1 INFO-VAX	Thu, 10 May 2001	Volume 2001 : Issue 260       Contents: Re: 7.2-2 availability. RE: ABS - backup service from NT to VMS device. Re: ABS - backup service from NT to VMS device. Re: ABS - backup service from NT to VMS device% Advanced Server Shares and ODS5 disks ) Re: Advanced Server Shares and ODS5 disks ) Re: Advanced Server Shares and ODS5 disks  Re: Bisynch  Re: Bisynch  Re: Bisynch  -- PROGRESS: Re: Compaq watch: radio ads and Inform - mostly good news.) Re: Conference: CETS-2001 - Conference CD ) Re: Conference: CETS-2001 - Conference CD ) Re: Conference: CETS-2001 - Conference CD ) Re: Conference: CETS-2001 - Conference CD ) Re: Conference: CETS-2001 - Conference CD ) Re: Conference: CETS-2001 - Conference CD ) Re: Conference: CETS-2001 - Conference CD ) Re: Conference: CETS-2001 - Conference CD " Cross Platform B/W VMS and WINDOWS- RE: email lists interesting to OpenVMS users? - Re: email lists interesting to OpenVMS users? # Re: Hobbyists: Free to a good home! * Re: How to repair corrupted ACCOUNTNG.DAT?* Re: How to repair corrupted ACCOUNTNG.DAT?1 Re: IBM announces MQSeries V5.1 for OpenVMS/Alpha 1 Re: IBM announces MQSeries V5.1 for OpenVMS/Alpha 1 Re: IBM announces MQSeries V5.1 for OpenVMS/Alpha 1 RE: IBM announces MQSeries V5.1 for OpenVMS/Alpha  Important news2 Re: mounting 2 disk with the same label in cluster2 Re: mounting 2 disk with the same label in cluster Re: OpenVMS with NetBeans 3.2 & RE: Oracle Parallel Server on OpenVMS? Print Accounting Re: Processor Affinity/ Re: Programming language usage on VMS systems ? / RE: Programming language usage on VMS systems ? / Re: Programming language usage on VMS systems ? / Re: Programming language usage on VMS systems ? / Re: Programming language usage on VMS systems ? / Re: Programming language usage on VMS systems ? % Re: Some more bits looking for a home  Re: SYSMAN and SMISERVER/ Re: The Internet, Tru64 and other unix variants / Re: The Internet, Tru64 and other unix variants / Re: The Internet, Tru64 and other unix variants / Re: The Internet, Tru64 and other unix variants / Re: The Internet, Tru64 and other unix variants / Re: The Internet, Tru64 and other unix variants / Re: The Internet, Tru64 and other unix variants / Re: The Internet, Tru64 and other unix variants / Re: The Internet, Tru64 and other unix variants / Re: The Internet, Tru64 and other unix variants / Re: The Internet, Tru64 and other unix variants / Re: The Internet, Tru64 and other unix variants 7 Re: VAX Just Stops - Not the Usual Suspects - follow up ' Re: VMS 7.2 BACKUP damaging system disk ' Re: VMS 7.2 BACKUP damaging system disk # Re: VMS 7.3 docs on OpenVMS webite?  vmsindex/query problemE Re: www.openvms.compaq.com doesn't display correctly in VMS Navigator E Re: www.openvms.compaq.com doesn't display correctly in VMS Navigator E Re: www.openvms.compaq.com doesn't display correctly in VMS Navigator E Re: www.openvms.compaq.com doesn't display correctly in VMS Navigator  [FUN] DCL minute of the evening # Re: [FUN] DCL minute of the evening # Re: [FUN] DCL minute of the evening # Re: [FUN] DCL minute of the evening # Re: [FUN] DCL minute of the evening # Re: [FUN] DCL minute of the evening # Re: [FUN] DCL minute of the evening   F ----------------------------------------------------------------------  # Date: Wed, 09 May 2001 22:43:14 GMT 2 From: "Zane H. Healy" <healyzh@shell1.aracnet.com> Subject: Re: 7.2-2 availability 5 Message-ID: <6KjK6.620$7O2.17522@typhoon.aracnet.com>   : Larry Kilgallen <Kilgallen@eisner.decus.org.nospam> wrote:I > Lots of people here will recommend 7.2-1 as being much better than 7.2. I > There are no such reports of field experience with 7.2-2.  If I were in * > your position, I would upgrade to 7.2-1.  F I tried to order the 7.2-1 update last week.  It's now only available < internally, however, I was able to order the 7.2-1H1 update.  L How different from 7.2-1 is 7.2-1H1?  I know there is support for additionalI hardware in -1H1.  I'm hoping there is no problem trying to install it on  older hardware.    			Zane    ------------------------------  % Date: Wed, 09 May 2001 17:04:03 -0400 : From: "Koska, John C. (LNG-MBC)" <John.C.Koska@bender.com>7 Subject: RE: ABS - backup service from NT to VMS device K Message-ID: <3D35AD137AAAD411A6BA0008C7B1B12D6DC01E@MBCALBEXC03.BENDER.COM>    > -----Original Message-----8 > From: Dave Gudewicz [mailto:david.gudewicz@abbott.com]' > Sent: Wednesday, May 09, 2001 9:38 AM  > To: Info-VAX@Mvb.Saic.Com 9 > Subject: Re: ABS - backup service from NT to VMS device  >  .  .    $ > Anybody else have comments on ABS? > 	 > Dave...  >   6 For my site, ABS v3.1A(196) is almost a fully working 3 solution.  We are still using a few workarounds to  6 get out our catalog reports printed to go off-site for7 disaster recovery purposes.  Otherwise, the reliability 7 and lack of human intervention has dropped considerably 7 with the past few builds.  We are backing up an OpenVMS  Alpha 7.2-1H1 system.   6 Granted, I have not done an complete disaster recovery4 yet with the tapes that have been cut from ABS.  But6 the savesets on the ABS backup tapes, seem to validate1 with SaveSet Manager, and I have done quite a few 4 file restores.  I expect to do an in-house disaster 6 recovery within a month or so.  So soonenough we will 7 find out, if the ABS backups can be restored.  (We are  5 still backing up system in parallel with a competing  6 product, that did past muster in an offsite DR test.)   7 I agree that the GUI is not perfect, but for my site it 6 was not that important since we wanted to automate the1 backups as much as possible.  We coded a few DCL  6 scripts for our Operators to init media, schedule and 9 run the ABS backups on-demand, and export the DR off-site 3 ABS backup tapes from a TL896. (Originally we media 6 duplicated our ABS backup tapes for shipment off-site,8 but we have since moved to an on-site backup along with 9 an off-site backup, and simply export the off-site backup  tapes to go off-site.)  ! Remaining issues for my site are:   2 1) Report generation when there is a lot of output8 2) MDMS INITIALIZE VOLUME /NOWAIT versus /WAIT and batch7    submitting a bunch of new media to init (ie: We find 4    init'ing tapes to be a bit difficult, if you want4    to init tapes across multiple tape drives, since 7    /NOWAIT is required.  And since /NOWAIT is required, 4    one has to be careful about issuing multiple MDMS+    INIT's or batch jobs which do the same.) 5 3) A few quirks in file restores, when multiple files     are to be restored.3 4) ABS REPORT SAVE /STORAGE not quite working right   8 All in all, ABS has come a long way in a short time from8 my perspective.  It looks like it will be a solid backup: product for OpenVMS very shortly, if not for most already.   :) jck
 John Koska Matthew Bender & Co., Inc. -"   a member of the LexisNexis Group
 1275 Broadway  Albany, NY  12204  USA  518-487-3255 JKoska@bender.com   * "I post personal opinion only, and all the* disclaimers one could imagine apply.  That( includes, I speak for myself only and my* views in no way represent my employer(s)."   ------------------------------  # Date: Thu, 10 May 2001 01:19:06 GMT ; From: "Curtis Rempel" <vmsguy.no.spam.here.please@home.com> 7 Subject: Re: ABS - backup service from NT to VMS device ; Message-ID: <e0mK6.17508$Hk4.304731@news1.rdc1.ab.home.com>   8 "Peter Weaver" <peter.weaver@stelco.ca> wrote in message. news:5YyJ6.235271$Z2.2516384@nnrp1.uunet.ca...H > "Curtis Rempel" <vmsguy.no.spam.here.please@home.com> wrote in message4 > news:etKI6.164$Hk4.13289@news1.rdc1.ab.home.com...    G > How do you handle your system drives? Aliases and Contiguous files on  > non-system drives?  ) We do image backups to another disk, i.e.   H $ BACKUP /IMAGE SYS$SYSDEVICE: $101$DGA409:[BACKUPS]SYSTEM.BCK /SAVE_SET /IGNORE=INTERLOCK   L for online images and of course the same thing without /IGNORE=INTERLOCK for standalone backups.   I Then the VMS NetBackup client is used to copy the backup disk contents (a * large RAID array) to the NetBackup server.  J If we need to restore, we pull the SYSTEM.BCK back to the backup array andE de-image it back to a different disk and boot from that disk instead.   B Yes, you do need another disk array for temporary storage, but theG enterprise benefits of NetBackup (multi-platform, parallel network card J throughput, drive multiplexing, integration with ORACLE and Exchange, etc)L far outweigh this minor inconvenience or "extra step".  And besides, disk is cheap these days.      > F > > Veritas support for the VMS client is out of the UK (Chertsey) and
 > Alan FayF > > is 110% top notch (thanks for all your extremely helpful and quick > responses  > > Alan!).  > E > Seconded, he is very quick to fix problems. I did get a bit nervous G > when I recieved a new image at 9:40 and noticed that the link time on C > it was 9:30 (after converting for time zones) - but the new image  > fixed the problem.   Doesn't surprise me at all :-)  * It has been my experience that in general:  : (Veritas NetBackup VMS support) = 1 / (Compaq ABS support)   ------------------------------  # Date: Thu, 10 May 2001 01:45:00 GMT ; From: "Curtis Rempel" <vmsguy.no.spam.here.please@home.com> 7 Subject: Re: ABS - backup service from NT to VMS device ; Message-ID: <womK6.17519$Hk4.307051@news1.rdc1.ab.home.com>   < "Dave Gudewicz" <david.gudewicz@abbott.com> wrote in message2 news:9dbh7n$rf9$1@fizban.fizban.pprd.abbott.com...L > We have spent some time with ABS, v3.newest, with VMS as the backup targetJ > only.  Seems OK but I have to agree with spotty support.  We also looked atI > VMS roadmaps and ABS *is* Compaq's solution with new stuff coming, etc.  ItH > would be nice if other GUIs were supported.  Don't mean to start a war here, J > but Motif is not "well known" and on the decline from what I understand.  J Supposedly there is a web interface in the works from what I have gathered8 through the CPQ rumour mill, but don't hold your breath.  L > We also looked at Veritas, and when we asked questions like "can it backupK > and restore a VMS system disk such that the restored disk is bootable" we J > got a blank stare as an answer.  Maybe we asked the wrong person.  Donno for  > sure.   F When we did our eval, I was at first a non-believer in NetBackup.  OurK Veritas rep was kind enough to arrange a conference call with Alan Fay, the H person responsible for the VMS client, in Chertsey, UK.  Our call lastedF approximately 1 hour and for the entire hour I wore my Mike Wallace 60J Minutes Backup hat and posed all the hard hitting questions.  Not only wasL Alan very personable, but he gave all the right answers and fully understoodJ precisely the issues and technical details of the matters being discussed.J After the call ended, I had reversed my viewpoint 180 degrees and all thatF was left was to do the eval to see it in action.  Veritas was somewhatF surprised that I had installed the server (on Tru64) and configured itI completely for the demo as they generally do that for the customer due to J the complexity involved.  When you've been through the grinder like I haveE for the past 3 years living, eating, sleeping, and breathing ABS (and I throwing up most of the time from it), doing the NetBackup install/config F was an absolute breeze.  The VMS client was a similar treat and workedI flawlessly.  Only one minor DNS issue tripped me up (I neglected to add a : reverse entry for the VMS box I was using for the client).  G I just re-read that last paragraph, you'd think I worked for Veritas or  something....  ;-)   > K > I believe TAR is the only backup format supported by Veritas.  Not saying : > thats bad, but that sure isn't native VMS backup format.  I No, and it doesn't really matter anyway, depending on how you look at it. L The VMS client just hands blocks of data over the wire to the server and theB server does whatever necessary to stream it across multiple drivesJ (multiplexing) in native UNIX tar format.  For the restore, the blocks areK handed back to the client which understands VMS storage (via the XQP).  So, K you can't just take a tape written using NetBackup and expect to load it in F a VMS tape drive and read it using VMS BACKUP.  This poses the obviousL question: what if my system won't boot so I can fire up the NetBackup clientK to do a restore?  Several options: keep periodic image save sets on another J disk array dedicated for this purpose and use standalone backup to restoreK from them.  Periodically copy those save sets to tape using VMS BACKUP to a L dedicated local DLT drive.  You could also use another VMS system to restoreJ the system disk to a temporary disk and then physically relocate that diskL (aren't snap-in StorageWorks disks great?) to the failed system.  In our SANK environment with many TB of storage, this works for us.  We also employ the K table top DLT approach, just in case.  If you're using 2 or 3 member shadow I sets or hardware based RAID5 for your system disk, a restore operation is L likely to be individual files due to operator error as opposed to a hardwareH failure.  In this case the VMS client restore operation is trivial.  ForK catastrophic hardware failures requiring an image restore, using one of the H other above approaches is the way to go.  Again, the VMS client is not aL "pure solution", however, the many enterprise benefits that NetBackup brings= to the table outweigh this apparent deficiency in my opinion.   L > Now this thread has started me thinking if ABS is/was the right way to go. > $ > Anybody else have comments on ABS? > 	 > Dave...  > : > "Peter Weaver" <peter.weaver@stelco.ca> wrote in message0 > news:5YyJ6.235271$Z2.2516384@nnrp1.uunet.ca...J > > "Curtis Rempel" <vmsguy.no.spam.here.please@home.com> wrote in message6 > > news:etKI6.164$Hk4.13289@news1.rdc1.ab.home.com... > > >...? > > > We have since killed ABS since Digital and now Compaq has  > > consistentlyJ > > > demonstrated that they cannot get their act together.  We have moved > > forward G > > > with Veritas NetBackup running on a Tru64 AlphaServer.  We tested  > > the VMS J > > > client support extremely rigorously and found no major deficiencies. > > I > > How do you handle your system drives? Aliases and Contiguous files on  > > non-system drives? > > H > > > Veritas support for the VMS client is out of the UK (Chertsey) and > > Alan FayH > > > is 110% top notch (thanks for all your extremely helpful and quick
 > > responses 
 > > > Alan!).  > > G > > Seconded, he is very quick to fix problems. I did get a bit nervous I > > when I recieved a new image at 9:40 and noticed that the link time on E > > it was 9:30 (after converting for time zones) - but the new image  > > fixed the problem. > >  > >  > > >...I > > > the software and ask for your money back - ALL of it.  Having grown  > > up with F > > > Digital and become accustomed to the stellar engineering efforts > > made by J > > > various product groups, it was an absolute nightmare to believe thatH > > > something like ABS was actually released in public as a product by > > Digital.A > > > I would have been personally embarrassed myself to have let  > > something like > > > ABS see the light of day.  > > H > > Compaq Analyze, DSN 2.2, DSN 3.0, ABS... How is this garbage gettingH > > through the door at Compaq? How do we get Compaq's attention so theyF > > know that we expect all the products on our VMS systems to work as > > well as the base OS itself?  > >  > >  >  >    ------------------------------  % Date: Wed, 09 May 2001 21:15:51 +0100 + From: Ray Swadling <ray@rgscomputing.co.uk> . Subject: Advanced Server Shares and ODS5 disks8 Message-ID: <609jftc8rbphcs3jru2ptm9kc25uffu9d8@4ax.com>   Hi,   D I'm experimenting with using Advanced server and ODS5 disks in my NT domain.   D I have a 18GB drive initialised as ODS5, and am trying to map shares onto the disk.    I'm getting errors as below.....   $ dir dka0:[000000]    Directory DKA0:[000000]   4 000000.DIR;1        BACKUP.SYS;1        BADBLK.SYS;1 BADLOG.SYS;14 BITMAP.SYS;1        CONTIN.SYS;1        CORIMG.SYS;1 INDEXF.SYS;19 SECURITY.SYS;1      vms.DIR;1           VMS_DATA01$.DIR;1  VMS_DATA01.DIR;1 VOLSET.SYS;1   Total of 13 files. $ adminV RGSCOMP\\RGS002> add share _share name: vms _path: dka0:[vms] - %PWRK-E-ERRADDSHARE, error adding share "VMS"o= -LM-E-NERR_UNKNOWNDEV, the device or directory does not existi   RGSCOMP\\RGS002>    D The disk dka0: is initialsed ODS5, and if I re-init the disk as ODS2; and create the directory and try to add the share it works.   C I can't find anything in the docs to suggest what the problem is soG can anyone shed any light??      Ray.   ------------------------------  # Date: Wed, 09 May 2001 23:59:23 GMT 1 From: bernie.milosavljevic@wesley.com.au (Bernie)n2 Subject: Re: Advanced Server Shares and ODS5 disks4 Message-ID: <vRkK6.21$911.3761@nsw.nnrp.telstra.net>  P About a year and a half ago I was experimenting with ODS5 disks using pathworks M and found problems creating shares. These were related to using root logical W. definitions but I can't remember the details.   O One thing to keep in mind with ODS5 is that it is difficult to go back to ODS2 eO because backups between the two are not compatible due to the long file names.  O The only way around it that I came up with were to create a pathworks share on  E an ODS2 disk, then do a share level copy, from one disk to the other.i   Hope this is useful. Bernie.BP In article <609jftc8rbphcs3jru2ptm9kc25uffu9d8@4ax.com>, ray@rgscomputing.co.uk  says...s >  >Hi, >cE >I'm experimenting with using Advanced server and ODS5 disks in my NT  >domain. > E >I have a 18GB drive initialised as ODS5, and am trying to map sharesc >onto the disk.e >r! >I'm getting errors as below.....t >  >$ dir dka0:[000000] >a >Directory DKA0:[000000] >e5 >000000.DIR;1        BACKUP.SYS;1        BADBLK.SYS;1g
 >BADLOG.SYS;1e5 >BITMAP.SYS;1        CONTIN.SYS;1        CORIMG.SYS;1e
 >INDEXF.SYS;1 : >SECURITY.SYS;1      vms.DIR;1           VMS_DATA01$.DIR;1 >VMS_DATA01.DIR;1g
 >VOLSET.SYS;1e >Y >Total of 13 files.e >$ admin >RGSCOMP\\RGS002> add sharer >_share name: vmse >_path: dka0:[vms]. >%PWRK-E-ERRADDSHARE, error adding share "VMS"> >-LM-E-NERR_UNKNOWNDEV, the device or directory does not exist >  >RGSCOMP\\RGS002>  >6 >nE >The disk dka0: is initialsed ODS5, and if I re-init the disk as ODS2 < >and create the directory and try to add the share it works. >nD >I can't find anything in the docs to suggest what the problem is so >can anyone shed any light?? >  >6 >Ray.t   ------------------------------  # Date: Thu, 10 May 2001 00:18:26 GMT . From: "Alphaman" <alphaman64@nixspam-home.com>2 Subject: Re: Advanced Server Shares and ODS5 disks9 Message-ID: <m7lK6.1622$I5.510606@news1.rdc1.tn.home.com>A  H Might you need "quotes" around that lower case directory name?  Note the2 error says it can't find the dir "VMS", not "vms".   Aarone --> Aaron Sakovich  http://members.home.net/sakovich/alphaman.html> Make April 15 just another day:        http://www.fairtax.org/H "The supersonic boom should hit you in just a few seconds." (Apollo 440)  6 Ray Swadling <ray@rgscomputing.co.uk> wrote in message2 news:609jftc8rbphcs3jru2ptm9kc25uffu9d8@4ax.com... > Hi,C >iF > I'm experimenting with using Advanced server and ODS5 disks in my NT	 > domain.  : ; > SECURITY.SYS;1      vms.DIR;1           VMS_DATA01$.DIR;1  :u	 > $ adminv > RGSCOMP\\RGS002> add share > _share name: vms > _path: dka0:[vms]-/ > %PWRK-E-ERRADDSHARE, error adding share "VMS" ? > -LM-E-NERR_UNKNOWNDEV, the device or directory does not exist7 :2 > Ray.   ------------------------------  # Date: Wed, 09 May 2001 17:54:12 GMTi1 From: "Mark D. Jilson" <jilly@clarityconnect.com>  Subject: Re: Bisynch2 Message-ID: <3AF9844C.968D95EC@clarityconnect.com>   John, E 	AT this point you need to call, on the phone, your local CSC and askoF for the MOD.  DSNlink won't cut it, Sales won't cut it, local FS won'tG cut it, only a telephone call to the CSC will do and then don't hang up C until you are satisfied that your issue is in the proper hands.  If ? you'll email me the SRQ of the DSNlink call I will research it.p   john nixon wrote:e > J > I have tried to log a call over DSN, but they claim I don't have suport.M > When I pointed out my contract number and the four line item numbers, whereGK > I am paying over $2,000 /year, they said they (DSN%DSNLINK) was the wrongo > group. >  > DSN%MOD never answered.1 > L > I am still waiting to hear back from DSN%CAD.  (I just sent them a message > today) > 2 > Our contract administrator is "looking into it". > M > I have heard back from my Compaq Sales Rep(but it was an automated, "out of-K > office" message) and his tech guy, and from my Compaq Authorized Resellero1 > and his tech guy.  They all said to log a call.A > E > By the "local service rep",  are you referring to our Field ServiceeL > Engineer?  I hadn't thouht of that because this is a software problem.  He' > is pretty sharp.  I will contact him.  > ? > "Sue Skonetski" <susan.skonetski@compaq.com> wrote in message , > news:0JbK6.45$fi2.1814@news.cpqcorp.net... > >p > >  > > Just a question.& > > Did you log a call for this issue? > >sJ > > For folks that may be new to the process.  My apoligies if everyone is > > already aware of this. > >eN > > You are much better off contacting your local service rep vs the sales rep > > (Digital or otherwise).e > >s= > > How the problem reporting structure works with-in Compaq.i > >tM > > Customer with a services contract logs a call. Once the problem report isrM > > logged and if the field can not answer it the problem will end up back at 8 > > VMS engineering in the Sustaining Engineering group. > >aK > > One of the good things about this procedure is that once the sustainingiJ > > engineering group fixes the bug it is then put into future releases of > VMS.' > > So the procedure helps VMS quality.  > >R > > Sue  > >  > >p5 > > "john nixon" <jnixon@cfl.rr.com> wrote in messageo= > > news:afDJ6.170899$fs3.28745768@typhoon.tampabay.rr.com...o< > > > If you remember the Bisynch thread a little while ago, > > >eK > > > Well, we bought the PBXDB for our Alphaserver.  We have a license.  I  > haveN > > > even lined up a source for some of the 8 port cards so that we only have > > toL > > > buy 4 Alphas to get my 26 ports instead of seven Alphas.  Now, we load > the C > > > software, and it rather rudely tells us that we have "BADVMS"sL > > > (%AXPBSCPTP-E-BADVMS,  this kit can only be installed on OpenVMS Alpha > > V6.2 > > > or V7.1) > > >eM > > > It says to contact your local Digital sales representative  (not Compaq.J > > > sales rep,but "Digital".   It was probably Bob Palmer that wrote the
 > > BADVMS > > > part). > > >GL > > > I wonder what I do now.  This was a big part of our VAX to Alpha port. > > Now L > > > that we have spent a rather large sum of money, I have to go tell them > > thatK > > > I did not research this well enough.   I sure hope I like living in a." > > > Sun/Solaris+Microsoft world. > > >5 > > >a > > >a > > >  > >  > >  > >-   -- 5D 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, 09 May 2001 20:49:10 +0100p+ From: "antonio.carlini" <arcarlini@iee.org>u Subject: Re: Bisynch' Message-ID: <3AF99F36.FC7C0F67@iee.org>-   Jan Vorbrueggen wrote: > / > "antonio.carlini" <arcarlini@iee.org> writes:e > F > > > I do not think that hacking VMSINSTAL to force it is the answer.L > > I recall that there is some way of forcing PCSI to install regardless ofI > > its opinion regarding the wisdom of so doing, but I don't recall how.n > L > AFAIR, VMSINSTAL does not give you the option if the callback is so coded.    + I still don't recall how to do it for PCSI.A   For VMSINSTAL it's a breeze.  : Run it with OPTIONS S (or something very similar to that).  : It will then unpack the .A saveset and SPAWN a subprocess.7 At this stage you can do what you like. I would suggestkA looking for KITINSTAL.COM in VMI$KWD (from memory, so hack arounda8 a little if it's somewhere else). Now edit it and change  the appropriate V7.1 for a V7.2.  7 Of course, once you do this you are 100% *on your own*.b5 Plenty of kits have these restrictions in to stop your1 crashing your system. (For example, plenty of oldi7 kits knew when they shipped on V5.5-2 that the upcomingk8 *major* exec reorg in V6 would fry them so the check was an important safety feature.)u  1 Personally I'd phone the relevant support centre  3 and keep escalating to people's managers and faxingi/ the invoice until someone takes notice and logsC, the call. Once it's logged you should get a * pretty rapid response (and if you're lucky& it won't be "oops, the SPD is wrong!".   Antonio    -- e   ---------------b- Antonio Carlini             arcarlini@iee.orge   ------------------------------  # Date: Thu, 10 May 2001 00:32:49 GMTa& From: "john nixon" <jnixon@cfl.rr.com>! Subject: Re: Bisynch  -- PROGRESS ? Message-ID: <RklK6.184576$fs3.31272454@typhoon.tampabay.rr.com>t  L I want to thank everyone that helped or tried to help me with my 2780/3780pe install on Alpha VMS 7.2-1  L I was finally able to get my support contract fixed so I could log a ticket.F Then it was a breeze to find out that there was a hidden update (V2.1)H locked away in a vault somewhere, which they were able to send me.  ThisJ version now installs on my V7.2-1 Alpha.  Now I have to round up the wanddL kit (which we should have in the office on CD).   Alsok, I am trying to findJ out if this will work on VMS 7.3.    As far as I know, Compaq has not madeF that announcement yet, but I would think that someone would know since' engineering has already shipped VMS 7.3c      < "Mark D. Jilson" <jilly@clarityconnect.com> wrote in message, news:3AF9844C.968D95EC@clarityconnect.com... > John,nF > AT this point you need to call, on the phone, your local CSC and askH > for the MOD.  DSNlink won't cut it, Sales won't cut it, local FS won'tI > cut it, only a telephone call to the CSC will do and then don't hang upaE > until you are satisfied that your issue is in the proper hands.  IfhA > you'll email me the SRQ of the DSNlink call I will research it.  >w > john nixon wrote:h > >nL > > I have tried to log a call over DSN, but they claim I don't have suport.I > > When I pointed out my contract number and the four line item numbers,e whereaG > > I am paying over $2,000 /year, they said they (DSN%DSNLINK) was they wrongO
 > > group. > >p > > DSN%MOD never answered.e > >rF > > I am still waiting to hear back from DSN%CAD.  (I just sent them a messagen
 > > today) > > 4 > > Our contract administrator is "looking into it". > >nL > > I have heard back from my Compaq Sales Rep(but it was an automated, "out ofD > > office" message) and his tech guy, and from my Compaq Authorized Reseller3 > > and his tech guy.  They all said to log a call.i > >aG > > By the "local service rep",  are you referring to our Field ServiceoJ > > Engineer?  I hadn't thouht of that because this is a software problem. He) > > is pretty sharp.  I will contact him.l > > A > > "Sue Skonetski" <susan.skonetski@compaq.com> wrote in messagea. > > news:0JbK6.45$fi2.1814@news.cpqcorp.net... > > >h > > >u > > > Just a question.( > > > Did you log a call for this issue? > > >aL > > > For folks that may be new to the process.  My apoligies if everyone is > > > already aware of this. > > > L > > > You are much better off contacting your local service rep vs the sales rep  > > > (Digital or otherwise).g > > >o? > > > How the problem reporting structure works with-in Compaq.  > > > L > > > Customer with a services contract logs a call. Once the problem report isL > > > logged and if the field can not answer it the problem will end up back at: > > > VMS engineering in the Sustaining Engineering group. > > >lB > > > One of the good things about this procedure is that once the
 sustainingL > > > engineering group fixes the bug it is then put into future releases of > > VMS.) > > > So the procedure helps VMS quality.e > > > 	 > > > Sues > > >M > > >g7 > > > "john nixon" <jnixon@cfl.rr.com> wrote in messagen? > > > news:afDJ6.170899$fs3.28745768@typhoon.tampabay.rr.com...r> > > > > If you remember the Bisynch thread a little while ago, > > > >sJ > > > > Well, we bought the PBXDB for our Alphaserver.  We have a license. I  > > haveK > > > > even lined up a source for some of the 8 port cards so that we onlys have > > > toI > > > > buy 4 Alphas to get my 26 ports instead of seven Alphas.  Now, wep load > > theKE > > > > software, and it rather rudely tells us that we have "BADVMS"sH > > > > (%AXPBSCPTP-E-BADVMS,  this kit can only be installed on OpenVMS Alphat
 > > > V6.2 > > > > or V7.1) > > > >dH > > > > It says to contact your local Digital sales representative  (not CompaqL > > > > sales rep,but "Digital".   It was probably Bob Palmer that wrote the > > > BADVMS > > > > part). > > > >oH > > > > I wonder what I do now.  This was a big part of our VAX to Alpha port.i	 > > > NowiI > > > > that we have spent a rather large sum of money, I have to go telln them
 > > > thatK > > > > I did not research this well enough.   I sure hope I like living inq ai$ > > > > Sun/Solaris+Microsoft world. > > > >b > > > >a > > > >h > > > >g > > >i > > >V > > >n >s > --F > Jilly - Working from Home in the Chemung River Valley - Lockwood, NY/ > - jilly@clarityconnect.com - Brett Bodine fani- > - Mark.Jilson@Compaq.com - since 1975 or sot- > - http://www.jilly.baka.com               -o >.   ------------------------------   Date: 9 May 2001 16:24:24 -0500t9 From: kaplow_r@eisner.encompasserve.org.mars (Bob Kaplow)sC Subject: Re: Compaq watch: radio ads and Inform - mostly good news.u3 Message-ID: <SFXKl+reth6V@eisner.encompasserve.org>   \ In article <874rvjpvzt.fsf@prep.synonet.com>, Paul Repacholi <prep@prep.synonet.com> writes:) > Alan Greig <a.greig@virgin.net> writes:e > B >> One place where VMS prominently *does not* feature is Winkler'sC >> article on "Where Compaq is Headed" (both editions - abbreviated6@ >> version in European edition). He talks about iPAQs, Proliant,E >> Windows, Alpha, Tru-64 but not OpenVMS (or Linux for that matter).u > D > I wonder if he is one of the group that help replace T64 with IIS?E > "We love our OSs so much we are ripping them out and replacing theme > with Billyshit."  I I've received SPAM Email from part of Compaq that suggested exactly that!a7 Note that it comes from notted SPAM site boomerang.com:d  8 Subject: FREE Shipping from Bequipt, The Accessory Store% From: Compaq <compaq@compaq-news.com>9) Reply-to: Compaq <compaq@compaq-news.com> + Errors-To: FK1503S9@concierge.boomerang.com % Date: Thu, 11 Jan 2001 11:01:25 -0800b   ...   I Compaq recommends Microsoft=AE Windows=AE 2000 Professional for business.t   ...n    ' I will NEVER buy anything from Bequipt!o   ------------------------------   Date: 9 May 2001 13:54:54 -0500 9 From: kaplow_r@eisner.encompasserve.org.mars (Bob Kaplow)i2 Subject: Re: Conference: CETS-2001 - Conference CD3 Message-ID: <9ALwQDoScynC@eisner.encompasserve.org>   _ In article <yV9K6.311$372.382868@typhoon2.ba-dsg.net>, "Jeff Killeen" <Jeff@IDM-IO.com> writes:i- > This is an FYI to all potential speakers..., > F > This year the conference CD with the session notes will be publishedM >>>>BEFORE<<< the conference.  That means there will be a requirement to havedN > your session materials submitted about mid August if you want them publishedJ > on the CD.  Any materials supplied after that point in time will only be > available via the web site.a > D > The primary goal here is to insure attendees with laptops have theG > presentations.  The secondary goal is to insure all presentations areeJ > available to attendees via the web right after the session is presented.  ) Makes me wish I had a laptop with a CD...   K Bit this is a good idea. One of the things I missed from the paper handoutsiH of years past was having them during the presentation. You never new forI sure if the speaker would get his slides on the CD. And you had to wait 2 I months for that really important presentation that you wanted to refereneM the week you got back.  K My only other comment would be to have a way to disseminate whatever eithersI didn't make the cut, or was updated between the cutoff and the symposium.n  K BTW, if it takes 2+ months to get the after-the-conference CD produced, how 9 is it going to happen in 2-3 weeks before the conference?a  I And for those of us who just can't go (I'm planning on it, and hoping for L it,  but never know what the corporate budget will be like by then), do makeJ the disks orderable as they have been in the past. I've at least been ableJ to talk my company into a small hunk of cash when they won't shell out for the whole week.e   ------------------------------  $ Date: Wed, 9 May 2001 14:43:50 -04002 From: "Sue Skonetski" <susan.skonetski@compaq.com>2 Subject: Re: Conference: CETS-2001 - Conference CD0 Message-ID: <pdgK6.66$fi2.2125@news.cpqcorp.net>  J I just spoke to the Web Master (Warren Sander) and the doc's will be up on the web site Saturday.  
 Warm Regards,n   Suei  - <steven.reece@quintiles.com> wrote in message2B news:OFDCE6B449.92DDE5CA-ON80256A47.005BA58E@qedi.quintiles.com... >f > L > If only it was a VMS CD.  Then someone might be able to justify a GS320 orJ > GS160 (for expandability) with lots of PCI cages with SCSI adapter cards in- > them connected to (unsupported) CD writers.c, > Sledgehammers and nuts spring to mind..... > :-)r >o > Hoff commented : > >>> I >   This distribution decision might provide the impetus to serve various K >   of the Compaq OpenVMS CETS2001 technical presentations via the website. J >   (Or to bring along the hardware necessary to replicate CD-R media. :-) > <<<r >a   ------------------------------  # Date: Wed, 09 May 2001 20:09:19 GMT.6 From: "Andy Bustamante" <A_C_Bustamante@earthlink.net>2 Subject: Re: Conference: CETS-2001 - Conference CDD Message-ID: <PthK6.2648$Ak7.257433@newsread1.prod.itd.earthlink.net>   Just to stir things up:Y  J Just how many people feel the need to carry laptops around?  My departmentJ shares one for travel (PowerTerm & e-mail) but I've never felt the need to1 lug it out of the hotel room and around the show.r  J Paper handouts allow for immediate notetaking without worrying about powerL problems.  If you want to service laptop users, how about network drops withI a DHCP server and an intranet web site (or read-only shared folder) wheres presentations can be staged.  G Dstributing current presentations should be the priority, waiting a few ' weeks for CD has never been a hardship.r   -- Andy Bustamanten Remove the ASCII 95s to replyR/ Jeff Killeen <Jeff@IDM-IO.com> wrote in message-0 news:yV9K6.311$372.382868@typhoon2.ba-dsg.net...      - > This is an FYI to all potential speakers...r >oF > This year the conference CD with the session notes will be publishedI > >>>BEFORE<<< the conference.  That means there will be a requirement tol haveD > your session materials submitted about mid August if you want them	 publishedsJ > on the CD.  Any materials supplied after that point in time will only be > available via the web site.e >eD > The primary goal here is to insure attendees with laptops have theG > presentations.  The secondary goal is to insure all presentations are J > available to attendees via the web right after the session is presented. > >v >  >C   ------------------------------  # Date: Wed, 09 May 2001 22:06:29 GMT-= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)E2 Subject: Re: Conference: CETS-2001 - Conference CD0 Message-ID: <009FBC3B.3BFAF269@SendSpamHere.ORG>  } In article <PthK6.2648$Ak7.257433@newsread1.prod.itd.earthlink.net>, "Andy Bustamante" <A_C_Bustamante@earthlink.net> writes:5 >Just to stir things up: >nK >Just how many people feel the need to carry laptops around?  My departmentoK >shares one for travel (PowerTerm & e-mail) but I've never felt the need to 2 >lug it out of the hotel room and around the show.  8 Another ASSumption that everybody has and uses a laptop.    --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM             wO city, n., 1. a place where trees are cut down and streets are named after them.p   ------------------------------   Date: 9 May 2001 18:10:28 -0500n9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)t2 Subject: Re: Conference: CETS-2001 - Conference CD3 Message-ID: <Vg0reJIf4HVg@eisner.encompasserve.org>e  } In article <PthK6.2648$Ak7.257433@newsread1.prod.itd.earthlink.net>, "Andy Bustamante" <A_C_Bustamante@earthlink.net> writes:e > Just to stir things up:S > L > Just how many people feel the need to carry laptops around?  My departmentL > shares one for travel (PowerTerm & e-mail) but I've never felt the need to3 > lug it out of the hotel room and around the show.y  H I have full access to a laptop purchased for developing PCMCIA software., It stays at home base and never leaves town.  L > Paper handouts allow for immediate notetaking without worrying about power > problems.    Exactly !!!s  > Furthermore, with paper there need be no quarrel over formats.   ------------------------------  # Date: Wed, 09 May 2001 22:30:33 GMTm& From: "Jeff Killeen" <Jeff@IDM-IO.com>2 Subject: Re: Conference: CETS-2001 - Conference CD6 Message-ID: <dyjK6.581$372.764112@typhoon2.ba-dsg.net>  H I will check but I believe the date will be 2 weeks before the Symposium   --        0 Jeff Killeen - www.Killeen.cc (All contact info)E =====================================================================   ? "Hoff Hoffman" <hoffman@xdelta.zko.dec.nospam> wrote in messagea* news:GaeK6.57$fi2.1893@news.cpqcorp.net...J > In article <9dbfku$r7a$1@fizban.fizban.pprd.abbott.com>, "Dave Gudewicz"# <david.gudewicz@abbott.com> writes:iK > :Thanks Jeff.  I think having the CD created *before* the conference is a  > :great idea. >hL >   Now for the downside of a mid-August deadline for a mid-September event. >oD >   Please do not expect to see current versions of my presentationsH >   included on the conference CD, nor the sessions of a number of otherH >   folks in OpenVMS Engineering.  While having the CD media immediatelyK >   available to the conference attendees is a certainly and clearly usefulaG >   -- and while I will certainly work to see the latest versions of mynG >   presentation materials included, the lead time can run afoul of thee >   nature of the business.e >AH >   Why?  Well, I am perpetually updating my slides.  For example, threeF >   of the technical presentations I usually perpetrate are being usedF >   this week in Lisbon and will be used again next week in London andH >   Zurich, and the presentation that I expect most folks are interestedJ >   in -- the technical update -- was updated twice in the past two weeks.H >   It may well be updated again later this week or early next.  (I haveI >   been known to update the presentation immediately prior to presentinge> >   it -- to incorporate important information just received.) >rE >   For more the static presentations, this pre-emptive approach will C >   clearly work just fine, and several of my presentations and the B >   presentations of other engineers would obviously qualify here. >.E >   As an obviously more costly alternative, I do know of conferences C >   that create the media at the end of the week of the conference.0 >pI >   This distribution decision might provide the impetus to serve variousgK >   of the Compaq OpenVMS CETS2001 technical presentations via the website.cJ >   (Or to bring along the hardware necessary to replicate CD-R media. :-) >I, >  --------------------------- pure personal# opinion ---------------------------S1 >    Hoff (Stephen) Hoffman   OpenVMS Engineeringn hoffman#xdelta.zko.dec.com >    ------------------------------  # Date: Wed, 09 May 2001 22:34:21 GMTf& From: "Jeff Killeen" <Jeff@IDM-IO.com>2 Subject: Re: Conference: CETS-2001 - Conference CD7 Message-ID: <NBjK6.2263$v66.803143@typhoon1.ba-dsg.net>e  L > Paper handouts allow for immediate notetaking without worrying about power > problems.t  F For 2002 would you be willing to pay an extra $50-75 for a paper copy?   --        . Jeff Killeen - Encompass CETS Business ManagerE =====================================================================    ------------------------------   Date: 9 May 2001 19:32:45 -0500 9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) 2 Subject: Re: Conference: CETS-2001 - Conference CD3 Message-ID: <AdgcgMLAZSR1@eisner.encompasserve.org>x  ` In article <NBjK6.2263$v66.803143@typhoon1.ba-dsg.net>, "Jeff Killeen" <Jeff@IDM-IO.com> writes:M >> Paper handouts allow for immediate notetaking without worrying about powerf >> problems. > H > For 2002 would you be willing to pay an extra $50-75 for a paper copy?  : Certainly, and I always did until you stopped offering it.  3 Of course I recognize that others feel differently.    ------------------------------  % Date: Thu, 10 May 2001 10:38:16 +0530-- From: "Prashanth" <Prashanth.SN@in.bosch.com>1+ Subject: Cross Platform B/W VMS and WINDOWSl6 Message-ID: <9dd7ka$7us$1@proxy.fe.internet.bosch.com>  H I have to develop an application Which interfaces b/w the VMS server and Windows workstation,K can any one tell me as to how to start with it, i have thought about Clienta Server technologysL but how is that i have to handle on the VMS server for servicing the windows client requests,) 1. Will the OS service my cleint requestsuE 2. Will the Ansi C compliers work there or i am not very clear pleasel suggest something.! 3. Should i use TCP/IP or Telnet.dK   the services i want to do the server may be compiling and linking and oned
 more thingK  if i open a file from the VMS  machine, how should i format it here on thec windows so as to view it properly.e   Thanks in Advance.   ------------------------------  % Date: Wed, 09 May 2001 11:37:56 -0700r! From: Tom Linden <tom@kednos.com>e6 Subject: RE: email lists interesting to OpenVMS users?9 Message-ID: <CIEJLCMNHNNDLLOOGNJIGEHCCIAA.tom@kednos.com>C  H Don't know if this is of interest, but some people still ues PL/I on VMS  9 PL1 (language) discussions [PL1-L@LISTSERV.DARTMOUTH.EDU]d   Tomo   > -----Original Message-----; > From: Hoff Hoffman [mailto:hoffman@xdelta.zko.dec.nospam]l( > Sent: Wednesday, May 09, 2001 10:01 AM > To: Info-VAX@Mvb.Saic.Comn4 > Subject: email lists interesting to OpenVMS users? >  >  > K >   I am looking to add a section containing various available email lists  J >   that may be of interest to OpenVMS users into the next edition of the  >   OpenVMS FAQ. > D >   I have addresses of some email lists (list servers) for various  > interest mI >   areas of OpenVMS users (eg: OSU webserver, vmsperl, purveyor, OpenVMS D >   ECO kit notifications), and have heard of other lists (eg: Rdb). > I >   Please post or email pointers to lists, and I will add them.  Thanks!o > 3 >  ---------------------------- #include <rtfaq.h> y > ----------------------------- 6 >       For additional, please see the OpenVMS FAQ --  > www.openvms.compaq.com    5 >  --------------------------- pure personal opinion r > ---------------------------u4 >    Hoff (Stephen) Hoffman   OpenVMS Engineering    > hoffman#xdelta.zko.dec.com >    ------------------------------  # Date: Wed, 09 May 2001 18:58:52 GMTs- From: goathunter@goatley.com (Hunter Goatley) 6 Subject: Re: email lists interesting to OpenVMS users?0 Message-ID: <3af99195.86303097@news.process.com>  N On Wed, 09 May 2001 17:01:07 GMT, hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) wrote:  L >  I have addresses of some email lists (list servers) for various interest H >  areas of OpenVMS users (eg: OSU webserver, vmsperl, purveyor, OpenVMSC >  ECO kit notifications), and have heard of other lists (eg: Rdb).e >c Here are the ones I run:  N   FSupdate@goatley.com                  VMS Freeware archive announcement listB   VMSnet-Internals@goatley.com		A two-way echo of vmsnet.internalsC   Alpha-IDS@goatley.com (*)             Alpha Internals discussionsr9   BLISSters@goatley.com (*)             BLISS discussionsuL   Info-MultiNet@process.com             MultiNet mailing list (news gateway)K   Info-TCPware@process.com              TCPware mailing list (news gateway)hH   Info-PMDF@process.com                 PMDF mailing list (news gateway)L   CHARON-VAX-Users@process.com          Discussion list for SRI's CHARON-VAXN   Info-Zip@wku.edu                      Info-Zip's Zip & UnZip discussion list  I To subscribe to any of the lists, the command SUBSCRIBE should be sent inf; the body of a mail message to the -request address like so:        list-request@hoste   For example:  !     BLISSters-request@goatley.comd  N (*) Alpha-IDS and BLISSters have been dormant for some time, but they do still
     exist.  J If anyone has any question about any of the lists, I can be reached at the address below.   Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/n9 goathunter@goatley.com     http://www.goatley.com/hunter/    ------------------------------  # Date: Thu, 10 May 2001 00:08:28 GMT . From: "Alphaman" <alphaman64@nixspam-home.com>, Subject: Re: Hobbyists: Free to a good home!9 Message-ID: <0_kK6.1608$I5.503894@news1.rdc1.tn.home.com>y  I In case you've been wondering why your email is bouncing -- the system iseE gone, claimed first thing this AM.  Sorry I couldn't post this noticef earlier.   Aaron  --> Aaron Sakovich  http://members.home.net/sakovich/alphaman.html> Make April 15 just another day:        http://www.fairtax.org/H "The supersonic boom should hit you in just a few seconds." (Apollo 440)  7 Alphaman <alphaman64@nixspam-home.com> wrote in message 6 news:lh2K6.81190$U4.17775056@news1.rdc1.tn.home.com...& > Hobbyists!  Your attention please... >lJ > Normally we sell these systems to system resellers/used dealers, but theJ > return is getting less and less.  So now, we'd like to offer this system up > to a hobbyist.   ------------------------------  % Date: Thu, 10 May 2001 00:26:29 +0100y4 From: John Laird <john@laird-towers.freeserve.co.uk>3 Subject: Re: How to repair corrupted ACCOUNTNG.DAT?s8 Message-ID: <7qjjft0f6pqv3bp4adm5sr8n5mf85etbdp@4ax.com>  A On Wed, 9 May 2001 13:59:03 +0000 (UTC), david20@alpha1.mdx.ac.ukn (D.Webb) wrote:   W >In article <3AF8D628.778D2512@ui.urban.org>, Jim Becker <jbecker@ui.urban.org> writes: F >>A crude hack that has worked for me before went something like this: >>  >>$ open /read bad accountng.bad# >>$ open /write good accountng.goodi
 >>$ again:& >>$ read /error=again /end=eof bad rec >>$ write good rec >>$ goto again >>$ eof:
 >>$ close badf >>$ close good >>E >>You lose data right around the corrupted spot, but at least you getS >>records after it.o >> >m	 >Why not m > 3 >ACCOUNT/BINARY/OUT=NEWACCOUNTNG.DAT  ACCOUNTNG.DAT:  H All of these approaches will recover data up to the point of corruption.E To expect RMS to continue and somehow "resynchronise" with the actualoG record structure later on, is asking for a minor miracle.  More likely,-H any application will throw up at the huge "record" encountered, and fromD then on will be interpreting garbage and good data as equal garbage.F You might luckily "land" on the start of a real record further on, but: at the risk of skipping large amounts of recoverable data.  H It's easy to patch the file at the last good block to prevent reading onE into the corrupted area, and those blocks can also be skipped (all bysE writing record lengths of -1 which indicates "no more records in this F block"), but the hard part is getting to the first decent record after- that, which will be part-way through a block.   G The only sensible way out is to write a small program.  It can walk the H file up to the point of corruption, and then use RFAs to position to the8 next uncorrupted record, and read on to the end of file.  E Before doing any of this, an ANAL/DISK is recommended - watch for anyaE multiply allocated blocks (ie. parts of the disk "owned" by more than 
 one file).    C If the contents of the file are regarded as worth keeping, considerp- creating a new file regularly (say midnight).u     	John  -- t
 John Laird   ------------------------------  % Date: Thu, 10 May 2001 01:50:04 -0400 ' From: Jim Becker <jbecker@ui.urban.org>a3 Subject: Re: How to repair corrupted ACCOUNTNG.DAT?k, Message-ID: <3AFA2C0C.DDB2D959@ui.urban.org>   John Laird wrote:i > C > On Wed, 9 May 2001 13:59:03 +0000 (UTC), david20@alpha1.mdx.ac.ukh > (D.Webb) wrote:m > Y > >In article <3AF8D628.778D2512@ui.urban.org>, Jim Becker <jbecker@ui.urban.org> writes: H > >>A crude hack that has worked for me before went something like this: > >>" > >>$ open /read bad accountng.bad% > >>$ open /write good accountng.goodt > >>$ again:( > >>$ read /error=again /end=eof bad rec > >>$ write good rec > >>$ goto again
 > >>$ eof: > >>$ close badb > >>$ close good > >>G > >>You lose data right around the corrupted spot, but at least you get  > >>records after it.- > >> > > 
 > >Why not > >25 > >ACCOUNT/BINARY/OUT=NEWACCOUNTNG.DAT  ACCOUNTNG.DATs > J > All of these approaches will recover data up to the point of corruption.G > To expect RMS to continue and somehow "resynchronise" with the actualaI > record structure later on, is asking for a minor miracle.  More likely,gJ > any application will throw up at the huge "record" encountered, and fromF > then on will be interpreting garbage and good data as equal garbage.H > You might luckily "land" on the start of a real record further on, but< > at the risk of skipping large amounts of recoverable data.  C Then minor miracles happen. More than once, I've seen valid recordse@ recovered _after_ the bad spot using the DCL read/write approachB outlined above. I rarely run across bad accounting files, but IIRCB that DCL read/write approach has recovered additional records eachF time I've used it. There is indeed loss of data in the neighborhood ofE the corrupted spot, as I mentioned. Whether the loss is more than can-- be afforded is a site-specific judgment call.   D I'm afraid I don't have a bad accounting file handy for the purposes of demonstrating.-  J > It's easy to patch the file at the last good block to prevent reading onG > into the corrupted area, and those blocks can also be skipped (all bytG > writing record lengths of -1 which indicates "no more records in this&H > block"), but the hard part is getting to the first decent record after/ > that, which will be part-way through a block.m > I > The only sensible way out is to write a small program.  It can walk thecJ > file up to the point of corruption, and then use RFAs to position to the: > next uncorrupted record, and read on to the end of file. > G > Before doing any of this, an ANAL/DISK is recommended - watch for anyaG > multiply allocated blocks (ie. parts of the disk "owned" by more thana > one file). > E > If the contents of the file are regarded as worth keeping, considere/ > creating a new file regularly (say midnight).  >  >         John > -- > John Laird   --
 Jim Becker+ The Urban Institute (http://www.urban.org/)n' Encompass (http://www.encompassus.org/)e. ESILUG (http://encompasserve.org/lugs/esilug/)   ------------------------------  $ Date: Wed, 9 May 2001 20:55:24 +0100, From: "Richard Maher" <maher_rj@hotmail.c*m>: Subject: Re: IBM announces MQSeries V5.1 for OpenVMS/Alpha1 Message-ID: <9dc79e$kq4$1@uranium.btinternet.com>r   Hi,    Credit where credit's due:-a  8 Well done to Sue (and all the others involved)!!!!!!!!!!  
 This is huge!r   Regards Richard Mahere  ; Sue Skonetski <susan.skonetski@compaq.com> wrote in message * news:zlbK6.43$fi2.1807@news.cpqcorp.net... >e >0 >a >0 > Dear Newsgroup >mC > Over the last months I've seen a number of questions on this listR
 concerning( > IBM's support for MQSeries on OpenVMS. >9> > IBM has just announced the availability of MQSeries V5.1 for OpenVMS/Alpha,D > V7.2 -1 or V7.3. Full details of the announcement can be found at:E > <http://www.ibm.com/software/ts/mqseries/messaging/openvmsv51/> andt >nL <http://www.ibm.com/software/ts/mqseries/messaging/openvmsv51/factsheet.html > >. >l > Warm Regards,f >  >  >d > Suel >i >u >f   ------------------------------  $ Date: Wed, 9 May 2001 16:21:05 -04002 From: "Sue Skonetski" <susan.skonetski@compaq.com>: Subject: Re: IBM announces MQSeries V5.1 for OpenVMS/Alpha0 Message-ID: <zEhK6.74$fi2.2239@news.cpqcorp.net>  L Thank you, we had a great team working on this and I will pass this along to, the team, since all I did was post the note.  
 Warm Regards,c   Suen    7 "Richard Maher" <maher_rj@hotmail.c*m> wrote in message-+ news:9dc79e$kq4$1@uranium.btinternet.com...n > Hi,m >s > Credit where credit's due:-. >8: > Well done to Sue (and all the others involved)!!!!!!!!!! >" > This is huge!e >h > Regards Richard Mahert > = > Sue Skonetski <susan.skonetski@compaq.com> wrote in messagea, > news:zlbK6.43$fi2.1807@news.cpqcorp.net... > >r > >e > >w > >m > > Dear Newsgroup > > E > > Over the last months I've seen a number of questions on this list  > concerning* > > IBM's support for MQSeries on OpenVMS. > >A@ > > IBM has just announced the availability of MQSeries V5.1 for > OpenVMS/Alpha,F > > V7.2 -1 or V7.3. Full details of the announcement can be found at:G > > <http://www.ibm.com/software/ts/mqseries/messaging/openvmsv51/> andt > >  >-L <http://www.ibm.com/software/ts/mqseries/messaging/openvmsv51/factsheet.html > > >. > >e > > Warm Regards,j > >R > >n > >c > > Sue0 > >C > >e > >D >e >g   ------------------------------  % Date: Wed, 09 May 2001 14:01:30 -0700e! From: Shane.F.Smith@Healthnet.comt: Subject: Re: IBM announces MQSeries V5.1 for OpenVMS/AlphaD Message-ID: <OFD612FD9B.AB28719E-ON88256A47.0072A3B5@foundation.com>  I Hiring Sue is arguably the best thing VMS marketing ever did. She's their  greatest asset, no question.   Shanep          > Richard Maher <maher_rj@hotmail.c*m> on 05/09/2001 12:55:24 PM  6 Please respond to Richard Maher <maher_rj@hotmail.c*m>   To:   Info-VAX@Mvb.Saic.Com  cc:e  ; Subject:  Re: IBM announces MQSeries V5.1 for OpenVMS/Alpha      Hi,o   Credit where credit's due:-   8 Well done to Sue (and all the others involved)!!!!!!!!!!  
 This is huge!o   Regards Richard Maher   ; Sue Skonetski <susan.skonetski@compaq.com> wrote in messagef* news:zlbK6.43$fi2.1807@news.cpqcorp.net... >e >v >  >  > Dear Newsgroup > C > Over the last months I've seen a number of questions on this listu
 concerning( > IBM's support for MQSeries on OpenVMS. > > > IBM has just announced the availability of MQSeries V5.1 for OpenVMS/Alpha,D > V7.2 -1 or V7.3. Full details of the announcement can be found at:E > <http://www.ibm.com/software/ts/mqseries/messaging/openvmsv51/> andn >lL <http://www.ibm.com/software/ts/mqseries/messaging/openvmsv51/factsheet.html   > >. >s > Warm Regards,k >  >d >a > Sue  >i >  >    ------------------------------  % Date: Wed, 09 May 2001 16:08:01 -0500h+ From: Christopher Smith <csmith@amdocs.com> : Subject: RE: IBM announces MQSeries V5.1 for OpenVMS/AlphaL Message-ID: <3B55D7F383B0D31197D9009027541CBF0D9D1E27@cmiexch1.cmi.itds.com>   > -----Original Message-----H > From: Shane.F.Smith@Healthnet.com [mailto:Shane.F.Smith@Healthnet.com]  @ > Hiring Sue is arguably the best thing VMS marketing ever did. 
 > She's theire > greatest asset, no question.   I tend to agree.   Regards,   Chrisg  ! Christopher Smith, Perl Developer  Amdocs - Champaign, IL   /usr/bin/perl -e '? print((~"\x95\xc4\xe3"^"Just Another Perl Hacker.")."\x08!\n");d '    ------------------------------   Date: 10 May 2001 02:20:12 GMT From: <ewcce@aol.com>t Subject: Important newst+ Message-ID: <9dctss$jvm$17065@news.ukr.net>t  \ Important new medical news for those suffering with genital herpes, cold sores and shingles! http://crazydude.com/newmedic    ------------------------------   Date: 9 May 2001 21:33:31 GMTf* From: bleau@umtof.umd.edu (Lawrence Bleau); Subject: Re: mounting 2 disk with the same label in cluster-) Message-ID: <9dcd3b$7ic$1@hecate.umd.edu>   k In article <OF4C693AB1.4705041F-ON80256A46.005586C3@qedi.quintiles.com>, steven.reece@quintiles.com writes:2  M Others have addressed this issue better than I can.  Allow me to add my $0.02t on this question, though:t  I >Can anyone say exactly what the DISK$volumelabel logicals are used for ?f  K I suspect that, back in the days of removable disk media, it could be quite K common for a user's data set to be mounted, at different points in time, on-P different drives, thus resulting in the use of different hardware device names. O The system operator or manager may not remember to change a system-wide logicalnD name each time (quite a chore with a lot of users, data sets, etc.).  O By having a logical name with a specific pattern that is derived from somethingoL on the media itself (in this case, the volume label), one has a bullet-proofM way of obtaining device independence.  Applications need not be changed; they D just reference DISK$label and they'll always get the correct device.  I Of course, with the advent of nonremovable drives, this feature is not as-K useful.  It could still be useful if a drive is going bad and another drive M (with a different device name) is put into service as a replacement while theIL first drive is fixed.  A BACKUP/IMAGE to the replacement drive would give itO the same label, and thus when mounted a DISK$label would be defined to point to,N the new drive.  Since this is a system manager operation, though, I doubt manyI users would be doing it, and so after the BACKUP/IMAGE changing a /systems2 logical name would be very little additional work.   Lawrence Bleau University of Maryland" Physics Dept., Space Physics Group 301-405-6223 bleau@umtof.umd.eduo   ------------------------------   Date: 9 May 2001 21:00:59 CDT = From: wayne@tachysoft.xxx.310887.killspam.0162 (Wayne Sewell)i; Subject: Re: mounting 2 disk with the same label in clusterm. Message-ID: <Po3T$g0BjRdZ@tachxxsoftxxconsult>  \ In article <874ruuztzo.fsf@prep.synonet.com>, Paul Repacholi <prep@prep.synonet.com> writes:    F > No amount of fideling with logicals etc will touch the real problem.                ^^^^^^^^?  K Haven't seen this term before.  Does this mean hanging out with our beardedh buddy in Havana?  :-)s   -- pO ===============================================================================EM Wayne Sewell, Tachyon Software Consulting  (281)812-0738  wayne@tachysoft.xxxm: http://www.tachysoft.xxx/www/tachyon.html and wayne.html  K change .xxx to .com in addresses above, assuming you are not a spambot  :-) O ===============================================================================aB Jed Clampett, checking into hotel: "This place got a cement pond?"+ Ellie May: "And do yuh let critters in it?"A   ------------------------------  % Date: Thu, 10 May 2001 01:37:04 +0200 ) From: Christof Brass <brass@infopuls.com>(& Subject: Re: OpenVMS with NetBeans 3.2, Message-ID: <3AF9D4A0.2E03DE03@infopuls.com>   Sue Skonetski wrote: > I > Folks I just got this and thought I would pass it on.  The URL works onh! > Netscape and Internet explorer.- >  > suec >  > -----Original Message----- > From: NetBeans% > Sent: Tuesday, May 08, 2001 2:49 PMOI > To: Java-Interest; NetBeans OVMS Intrst; NetBeans OVMS; OpenVMS Systems- > Software Group/ > Subject: Announcing NetBeans 3.2 for OpenVMS!  > < > Java Development Comes of Age on OpenVMS with NetBeans 3.2D > Now you can develop Java applications on and for OpenVMS using theL > open-source leader in integrated GUI development. This is your opportunityG > to get in at the beginning of a new phase of OpenVMS's evolution as anM > leading-edge eBusiness platform. Although NetBeans is currently exclusivelysF > a Java IDE, development is underway to extend the languages NetBeans8 > supports to C, C++, Fortran, Cobol, Pascal, and Basic. > D > By evaluating NetBeans, you can make a difference to the future ofG > application development tools for OpenVMS. And, depending on what theuK > NetBeans for OpenVMS team hears back from you, the possibility exists fornE > creating add-in modules with a variety of features that would be ofa6 > particular value to the OpenVMS developer community. > / > Download NetBeans for OpenVMS Free of Charge!  > > > http://www.openvms.compaq.com/openvms/products/ips/netbeans/  ; Thanks! This is a major step forward since this open-source = "product" received several good comments in computer magazinsc comparing IDE for Java.M   ------------------------------  % Date: Wed, 09 May 2001 15:16:23 -0500c+ From: "Main, Kerry" <Kerry.Main@compaq.com>w/ Subject: RE: Oracle Parallel Server on OpenVMS?pR Message-ID: <DC4745D1A85CA04180C83CDC706A9D180D95E2@cthexc02.americas.cpqcorp.net>   Robert,   H re: Oracle Rdb - my personal preference as it understands and is tightlyH integrated with OpenVMS clustering. In addition, the latest version  (orK about to be released) V7.1 has been optimized to take advantage of features 6 like Galaxy - shared memory between OS partitions etc.   re: Oracle Server v8.1.x:   J Oracle V8.1.7 Server (Oracle 8i Release II) and OPS is currently availableJ and supported for OpenVMS V7.2-1. My understanding is that there are plans1 to certify V8.1.7 with OpenVMS V7.3 soon as well.a  H The issue with Oracle OPS and multi-site access is that it requires fullJ read-write disk access (active-active) from each server at each site. ThisJ is not the type of thing one can run over a stateless type of data sharing connection.   G For many OS's, this is a considerable challenge. For OpenVMS it is not.n  L We are currently working with a Customer locally on an active-active, OracleB V8.1.7 (with OPS) multi-site SAN cluster running OpenVMS V7.2-1 (+D applicable patches). Their datacenters are approximately 8km apart.   I Their plan is to have Windows NT/UNIX application servers (as dictated byeJ their business folks) accessing (SQLnet or Net8) high availability, highly+ scalable, back end Oracle V8.1.7 databases.e  F The Customer has received a letter of full support from Oracle on this multi-tier strategy.  J As a fyi, OpenVMS active-active, multi-site SAN clusters are now supportedL up to 100KM apart. Additional information on this and other SAN related info can be found at:6 http://www.openvms.compaq.com/openvms/fibre/index.html     Regards,  
 Kerry Main Senior Consultantf Compaq Canada Inc. Professional Services9 Voice: 613-592-4660d Fax  :  819-772-7036 Email: Kerry.Main@Compaq.com     -----Original Message-----9 From: Robert Schmoelzer [mailto:robert.schmoelzer@aon.at]a Sent: May 8, 2001 5:29 PMe To: Info-VAX@Mvb.Saic.Comi+ Subject: Oracle Parallel Server on OpenVMS?     	 Hi there,d  > We are currently evaluating a solution for a high availability5 database solution which should run on a GS80 cluster.l   Options: -RDB on OpenVMS 8     seems technically great but limited support by local6     Compaq and Oracle people; Oracle says, dont use it'     for compatibility reasons (with 8i)   ( -Oracle Parallel Server on Tru64 cluster6     Oracle and Compaq (sales) prefer this, since it is4     8i "compatibel", but there seems to be no one at5     the local Q and Oracle available who want to taken0     responsibility for installation and support.  # - Oracle Parallel Server on OpenVMSd     is it available???9     Oracle says no, but from a consulting company i heardo#     there is an OPS version for VMSf4     does anybody know more, use it or has some info?   what do you people think about -performance
 -availabilityf
 -managebility  -support -futuren -... of the different possibilities     regardso Robert   ------------------------------  % Date: Wed, 09 May 2001 14:11:13 -0700m. From: Jack Trachtman <Jack.Trachtman@vmmc.org> Subject: Print Accountingc' Message-ID: <3AF9B270.2229528@vmmc.org>t  G I am trying to write a script that will relate print utilization to theg batcht: jobs requesting the printing.  The Accounting file's PrintC record's Process ID field seems to be the ID of the batch job which 
 issued the6 PRINT cmd, but the job would have exited by the time I run my script.  F Is there any way to connect an Accounting file Print entry with either the name1 or (preferably) the batch-que entry number of thed job which issued the Print cmd?a   thanks   ------------------------------  # Date: Wed, 09 May 2001 18:22:37 GMT-2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: Processor Affinityb0 Message-ID: <NVfK6.64$fi2.2137@news.cpqcorp.net>  i In article <4ddK6.1911$VX.126461@news.uswest.net>, "Michael D. Ober" <mdo.@.wakeassoc.com.nospam> writes: E :Under WinNT, a process can be switched between processors while it'soK :running.  Is this true under VMS 7.2-1?  If so, I need a source reference @ :to show my boss.s  E   This rescheduling feature was first available with ASMP (AsymmetriciF   Multiprocessing) in the (as it was then known) VAX/VMS V3.0 release D   (1982).  This feature has subsequently been made available in all C   ASMP releases and in the ASMP replacement known as SMP (Symmetric D   Multiprocessing) starting with the availability of SMP in the (as 3   it was then known) VAX/VMS V5.0 release (1988).     B   Like most other operating systems, OpenVMS prefers to rescheduleC   processes onto the same processor -- arbitrary process preemptionhA   and rescheduling on another processor invalidates the processorsI   caches -- and OpenVMS also prefers to perform the process rescheduling mG   interrupts only when necessary -- if there are no candidate processes I   that might preempt the process, why bother reschedulinging the process.g  F   We have subsequently enhanced scheduling in various areas, including@   the release of the class scheduler API (1993), and current andB   on-going work on process operations in NUMA environments.  ThereB   have been a number of other enhancements to scheduling, as well.  F   As for information sources, this is documented in the Internals and D   Data Structures Manual, and in the Special Processing EnvironmentsD   Chapter (SMP) of the System Manager's Manual, specifically in the B   Understanding Multiprocessing section.  Also please see the URL:  M http://www.openvms.compaq.com/doc/72final/6491/6491pro_004.html#openvms_scheda    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Wed, 09 May 2001 18:28:48 GMT B From: Simon Clubley <simon_clubley@remove_me.excite.com-Earth.UFP>8 Subject: Re: Programming language usage on VMS systems ?6 Message-ID: <A%fK6.8082$vg1.669385@www.newsranger.com>  ( On 9 May 2001 08:51:26 -0500, in article? <sgL4hxQtUL3v@eisner.encompasserve.org>, Larry Kilgallen wrote:e >i| >In article <QMaK6.7475$vg1.610094@www.newsranger.com>, Simon Clubley <simon_clubley@remove_me.excite.com-Earth.UFP> writes:J >> I have been curious for a while about which languages are in use on VMSF >> systems and what percentage of systems these languages run on. Does< >> anybody have an idea of what this distribution might be ? >> pF >> It would be interesting to know as well what the distribution is inC >> the installed base (VAX and/or Alpha) as well as in new systems.t >> t >// >I would expect more precision from an Ada fan.u >r   Noted :-), but see below.s  C >By "language use", do you mean execution or compilation machines ?9< >If execution, do you want to include third party products ? >How about layered products ?u  H If it's a programming language, regardless of source, then it qualifies.  C I left the definition vague because I wanted to see which languageseI people mentioned. I also wanted to see if people counted end user systemsrK as using the language. My viewpoint is that if, for example, 5 machines aresI used for development and the application developed is deployed across 200m6 machines, then the language is in use on 205 machines.  ? >Do you want a raw count of machines, or hours per language per. >machine per year ?a< >Would that be absolute hours or percentage of total hours ? >   K I was not looking for anything like this level of detail. The response fromw2 Arne was the kind of thing that I was looking for.  @ >Presumably you don't want to count languages that are installed> >but never used (if it is only compilation that concerns you). >l  L This however did not occur to me as I only install products that I intend to  use. Thanks for pointing it out.   >======= >mA >I hope the above can be of help in formatting your actual surveya@ >questions.  Figuring out how to get back a scientific sample is >well beyond my capabilities.h  D You are correct that this is not a scientific sample. :-) I was just6 interested in which selection of languages are in use.   Simon.   -- *; Simon Clubley, simon_clubley@remove_me.excite.com-Earth.UFPmJ 'The statement that "it can never happen" is not an acceptable programmingH approach. You must assume it can happen and be in control when it does.'=           -- Ada 95 Quality and Style Guidelines, US DoD AJPO>   ------------------------------  % Date: Wed, 09 May 2001 11:35:55 -0700e! From: Tom Linden <tom@kednos.com> 8 Subject: RE: Programming language usage on VMS systems ?9 Message-ID: <CIEJLCMNHNNDLLOOGNJIAEHCCIAA.tom@kednos.com>i   PL/I 80% C    20%  9 But, of course, we do the PL/I compiler for VMS and Tru64f   Tomp   > -----Original Message----- > From: Simon Clubleya7 > [mailto:simon_clubley@remove_me.excite.com-Earth.UFP]t( > Sent: Wednesday, May 09, 2001 11:29 AM > To: Info-VAX@Mvb.Saic.Com.: > Subject: Re: Programming language usage on VMS systems ? >  >.* > On 9 May 2001 08:51:26 -0500, in articleA > <sgL4hxQtUL3v@eisner.encompasserve.org>, Larry Kilgallen wrote:- > > ? > >In article <QMaK6.7475$vg1.610094@www.newsranger.com>, Simonh@ > Clubley <simon_clubley@remove_me.excite.com-Earth.UFP> writes:L > >> I have been curious for a while about which languages are in use on VMSH > >> systems and what percentage of systems these languages run on. Does> > >> anybody have an idea of what this distribution might be ? > >>H > >> It would be interesting to know as well what the distribution is inE > >> the installed base (VAX and/or Alpha) as well as in new systems.j > >> > >B1 > >I would expect more precision from an Ada fan.i > >  >d > Noted :-), but see below.e >dE > >By "language use", do you mean execution or compilation machines ? > > >If execution, do you want to include third party products ? > >How about layered products ?6 >$J > If it's a programming language, regardless of source, then it qualifies. >lE > I left the definition vague because I wanted to see which languages K > people mentioned. I also wanted to see if people counted end user systemsl@ > as using the language. My viewpoint is that if, for example, 5 > machines areK > used for development and the application developed is deployed across 200m8 > machines, then the language is in use on 205 machines. >tA > >Do you want a raw count of machines, or hours per language perR > >machine per year ? > > >Would that be absolute hours or percentage of total hours ? > >a >0? > I was not looking for anything like this level of detail. The  > response froma4 > Arne was the kind of thing that I was looking for. >5B > >Presumably you don't want to count languages that are installed@ > >but never used (if it is only compilation that concerns you). > >h >.B > This however did not occur to me as I only install products that
 > I intend tor" > use. Thanks for pointing it out. >t
 > >======= > > C > >I hope the above can be of help in formatting your actual surveyeB > >questions.  Figuring out how to get back a scientific sample is > >well beyond my capabilities.r >.F > You are correct that this is not a scientific sample. :-) I was just8 > interested in which selection of languages are in use. >t > Simon. >  > --= > Simon Clubley, simon_clubley@remove_me.excite.com-Earth.UFPoL > 'The statement that "it can never happen" is not an acceptable programmingJ > approach. You must assume it can happen and be in control when it does.'? >           -- Ada 95 Quality and Style Guidelines, US DoD AJPOu >n   ------------------------------   Date: 9 May 2001 17:02:41 GMT@3 From: gartmann@immunbio.mpg.de (Christoph Gartmann)N8 Subject: Re: Programming language usage on VMS systems ?0 Message-ID: <9dbt7h$dtq$1@n.ruf.uni-freiburg.de>   Hi,e  M We use 70% Pascal, 15% Fortran and 15% C. And we logged several calls becausenK of C, very few because of Fortran and perhaps one or two because of Pascal.    Regards,    Christoph Gartmann   H -- --------------------------------------------------------------------+H | Max-Planck-Institut fuer      Phone   : +49-761-5108-464   Fax: -452 |H | Immunbiologie                                                        |H | Postfach 1169                 Internet: gartmann@immunbio.mpg.de     |H | D-79011  Freiburg, FRG                                               |H +--------- http://www.immunbio.mpg.de/home/english/menue.html ---------+   ------------------------------  $ Date: Wed, 9 May 2001 15:05:33 -04000 From: "Syltrem" <syltrem@videotron.ca.spammenot>8 Subject: Re: Programming language usage on VMS systems ?8 Message-ID: <ZqgK6.25273$WF.1257347@weber.videotron.net>   We use mostly Powerhouse (70%)I Then Basic (10%), Cobol (10%), Progress (yeuk - 1%) and now starting with  Gembase (remaining %).K For Progress and Gembase we have 3rd party applications. I have not counted ) the programs in there, just the in-house.a  8 There is 3-4 Fortran programs around, and as many macro.  D Oh! If DCL is counted, it's used everywhere to make the others work!   --   Syltremm; http://pages.infinit.net/syltrem (OpenVMS related web site)i> To reply to myself directly, remove .spammenot from my address  K "Simon Clubley" <simon_clubley@remove_me.excite.com-Earth.UFP> a crit dansa< le message news: QMaK6.7475$vg1.610094@www.newsranger.com...I > I have been curious for a while about which languages are in use on VMS E > systems and what percentage of systems these languages run on. Doesu; > anybody have an idea of what this distribution might be ?e >gE > It would be interesting to know as well what the distribution is inpB > the installed base (VAX and/or Alpha) as well as in new systems. > H > [I have no specific reason for the information, I am just interested.] >  > Thanks for any information,n >i > Simon. >. > --= > Simon Clubley, simon_clubley@remove_me.excite.com-Earth.UFPsL > 'The statement that "it can never happen" is not an acceptable programmingJ > approach. You must assume it can happen and be in control when it does.'? >           -- Ada 95 Quality and Style Guidelines, US DoD AJPOa   ------------------------------   Date: 9 May 2001 18:19:28 -0500n9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)=8 Subject: Re: Programming language usage on VMS systems ?3 Message-ID: <l52whR8eKh+Z@eisner.encompasserve.org>l  { In article <A%fK6.8082$vg1.669385@www.newsranger.com>, Simon Clubley <simon_clubley@remove_me.excite.com-Earth.UFP> writes:h* > On 9 May 2001 08:51:26 -0500, in articleA > <sgL4hxQtUL3v@eisner.encompasserve.org>, Larry Kilgallen wrote:p >>} >>In article <QMaK6.7475$vg1.610094@www.newsranger.com>, Simon Clubley <simon_clubley@remove_me.excite.com-Earth.UFP> writes:t    D >>By "language use", do you mean execution or compilation machines ?= >>If execution, do you want to include third party products ?r >>How about layered products ? > J > If it's a programming language, regardless of source, then it qualifies.  3 All VMS machines run lots of Macro-32 and Bliss-32.OK Alpha VMS machine also run considerable C, with some Bliss-64 and Macro-64.O? All VMS machines running a relatively recent version of VMS rune? a smidgen of Ada, for instances processing breakin evasion.  Ife5 they  do too much of that, it is a Bad Thing(tm). :-)u  E > I left the definition vague because I wanted to see which languagesEK > people mentioned. I also wanted to see if people counted end user systemsCM > as using the language. My viewpoint is that if, for example, 5 machines areaK > used for development and the application developed is deployed across 200.8 > machines, then the language is in use on 205 machines.  ? On my own company machines I program in Ada, Scan and Bliss-32.    ------------------------------  # Date: Thu, 10 May 2001 00:26:20 GMTt$ From: Scott Vieth <svieth@wi.rr.com>8 Subject: Re: Programming language usage on VMS systems ?) Message-ID: <3AF9E0A5.A2B220CB@wi.rr.com>c  K None of the responses have yet mentioned the M (mumps) family: ISM, DSM and  Cache.  B There are a lot of shops writing code in the various flavors of M.   -Scott   Simon Clubley wrote:  I > I have been curious for a while about which languages are in use on VMSKE > systems and what percentage of systems these languages run on. Doesp; > anybody have an idea of what this distribution might be ?2 >eE > It would be interesting to know as well what the distribution is incB > the installed base (VAX and/or Alpha) as well as in new systems. > H > [I have no specific reason for the information, I am just interested.] >i > Thanks for any information,u >  > Simon. >  > --= > Simon Clubley, simon_clubley@remove_me.excite.com-Earth.UFPxL > 'The statement that "it can never happen" is not an acceptable programmingJ > approach. You must assume it can happen and be in control when it does.'? >           -- Ada 95 Quality and Style Guidelines, US DoD AJPOu   ------------------------------  % Date: Thu, 10 May 2001 06:13:30 +0100eQ From: Aidan Karley <aidan@karley.go-and-spam-me-you-sweaty-toads.freeserve.co.uk>f. Subject: Re: Some more bits looking for a homeY Message-ID: <VA.000001b1.13bcff1b@karley.go-and-spam-me-you-sweaty-toads.freeserve.co.uk>s  ; In article <SdTJ6.38$UT2.4542@wards>, Adrian Lumsden wrote:n   > " > 2   DEPCA PRO Ethernet interface/ > 1   DE203-PA ISA Ethernet Interface, 10Base2, 6 >     includes manual and driver floppy. Boxed as new.  +      NEWBIE ALERT! Dumb question coming up.TF      Are these cards that are compatible with a PDP-11? I've recently G acquired a PDP-11 (details still unsure, but deeply nicotine stained),  F which I have a hankering to use as a firewall between my home network C and the Internet (there's a weird sort of symmetry to that). I was eH expecting to have to set up one of the Linux boxen to act as a terminal C to the PDP and then telnet into the Linux box from the rest of the aG network, but I think it might be easier to do if I had direct Ethernet   into the PDP. I think. Maybe.f  H > These available for free; location Leicester, UK; you collect or cover > shipping. I      What's the weight of the package. If I knew that I'd be able to ask SD the P.O. what the shipping would be and post you a self-addressed &  stamped label.     > and finally an oddball:- > 5 > 1   Adaptec 2842 VL Bus High Performance Bus Mastere6 >     VL-to-Fast SCSI Host adapter with EZSCSI 3.1 kit3 >     (Yep! I used to have a PC with VL bus in it!)  > D      Curses! I gave away my last VL-bus MoBo a couple of months ago.   Aidan Karley	 Aberdeen,O	 Scotland.eB Message written at Wed, 09 May 2001 13:08 +0100, but posted later.   ------------------------------  % Date: Thu, 10 May 2001 09:55:18 +1000 - From: "Dick Adams" <adams.dick.rc@bhp.com.au>b! Subject: Re: SYSMAN and SMISERVER 4 Message-ID: <9dclh3$rvr$1@gossamer.itmel.bhp.com.au>   Brian,     DECNET Phase IVo  G     On the node in question there is no object SMISERVER whereas on the?I other nodes there is, this certainly seems to be the difference. Now withsJ this information I have looked at the system more carefully and found thatK SMISERVER is indeed running, but it's name is of the form NET_xxxx. Is thisl started at system startup?  4 Thanks to everyone else who offered suggestions too.   Dick  H Brian Schenkenberger, VAXman- <system@SendSpamHere.ORG> wrote in message* news:009FBBDE.5BA7BB2D@SendSpamHere.ORG...C > In article <9dakrg$3i2$1@gossamer.itmel.bhp.com.au>, "Dick Adams" " <adams.dick.rc@bhp.com.au> writes:L > >Thanks Brian but I'm afraid the online docs are not able to enlighten me. > >c$ > >To rephrase my original question.L > >I don't understand why SYSMAN commands work to all the nodes which do not= > >have SMISERVER running (and aren't in cluster) except one?s > >e > >Some more information.uJ > >When I SET ENVIRONMENT /NODE=(node1,...) it prompts me for the password toL > >use on remote nodes, I can sign on to all the nodes using the same useridL > >and password. When I issue a command such as DO SHOW TIME, it executes onD > >all nodes until it get to the one in question, it then fails with > >D@ > >"SYSTEM-F-INVLOGIN, login information invalid at remote node" > >4H > >On the target node the accounting records show a login failure with a useridJ > >of ILLEGAL and final status text of "%LOGIN-F-NOSUCHUSER, no such user" > > L > >This is not actually causing me a problem as I can start up SMISERVER andJ > >everything works as expected, but, why is one node different to all the
 > >others? >e9 > DECnet phase IV or V running on the problematic target?  >1* > Look at the SMISERVER object definition: >a > Phase IV:l1 > $ MCR NCP SHOW OBJECT SMISERVER CHARACTERISTICSw >s
 > Phase V:A > $ MCR NCL SHOW NODE 0 SESSION CONTROL APPLICATION SMISERVER ALLh CHARACTERISTICSt >  > --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 afterp them.o   ------------------------------  % Date: Wed, 09 May 2001 18:35:19 +0100a0 From: andrew harrison <andrew.nospam@uk.sun.com>8 Subject: Re: The Internet, Tru64 and other unix variants* Message-ID: <3AF97FD7.5F804E60@uk.sun.com>   Christof Brass wrote:s >   @ > SUN will sooner or later replace Solaris by Linux. You can see > signs of this already.   Ohh and what signs are they ??  = As a matter of interest what do you think linux actually is ?r   Let me help you.  7 Its a UNIX like kernel and not much else. The utilitiest7 the libraries etc which make up the bulk of what you orr3 I think of as an OS have come from GNU, commercial n% donations or other freeware projects.e  ' And what is so attractive about Linux.    3 Its small, its fast and it has good device support h2 for PC's and devices that you can attach to PC's.  But its only really a kernel.w  1 So what happens when you add proper SMP support, #0 threads, kernel pre-emption, real-time, a proper/ JFS, resource management, hot plug support etc  - etc which some if not most commercial UNIX's  	 have now.d  . You get a much bigger, slower etc kernel that , suddenly looks rather like Solaris or HP-UX  or AIX.   - At this point why would Sun for example want F, to substitute the Solaris kernel for a Linux
 kernel ???   Regards- Andrew Harrisong Enterprise IT Architecto   ------------------------------    Date: 10 May 2001 01:08:00 +0800, From: Paul Repacholi <prep@prep.synonet.com>8 Subject: Re: The Internet, Tru64 and other unix variants- Message-ID: <87heyubgcv.fsf@prep.synonet.com>o  2 andrew harrison <andrew.nospam@uk.sun.com> writes:  C > Incedentally you seem to missunderstand what Linux is, it isn't aaE > new OS it is simply a new UNIX like kernel that makes extensive usewD > of the the GNU and other freeware utilities to provide a UNIX like@ > environment. Almost all of what people think of as being Linux > existed before Linux.a   Oh, yeah...s   Kernel           Noe File system      Nof Compiler         No [1]o Shell (Bash)     Yes Network stack    Nos  0 [1] The current GCC is EGCS, not the pre 92 gcc.     -- c< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.H Spam-To: uce@ftc.gov,enforcement@sec.gov,sness@fcc.gov,hfurchtg@fcc.gov,#   mpowell@fcc.gov,gtristan@fcc.gov     ------------------------------  % Date: Wed, 09 May 2001 19:29:35 +0100i  From: steven.reece@quintiles.com8 Subject: Re: The Internet, Tru64 and other unix variantsH Message-ID: <OFD955FD2A.6140210D-ON80256A47.00655475@qedi.quintiles.com>  K A much bigger, slower etc kernel that suddenly looks rather like Solaris ors HP-UX or AIX eh?  K In that case it should be an afternoon's work to port Solaris, AIX or HP-UXt# to a decent platform such as Alpha.o  	 It's not?a( Well how can they be so much alike then? :-)) Steve.     Andrew Harrison wrote: >>>c& And what is so attractive about Linux.  2 Its small, its fast and it has good device support1 for PC's and devices that you can attach to PC's.u But its only really a kernel.   0 So what happens when you add proper SMP support,0 threads, kernel pre-emption, real-time, a proper. JFS, resource management, hot plug support etc, etc which some if not most commercial UNIX's	 have now.e  - You get a much bigger, slower etc kernel thatf+ suddenly looks rather like Solaris or HP-UXo or AIX., <<<n   ------------------------------  % Date: Wed, 09 May 2001 11:30:43 -0700 ! From: Shane.F.Smith@Healthnet.comm8 Subject: Re: The Internet, Tru64 and other unix variantsD Message-ID: <OF3C832786.C8AC70F6-ON88256A47.00659DB1@foundation.com>  J Oh no! An Andrew Harrison vs. Christof Brass thread! Run for the hills!!!! ;-P    Shaner          D andrew harrison <andrew.nospam@uk.sun.com> on 05/09/2001 08:04:05 AM  < Please respond to andrew harrison <andrew.nospam@uk.sun.com>   To:   Info-VAX@Mvb.Saic.Com  cc:o  9 Subject:  Re: The Internet, Tru64 and other unix variants      Christof Brass wrote:- > = > Running Tru64 and/or Linux apps on VMS won't happen. Do youe= > realise the problems some UNIX flavours have in running WNToB > apps? What could and should be accomplished is to have something@ > like an aBCS (analog to iBCS) which would help in running UNIX1 > flavoured Alpha apps on all UNIX flavoured OSs.y >l  7 Rubbish its already happening. What do you think Oraclew8 use as their development platform ?? it isn't Windows NT: or OpenVMS. How about all the GNU utilities that have been ported to OpenVMS ??  = You windows point is also rubbish, there is a huge differenceo: between say Tru64/Solaris/Linux which have well documented7 API's with easy access to source code for more detailed 8 scrutiny and Windows, which is not as well documented as7 it should be, which changes rapidly and which has majorl: apps that use the undocumented/poorly documented features.  9 Incedentally you seem to missunderstand what Linux is, itw= isn't a new OS it is simply a new UNIX like kernel that makesi9 extensive use of the the GNU and other freeware utilitiesr= to provide a UNIX like environment. Almost all of what peoplei- think of as being Linux existed before Linux.e  ? > To run VMS apps on UNIX or vice versa wouldn't be a good ideaaB > because it's simply not necessary. There are much better ways to > have the same effect.u > A > Dropping Tru64 is not a problem for the Alpha. Because VMS onlym> > runs on Alpha, dropping Alpha means dropping VMS. This won't9 > happen in the foreseeable future (if something like thelA > foreseeable future exists at all). Or VMS has to be ported to ao > different HW than Alpha.  1 I don't think that Compaq will drop tru64 anytimea4 soon, but if they did then it would be the beginning5 of the end for VMS because it would no longer benefite3 from the economies of scale that it does by sharingt  the same HW platform with Tru64.   Regardsl Andrew Harrisont Enterprise IT Architecte   ------------------------------  % Date: Wed, 09 May 2001 19:45:03 +0100.0 From: andrew harrison <andrew.nospam@uk.sun.com>8 Subject: Re: The Internet, Tru64 and other unix variants* Message-ID: <3AF9902F.B0DEF0D7@uk.sun.com>   Paul Repacholi wrote:  > 4 > andrew harrison <andrew.nospam@uk.sun.com> writes: > E > > Incedentally you seem to missunderstand what Linux is, it isn't aaG > > new OS it is simply a new UNIX like kernel that makes extensive use F > > of the the GNU and other freeware utilities to provide a UNIX likeB > > environment. Almost all of what people think of as being Linux > > existed before Linux.e > 
 > Oh, yeah...  >  > Kernel           No   . FreeBSD existed before Linux and ran basically the same utilites.   > File system      Nos  1 Since when has a filesystem not been part of the s kernel.A   > Compiler         No [1]- > Shell (Bash)     Yes > Network stack    No   - Since the Linux IP stack is based on the BSD n2 one which vastly predates Linux and was available / for a multitude of platforms including FreeBSD.i  2 So we have an OS that uses a written from scratch 3 kernel, with a written from scratch filesystem, bute1 which relies either directly in the case of most c/ of the utilites or indirectly as in the case ofn) the IP stack on other freeware projects. -   Regards- Andrew Harrison- Enterprise IT Architectu   ------------------------------    Date: 10 May 2001 03:03:44 +0800, From: Paul Repacholi <prep@prep.synonet.com>8 Subject: Re: The Internet, Tru64 and other unix variants- Message-ID: <878zk6bazz.fsf@prep.synonet.com>,  2 andrew harrison <andrew.nospam@uk.sun.com> writes:  3 > So what happens when you add proper SMP support, e2 > threads, kernel pre-emption, real-time, a proper1 > JFS, resource management, hot plug support etc r   You would be half way to VMS.o   -- j< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.e@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.H Spam-To: uce@ftc.gov,enforcement@sec.gov,sness@fcc.gov,hfurchtg@fcc.gov,#   mpowell@fcc.gov,gtristan@fcc.gov ,   ------------------------------  % Date: Thu, 10 May 2001 02:18:32 +0200a) From: Christof Brass <brass@infopuls.com>e8 Subject: Re: The Internet, Tru64 and other unix variants, Message-ID: <3AF9DE58.A78BE75A@infopuls.com>   andrew harrison wrote: >  > Christof Brass wrote:t > > ? > > Running Tru64 and/or Linux apps on VMS won't happen. Do you-? > > realise the problems some UNIX flavours have in running WNTwD > > apps? What could and should be accomplished is to have somethingB > > like an aBCS (analog to iBCS) which would help in running UNIX3 > > flavoured Alpha apps on all UNIX flavoured OSs.a > >h > 9 > Rubbish its already happening. What do you think Oracle : > use as their development platform ?? it isn't Windows NT< > or OpenVMS. How about all the GNU utilities that have been > ported to OpenVMS ??  @ Not re-compiled apps - like iBCS having aBCS would allow running@ apps unmodified in binary form on all Alpha based UNIX flavours.; This is *not* porting. Please read my posts more carefully.  Thanks.   ? > You windows point is also rubbish, there is a huge differencem< > between say Tru64/Solaris/Linux which have well documented9 > API's with easy access to source code for more detailed-: > scrutiny and Windows, which is not as well documented as9 > it should be, which changes rapidly and which has majorn< > apps that use the undocumented/poorly documented features.  8 The problems of running WNT apps on UNIX don't come from> undocumented API parts of WNT, they don't come from apps using? undocumented calls and they also don't come from the mistakenlyM, stated "fact" of a rapidly changing WNT API.: If the undocumented parts were there reason only apps from; companies that have access to information about those callsm would make trouble.b< The API is documented enough and for this type of programmer5 it's easy to reverse engineer what they want to know.?< Adding calls to the API is not a big deal as long as the app= doesn't use them. This problem isn't simply not there because ; you target your effort to one specific version of a program 7 which won't change during that development phase of thea "emulator".n@ From your post I have to conclude that you don't have experience+ in implementing a binary standard emulator.   < The by far most important reason for the trouble is the huge difference in architecture.t  ; > Incedentally you seem to missunderstand what Linux is, itu? > isn't a new OS it is simply a new UNIX like kernel that makes ; > extensive use of the the GNU and other freeware utilitiesu? > to provide a UNIX like environment. Almost all of what peoplee/ > think of as being Linux existed before Linux.n  ? I have simply no clue on how you came to that impression that Is? seem to misunderstand what Linux is. May I refer you to severali? posts about the Linux topic I made in cov within the last eighta= months. You will see that I very precisely know what Linux isi and what it is *not*.P  A > > To run VMS apps on UNIX or vice versa wouldn't be a good idea2D > > because it's simply not necessary. There are much better ways to > > have the same effect.  > >.C > > Dropping Tru64 is not a problem for the Alpha. Because VMS only.@ > > runs on Alpha, dropping Alpha means dropping VMS. This won't; > > happen in the foreseeable future (if something like the C > > foreseeable future exists at all). Or VMS has to be ported to aa > > different HW than Alpha. > 3 > I don't think that Compaq will drop tru64 anytime 6 > soon, but if they did then it would be the beginning7 > of the end for VMS because it would no longer benefitI5 > from the economies of scale that it does by sharing " > the same HW platform with Tru64.  ? Thanks for repeating this BS which you posted already and whichy@ could sound smart on first sight. You seem not to understand the= mechanics of SW development. Dropping Tru64 (I appoligise foru@ the Tru64 fans) is one of the best things that can happen to VMS= because it will allow to concentrate development and increase-? the consolidated revenue much more than with this "distributed"0? VMS / Tru64 strategy. BTW I personally have the deep impressiona; that you also don't understand the market. Ever heard about ) branding, competition and selling factor?a  	 > Regardsp > Andrew Harrisona > Enterprise IT Architecta   ------------------------------  % Date: Thu, 10 May 2001 02:22:48 +0200 ) From: Christof Brass <brass@infopuls.com>E8 Subject: Re: The Internet, Tru64 and other unix variants+ Message-ID: <3AF9DF58.F6AF602@infopuls.com>R   Paul Repacholi wrote:p > 4 > andrew harrison <andrew.nospam@uk.sun.com> writes: > E > > Incedentally you seem to missunderstand what Linux is, it isn't aiG > > new OS it is simply a new UNIX like kernel that makes extensive usehF > > of the the GNU and other freeware utilities to provide a UNIX likeB > > environment. Almost all of what people think of as being Linux > > existed before Linux.P > 
 > Oh, yeah...t >  > Kernel           Nos > File system      Not > Compiler         No [1]  > Shell (Bash)     Yes > Network stack    No  > 2 > [1] The current GCC is EGCS, not the pre 92 gcc. >  > --> > Paul Repacholi                               1 Crescent Rd.,9 > +61 (08) 9257-1001                           Kalamunda..B >                                              West Australia 60760 > Raw, Cooked or Well-done, it's all half baked.J > Spam-To: uce@ftc.gov,enforcement@sec.gov,sness@fcc.gov,hfurchtg@fcc.gov,$ >   mpowell@fcc.gov,gtristan@fcc.gov   Interesting post.e: The question is now how it could be explained that someone< states that Linux isn't a new OS but everbody knows that the= Linux kernel has been newly developed. We need a criterion to0; decide which parts have to be newly developed to be able toa+ seriously claim that something is a new OS.D' Another question: is Solaris a new OS??u   ------------------------------  % Date: Thu, 10 May 2001 02:31:28 +0200b) From: Christof Brass <brass@infopuls.com> 8 Subject: Re: The Internet, Tru64 and other unix variants, Message-ID: <3AF9E160.E4E4806C@infopuls.com>   andrew harrison wrote: >  > Paul Repacholi wrote:f > >i6 > > andrew harrison <andrew.nospam@uk.sun.com> writes: > > G > > > Incedentally you seem to missunderstand what Linux is, it isn't aiI > > > new OS it is simply a new UNIX like kernel that makes extensive usesH > > > of the the GNU and other freeware utilities to provide a UNIX likeD > > > environment. Almost all of what people think of as being Linux > > > existed before Linux.e > >e > > Oh, yeah...i > >a > > Kernel           Noc > 0 > FreeBSD existed before Linux and ran basically > the same utilites.  > It's strange that someone reveals basic misunderstanding of OS? design and especially Linux "design" who stated that some other   didn't understand what Linux is.  @ BSD and System V existed before Solaris. Is Solaris a new OS? Is it an OS at all?: You completely missed the point. Linux has a newly written> kernel. What has this "ran bascially the same utilities" to do: with the kernel beeing a new one? To recall the situation:@ "kernel no" means that the kernel hasn't been there before Linus Thorvalds developed it.t   >  > > File system      NoM > 2 > Since when has a filesystem not been part of the	 > kernel.   > Since when has it been part of? Ever heard of minix, ext2, fat9 etc.? All part of the kernel? Very interesting, very veryN interesting.   > > Compiler         No [1]t > > Shell (Bash)     Yes > > Network stack    NoI > . > Since the Linux IP stack is based on the BSD3 > one which vastly predates Linux and was available 1 > for a multitude of platforms including FreeBSD.g  < Are you sure? I'm not that sure. I seem to remember that the? stack has at least been rewritten. But honestly you wrote a lot 1 of very strange things about your OS architectureP? (mis)understanding in your last few posts to this thread that Ip= decide not check this in detail because I simply be sure thatC$ you are probably also wrong on this.  3 > So we have an OS that uses a written from scratch 5 > kernel, with a written from scratch filesystem, but02 > which relies either directly in the case of most1 > of the utilites or indirectly as in the case ofa* > the IP stack on other freeware projects.   What's the point?>  	 > RegardsB > Andrew Harrison  > Enterprise IT Architectr   ------------------------------  % Date: Thu, 10 May 2001 02:35:57 +0200h) From: Christof Brass <brass@infopuls.com>r8 Subject: Re: The Internet, Tru64 and other unix variants, Message-ID: <3AF9E26D.FBE1C267@infopuls.com>   Jan Vorbrueggen wrote: > - > Christof Brass <brass@infopuls.com> writes:  > B > > SUN will sooner or later replace Solaris by Linux. You can see > > signs of this already. > H > You can? I cannot. It would imply a singular loss of control by Sun to< > do that, and not least a significant investment write-off. > 
 >         Janr  9 Yes and no. SUN is investing substantailly into the Linuxu@ business. Solaris is open source. From what I see SUN will merge= both. There is no point in "loss of control" because Linux iso> open source. Having a Linux distribution from SUN for SPARC is; the best what SUN can offer - full control, minimum effort,o maximum effect.    ------------------------------  % Date: Thu, 10 May 2001 02:46:27 +0200i) From: Christof Brass <brass@infopuls.com>-8 Subject: Re: The Internet, Tru64 and other unix variants, Message-ID: <3AF9E4E3.22C49A75@infopuls.com>   andrew harrison wrote: >  > Christof Brass wrote:> > >j > B > > SUN will sooner or later replace Solaris by Linux. You can see > > signs of this already. >   > Ohh and what signs are they ??  > Investment into Linux. Dropping CDE in favour of KDE or Gnome. Making Solaris open source.   ? > As a matter of interest what do you think linux actually is ?  >  > Let me help you. > 9 > Its a UNIX like kernel and not much else. The utilitiess9 > the libraries etc which make up the bulk of what you oro4 > I think of as an OS have come from GNU, commercial' > donations or other freeware projects.1   Sorry, not much help.   ( > And what is so attractive about Linux. > 4 > Its small, its fast and it has good device support3 > for PC's and devices that you can attach to PC's.9 > But its only really a kernel.   8 I would add much different things to Linux attractivity.9 But what is the point in your "its only really a kernel"?   2 > So what happens when you add proper SMP support,2 > threads, kernel pre-emption, real-time, a proper0 > JFS, resource management, hot plug support etc. > etc which some if not most commercial UNIX's > have now.i  : I completely disagree. SMP support is beeing added, kernel? threads are there, both without blowing out the kernel. What do.7 you mean exactly with kernel pre-emption? Solaris isn'tu@ "real-time". JFS is there. Again I severly doubt that the points? you mention make the kernel bigger or slower or whatever worse.U? It might have made Solaris worse because SUN hire cheap collegec8 finishers and fire them after a few years when they want@ substantial more money. This won't work for quality. (source: Ed? Yourdan - Decline and Fall of the American Programmer.) Solarisv? 2.x with x smaller 5 was a piece of shit, every single version; ; and a collection of patches. Why not drop it and use Linux?,  / > You get a much bigger, slower etc kernel thatu- > suddenly looks rather like Solaris or HP-UX 	 > or AIX.   < No it won't, unless you do it without planing, designing and< with student programmers like SUN (and don't forget the well: trained enterprise architects who supply the most valuable contributions).0  . > At this point why would Sun for example want. > to substitute the Solaris kernel for a Linux > kernel ???  > Because its cheaper. There is no point in having your own UNIX@ flavour with minuscle market share. You make money with e.g. HW.8 Sorry I can't continue it would end up in a lesson about	 business.o  	 > Regardso > Andrew Harrison: > Enterprise IT Architectt   ------------------------------  % Date: Thu, 10 May 2001 02:49:49 +0200 ) From: Christof Brass <brass@infopuls.com> 8 Subject: Re: The Internet, Tru64 and other unix variants, Message-ID: <3AF9E5AD.476C6E94@infopuls.com>   Jan Vorbrueggen wrote: > - > Christof Brass <brass@infopuls.com> writes:N > 1 > > There is no point in having UNIX apps on VMS.  > K > I suggest you first define what a "UNIX app" is - possibly in contrast to > > a "VMS app" - and then we might have a basis for discussion. > 
 >         Jann  ; Simply an app which uses mostly UNIX system calls and whichr? doesn't take full advantage of the VMS features like RMS files.t0 Poorly ported UNIX apps which should run on VMS.   ------------------------------  # Date: Thu, 10 May 2001 00:28:09 GMT6) From: rob.buxton@wcc.govt.nz (Rob Buxton) @ Subject: Re: VAX Just Stops - Not the Usual Suspects - follow up1 Message-ID: <3af9dedd.620332020@news.wcc.govt.nz>t  C When I first posted this I reported I had seen a message similar toh@ what you're suspecting. It has appeared a couple of times on the5 screen of the AlphaStation in the cluster that dies. i   Message was:  F arp, illegal IRP address 255.255.255.255 used by hardware 00-80-64-15!  C Alas, it seemed to be shortened. I've looked through the arp tablestA but haven't found a hardware address that matches the part given.n  ? I've not seen this message on the consoles of any of the VAXes.   F Alas, the Alpha Screen often just locks. as it's clustered with one ofE the errant MV3100s, it doesn't always allow me to view what been sentV to the console.o  B So, the issue does seem to be IP related. Just odd why it hits the( MV3100s and not the VAX4000s or Alphas.   . On 09 May 2001 10:50:22 +0200, Jan Vorbrueggen8 <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote:  , >rob.buxton@wcc.govt.nz (Rob Buxton) writes: >  >> Here's the update.  >[snip]t >bN >Sounds very much like a multicast packet ((R)ARP?) causing a HALT instructionM >to be executed in kernel mode. The console(s) should say/have said so, BTW -a
 >did they? >r >	Jan    ------------------------------  % Date: Wed, 09 May 2001 11:13:16 -0700 + From: "Barry Treahy, Jr." <treahy@mmaz.com>o0 Subject: Re: VMS 7.2 BACKUP damaging system disk( Message-ID: <3AF988BC.8A046DDD@mmaz.com>  ! steven.reece@quintiles.com wrote:>  O > What's the command line that you're using for the backup and for the restore?: > K > Whilst I appreciate your experience and knowledge, it's easy to trip over J > things and be so deep into a problem that you can't see the wood for the > trees.   You are absolutely correct.   N I just did two on-line disk to disk backups, with logging off and this is what
 I've seen:  C V4500$ backup/nolog/nofast/noalias/ignore=(noback,inter) $145$dia0:  $145$dia1:/imageN %BACKUP-W-AFNOTSAVED, alias file $145$DIA0:[SYS0]SYSCOMMON.DIR;1 was not savedN %BACKUP-W-AFNOTSAVED, alias file $145$DIA0:[SYSEXE]SYSBOOT.EXE;1 was not savedA V4500$ backup/nolog/nofast/alias/ignore=(noback,inter) $145$dia0:c $145$dia1:/imageN %BACKUP-W-AFNOTSAVED, alias file $145$DIA0:[SYS0]SYSCOMMON.DIR;1 was not saved V4500$  O Now I can see why the first failed to boot because the alias to the SYSBOOT was K broke.  What I do not understand is why the second backup still ignored theeL SYSCOMMON aliases, thereby producing a bad VMS tree, but more importantly itM ALSO failed to recreate the SYSBOOT aliases even though it did not report anyo errors:p  ; V4500$ dir/file $145$dia0:[000000...]sysboot.exe,$145$dia1:0  + Directory $145$DIA0:[SYS0.SYSCOMMON.SYSEXE]<   SYSBOOT.EXE;1        (113,1,0)   Total of 1 file.   Directory $145$DIA0:[SYSEXE]   SYSBOOT.EXE;1        (113,1,0)   Total of 1 file.  ' Directory $145$DIA0:[VMS$COMMON.SYSEXE]    SYSBOOT.EXE;1        (113,1,0)   Total of 1 file.   Directory $145$DIA1:[SYSEXE]   SYSBOOT.EXE;1        (113,1,0)   Total of 1 file.  & Grand total of 4 directories, 4 files.  N This does appear to be a major flaw in the on-line BACKUP keeping in mind that2 stand-alone version does not exhibit this problem.  # Any and all comments appreciated...    Barry    --  ? Barry Treahy, Jr  *  Midwest Microwave  *  Vice President & CIOm  A E-mail: Treahy@mmaz.com * Phone: 480/314-1320 * FAX: 480/661-7028    ------------------------------  # Date: Wed, 09 May 2001 18:19:52 GMT 8 From: hammond@not@peek.ppb.cpqcorp.net (Charlie Hammond)0 Subject: Re: VMS 7.2 BACKUP damaging system disk0 Message-ID: <cTfK6.63$fi2.2175@news.cpqcorp.net>  ) In article <3AF97871.AF481058@mmaz.com>, o- "Barry Treahy, Jr." <treahy@mmaz.com> writes:  ..D >This ... does not determine if the proper aliases  from VMS$COMMON G >to the site specific roots of [sysX.syscommon] exist or are correct.  u ..  5 Although this does not appear to be the problem here,a. A quick test for correct root aliases is this:  #     Execute the following commands:l  E $ DIRECTORY/FILE_ID/NOHEADING/NOTRAILING <dev>:[000000]VMS$COMMON.DIRoB $ DIRECTORY/FILE_ID/NOHEADING/NOTRAILING <dev>:[SYS*]SYSCOMMON.DIR     ?     If you are booted booted from the system disk, you can use n"     SYS$SYSDEVICE for the "<dev>".  4     All of the reported FILE IDs should be the same.  @ It is possible that there could be problems which this test will? not detect, but if it fails -- i.e. if the file ids are NOT alln* the same -- you have a mess on your hands.  @ BTW, it is an EXCELLENT idea to do this test before upgrading anE OpenVMS system disk.  (Alpha or VAX)  If the direcory alias structuree? is not correct, the upgrade process can make a real mess of thee system disk.  G BTW, "mess" is a technical term; the definition is left as an exercise.o  @ And while I'm at it -- *ALWAYS* have a known good backup of your system disk before upgrading.n   -- tK     Charlie Hammond -- Compaq Computer Corporation -- Pompano Beach  FL USAtH        (hammond@not@peek.ppb.cpqcorp.net -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  $ Date: Wed, 9 May 2001 16:19:05 -04002 From: "Sue Skonetski" <susan.skonetski@compaq.com>, Subject: Re: VMS 7.3 docs on OpenVMS webite?0 Message-ID: <HChK6.73$fi2.2221@news.cpqcorp.net>  L According to Warren Sander the OpenVMS Web master the docs will be up on the Web site Saturday.   Sue     8 "Ryan Moore" <rmoore@rmoore.dyndns.org> wrote in message6 news:Pine.LNX.4.31.0105081837100.1566-100000@jaipur...I > When are the VMS 7.3 docs going to show up on the OpenVMS documentationeL > website (http://www.openvms.compaq.com:8000/index.html)?  It would be niceK > to start reading the Release Notes and New Features manuals before we getl > our hardcopies.u >r > -Ryant >n   ------------------------------  + Date: Wed, 09 May 2001 13:56:12 -0600 (MDT)x" From: John Nebel <nebel@csdco.com> Subject: vmsindex/query problemeG Message-ID: <Pine.OSF.4.21.0105091323250.12024-100000@athena.csdco.com>   G I have been testing vmsindex and query to be used for a web site searcha and it works pretty well.t   Searches of the form:    court and (supreme or burrito)  / fail although the docs give that as an example.e   court (supreme or burrito) s  I does work as intended.  I must be doing something wrong as query has beenoI around for a long time and should be fine.  Can anyone set me straight ont this?l   Thanks,s  
 John Nebel   ------------------------------  # Date: Wed, 09 May 2001 19:57:38 GMT40 From: sander@vmsbiz.enet.dec.com (Warren Sander)N Subject: Re: www.openvms.compaq.com doesn't display correctly in VMS Navigator0 Message-ID: <SihK6.70$fi2.2186@news.cpqcorp.net>   |>E |>The VMS server is a lot better than the idiots across at the Q. Andf= |>thank you for checking, there are just so many who don't...  |>N That is a constant issue with them. They live in a MS world. I'm not going to N say bad or good it's just what is. They use lots of tools and very few of them actually know HTML at all. a   |>E |>> In fact there is NO 'microsoft' type product being used by me forl. |>> web management, creation or anything else. |>F |>Oh? so who is the tooth fairy who drops all the PPT files under yourF |>pillow? Or the COE presentations? I suspect you have missed a few...  J I didn't save I don't have intel systems. I do and I do use Powerpoint etcI to do 'save-as a web page' for presentations. That's why you can see them G with 3.03b. You can download the presentations also. Notice that those  N presentations are a lot different than what you get with vanilla powerpoint. IL did a lot of hacking to put in the headers and trailers etc. And most of allH number the slides not just put '.' next to each slide title. I also get P a lot of word documents that often it's easier to save as text and format myselfO rather than mess around with html files that word makes that are about 10 timesw bigger than they need to be...   |>E |>So how many of these non NS3 hits are from VMS systems? Perhaps VMStF |>might be a metric of some small value on a VMS web site. We allreadyA |>know it counts for nothing to the idiots at the compaq main webw	 |>server.e |>  G I use webtrends to do my stats. Unfortunatly webtrends don't catagorizesG OpenVMS, or Tru64 Unix as seperate listings. They dump them into Other.l= I've sent a request in but haven't heard an answer from them.1     |>G |>Simple. Give them a Vax, current VMS and the *STANDARD* browser. HookEK |>it up to a 28K serial line. Turn off Jave and JS, or implement a complete & |>protected subsystem for the browser. |>H They have an Alpha. We sent a system down and one of the contactors set L it up with OpenVMS and Tru64 and Linux. They did do some stuff. Unfortunatly/ with the layoff's that contractor is gone now. e -- cB ------------------------------------------------------------------6 Warren Sander                        OpenVMS MarketingD Compaq Computer Corporation          Work:  warren.sander@compaq.comE 200 Forest Street MR01-3/J1          Personal: sander@ma.ultranet.comt3 Marlboro, MA 01752                   (508) 467-4875b6    My opinions are my own and I only speak for myself -          Read http://www.openvms.compaq.com/ -B ------------------------------------------------------------------   ------------------------------  $ Date: Wed, 9 May 2001 17:00:14 -0400; From: "Brian Tillman" <tillman_brian@notnoone.notnohow.com>CN Subject: Re: www.openvms.compaq.com doesn't display correctly in VMS Navigator$ Message-ID: <3af9afe0$1@news.si.com>  G >I dropped by the top level VMS page for the first time in a long whilen
 >today, at > ! >  http://www.openvms.compaq.com/  >XG >using the Navigator 3.03 program on VMS.  And it popped up a file saveeK >window for a file called "survey83.js". Javascript was enabled.  It didn't G >say "error" but I'm going to call it that, since there's absolutely noh+ >point in saving a javascript file to disk.   H I visited this site just because of your posting.  I used Navigator GoldL V3.03 on an OpenVMS VAX system.  I got a survey pop-up that asked me several; questions.  At no time did Navigator ask me to save a file.d  # In other words, "it worked for me".m --A Brian Tillman                   Internet: tillman_brian at si.comoA Smiths Aerospace                          tillman at swdev.si.comh= 3290 Patterson Ave. SE, MS      Addresses modified to preventk< Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@"8        This opinion doesn't represent that of my company   ------------------------------  " Date: Wed, 9 May 2001 20:56:59 GMT/ From: "Richard L. Dyson" <rick-dyson@uiowa.edu> N Subject: Re: www.openvms.compaq.com doesn't display correctly in VMS Navigator) Message-ID: <3AF968CB.3EFE34F4@uiowa.edu>N   Warren Sander wrote:G > |>So how many of these non NS3 hits are from VMS systems? Perhaps VMSIH > |>might be a metric of some small value on a VMS web site. We allreadyC > |>know it counts for nothing to the idiots at the compaq main webu > |>server.g > |> > I > I use webtrends to do my stats. Unfortunatly webtrends don't catagorizebI > OpenVMS, or Tru64 Unix as seperate listings. They dump them into Other.y? > I've sent a request in but haven't heard an answer from them.o  E    Analog catalogs OpenVMS OS browsers (VAX and Alpha, separately)...   G    It runs everywhere (including OpenVMS) and handles about all the log  types.   	http://www.analog.cx/   or/ 	http://www-pi.physics.uiowa.edu/~dyson/analog/h   Rick -- tH Richard L. Dyson                                    rick-dyson@uiowa.eduH  _   _      _____                http://www-pi.physics.uiowa.edu/~dyson/H | | | |    |_   _|   Systems Analyst                     O: 319/335-1879H | | | | of   | |     The University of Iowa            FAX: 319/335-17536 | \_/ |     _| |_    Department of Physics & Astronomy-  \___/     |_____|   Iowa City, IA 52242-1479-   ------------------------------  % Date: Thu, 10 May 2001 01:30:17 +0200t) From: Christof Brass <brass@infopuls.com> N Subject: Re: www.openvms.compaq.com doesn't display correctly in VMS Navigator, Message-ID: <3AF9D309.7FCE4CA1@infopuls.com>   Warren Sander wrote: > L > The web server is configured correctly. I didn't get the mime type updatedU > on the external server. I did get it on my internal copy and I DO TEST EVERY SINGLEuG > PAGE ON THE SERVER WITH NETSCAPE 3.03B with Java. I don't use the vaxr > version anymore. > M > I don't know where the 'microsoft' comment came from. The server is and has Q > been Purveyor for OpenVMS for over 5 years. It is running on an OpenVMS system.  > S > In fact there is NO 'microsoft' type product being used by me for web management,m > creation or anything else. > M > I use EDT and/or TPU to edit my pages. I have for 5 years and while I mightFU > be moving to newer technology over the summer I still plain to make sure everything  > works with Netscape 3.03b. > O > And also fyi.. On a monthly average the web site gets around 4.5 million hitsnK > per month. Of those hits, more than 3 million come from IE type browsers,KL > and 800,000 from netscape type browers (the rest are various other browers > including Opera, Mosaic etc. > G > Of the 800,000 hits from netscape 3.x versions are about 60,000 hits.  > N > That's less than 10% of the Netscape hits and 1.5% of the total hits for the > month. > P > There's lots of things that we can't do on the site in order to make sure thatS > 3.03b continues to work. This is for the OpenVMS site only not all of compaq.com.n > O > I added in the new survey in order to try to get some OpenVMS customer voices M > into the folks in Houston. I didn't make up the survey and have very littletQ > input into it but I felt it was better to have you all have the ability to give O > some feedback and represent the OpenVMS customer base to the compaq.com folkseT > rather than not do anything at all. Yes the survey JS etc were made for and testedN > only on IIS type servers and IE browers. I had to make a couple of tweaks toQ > make Purveyor spit the file out correctly (one of which was the .js mime type).. > 	 > -warrent   [SNIP]  4 Thanks. Continue your good work and keep your style.  @ I also write HTML mostly with a plain text editor. This ensures = efficient HTML coding and slim pages which load fast on everyl browser.   > --D > ------------------------------------------------------------------8 > Warren Sander                        OpenVMS MarketingF > Compaq Computer Corporation          Work:  warren.sander@compaq.comG > 200 Forest Street MR01-3/J1          Personal: sander@ma.ultranet.comr5 > Marlboro, MA 01752                   (508) 467-4875d7 >    My opinions are my own and I only speak for myselfl. >          Read http://www.openvms.compaq.com/D > ------------------------------------------------------------------   ------------------------------  % Date: Wed, 09 May 2001 19:57:09 +0200 - From: Didier Morandi <didier.morandi@ubs.com>s( Subject: [FUN] DCL minute of the evening' Message-ID: <3AF984F5.30862603@ubs.com>o  9 (from Nick Barnes, UBS Zrich. Posted without permission)e  B OK, I've got a problem which you may want to have a go at solving.  K I've got a plain text file which contains two columns. Column 1 contains a eP primary key and is used for looking up in order to return the data in column 2. N The data in column 1 is variable length and may be several hundred characters  long.0   e.g.   321,"Nicholas Barnes"a 10s1,"Jo Brown"s 34d45,"Tom Smith":( xdbvdafgdm4657hemt7utyb517,"Fred Bloggs"O 24sdf6b5sh46nd2nf4uj3fu5mui1l4j65768413154574687689617135431876981651,"Sam Fox"  .i .f .a  ' The file is several hundred lines long.h  M I need to find a way of performing multiple lookups on this file from DCL in aP the fastest time possible. Ideally, one hundred lookups should take, at most, a  second.s  P I would prefer the program doing the lookup should do no more than, for example:   	@lookup "321"   and have the symbolt  # 	lookup_result == "Nicholas Barnes"t  	 returned.e  H The challenge, therefore, is who can come up with the fastest method of M performing this lookup. I don't expect you to write code (though if you can,   that'd be a bonus).   	 Have fun.h  	 Nicholas.c   ------------------------------  % Date: Wed, 09 May 2001 20:31:14 +0200I- From: Didier Morandi <didier.morandi@ubs.com>a, Subject: Re: [FUN] DCL minute of the evening' Message-ID: <3AF98CF2.E62FB606@ubs.com>n   Didier Morandi wrote:h > ; > (from Nick Barnes, UBS Zrich. Posted without permission)s > D > OK, I've got a problem which you may want to have a go at solving. ../...   A solution:r   $!+e $! FTMYD.COM! $! Faster Than Me You Die dot coms $!
 $! <input>1 $! a variable length key from 1 to 255 chars long  $! $! <output>f" $! the data matching the given key $! $! <side effects>  $! Nick will be amazed (I hope)r $! $! Revision historya $!( $! Version Date        Author     actionO $! ------- ----------- ---------- --------------------------------------------- * $! V1.0-0  09-may-2001 D. Morandi creation $!- 
 $ set noon $ say = "write sys$output"   $ type/page sys$inputr $DECKu   Hallo zm,e  N This (fast'n fancy) procedure will fetch data via a variable length  key. The N actual  record is a random text composed of the key plus additional text. The M TEXT.TXT  file is 5000 lines long and has a random key length. Actually it iseM NOT  an indexed file, just plain text with a random position comma delimitingaM the "key". For your information, the random position is f$ext(21,2,f$time()).m  M Select  now  a  key among the ones which will be displayed on the next screenDM using  CUT/PASTE.  The  key  string goes from char 1 to the comma. You may or N may not select the comma, Then paste the key when prompted for it, then press  RETURN.p   $EODG $ inq dummy "Press <RETURN> to start"	!the text is 5000 lines selected o@ $ search/nohighlight test.txt bb,cc	! from lib/ext=* helplib.hlb2 $ inq key "Enter key"			! then sorted/merged/nodupG $ if key .eqs. "" then goto HUH		!  then spaces removed and 1 "," addedu $!-------------------- $ start_time=f$time()4 $ key = key - ",":$ $ define/user sys$output result.temp $ define/user sys$error nl:D $ search test.txt "''key',"- $ open/read ch result.temp $ read ch line
 $ close ch% $ lookup_symbol=f$element(1,",",line)t $ end_time=f$time(), $!-------------------- $ type/page nl:C $ say "Your key was:" 	 $ say keyI $ say "". $ say "Key length is ",f$length(key)," chars." $ say "" $ say "your data are:" $ say lookup_symbol  $ say ""G $ say "data length is ",f$length(lookup_symbol)+f$length(key)," chars."g $ say "" $ say "The whole record was:"o
 $ say line $ dele_ result.temp;*o $ say ""( $ say "The query started on ",start_time& $ say "and ended on         ",end_time $ say "" $ say ""6 $ say "Have a good day. Thank you for DCLing with us." $ say "" $ exit $HUH:. $ say "" $ say "Hahaha."O $ exit  K It seems that this solution goes much faster than creating a 255 chars long.F primary key padded with spaces, then convert/fdl the file then then...  N John McLean did faster reading the file in memory but this doesn't work with a 100 000 lines file.   ; It also appears that a read/key is faster for a small file.a   Enjoy the challenge.   D.   ------------------------------   Date: 9 May 2001 19:30:33 GMT 2 From: mathog@seqaxp.bio.caltech.edu (David Mathog), Subject: Re: [FUN] DCL minute of the evening, Message-ID: <9dc5sp$2de@gap.cco.caltech.edu>  W In article <3AF984F5.30862603@ubs.com>, Didier Morandi <didier.morandi@ubs.com> writes:2: >(from Nick Barnes, UBS Zrich. Posted without permission) >eC >OK, I've got a problem which you may want to have a go at solving.n >(L >I've got a plain text file which contains two columns. Column 1 contains a Q >primary key and is used for looking up in order to return the data in column 2. oO >The data in column 1 is variable length and may be several hundred characters e >long. >p >e.g.s >  >321,"Nicholas Barnes" >10s1,"Jo Brown" >34d45,"Tom Smith") >xdbvdafgdm4657hemt7utyb517,"Fred Bloggs"hP >24sdf6b5sh46nd2nf4uj3fu5mui1l4j65768413154574687689617135431876981651,"Sam Fox" >..s >..: >..  >x( >The file is several hundred lines long.  F Ok, the file is small, but the primary key is very poorly designed forL rapid retrieval.  Let's hope that strings containing "," are not valid keys.I You might want to start by redoing the data format so that you have threehV fields, a fixed length hashed key, the full key, and the data.  If you choose the hashK properly ( big enough) you'll get no duplicates.  Then you could store the -> data in an indexed file and do retrievals using hashed keys.     >uN >I need to find a way of performing multiple lookups on this file from DCL in Q >the fastest time possible. Ideally, one hundred lookups should take, at most, a t >second.  I That's pushing it for DCL, or actually any other language, if this has tocK be read from disk and no file caching is involved.  You're only allowing 10-G ms/retrieval which is perilously close to the seek time for some disks.nH Plus each operation one will need to retrieve the DCL procedure too.  IfA you can put both the data and the DCL procedure on a RAMdisk thisn constraint will go away. l   >nQ >I would prefer the program doing the lookup should do no more than, for example:- >- >	@lookup "321"e >9 >and have the symbol > $ >	lookup_result == "Nicholas Barnes" >m
 >returned.  F Here's a skeleton, fill in the blanks, most notably, the hash function+ you want to use for making the primary key.d   $e $ call make_index_key P11 $ open/read ifil: RAMDISK:[directory]datafile.idx I $ read/key='calculated_hash'/error=handlethis/end=handlethis ifil: string.
 $ close ifil:0! $ realkey=f$element(0,",",string)t* $ if(realkey .nes. P1)then goto handlethat* $ lookup_result == f$element(1,",",string) $ exit
 $ handlethis:n $   close ifil:u
 $ handlethat:o! $   lookup_result == "____BAD___"d $ exit u    K It would probably be easier to do this in some compiled language, and then -9 you could install the program to pick up a bit of speed. M   Regards,   David Mathog mathog@caltech.edu? Manager, sequence analysis facility, biology division, Caltech 2   ------------------------------  # Date: Thu, 10 May 2001 00:50:53 GMTe. From: "Alphaman" <alphaman64@nixspam-home.com>, Subject: Re: [FUN] DCL minute of the evening9 Message-ID: <NBlK6.1637$I5.539371@news1.rdc1.tn.home.com>v  K I am assuming that your primary keys are unique.  If not, loop-ify the READ.$ loop to return each possible result.  K The concept is to "create" a key on the fly -- just restrict your search totJ the user entered data with a comma appended to it, where the key length isD the length of the user entered data + 1 (the comma) and the starting? position is column 1.  The slowest part of this is probably the'K open/read/close of the temporary data file.  It might be much faster if youc) run the output of the search command in ae  C $ pipe (search... | (inquire temp ; def /job lookup_results &temp )h  L Then just parse the logical's contents instead of the RECORD symbol obtained/ through the opening & closing of the DATA file.r  
 I-win.Com:   $       set noon  $       if p1 .eqs. "" then exit $       p1 = p1 + ","  $       P1Length = f$length(p1)s  $       search names.dat "''p1'"3 /key=(pos=1,size='P1Length')/output=TempResults.datt& $       open/read Data TempResults.dat% $       read /end=AllDone Data Recorde2 $       Lookup_Result == 'f$element(1,",",Record)'& $       write sys$output Lookup_Result
 $ AllDone: $       close Data  $       delete TempResults.dat;* $       exit  $ Thanks for all the good ideas, guys!   BTW, what's my prize?h Aaron  --> Aaron Sakovich  http://members.home.net/sakovich/alphaman.html> Make April 15 just another day:        http://www.fairtax.org/H "The supersonic boom should hit you in just a few seconds." (Apollo 440)  8 Didier Morandi <didier.morandi@ubs.com> wrote in message! news:3AF984F5.30862603@ubs.com...f; > (from Nick Barnes, UBS Zrich. Posted without permission)r > D > OK, I've got a problem which you may want to have a go at solving. >aL > I've got a plain text file which contains two columns. Column 1 contains aG > primary key and is used for looking up in order to return the data inm	 column 2.uD > The data in column 1 is variable length and may be several hundred
 characters > long.m >  > e.g. >v > 321,"Nicholas Barnes"d > 10s1,"Jo Brown"n > 34d45,"Tom Smith"d* > xdbvdafgdm4657hemt7utyb517,"Fred Bloggs"L > 24sdf6b5sh46nd2nf4uj3fu5mui1l4j65768413154574687689617135431876981651,"Sam Fox" > .a > .w > .o >h) > The file is several hundred lines long.r >tK > I need to find a way of performing multiple lookups on this file from DCLf inI > the fastest time possible. Ideally, one hundred lookups should take, at  most, ah	 > second.a >eI > I would prefer the program doing the lookup should do no more than, forT example: >x > @lookup "321"e >u > and have the symbolt >l$ > lookup_result == "Nicholas Barnes" >- > returned.- >-I > The challenge, therefore, is who can come up with the fastest method of I > performing this lookup. I don't expect you to write code (though if you: can, > that'd be a bonus).0 >o > Have fun.R >3 > Nicholas.P   ------------------------------  % Date: Wed, 09 May 2001 22:27:08 -0400)- From: JF Mezei <jfmezei.spamnot@videotron.ca>y, Subject: Re: [FUN] DCL minute of the evening+ Message-ID: <3AF9FC78.87C8A91@videotron.ca>-   > > 321,"Nicholas Barnes"- > > 10s1,"Jo Brown"- > > 34d45,"Tom Smith"-  J If the data is fairly static, I would consider converting the data into anI indexed file. You can use F$ELEMENT to extract the key and value and thenaG write a record to the indexed file (making the key size large enough tod contain the longest key.  M Then when you need to use the data ,it is just a single READ command, and you.? can then extract the value from a fixed location in the record.i   ------------------------------  # Date: Thu, 10 May 2001 05:41:53 GMTr From: Dirk Munk <munk@home.nl>, Subject: Re: [FUN] DCL minute of the evening' Message-ID: <3AFA2A20.DB5BF745@home.nl>g  R You may have a big problem, since (AFAIK) the maximum record lenght you can use in DCL is 255 bytes.    Didier Morandi wrote:y  ; > (from Nick Barnes, UBS Zrich. Posted without permission)  >nD > OK, I've got a problem which you may want to have a go at solving. >dL > I've got a plain text file which contains two columns. Column 1 contains aQ > primary key and is used for looking up in order to return the data in column 2.eO > The data in column 1 is variable length and may be several hundred characters  > long.  >p > e.g. >i > 321,"Nicholas Barnes"o > 10s1,"Jo Brown"  > 34d45,"Tom Smith" * > xdbvdafgdm4657hemt7utyb517,"Fred Bloggs"Q > 24sdf6b5sh46nd2nf4uj3fu5mui1l4j65768413154574687689617135431876981651,"Sam Fox": > .w > .e > .c >a) > The file is several hundred lines long.g >rN > I need to find a way of performing multiple lookups on this file from DCL inQ > the fastest time possible. Ideally, one hundred lookups should take, at most, ai	 > second.m > R > I would prefer the program doing the lookup should do no more than, for example: >h >         @lookup "321"e >r > and have the symbole >e, >         lookup_result == "Nicholas Barnes" >c > returned.  >nI > The challenge, therefore, is who can come up with the fastest method ofeN > performing this lookup. I don't expect you to write code (though if you can, > that'd be a bonus).A >n > Have fun.r >l > Nicholas.I   ------------------------------  % Date: Thu, 10 May 2001 07:48:35 +0200.- From: Didier Morandi <didier.morandi@ubs.com>i, Subject: Re: [FUN] DCL minute of the evening' Message-ID: <3AFA2BB3.50CCF39C@ubs.com>    Alphaman wrote:d >  > BTW, what's my prize?_  N Well, it is pleasant and funny to notice than among three different solutions,N we got here the three solutions we implemented internally at UBS. Mine was theM one I posted: search, John's was the HASH solution and Nick the FDLed file...f  N Your prize is a bottle of French Burgundy "Clos Les Mouches 1979" but you have3 to pick it up in Zrich on the 18th of May only :-)1   D.   ------------------------------   End of INFO-VAX 2001.260 ************************