyN has accomodated this need. It's the much and unjustly maligned goto statement.  V How many instances of convoluted and incomprehensible nested blocks filled with armiesT of boolean flags have you encountered over the last ten years. And how many of theseY instances were designed to accomplish, using goto-less block structured logic, what couldk\ have been coded much more concisely with a single statement label and a few very appropriate goto statements?  W That's just lunacy and it's the result of brainwashing, not a basic weakness in the PL.e[ Hell, I'll bet +90% of the C/C++ programmers under the age of 35 don't even know the syntaxo for coding a goto statement.  [ Akin to what Brian said previously about the PL helping you shoot yourself in the foot, our- basic programmer training with it's extreme emphasis on avoiding certain specific bad programming practices has helped as well.e> It's given us programmer's who mis-use the favored features of\ their language of choice rather than use a "tainted" construct much more aptly suited to the
 need at hand.e   > just my opinions,b     >Rick Cadruvi...  	 Mine too,   	 Dan Allen    ------------------------------  % Date: Wed, 28 Feb 2001 09:01:55 -0500c& From: Donald G Plugge <plugge@usa.net> Subject: GTK+ on VMS' Message-ID: <3A9D04D1.8481C67B@usa.net>r   Graphical Developmente  H I've gone around and around on this issue for our in-house software.  IfB you ever want to port your VMS graphical software packages over toE another platform (Mac, PC, Unix), then Motif doesn't appear to be the  best option.  G Recently I download the GTK+ library from the compaq web site and I wastH able to get the package installed and the demos running.  The library is? running on many platforms and appears to have a good following,- especially by the Linux people.-  H Would anyone be interested in opening a channel of discussion in settingC up a working development environment for GTK+? Perhaps via personaleB email.  Or perhaps some of you are already working with a group of VMS/GTK+ developers?  B My current in-house software supports graphics from old tektronixsE terminals to PCX serving on the PC platform using X Windows.  We alsoRG create HPGL and PostScript hardcopies.  These two in-house projects areeC currently in maintenance mode.  I'd like to prototype with GTK+ andvD bounce configuration issues off other VMS developers.  Perhaps otherE would like to discuss issues such as MMS, the logical environment andaC DEC C/C++ as it relates to the GTK+ library.  Or even other graphic 9 packages like wxWindows, PGPLOT, Qt or Perl/Tk or Tcl/Tk.    Donald G Plugge	   ------------------------------  # Date: Wed, 28 Feb 2001 16:04:29 GMTO1 From: "Mark D. Jilson" <jilly@clarityconnect.com> L Subject: Re: How to get Polycenter Performance Advisor Graphs to a web page?2 Message-ID: <3A9D2232.BDD68670@clarityconnect.com>  H And Linux (Intel) very shortly if not already.  To test drive PAWZ start: at http://www.testdrive.compaq.com/software/pawz/pawz.html     Ian Parker wrote:N > G > It may be worthwhile to look at Compaq's PAWZ product.  It has a web-fF > based display and, IIRC, supports VMS data collection as well Tru64, > Solaris, etc.e >  > -- > Ian Parker   -- cD Jilly	- Working from Home in the Chemung River Valley - Lockwood, NY0 	- jilly@clarityconnect.com			- Brett Bodine fan. 	- Mark.Jilson@Compaq.com			- since 1975 or so, 	- http://www.jilly.baka.com               -   ------------------------------  % Date: Wed, 28 Feb 2001 11:58:33 -0500r- From: "Peter Weaver" <peter.weaver@stelco.ca>sY Subject: Re: Installing Availability Manager V1.4 on NT 4.0 breaks ReflectionsV8.0.2 V8.0i4 Message-ID: <y7an6.135202$Z2.1822050@nnrp1.uunet.ca>  < "Steven P. Underwood" <StevenU@POBoxes.com> wrote in message+ news:3a9c7992.17318301@news.telocity.com...  >...E > I have Compaq Analyze running on my ES40 that I downloaded from thelE > Compaq web site (after being pointed to it by support).  I guess itrF > works, but it is WAY to unixy for my liking and where is the /since=	 > switch.s >tB > I used to be able to figure out what analyze/error was trying to tell? > me, now I simply call tech support and have them tell me whatu command D > they want me to run and let them analyze it.  If they want to make it? > harder on me, that is what I have a hardware support contractiF > (extended to 3 years) for.  I have yet to get any useful information> > out of it that can help me determine if the tape errors I am. > experiencing are hard/soft/media/etc errors.  F I have Compaq Analyze running on our test ES40 Galaxy. I refuse to putD it on our production ES40. Compaq has agreed that if we ever need toF analyze our error log on the production ES40 then we can ship them the9 ERRLOG.SYS and they will run Compaq Analyze on it for us..  B I do not know if GOOGLE goes back far enough, but I posted severalE notes here about Compaq Analyze last March. Here is part of one of myhF posts and one of the major reasons I will not put Compaq Analyze on my production machines;    = I have been connected to HTTP://nodename:7902 since some timen; yesterday morning. I am on a page that is titled "localhostdA (OpenVMS)." Every 10 minutes a new timestamp entry appears on theh screen.t  : After a while the following items appeared in the browser;  @ 123.      ds20.errlog reports Problem Found: Memory Channel Link. Transmit Error at Wed Mar 29 13:53:19 EST 2000E 124.      ds20.errlog reports Problem Found: Control Packet Heartbeat - Timeout Error at Wed Mar 29 13:53:19 EST 2000eF 125.      ds20.errlog reports Problem Found: Memory Channel Phase Lock* Loop Error at Wed Mar 29 13:53:19 EST 2000 126.      ds20.errlog.  D Where entry 126 came from I do not know. It mentions "DS20" but I amD on an ES40. Entries 123 through 125 all show me that the nodename isC "sabl28," I do not have any node name "sabl28" anywhere that I knownF of. All three of these entries report "Local Time of Event: Thu, 8 Apr8 1999 15:45:38 -0400" (yes, 1999 that was not a typo) yet: SYS$ERRORLOG:ERRLOG.SYS has a created date of 14-MAR-2000.  . When I click on entry 126 I see the following;  D 1.  Tru64 UNIX CAM Event at Thu Apr 08 15:43:17 EDT 1999 from sabl28F 2.  Tru64 UNIX Asychronous Device Attention at Thu Apr 08 15:45:38 EDT 1999 from sabl28B 3.  Tru64 UNIX Panic ASCII Message at Thu Apr 08 15:45:38 EDT 1999 from sabl28aF 4.  Tru64 UNIX Asychronous Device Attention at Thu Apr 08 15:45:40 EDT 1999 from sabl28  C As far as I am aware, we do not have any Tru64 machines anywhere on-E this network. Where are these errors coming from? - Wait a minute. AsgB I was creating this message I decided to take a look at something. Yesterday morning I renamedr9 SYS$SYSDEVICE:[SYS0.COMPAQ.SVCTOOLS.CA]EXAMPLES.DIR to benB SYS$SYSDEVICE:[SYS0.COMPAQ.SVCTOOLS.CA]EXAMPLES_KEEP.DIR because IB noticed that CA ANALYZE keep trying to read in the examples ratherD than default to the correct errorlog. I have verified that am on the: node that is in the SYS0 tree. I just took a look and saw;  ? Directory SYS$SYSDEVICE:[SYS0.COMPAQ.SVCTOOLS.CA.EXAMPLES_KEEP]i  5 DS20.ERRLOG;1              4   2-FEB-2000 15:29:50.00k   Total of 1 file, 4 blocks.  C Dumping that file shows me the text "sabl28." So I must assume that.F somehow CA decided to read in an example errorlog and show it to me asF real data! PLEASE TELL ME THAT SOMEHOW I HAVE A PRE-BETA COPY OF CA ONA THIS SYSTEM AND THAT THIS WILL **NEVER** HAPPEN ONCE THE SHIPPINGn PRODUCT IS INSTALLED!    ------------------------------  % Date: Wed, 28 Feb 2001 13:17:08 +0100 1 From: "Tomasz Dryjanski" <tdryjanski@hotmail.com>n" Subject: Invalid mail notification. Message-ID: <97iq89$30pv$1@news2.ipartners.pl>   Hi,eJ I have the following problem: once I had 37 unread mail messages. I made aG full system disk backup for some purposes and then read and deleted allo	 messages. 5 Then I had to restore the system disk and now I have:  PROMIS::$ mail   You have 37 new messages.d  
 MAIL> read  , %MAIL-E-NOTEXIST, folder MAIL does not exist   MAIL>e   Besides mail works OK. What should I do to fix it?a  	 Regards -  T. D.    ------------------------------  % Date: Wed, 28 Feb 2001 13:23:09 -0000 * From: "Richard Brodie" <R.Brodie@rl.ac.uk>& Subject: Re: Invalid mail notification+ Message-ID: <97iu3v$phs@newton.cc.rl.ac.uk>e  < "Tomasz Dryjanski" <tdryjanski@hotmail.com> wrote in message( news:97iq89$30pv$1@news2.ipartners.pl...  C > I have the following problem: once I had 37 unread mail messages.  > What should I do to fix it?m  @ FAQ: http://www.openvms.compaq.com/wizard/openvms_faq.html#MAIL7   ------------------------------    Date: 28 Feb 2001 08:45:42 -0500- From: koehler@encompasserve.org (Bob Koehler)t& Subject: Re: Invalid mail notification3 Message-ID: <eyi5awv6wf46@eisner.encompasserve.org>a  b In article <97iq89$30pv$1@news2.ipartners.pl>, "Tomasz Dryjanski" <tdryjanski@hotmail.com> writes: > You have 37 new messages.  >  > MAIL> read > . > %MAIL-E-NOTEXIST, folder MAIL does not exist >    From the FAQ:  read/newe  F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences Corporationt= NASA GSFC Flight Software       | Federal Sector, Civil GrouppE                                 | please remove ".aspm" when replyingm   ------------------------------   Date: 28 Feb 2001 14:20:51 GMT3 From: gartmann@immunbio.mpg.de (Christoph Gartmann) & Subject: Re: Invalid mail notification0 Message-ID: <97j1g3$rt5$1@n.ruf.uni-freiburg.de>  b In article <97iq89$30pv$1@news2.ipartners.pl>, "Tomasz Dryjanski" <tdryjanski@hotmail.com> writes:K >I have the following problem: once I had 37 unread mail messages. I made asH >full system disk backup for some purposes and then read and deleted all
 >messages.6 >Then I had to restore the system disk and now I have: >PROMIS::$ maile >d >You have 37 new messages. >. >MAIL> read  >o- >%MAIL-E-NOTEXIST, folder MAIL does not exist, >o >MAIL> >  >Besides mail works OK.n >What should I do to fix it?  # http://eisner.decus.org/vms/faq.html   See "MAIL 7"   Regards,    Christoph Gartmann   H -- --------------------------------------------------------------------+H | Max-Planck-Institut fuer      Phone   : +49-761-5108-464   Fax: -452 |H | Immunbiologie                                                        |H | Postfach 1169                 Internet: gartmann@immunbio.mpg.de     |H | D-79011  Freiburg, FRG                                               |H +--------- http://www.immunbio.mpg.de/home/english/menue.html ---------+   ------------------------------  % Date: Wed, 28 Feb 2001 11:46:39 -0600-% From: "steve smith" <ssmith@lbpc.com>F" Subject: Issue dcl command from NT% Message-ID: <8Qan6.40$DQ2.422@client>u  L Is there any way to have my NT server issue command on a Alphavms system run version 4.2 of ucx   -- Steve Smith3 Manager Technical Services Information Technology Law Bulletin Publishing Co.> (312)644-7807 Ext.137@ (312)644-4355 Fax. ndelise@LBPC.com http://www.lawbulletin.com   ------------------------------    Date: 28 Feb 2001 13:23:39 -0500+ From: young_r@encompasserve.org (Rob Young)b& Subject: Re: Issue dcl command from NT3 Message-ID: <8iWY5MdkzMd2@eisner.encompasserve.org>o  M In article <8Qan6.40$DQ2.422@client>, "steve smith" <ssmith@lbpc.com> writes:aN > Is there any way to have my NT server issue command on a Alphavms system run > version 4.2 of ucx >   D 	Yes.   You can RSH from NT (RSH showed up in later service packs I ? 	believe.  At one time I went and got a third party RSH for an I  	NT box that did an RSH to VMS).   				Robg   ------------------------------  % Date: Wed, 28 Feb 2001 19:35:21 +0100i  From: Paul Sture <paul@sture.ch>& Subject: Re: Issue dcl command from NT+ Message-ID: <VA.000002d5.90980bd6@sture.ch>n  8 In article <8Qan6.40$DQ2.422@client>, Steve smith wrote:' > From: "steve smith" <ssmith@lbpc.com>a > Newsgroups: comp.os.vmse$ > Subject: Issue dcl command from NT' > Date: Wed, 28 Feb 2001 11:46:39 -0600b > N > Is there any way to have my NT server issue command on a Alphavms system run > version 4.2 of ucx >e? Interactively, the following will work with Alpha or VAX TCPIP:o  +  rexec vms_hostname -l username vms_commanda,  rsh   vms_hostname -l username vms_command   I rexec asks for the password on my system, rsh has the connection refused.   O You should look up the NT help for rexec and rsh in StartHelpNT Commands for h a fuller explanation.a > --
 > Steve Smith  > Manager Technical Services > Information Technology > Law Bulletin Publishing Co.  > (312)644-7807 Ext.137d > (312)644-4355 Fax- > ndelise@LBPC.com > http://www.lawbulletin.com >3   ___5
 Paul Sture Switzerlandr   ------------------------------    Date: 28 Feb 2001 09:13:26 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>2 Subject: Re: Janitor fixes 90LH Message-ID: <y43dcz1azd.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  3 "Mark D. Jilson" <jilly@clarityconnect.com> writes:r  F > Working at a financial institution in NYC in one of it's undergroundI > computer rooms.  We could 'see' the subway trains go by on our monitors"I > ;*)  Very interesting visual effects but I never heard of this actuallyd > bothering anything.   G When they first tested a new accelerator at the University of Bonn, the M power cables were layed once around the ring instead of twice in compensatingtF directorions. Then the power to the magnets was ramped up and down, asI accelerators are wont to do, in a five-second rhythm. All displays in theoJ whole building, up to the fourth (US) floor, moved to the side and back in+ concert. Quite amusing for a short while...e   Janb   ------------------------------    Date: 28 Feb 2001 09:10:42 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>i Subject: Re: Janitor fixes 90LH Message-ID: <y466hv1b3x.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  # Shane.F.Smith@Healthnet.com writes:   M > Then there was the PDP 11/73 (I think - it was the one with the switches oniK > the front) at a company I used to work for, that crashed reliably at 5pm,aM > but only if unattended. The company didn't have space for it in the machine-M > room so had put it in the corridor outside. When the system manager watched E > from hiding at 5pm, he saw the secretaries leaving for the day, andd0 > trailing their fingers along the switches.....  > What, the switches weren't disabled? How careless can you get!   	Jan   ------------------------------  % Date: Wed, 28 Feb 2001 12:19:33 -0500i- From: "Peter Weaver" <peter.weaver@stelco.ca>  Subject: Re: Janitor fixes 90L4 Message-ID: <dran6.135219$Z2.1822275@nnrp1.uunet.ca>  C "Jan Vorbrueggen" <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>  wrote in messageB news:y43dcz1azd.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de...5 > "Mark D. Jilson" <jilly@clarityconnect.com> writes:  >i< > > Working at a financial institution in NYC in one of it's undergrounddB > > computer rooms.  We could 'see' the subway trains go by on our monitorsB > > ;*)  Very interesting visual effects but I never heard of this actually > > bothering anything.  >eE > When they first tested a new accelerator at the University of Bonn,l thetB > power cables were layed once around the ring instead of twice in compensatingE > directorions. Then the power to the magnets was ramped up and down,: asD > accelerators are wont to do, in a five-second rhythm. All displays in theD > whole building, up to the fourth (US) floor, moved to the side and back inc- > concert. Quite amusing for a short while...1 >3 > Jano  F In a previous life I worked at a company with an Electric Arc Furnace.B We had a VT100 about 10 metres away. When the furnace fired up theC left side of the screen would compress down to a single pixel. When @ the furnace was finished then the screen would return to normal.@ VT220's or above did not last through one shift out there so theE technicians always had to make sure we had at least one working VT100 D in the spares. When the new management came in they said they wantedF to get rid of the "legacy" terminals and give everyone a PC. SomewhereF out there is a few tonnes of steel with a Pentium 133 somewhere in it.. And there is still a VT100 by that furnace. ;)   ------------------------------  % Date: Wed, 28 Feb 2001 19:54:11 +0100l" From: "Hans Vlems" <hvlems@iae.nl> Subject: Re: Janitor fixes 90L( Message-ID: <97jhcu$e11$1@news.IAEhv.nl>  B The reason I ended up is that I studied analytical chemistry, more specifically spectroscopy.J This was one of the first disciplines in chemistry that started to connect computers toL analytical equipment. The earlier systems came with a teletype (ASR33 ring a bell?).oJ Then video terminals became quickly popular but they were expensive, for a small team anyway.J Then our little group was able to buy a new spectrometer and a second hand video terminal.sL They arrived both at the same time. After the spectrometer was installed and
 calibratedJ we turned on the video terminal (a Beehive 152). Nothing. It worked at the
 shop where wea had it tested.H Now I should tell you that our kind of spectroscopy was NMR and this the first one that came withL a hefty superconducting magnet... The terminal worked all right but it wrote its output about9 5 meters to its left, where the 7 Tesla magnet stood ....-I The terminal was turned and in popped the text, a little out of focus butG that was easily corrected.   Hans  . David D Miller heeft geschreven in bericht ... >Folks >iI >Time to lighten up.  I'm sure you've all had experiences similar to thisu >... >rI >One of my 90L term server ports was acting up.  Before I had a chance to K >check it out the evening shift came in.  The janitor plugged his vacuum inoH >the same power strip that the 90L was plugged into and blew the circuitH >breaker.  He of course quietly found another receptacle to use.  In theG >morning the circuit breaker was reset and (sure enough) the 90L is now- >behaving itself.  >u% >Care to share your low-tech stories?d >  >dave. >h   ------------------------------  % Date: Wed, 28 Feb 2001 10:40:58 -0500K% From: "Islandco" <sales@islandco.com>r" Subject: Re: Low cost workstations/ Message-ID: <t9q68rip3oqbdf@corp.supernews.com>e  
 Good point  E We can get the QL-MT1AE-6X 12 Unit Alphastation Base license for $900 6 We can pass this on at Cost to help us sel the systems   Davidc   -- Island Computers US Corporationo 2700 Gregory Street 	 Suite 150n Savannah GA 31404t Tel: 912 447 6622K Fax: 912 201 0096i sales@islandco.com www.islandco.com  C This message and any files transmitted with it are confidential andaJ may be privileged and/or subject to the provisions of privacy legislation.H They are intended solely for the use of the individual or entity to whomE they are addressed. If the reader of this message is not the intendede
 recipient,G please notify Island Computers US Corp immediately and then delete this: message.I You are notified that reliance on, disclosure of, distribution or copying3 of this message is prohibited.    B "David J. Dachtera" <djesys.nospam@earthlink.net> wrote in message' news:3A9C5118.A15A5C2A@earthlink.net...f > Arne Vajhj wrote: > >a > > JF Mezei wrote: K > > > In support of David Dachtera's fight for low cost computing, it seemsa that SUN > > > has become our ally. > > >e. > > > http://www.sun.com/desktop/sunblade1000/ > > > F > > > AP news reported that those workstations will sell for $1000 US. > >o< > > If you want a low-cost Alpha then go to www.islandco.com# > > and click on "Low-Cost Alpha" !o >oH > Are they selling low-cost OpenVMS (commercial) licenses as well (under > $700 US)?e >e > -- > David J. Dachtera0 > dba DJE Systemsa > http://www.djesys.com/ >s< > Unofficial Affordable OpenVMS Home Page and Message Board:! > http://www.djesys.com/vms/soho/s >-H > This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings > is to be expected. >nB > Feel free to exercise your rights of free speech and expression. >CH > However, attacks against individual posters, or groups of posters, are > strongly discouraged.d   ------------------------------    Date: 28 Feb 2001 12:51:42 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)d" Subject: Re: Low cost workstations3 Message-ID: <AOlyRfc3QIh6@eisner.encompasserve.org>   W In article <t9q68rip3oqbdf@corp.supernews.com>, "Islandco" <sales@islandco.com> writes:y > Good point > G > We can get the QL-MT1AE-6X 12 Unit Alphastation Base license for $900n8 > We can pass this on at Cost to help us sel the systems  4 I think that would be an excellent competitive move,C (even though I am not currently in the market for another machine).e   ------------------------------  % Date: Wed, 28 Feb 2001 06:20:00 -0800o1 From: "Farrell, Michael" <MFarrell@voltdelta.com>t8 Subject: RE: My name is Dean, and I'm a [Un*x|NT] hater.Q Message-ID: <7DF45F22D904D31192EE00805F578DF2018BC77E@ny-exchange1.maintech1.com>n  ' 	Dean Woodward [deanw@rdrop.com] wrote:s 	[snip]  >  .  You see, in Un*x,rL > > the program never gets to know what the actual command line was, even if > > it wanted to.  > C 	This brings up an interesting question I've always wondered about.   E 	In DCL, is it possible for the called DCL procedure to know the namea- of the DCL symbol that was used to call it.  h 	E.G.: 	$  myproc = "@my_dcl_proc"  	$ myproc arg1 arg2p  ? 	And inside my_dcl_proc.com one could determine that the symbold "myproc" was what was used tos 	invoke my_dcl_proc.com.  C 	If it is possible to do this, could it also be done from inside an02 executable image written in a compilable language?   ------------------------------  % Date: Wed, 28 Feb 2001 09:27:17 -0500.2 From: rdeininger@mindspring.com (Robert Deininger)8 Subject: Re: My name is Dean, and I'm a [Un*x|NT] hater.L Message-ID: <rdeininger-2802010927170001@user-2ive673.dialup.mindspring.com>  
 In articleF <7DF45F22D904D31192EE00805F578DF2018BC77E@ny-exchange1.maintech1.com>,2 "Farrell, Michael" <MFarrell@voltdelta.com> wrote:  L         In DCL, is it possible for the called DCL procedure to know the name/ > of the DCL symbol that was used to call it.  t >         E.G.:y% >         $  myproc = "@my_dcl_proc" e >         $ myproc arg1 arg2 > H >         And inside my_dcl_proc.com one could determine that the symbol > "myproc" was what was used tol! >         invoke my_dcl_proc.com.c   I don't know a way to do this.    L >         If it is possible to do this, could it also be done from inside an4 > executable image written in a compilable language?  G An executable can get the command line.  See the CLI$ utility routines.m   --   Robert Deininger rdeininger@mindspring.com-   ------------------------------  # Date: Wed, 28 Feb 2001 15:08:24 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)n8 Subject: RE: My name is Dean, and I'm a [Un*x|NT] hater.0 Message-ID: <009F84F6.D9E12977@SendSpamHere.ORG>   In article <7DF45F22D904D31192EE00805F578DF2018BC77E@ny-exchange1.maintech1.com>, "Farrell, Michael" <MFarrell@voltdelta.com> writes: ( >	Dean Woodward [deanw@rdrop.com] wrote: >	[snip] >>  .  You see, in Un*x,M >> > the program never gets to know what the actual command line was, even ifO >> > it wanted to. i >> SD >	This brings up an interesting question I've always wondered about. >4F >	In DCL, is it possible for the called DCL procedure to know the name. >of the DCL symbol that was used to call it.   >	E.G.:o >	$  myproc = "@my_dcl_proc" t >	$ myproc arg1 arg2 >M@ >	And inside my_dcl_proc.com one could determine that the symbol >"myproc" was what was used to >	invoke my_dcl_proc.com.D > D >	If it is possible to do this, could it also be done from inside an3 >executable image written in a compilable language?m  1 $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE CTL$AG_CLIDATA9, $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE PPD$L_PRC2 $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE PRC_L_RECALLPTR $s1 $ PRC = 'F$fao("%x!@XL",CTL$AG_CLIDATA+PPD$L_PRC)r. $ NXT = 'F$fao("%x!@XL",PRC+PRC_L_RECALLPTR)-1" $ CMD = NXT-'F$fao("%x!@XB",NXT)-1( $ WRITE SYS$OUTPUT "''F$fao("!AC",CMD)'"     --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM             iO city, n., 1. a place where trees are cut down and streets are named after them.-   ------------------------------    Date: 28 Feb 2001 16:31:46 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>a8 Subject: Re: My name is Dean, and I'm a [Un*x|NT] hater.H Message-ID: <y4d7c2lt7h.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  ? system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes:h  3 > $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE CTL$AG_CLIDATA . > $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE PPD$L_PRC4 > $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE PRC_L_RECALLPTR > $ 3 > $ PRC = 'F$fao("%x!@XL",CTL$AG_CLIDATA+PPD$L_PRC)o0 > $ NXT = 'F$fao("%x!@XL",PRC+PRC_L_RECALLPTR)-1$ > $ CMD = NXT-'F$fao("%x!@XB",NXT)-1* > $ WRITE SYS$OUTPUT "''F$fao("!AC",CMD)'"  - Wonderful what FAO and DCL allow you to do...D   	Jan   ------------------------------  # Date: Wed, 28 Feb 2001 16:31:06 GMTA= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) 8 Subject: Re: My name is Dean, and I'm a [Un*x|NT] hater.0 Message-ID: <009F8502.6725D39B@SendSpamHere.ORG>   In article <y4d7c2lt7h.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>, Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> writes:@ >system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes: >-4 >> $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE CTL$AG_CLIDATA/ >> $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE PPD$L_PRCj5 >> $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE PRC_L_RECALLPTR$ >> $4 >> $ PRC = 'F$fao("%x!@XL",CTL$AG_CLIDATA+PPD$L_PRC)1 >> $ NXT = 'F$fao("%x!@XL",PRC+PRC_L_RECALLPTR)-1o% >> $ CMD = NXT-'F$fao("%x!@XB",NXT)-1n+ >> $ WRITE SYS$OUTPUT "''F$fao("!AC",CMD)'". >m. >Wonderful what FAO and DCL allow you to do... >a >	Jan   L ... and a little SYMBOL!  You can't get at first three symbol values easily  without it.O --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM,            dO city, n., 1. a place where trees are cut down and streets are named after them.x   ------------------------------    Date: 28 Feb 2001 18:03:15 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>-8 Subject: Re: My name is Dean, and I'm a [Un*x|NT] hater.H Message-ID: <y44rxe68q4.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  ? system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes:   0 > >Wonderful what FAO and DCL allow you to do...N > ... and a little SYMBOL!  You can't get at first three symbol values easily 
 > without it..  M Agreed, though the first two aren't changing much, I think. The CLI stuff can-K change with every release, and some EXE etc stuff can change with a reboot,c& right? Those are the really hard ones.  J But it's really fun how one can do such a base thing as passing an address right past DCL into $FAO.p   	Jan   ------------------------------    Date: 28 Feb 2001 18:04:31 +0100* From: eplan@kapsch.net (Peter LANGSTOEGER)8 Subject: Re: My name is Dean, and I'm a [Un*x|NT] hater.* Message-ID: <3a9d2f9f$1@news.kapsch.co.at>   In article <y4d7c2lt7h.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>, Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> writes:@ >system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes:4 >> $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE CTL$AG_CLIDATA/ >> $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE PPD$L_PRC 5 >> $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE PRC_L_RECALLPTRn >> $4 >> $ PRC = 'F$fao("%x!@XL",CTL$AG_CLIDATA+PPD$L_PRC)1 >> $ NXT = 'F$fao("%x!@XL",PRC+PRC_L_RECALLPTR)-1A% >> $ CMD = NXT-'F$fao("%x!@XB",NXT)-1 + >> $ WRITE SYS$OUTPUT "''F$fao("!AC",CMD)'". > . >Wonderful what FAO and DCL allow you to do...   Come on,G without VAXman's SYMBOL package, DCL would not be of great help here...I   -- l< Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888 < <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"   ------------------------------    Date: 01 Mar 2001 01:03:07 +0800, From: Paul Repacholi <prep@prep.synonet.com>8 Subject: Re: My name is Dean, and I'm a [Un*x|NT] hater.- Message-ID: <87d7c2oi44.fsf@prep.synonet.com>m  ' Dean Woodward <deanw@rdrop.com> writes:n  1 > Some days, it doesn't pay to get out of bed...     ...   E > I telnet'ed to my (small, local, run by a friend) ISP today to findxB > out when my account's due, and chanced across a file named '-l'.  F > Those of you who know anything about Un*x at all can sympathize; forE > the rest of you, here's a lesson in user shell design.  You see, in4C > Un*x, the program never gets to know what the actual command line_E > was, even if it wanted to.  So when I type 'rm -i *l', rm (the un*x B > delete program) sees rm -i -l [...] and knowing that -l is a badE > switch, happily barfs a perfectly useless and cryptic error messageg > and dies:1  D Once, long ago in a poor crippled universe, that is being constantly& GUIed and re-released under the GPL...  E One of the knikier shells way back when had the feature of re-scaning @ the expanded command line. And on this on, was 'dir' a ls mongolA offspring that did not show ANY files with non alphanumeric first 7 char. Don't confuse the luzer, OK... ls was also wierd.   D Now one day, on of the unix gods typed 'dir' in the root. Odd, thereD is a file call 'rm' in there...??? So he typed 'ls *'... And waited,B and waited, and noticed the DECwriter was very busy in the machine room...o  C Some BOFH had added the files '&' 'rm' '-Rf' and '*'. So the rescan C expanded the command again to 'ls & rm -Rf *' and shoved the second F result un-rescaned ( feature of the & they figured ) init rm maw.  Oh,  and rm knew how to glob input...  A They stopped running experimental UI SW and the main machine soon. after that.u   And so it goes...J   -- f< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.s@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  % Date: Wed, 28 Feb 2001 12:18:41 -0500 1 From: "Farrell, Michael" <MFarrell@voltdelta.com>-8 Subject: RE: My name is Dean, and I'm a [Un*x|NT] hater.Q Message-ID: <7DF45F22D904D31192EE00805F578DF2018BC784@ny-exchange1.maintech1.com>e  J Thanks for this.  Now, my only problem may be that I don't hae your SYMBOL package.  How can I get it?r   Thanks again   Mike Farrell   > -----Original Message-----> > From:	system@SendSpamHere.ORG [SMTP:system@SendSpamHere.ORG]- > Sent:	Wednesday, February 28, 2001 10:08 AMx > To:	Info-VAX@Mvb.Saic.Com5: > Subject:	RE: My name is Dean, and I'm a [Un*x|NT] hater. >  > In articleH > <7DF45F22D904D31192EE00805F578DF2018BC77E@ny-exchange1.maintech1.com>,5 > "Farrell, Michael" <MFarrell@voltdelta.com> writes:"* > >	Dean Woodward [deanw@rdrop.com] wrote:
 > >	[snip] > >>  .  You see, in Un*x,L > >> > the program never gets to know what the actual command line was, even > if > >> > it wanted to. o > >>  F > >	This brings up an interesting question I've always wondered about. > > H > >	In DCL, is it possible for the called DCL procedure to know the name0 > >of the DCL symbol that was used to call it.  	 > >	E.G.:a > >	$  myproc = "@my_dcl_proc" n > >	$ myproc arg1 arg2 > >sB > >	And inside my_dcl_proc.com one could determine that the symbol  > >"myproc" was what was used to > >	invoke my_dcl_proc.com.  > >eF > >	If it is possible to do this, could it also be done from inside an5 > >executable image written in a compilable language?1 > 3 > $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE CTL$AG_CLIDATAw. > $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE PPD$L_PRC4 > $ SYMBOL/SET/LOCAL/DEPTH/EXECUTIVE PRC_L_RECALLPTR > $ 3 > $ PRC = 'F$fao("%x!@XL",CTL$AG_CLIDATA+PPD$L_PRC)W0 > $ NXT = 'F$fao("%x!@XL",PRC+PRC_L_RECALLPTR)-1$ > $ CMD = NXT-'F$fao("%x!@XB",NXT)-1* > $ WRITE SYS$OUTPUT "''F$fao("!AC",CMD)'" >  >  > --4 > VAXman- OpenVMS APE certification number: AAA-0001 > VAXman(at)TMESIS(dot)COM
 >            eK > city, n., 1. a place where trees are cut down and streets are named afterg > them.o   ------------------------------  % Date: Wed, 28 Feb 2001 12:16:53 -0500o2 From: "Sue Skonetski" <susan.skonetski@compaq.com>' Subject: OpenVMS Technical Update Day's-8 Message-ID: <uoan6.222$dl6.2667@gazette.loc1.tandem.com>   Washington DC and Dallas  E Here is the information for the    Washington DC    OpenVMS Technicalh Update - March 27-28  @ Registration is filling quickly for DC. http://compaq-signup.com  L Here is the information for the   Dallas     OpenVMS Technical Update, April 18-20r  E The Dallas Ft Worth Compaq User Group (The DFWCUG) will be once again3C hosting the DFWdays 2001 Compaq Seminar April 18th, 19th, and 20th.h  L (Note: This will be held in the middle of the week -- unlike last year whereF we held it on the weekend...).  The agenda is not complete for the VMS/ sessions however it will be very similar to DC.s  A Registration and other schedule information is on the website at:e   http://dfwdays.dfwcug.org/  
 Warm Regards,a     Suet   ------------------------------    Date: 28 Feb 2001 09:14:54 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>h8 Subject: Re: PCSI kits - where was a specific kit built?H Message-ID: <y4zof7z0jl.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  : hammond@not@peek.ppb.cpqcorp.net (Charlie Hammond) writes:  J > Neither of these will get you the exact PRODUCT PACKAGE command that was > used, but it will be a start.g  F Looks like an oversight to me - BACKUP has listed that from day one...   	Jan   ------------------------------  # Date: Wed, 28 Feb 2001 09:23:18 GMTk, From: Didier Morandi <Didier.Morandi@gmx.ch>8 Subject: Re: PCSI kits - where was a specific kit built?& Message-ID: <3A9CC39B.924C7B47@gmx.ch>   Charlie Hammond wrote: > I > See PRODUCt LIST.  (This works only for sequential format kits, but forwP > reference formate kits you already have all the files in a directory hierarchy
 > anyway.)  E PRODUCT LIST gives you the contents of the "file" section of the PDF.eE You can't have information on the original location of the kit files.    D.   ------------------------------  # Date: Wed, 28 Feb 2001 17:10:20 GMTc8 From: hammond@not@peek.ppb.cpqcorp.net (Charlie Hammond)8 Subject: Re: PCSI kits - where was a specific kit built?8 Message-ID: <0ian6.221$dl6.2722@gazette.loc1.tandem.com>   In article <y4zof7z0jl.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>, Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> writes:; >hammond@not@peek.ppb.cpqcorp.net (Charlie Hammond) writes:c >iK >> Neither of these will get you the exact PRODUCT PACKAGE command that wase  >> used, but it will be a start. >fG >Looks like an oversight to me - BACKUP has listed that from day one...e  ? Well, to be fair, BAKUP is not a software installation utility.t# It captures this for other reasons.e  G But... It is not an unreasonable request for PRODUCT PACKAGE to capturel9 this, perhaps as comments in the Product Definition File.i  K I have called this to the attention of the POLYCENTER Software Installation  (PCSI) utility developers.   -- tK     Charlie Hammond -- Compaq Computer Corporation -- Pompano Beach  FL USAaH        (hammond@not@peek.ppb.cpqcorp.net -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------    Date: 28 Feb 2001 09:29:02 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>c) Subject: Re: Possible security hole in...oH Message-ID: <y4r90jyzw1.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  4 hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) writes:  L >   The interest of Sun Marketing in OpenVMS Security has been instrumental J >   in my being able to light a fire under a few of the appropriate local M >   security keisters, and I'd like to thank Mr Harrison his assistance here.e   8-).  G What is a "keister", incidentally? My trusty old copy of webster, stillo4 answering from a NeXT cube, will know nothing of it.   	Jan   ------------------------------  % Date: Wed, 28 Feb 2001 08:44:11 -0600 * From: WILLIAM WEBB <WWEBB1@email.usps.gov>) Subject: RE: Possible security hole in... - Message-ID: <0033000017371292000002L022*@MHS>   H =0AKeister is Yiddish for what our English brethren refer to as a "bum"= .e   > -----Original Message-----1 > From: Info-VAX-Request@Mvb.Saic.Com at INTERNETt, > Sent: Wednesday, February 28, 2001 3:54 AM8 > To: Webb, William W; Info-VAX@Mvb.Saic.Com at INTERNET+ > Subject: RE: Possible security hole in...c >e >c6 > hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) writes: >r; > >   The interest of Sun Marketing in OpenVMS Security has  > been instrumental09 > >   in my being able to light a fire under a few of thed > appropriate localR> > >   security keisters, and I'd like to thank Mr Harrison his > assistance here. >5 > 8-). >7: > What is a "keister", incidentally? My trusty old copy of > webster, still6 > answering from a NeXT cube, will know nothing of it. >s
 >      Jan >=   ------------------------------  % Date: Wed, 28 Feb 2001 19:27:25 +0010 % From: paddy.o'brien@zzz.tg.nsw.gov.aud= Subject: Re: Problem compiling Fortran program with /SEPARATEe5 Message-ID: <01K0NL7VXCS200AENZ@tgmail.tg.nsw.gov.au>   	 Lawrence,h  K (I'm having a few problems with my VMSmail line width, so quoting is weird d- -- but then I've seen that in other's posts).d  E >I'm having a problem with a legacy program I am porting.  I'm on an o >OpenVMS AXPH >7.1-2 system running Compaq Fortran V7.2-780.  The program in question 	 >has beenlF >running for some time.  It has a large number of modules in a single 	 >file.  I.I >was in the process of separating the modules into several files when thei5 >program stopped working; ie, produced bad results.  n  H This is a very old version.  My current version is V7.3-965.  This is a G year old and may have been superceded.  Exotic or erotic problems have e? been fixed -- I helped to find some -- and some really sexy :-)s  J >I narrowed it down to a single module (there may be more than one) that,  >whenwJ >as part of the single source file, the program works, but when separated,F >fails.  I did a DIFF on the source file, and there was no difference  >between	 >the two.i >tF >I started looking at storage maps, and there were very different.  I  >decided to D >use the /SEPARATE_COMPLIATION qualifier, and that is what made the  >difference. >aK >When I compile the single source file with /SEPARATE, the program fails.  g >WhennG >I omit the qualifier (/NOSEPARATE is the default), it works.  (It had   >beeno< >compiled without the /SEPARATE qualifier in the past, btw.) >-K >Does anyone have any idea what to look for that would cause this problem? . > IBG >can post the source code if you wish, but let's take this a step at a U >time.  E I assume from your description that the problem is in the results of DI running rather than the compilation.  Do you just get different results, l or an access violation?d  K Why not write to Steve Lionel's team: fortran@compaq.com (IIRC)?  Or often YH Steve is amenable to you writing directly to him -- an exercise for the  reader.w  E The maps will be different because /NOSEP tries to inline small code  0 snippets.  "Small" is indeterminate to us users.  I I would assume some problem with the argument/dummy argument lists which y6 the in-lining rectifies for you.  I don't really know.  F In lieu of your discussing it with the real experts (Steve's team), I K would be happy to spend a bit of time looking at your code this weekend if eI you care to mail me the source privately.  I have some sort of expertise nD in this area as I (my team) have beta tested since the inception of # Digital's (Compaq's) F90 compilers.E   Regards, Paddy   Paddy O'Brien, Transmission Development, 
 TransGrid, PO Box A1000, Sydney South,  NSW 2000, Australia    Tel:   +61 2 9284-3063 Fax:   +61 2 9284-3050& Email: paddy.o'brien@zzz.tg.nsw.gov.au  F Either "\'" or "\s" (to escape the apostrophe) seems to work for most  people, ; but that little whizz-bang apostrophe gives me little spam.X   ------------------------------    Date: 28 Feb 2001 10:02:52 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>-= Subject: Re: Problem compiling Fortran program with /SEPARATEgH Message-ID: <y4g0gzyybn.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  , bleau@umtof.umd.edu (Lawrence Bleau) writes:  M > I started looking at storage maps, and there were very different. I decidedaH > to use the /SEPARATE_COMPLIATION qualifier, and that is what made the 
 > difference.E  M Sounds like unmatched commons, or an interface problem (mismatched arguments)LL that the compiler can see and correct when doing a whole-source compilation,N but not in a seperate compilation (which is what /SEPERATE does - the compiler: behaves as if it looses its memory between program units).  N If, as you say, the storage maps are different, see which PSECT (which has theL name of the common it stores) are different. Then look at the definitions of+ those commons, and check their consistency.    	Jan   ------------------------------   Date: 28 Feb 2001 13:16:25 GMT! From: markdiaz@aol.com (MarkDiaz) = Subject: Re: Problem compiling Fortran program with /SEPARATEc: Message-ID: <20010228081625.23645.00000322@ng-mg1.aol.com>  I >I narrowed it down to a single module (there may be more than one) that,n >wheneJ >as part of the single source file, the program works, but when separated, >fails.   K My two cents: If you are porting from VAX, look for a character variable or$G string literal which gets passed by descriptor (default on AXP), but is'N received by reference. The VAX linker tried to guess this based on informationO in the object module. Compiling /NOSEPARATE lets the compiler use info from theo' other modules, where /SEPARATE doesn't.i  	 Mark Diaz  markdiaz@aol.com   ------------------------------  % Date: Wed, 28 Feb 2001 09:43:04 -0500t, From: Steve Lionel <Steve.Lionel@compaq.com>= Subject: Re: Problem compiling Fortran program with /SEPARATEc8 Message-ID: <vd3q9tgm5gqt7t1unq3kor9t8l410ruo9o@4ax.com>  A On 27 Feb 2001 23:58:09 GMT, bleau@umtof.umd.edu (Lawrence Bleau)b wrote:  P >I'm having a problem with a legacy program I am porting.  I'm on an OpenVMS AXPP >7.1-2 system running Compaq Fortran V7.2-780.  The program in question has beenN >running for some time.  It has a large number of modules in a single file.  II >was in the process of separating the modules into several files when thee5 >program stopped working; ie, produced bad results.  s > N >I narrowed it down to a single module (there may be more than one) that, whenJ >as part of the single source file, the program works, but when separated,M >fails.  I did a DIFF on the source file, and there was no difference between 	 >the two.  > P >I started looking at storage maps, and there were very different.  I decided toO >use the /SEPARATE_COMPLIATION qualifier, and that is what made the difference.a >hO >When I compile the single source file with /SEPARATE, the program fails.  When.K >I omit the qualifier (/NOSEPARATE is the default), it works.  (It had beeni< >compiled without the /SEPARATE qualifier in the past, btw.) >aM >Does anyone have any idea what to look for that would cause this problem?  IrL >can post the source code if you wish, but let's take this a step at a time.  F I would suggest installing a current version of the compiler - the one1 you have is two years old.  7.4 is available now.   E /SEPARATE_COMPILATION dramatically alters how the compiler operates -%C it disables interprocedural analysis and optimization.  It could be = that you have an error in your source that is revealed by thew different optimization.   > If you'd like us to take a look at it, send me the source (andE instructions for reproducing the problem.)  But I would suggest first & trying it with a more recent compiler.    - Steve Lionel (mailto:Steve.Lionel@compaq.com)< Fortran Engineeringl& Compaq Computer Corporation, Nashua NH  6 Compaq Fortran web site: http://www.compaq.com/fortran   ------------------------------  % Date: Wed, 28 Feb 2001 17:42:59 +00007+ From: "antonio.carlini" <arcarlini@iee.org>eL Subject: Re: Sun Blade 100 with a VMS monitor ? (was: Low cost workstations)' Message-ID: <3A9D38A3.615E85FF@iee.org>n   Larry Kilgallen wrote:G > While I have a couple of VRT19s, I also have some of the more generic F > monitors.  It is reassuring to know that Sun does not have their ownG > unique way of doing it (like Sbus), as I have read a lot lately about & > special Sun cables for KVM switches.  8 As it so happens, I have the aforementioned Sun Ultra 5,6 a generic PC and a VAXstation 4000 all on my desk (for9 suitable values of "on" obviously ...) and all displayinga on the same monitor. .  5 The VAXstation goes in via BNC synch-on-green and hasr its own keyboard and mouse.n  1 The PC keyboard, mouse and video go via a genericf KVM switch.   . The Sun video goes via the KVM switch but the ( keyboard and mouse don't ... in fact the& Sun mouse plugs into the Sun keyboard,. which then plugs into the back of the Ultra 5.  / So although the Sun video is SVGA, the keyboard , is not a PC-style keyboard and will not work+ via my KVM. I don't know what name (if any)N2 the Sun combined mouse-keyboard interface goes by.  . I assume that it would be possible to purchase/ KVMs suitable for Suns, and possibly KVMs whicho, could cope with some combination of Suns and& PCs. I just don't have any such beast.     Antonion   -- e   --------------- - Antonio Carlini             arcarlini@iee.org    ------------------------------  % Date: Wed, 28 Feb 2001 11:06:18 -0500m2 From: rdeininger@mindspring.com (Robert Deininger) Subject: Re: TCPIP 5.0L Message-ID: <rdeininger-2802011106180001@user-2ive673.dialup.mindspring.com>  L In article <dlxm6.23$H61.516@client>, "steve smith" <ssmith@lbpc.com> wrote:  L > I need to upgrade from 4.2 to 5.0 of tcpip, before going to AlphaVMs 7.2-1N > on my ES40. Besides the patch for 5.0, is anyone aware of anything else that5 > I need to check into before performing the upgrade.aM > I plan to backup the system disk to a spare disk, boot off of the spare and-+ > upgrade it to tcpip 5.0, as a safety net.   I Think about V5.1 instead.  It's just newly released, adds significant newa# features, and may be more reliable..   -- g Robert Deininger rdeininger@mindspring.com8   ------------------------------   Date: 27 Feb 2001 20:28:18 CDT= From: wayne@tachysoft.xxx.344932.killspam.0142 (Wayne Sewell)eN Subject: the haunted vt100 (was Re: a little humor for this so often humorless. Message-ID: <L8V8hJMvm5UH@tachxxsoftxxconsult>  H Back in the dawn of the vax, I was working at a defense contractor.  OurM software project consisted of about 70 programmers working on what eventuallyaO became a couple of million lines of vax pascal code.  This was circa 1980, back L in the days when all access to the system was via vt100.  There was no X, no= windoze (billy had not yet devastated the computer industry).   K A friend of mine was having weird problems with his vt100.  For no apparentfJ reason, the words "Leave me alone!!" would suddenly appear on his terminalM while he was trying to work.  Apparently his vt100 did not like for people toC type on its keyboard.   E At first he thought somebody was sending broadcast messages, but "set M term/nobroad" had no effect.  He tried to trap one of the messages at the IRP_L level (this guy was *not* an idiot; he and I were the only two people in theO company who knew vms internals).  He tried to intercept the messages at severalaF different levels, but still they kept appearing on the screen.  It wasL infrequent enough that he could do his work, but happened often enough to be	 annoying.L  H I was in his office one day and saw the "Leave me alone!!" several timesM myself.  At the time, he was typing a file.  The bogus message appeared aboutf every ten lines.  L Something hit me.  "Is that file pure ascii?  Are there any control codes in it?"  N "I don't know," he said.  "It's supposed to be a text file, but there might be control codes."S  + "Do a control-E on the keyboard right now."e  5 Sure enough, "Leave me alone!!" appeared instantly.     N It was the answerback message.  This is a string that lives inside the vt100'sL nonvolatile memory.  When a control-e is received by the keyboard, the vt100N responds with the answerback.  This is an artifact that harks all the way backL to the days of teletype.   It was a means of identification.  You programmedM the terminal's location or some other identifying string.   Amazingly enough, O you can even find the answerback in decterms (now *that* is emulation).  From aV2 decterm, go to Options/General and you can set it.  O We never did find out who programmed an answerback into his terminal.  It couldCN have been done long before it was even assigned to him.  Once set, it is thereO until explicitly cleared.  In any case, after going into setup and clearing theg= answerback, his vt100 never again told him to leave it alone.e         --  O ===============================================================================pM Wayne Sewell, Tachyon Software Consulting  (281)812-0738  wayne@tachysoft.xxx-: http://www.tachysoft.xxx/www/tachyon.html and wayne.html  K change .xxx to .com in addresses above, assuming you are not a spambot  :-)sO ===============================================================================oO Dean Wormer to Flounder: "Fat, drunk, and stupid is no way to go through life."w   ------------------------------  % Date: Wed, 28 Feb 2001 12:49:37 -0500 - From: "Richard D. Piccard" <piccard@ohio.edu>o' Subject: Re: UCX/TCPIP 5.1 for Hobbyistl( Message-ID: <3A9D3A2D.8D708479@ohio.edu>  E I suspect they just didn't test it with any VAX that had less than 32  MBytes.r  +                                         RDP     
 trevor wrote:s   > Hi > E > I'd like to try TCPIP services 5.1 on my hobbyist machines at home.:G > Trouble is at work I only have Alpha's now and (even worse) at home IuD > only have VAX'en, so I don't have access to the appropriate media. >cI > Would some kind soul be prepared to copy or loan me the media for TCPIPiI > services 5.1 for VAX.  I have a hobbyist license.  I'm in the UK, and Io- > can read TK50, TK70, TLZ06 and CD-ROM.  Ta.- >-F > By the way I notice that the TCPIP 5.1 SPD says the Required minimumJ > memory for VAX is 32MB.  I think this will preclude a significant numberJ > of hobbyist class machines running it, e.g VAX station 4000 VLC 24MB maxI > memory IIRC.  I wonder if this really is the Minimum Requirement, or ifJJ > perhaps operation is possible with 24MB but at reduced performance.  AnyH > thoughts ?  Are many people using 5.1 yet ?  Is there an install check > for memory ? >yI > It seems to me that a PDP-11 with 256kB of RAM used to be able to do antH > awful lot that you could'nt possibly do now unless you have buckets ofC > the stuff.  Yes I know chalk and cheese, but I have seen the termiG > bloatware used with regard M$oft products and I would hate to see thes? > same term levelled at VMS, after all what does VMS stand for.c >e > Regards, TrevorD >n  > trevor_deja AT my-deja DOT com   --B ==================================================================B Dick Piccard                           Academic Technology ManagerB piccard@ohio.edu                                 Computer ServicesB http://oak.cats.ohiou.edu/~piccard/                Ohio University   ------------------------------  % Date: Wed, 28 Feb 2001 09:57:37 +0000-% From: Alan Greig <a.greig@virgin.net> R Subject: Re: US telephone vote on SPAM in comp.os.vms (was: High Search Engine...)8 Message-ID: <dnip9t8ofrmvmbb2oir664pf3a3b54807m@4ax.com>  B On Tue, 27 Feb 2001 12:03:52 +0000, Steve.Spires@yellowpages.co.uk wrote:  M >Contact:   Tel: 3063  -  IS - Infrastructure, 1st Floor, Bridge Street Plazat >a >e >Really? Is that true? > L >It doesn't work like that in the UK, because many agencies thought it wouldL >disuade people from calling them [Childline, Samaritans, Rape Crisis, CrimeI >Stoppers etc] so do the respective US agencies not use freecall numbers?n   Steve,  ? Agencies such as the above on UK 0800 numbers *do* see incoming-B caller-id even if withheld. Where it doesn't appear is on your own itemized bill.      
 >Steve Spiresj >: >: >s >lL >moroney@world.std.spaamtrap.com (Michael Moroney) on 27/02/2001 07:03:19 AM >t! >To:        Info-VAX@Mvb.Saic.Comb, >cc:         (bcc: Steve Spires/YellowPages)P >From:      moroney@world.std.spaamtrap.com (Michael Moroney), 27 February 2001, >           7:03 a.m.0 >3K >Re: US telephone vote on SPAM in comp.os.vms (was: High Search Engine ...)R >P >i >  >u: >"David J. Dachtera" <djesys.nospam@earthlink.net> writes:I >>> :So I would urge everyone (in the US) who wants a SPAM-free newsgroupcI >>> :to call _both_ numbers and express that sentiment.  The spammer willd! >>> :cheerfully pay for the call.a >>> P >>>   You assume that this is not something that is targeting the owner of theseN >>>   toll-free numbers -- by calling the number, you could be costing someoneM >>>   that the spammer has targeted the money, and not costing the spammer...  >nF >>You may want to precede the number with the caller-id block code, if1 >>your local telephone provider has this service.u >cB >>In Metro Chicago (Ameritech), dial *67 first to block caller id. >0G >Caller ID block does *NOT* block your number when calling 800 numbers!e- >Remember this before doing something stupid.n >s >-Mike >- >t   -- Alan   ------------------------------    Date: 01 Mar 2001 00:37:34 +0800, From: Paul Repacholi <prep@prep.synonet.com>! Subject: VMS wanted list. Take IIr0 Message-ID: <87hf1eojap.fsf_-_@prep.synonet.com>  4 hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) writes:  A > In article <871ysma6dz.fsf_-_@prep.synonet.com>, Paul Repacholio! > <prep@prep.synonet.com> writes:   B > :Hum, perhaps it is time for a 'what do folks want on VMS' list.   >   Um, (respectfully) no...  E >   Folks around here do have a good idea of the common requests -- It? >   have been working to make sure commonly-requested tools are  >   available...  A >   Perhaps it is time for a `how can we help you port it?' list,r >   instead...  ; >   Like most any commercial software organization, OpenVMSkE >   Engineering does not have the cycles available to port everythinghB >   for everyone...  (I am hesitant to offer to coordinate such anB >   effort, as I'm already working regularly in this area with the >   Freeware...)  : I was not thinking of an open ended winge list to ZK Hoff.  B Everyone had their own list of 'must haves', wants, and nice extraA stuff. The common feature is that they are all different... So ite? would be nice if we had some idea of what others wanted, and ifR someone else was working on it.   F It would be even nicer if we started to do some design and engineeringC *before* people started cutting code... You know, those things thath( under lie WHY we like VMS to start with.  5 For VMS itself, there is the traditional SIR process.    This is a parallel to that.t  > So let see if we can do it, and do it well. BTW, that includes6 ruthless editing of Subject: and no topic drift folks.  D So let's start with the wishlist process. NO 'I want...' yet please.9 This is about how we conordinate and plan the whole mess.d   -- t< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.r@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------    Date: 28 Feb 2001 15:03:28 +0100* From: eplan@kapsch.net (Peter LANGSTOEGER) Subject: Re: VX1 and PC164LX* Message-ID: <3a9d0530$1@news.kapsch.co.at>  v In article <DZym6.44421$3%5.332928@e420r-sjo3.usenetserver.com>, Eric Dittman <dittman@narnia.int.dittman.net> writes:6 >Has anyone tried using a VX1 board (with the required5 >patches) on a PC164LX?  I've tried and I can get the 6 >display up, but the system is so incredibly slow that6 >logging in on a serial line takes about five minutes,1 >followed by about five minutes to do a SHOW SYS,e9 >followed by a STOP/ID= on the DECwindows SERVER process,t, >at which time the system returns to normal. > : >If anyone has any ideas, let me know.  This is on V7.2-1.  ) Sounds as if you haven't enough memory....   -- f< Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888a< <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"   ------------------------------  % Date: Wed, 28 Feb 2001 10:55:05 -0500a% From: "Islandco" <sales@islandco.com>i Subject: Re: VX1 and PC164LX/ Message-ID: <t9q73ajglme449@corp.supernews.com>d   Eric  K I tried sending you the mail delivery error message and got a mail deliveryo) error on the error that I was sending youh  	 Go figureE   DT   -- Island Computers US Corporationo 2700 Gregory Streetn	 Suite 150i Savannah GA 31404s Tel: 912 447 6622s Fax: 912 201 0096  sales@islandco.com www.islandco.com  C This message and any files transmitted with it are confidential andsJ may be privileged and/or subject to the provisions of privacy legislation.H They are intended solely for the use of the individual or entity to whomE they are addressed. If the reader of this message is not the intendeds
 recipient,G please notify Island Computers US Corp immediately and then delete thisn message.I You are notified that reliance on, disclosure of, distribution or copying  of this message is prohibited.    @ "Eric Dittman" <dittman@narnia.int.dittman.net> wrote in message: news:DZym6.44421$3%5.332928@e420r-sjo3.usenetserver.com...7 > Has anyone tried using a VX1 board (with the requiredb6 > patches) on a PC164LX?  I've tried and I can get the7 > display up, but the system is so incredibly slow that 7 > logging in on a serial line takes about five minutes,e2 > followed by about five minutes to do a SHOW SYS,: > followed by a STOP/ID= on the DECwindows SERVER process,- > at which time the system returns to normal.n >c; > If anyone has any ideas, let me know.  This is on V7.2-1.t > -- > Eric Dittman > dittman@dittman.net3 >9   ------------------------------  # Date: Wed, 28 Feb 2001 15:40:17 GMT(- From: goathunter@goatley.com (Hunter Goatley)e* Subject: WKU FILESERV: Updated DSNLINK_NEW1 Message-ID: <3a9d1bc3.473008360@swen.process.com>r  9 The following package has been updated on ftp.wku.edu ands its mirrors:      o DSNLINK_NEW  (Updated)n  < 	DSNLINK_NEW V5.8 is a DCL procedure that can be run nightly; 	to mail all new articles added to the DSNlink databases to ; 	a list of users, as well as automatically retrieve desired ? 	articles.  Written by Dan Wing and enhanced by several others,i> 	this version of DSNLINK_NEW includes bug fixes and works with 	DSNlink V3.0 (as did V5.7).  1 You can find DSNLINK_NEW using one of these URLs:    http://www2.wku.edu/fileserv/g  . ftp://ftp.wku.edu/vms/fileserv/dsnlink_new.zip6 ftp://ftp.process.com/wku/vms/fileserv/dsnlink_new.zip? http://www.tmk.com/ftp/ftp-wku-edu/vms/fileserv/dsnlink_new.zipc: ftp://ftp.tmk.com/ftp-wku-edu/vms/fileserv/dsnlink_new.zip  5 These mirrors should be updated in the next 24 hours:h  E           ftp.vms.stacken.kth.se, under [.MIRRORS.WKU.VMS.FILESERV]. a8           ftp.ctrl-c.liu.se, under [.WKU.VMS.FILESERV].            ftp.riken.go.jp :           ftp.vsm.com.au, under kits and kits/decwindows. :           www.vsm.com.au/ftp/, via the WWW instead of FTP.     Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/ 9 goathunter@goatley.com     http://www.goatley.com/hunter/r   ------------------------------  # Date: Wed, 28 Feb 2001 15:18:16 GMTc- From: goathunter@goatley.com (Hunter Goatley) 4 Subject: WKU FILESERV: Updated MMK, plus WASD mirror1 Message-ID: <3a9d15de.471498759@swen.process.com>s  9 The following package has been updated on ftp.wku.edu and  its mirrors:      o DSNLINK_NEW  (Updated)e  < 	DSNLINK_NEW V5.8 is a DCL procedure that can be run nightly; 	to mail all new articles added to the DSNlink databases to ; 	a list of users, as well as automatically retrieve desired ? 	articles.  Written by Dan Wing and enhanced by several others,5> 	this version of DSNLINK_NEW includes bug fixes and works with 	DSNlink V3.0 (as did V5.7).  1 You can find DSNLINK_NEW using one of these URLs:3   http://www2.wku.edu/fileserv/r  . ftp://ftp.wku.edu/vms/fileserv/dsnlink_new.zip6 ftp://ftp.process.com/wku/vms/fileserv/dsnlink_new.zip? http://www.tmk.com/ftp/ftp-wku-edu/vms/fileserv/dsnlink_new.zipq: ftp://ftp.tmk.com/ftp-wku-edu/vms/fileserv/dsnlink_new.zip  5 These mirrors should be updated in the next 24 hours:o  E           ftp.vms.stacken.kth.se, under [.MIRRORS.WKU.VMS.FILESERV]. t8           ftp.ctrl-c.liu.se, under [.WKU.VMS.FILESERV].            ftp.riken.go.jp :           ftp.vsm.com.au, under kits and kits/decwindows. :           www.vsm.com.au/ftp/, via the WWW instead of FTP.     Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/e9 goathunter@goatley.com     http://www.goatley.com/hunter/t   ------------------------------  # Date: Wed, 28 Feb 2001 10:41:35 GMTs, From: Didier Morandi <Didier.Morandi@gmx.ch>. Subject: [info] Telnet/Rlogin and 7.2-1 broken& Message-ID: <3A9CD5F4.C206C283@gmx.ch>  F We just discovered that the xwindow sessions when logged on VMS didn'tC work on a SUN station since we upgraded an Alpha from 7.1 to 7.2-1.   ; The error message is "can't open display" after the "createt5 /terminal=decterm /detach" command at VMS login time.a  C We found that the command "setenv term dtterm" had to supersede then- "setenv term xterm" command to work properly.-   Fyi.   D.   ------------------------------    Date: 28 Feb 2001 17:18:47 +0100* From: eplan@kapsch.net (Peter LANGSTOEGER)2 Subject: Re: [info] Telnet/Rlogin and 7.2-1 broken* Message-ID: <3a9d24e7$1@news.kapsch.co.at>  U In article <3A9CD5F4.C206C283@gmx.ch>, Didier Morandi <Didier.Morandi@gmx.ch> writes:rG >We just discovered that the xwindow sessions when logged on VMS didn't0D >work on a SUN station since we upgraded an Alpha from 7.1 to 7.2-1.  " What did you upgrade in addition ? UCX V4 to TCPIP V5 ?* DECwindows-MOTIF to DECwindows-MOTIF/CDE ?  < >The error message is "can't open display" after the "create6 >/terminal=decterm /detach" command at VMS login time. >lD >We found that the command "setenv term dtterm" had to supersede the. >"setenv term xterm" command to work properly.  K This seems to be an effect of a move to CDE. Therefor no problem of VMS ;-)a  G My general suggestion is: Upgrade the layered products when 1) you wantdH to use their new features or 2) you need them for the new opsys version.C But upgrade them _before_ you upgrade the opsys (if they run on thes> old opsys version of course - unlike DECnet)! You don't reallyA save time when you upgrade all at once and have no idea where newu  features/behaviours come from...   -Peter  B PS: If you upgrade UCX to TCPIP, go directly to V5.1 or eventuallyK to V5.0-11 (aka V5.0A) ECO2. Older Versions of TCPIP like V5.0-9 (aka V5.0)oI and V5.0-10 (aka V5.0A !!) are known to be buggy (eg. a bug in the TELNETt+ server, and this one might also effect you)a   -- l< Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888C< <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"   ------------------------------  # Date: Wed, 28 Feb 2001 11:00:12 GMTi, From: Didier Morandi <Didier.Morandi@gmx.ch>4 Subject: [Q] concatenating executable files is OK???& Message-ID: <3A9CDA52.902A30F7@gmx.ch>   How come  0 ISLKP1> copy sfx_axp.exe,SPLAXP015.ZIP spell.exe ISLKP1> r spellhG UnZipSFX 5.32 of 3 November 1997, by Info-ZIP (Zip-Bugs@lists.wku.edu). %   inflating: splaxp015.a             L ISLKP1>    could work?t   D.D (where is my Kenah/Goldenberg? Did someone see my Kenah/Goldenberg?)   ------------------------------  % Date: Wed, 28 Feb 2001 11:34:42 -0000 * From: "Richard Brodie" <R.Brodie@rl.ac.uk>8 Subject: Re: [Q] concatenating executable files is OK???+ Message-ID: <97inok$mn6@newton.cc.rl.ac.uk>o  Z "Didier Morandi" <Didier.Morandi@gmx.ch> wrote in message news:3A9CDA52.902A30F7@gmx.ch...
 > How come >a2 > ISLKP1> copy sfx_axp.exe,SPLAXP015.ZIP spell.exe > ISLKP1> r spellv
 > could work?   H The image header is read from the first file, and all the image sections@ mapped. There is nothing describing the second part so the image activator ignores it.I   ------------------------------    Date: 28 Feb 2001 15:58:49 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>f8 Subject: Re: [Q] concatenating executable files is OK???H Message-ID: <y4ofvmluqe.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  , "Richard Brodie" <R.Brodie@rl.ac.uk> writes:  4 > > ISLKP1> copy sfx_axp.exe,SPLAXP015.ZIP spell.exe > > ISLKP1> r spellm > > could work?sJ > The image header is read from the first file, and all the image sectionsB > mapped. There is nothing describing the second part so the image > activator ignores it.=  K ...and the unpack program then reads the file name of itself from the image N activator data in the CTL region (undocumented, but unlikely to be broken) andK the executable's length from the copy of the image header stashed there. ItoM then knows where the "real" data in the image begins. A secondary requirement=K is that, since images a fixed-512 files, as (presumably) are ZIP files, theFN append works properly no matter where in the last block the image really ends.   	Jan   ------------------------------  % Date: Wed, 28 Feb 2001 13:06:40 -0500c- From: "Richard D. Piccard" <piccard@ohio.edu>P8 Subject: Re: [Q] concatenating executable files is OK???( Message-ID: <3A9D3E2C.806B548A@ohio.edu>  P In the mid-1980s, in my former job, we had students who tried to hide executableO games from us by changing the file name and length using this trick.  It didn't  work,y   $ SEARCH   was our friend!l  #                                 RDPd     Jan Vorbrueggen wrote:  . > "Richard Brodie" <R.Brodie@rl.ac.uk> writes: >p6 > > > ISLKP1> copy sfx_axp.exe,SPLAXP015.ZIP spell.exe > > > ISLKP1> r spelli > > > could work?vL > > The image header is read from the first file, and all the image sectionsD > > mapped. There is nothing describing the second part so the image > > activator ignores it.  >uM > ...and the unpack program then reads the file name of itself from the imagenP > activator data in the CTL region (undocumented, but unlikely to be broken) andM > the executable's length from the copy of the image header stashed there. It O > then knows where the "real" data in the image begins. A secondary requirementuM > is that, since images a fixed-512 files, as (presumably) are ZIP files, theyP > append works properly no matter where in the last block the image really ends. > 
 >         Jany   --B ==================================================================B Dick Piccard                           Academic Technology ManagerB piccard@ohio.edu                                 Computer ServicesB http://oak.cats.ohiou.edu/~piccard/                Ohio University   ------------------------------  # Date: Wed, 28 Feb 2001 12:01:28 GMT=, From: Didier Morandi <Didier.Morandi@gmx.ch>& Subject: [Request] f$element delimiter& Message-ID: <3A9CE8AF.9680B88C@gmx.ch>  G I would like f$element to allow more than one character as a delimiter..  0 For example, I would like to be able to do this:   $ sh symb line line = "$ cmd=="@myproc"",8 $ proc_file = f$element(1,"==",line) - "@" - """" - """"   D.   ------------------------------  % Date: Wed, 28 Feb 2001 14:52:08 +0100c From: zessin@decus.deD* Subject: RE: [Request] f$element delimiter* Message-ID: <009F851E.7C941D7A.3@decus.de>   idier Morandi wrote:I > I would like f$element to allow more than one character as a delimiter.r >e2 > For example, I would like to be able to do this: >  > $ sh symb line > line = "$ cmd=="@myproc""a: > $ proc_file = f$element(1,"==",line) - "@" - """" - """"  / Don't know about DCL, but Python can do it now:r   >>> s='$ cmd=="@myproc"' >>> l = s.split ('==') >>> print ln ['$ cmd', '"@myproc"'] >>>s   The result is a list. Also:d   >>> s='$ cmd=="@myproc"  ===x' >>> l = s.split ('==') >>> print ls ['$ cmd', '"@myproc"  ', '=x'] >>>S  ( See: http://www.decus.de/~zessin/python/   --
 Uwe Zessin   ------------------------------   End of INFO-VAX 2001.118 ************************