1 INFO-VAX	Thu, 09 Nov 2000	Volume 2000 : Issue 626       Contents: Re: A DEC history site? " Accessing a OpenVMS server from NT& Re: Accessing a OpenVMS server from NT& Re: Accessing a OpenVMS server from NT! Re: Alpha Hardware Experts...Help - Re: Backup written on TZ87 to be read on TZ86 - Re: Backup written on TZ87 to be read on TZ86  Re: Bug in PIPE: please check. Re: Cluster Problem  Re: Cluster Problem  Re: Cluster Problem  Re: Cluster Problem  Re: Cluster Problem  Re: Cluster Problem  Re: Cluster Problem  Re: Cluster Problem  Re: Curious printing problem Curious printing problem DECforms...  Re: DECforms... 1 FA: Reflection4 For Win DEC VT340/300 ReGIS/Sixel @ Re: Fatal:  Insufficient virtual memory to continue compilation.@ Re: Fatal:  Insufficient virtual memory to continue compilation.@ Re: Fatal:  Insufficient virtual memory to continue compilation.# Fibre Channel Encapsulation/Gateway ) IP-Verbindung von zu Hause in Deutschland , Re: login prompts inhibited with logins=0 ??, Re: login prompts inhibited with logins=0 ??( Re: NCP commands in the 7.2 Cluster doc?( Re: NCP commands in the 7.2 Cluster doc? Re: OpenVMS on 5300  Re: OpenVMS on 5300 = Re: Oracle/VMS performance problems with Multithreaded Server  Re: PERL on VMS   Re: Porting from Unix to OpenVMS  Re: Porting from Unix to OpenVMS Re: RECALL question/suggestion.  RE: RECALL question/suggestion.  Re: RECALL question/suggestion.  Re: RECALL question/suggestion.  Re: remote disconnect on telnet  Re: SHOW PROCESS  Re: Starting VMS install with cd/ Re: Suggestion for DCL scripts which open files / Re: Suggestion for DCL scripts which open files  Re: Sun "uptime" belly laugher+ Re: Transfering users via FTP & email lost.  Re: Use P1 in PIPE! Re: VAXstation 4000-90 Error Code * VMS Fortran question: entering hex numbers. Re: VMS Fortran question: entering hex numbers. Re: VMS Fortran question: entering hex numbers' Re: VMS VMAIL, SMTP transports, headers ' Re: VMS VMAIL, SMTP transports, headers ) Re: Where's C's PRV$M_READALL & _GRPPRV ? # Re: Where's the Y2K compliance page # Re: Where's the Y2K compliance page # Re: Where's the Y2K compliance page # Re: Where's the Y2K compliance page ) Re: www.northernlight.com using Solaris ?   F ----------------------------------------------------------------------  + Date: Wed, 08 Nov 2000 17:15:53 -0500 (EST) . From: Mark Fearer <mfearer@mail.fearernet.com>  Subject: Re: A DEC history site?H Message-ID: <Pine.LNX.4.21.0011081714480.3614-100000@mail.fearernet.com>   > Didier Morandi wrote: K > > Where can I find a site featuring the DEC history? With pictures of the ( > > Mass. plants, the Mill, people, etc. > >      This was *sort* of interesting:  4 	http://www.digital.com/timeline/timeline-57-61.html   -  Mark Fearer    ------------------------------   Date: 08 Nov 2000 23:57:56 GMT# From: casinoop2@aol.com (CasinoOp2) + Subject: Accessing a OpenVMS server from NT : Message-ID: <20001108185756.29203.00000881@ng-cs1.aol.com>  O Does anyone know what is required to access an OpenVMS file server from Windows N NT? What I want to do is map a network drive to a specific VMS directory. WhenM I try to search the network (Network Neighborhood/ Entire Network/ Netware or K Compatible Network) all I get is "Unable to browse the network. The network J path was not found.". What am I not thinking of? What is meant by "networkM path" anyway? I have installed Pathworks and the server is properly providing N the licenses for pwxxwinat07.01, and pwlmxxxca07.02. I can login to the serverO with Reflection X or PowerTerm 525. I can transfer files using ftp. But for the O life of me I cannot map a drive. There must be someone out there that can point  me in the right direction.     Clark Calkins, engineer  Schafer Corporation  ccalkins@schafercorp.com   ------------------------------  % Date: Wed, 08 Nov 2000 16:11:39 -0800 ! From: Shane.F.Smith@Healthnet.com / Subject: Re: Accessing a OpenVMS server from NT D Message-ID: <OF66D6D3BD.FBD79A81-ON88256992.000071CE@foundation.com>  K Check your network card, drivers and connection on the PC. That is a common K symptom of the network card and/or driver having a problem. As a veteran of H countless Quake LAN games, setting up the networks on the fly, I've seenI this many times. You might also want to check you've got the file sharing 5 loaded by trying to share out one of the PC's drives.    Shane           7 casinoop2@aol.com (CasinoOp2) on 11/08/2000 03:57:56 PM    To:   Info-VAX@Mvb.Saic.Com  cc:   , Subject:  Accessing a OpenVMS server from NT    G Does anyone know what is required to access an OpenVMS file server from  Windows I NT? What I want to do is map a network drive to a specific VMS directory.  WhenJ I try to search the network (Network Neighborhood/ Entire Network/ Netware orK Compatible Network) all I get is "Unable to browse the network. The network J path was not found.". What am I not thinking of? What is meant by "networkC path" anyway? I have installed Pathworks and the server is properly 	 providing G the licenses for pwxxwinat07.01, and pwlmxxxca07.02. I can login to the  serverK with Reflection X or PowerTerm 525. I can transfer files using ftp. But for  the I life of me I cannot map a drive. There must be someone out there that can  point  me in the right direction.     Clark Calkins, engineer  Schafer Corporation  ccalkins@schafercorp.com   ------------------------------  # Date: Thu, 09 Nov 2000 00:59:31 GMT 1 From: Bill Hall <bill.hall@nav-international.com> / Subject: Re: Accessing a OpenVMS server from NT ) Message-ID: <8ucsth$ljf$1@nnrp1.deja.com>   D In article <OF66D6D3BD.FBD79A81-ON88256992.000071CE@foundation.com>,$   Shane.F.Smith@Healthnet.com wrote: > F > Check your network card, drivers and connection on the PC. That is a commonB > symptom of the network card and/or driver having a problem. As a
 veteran ofE > countless Quake LAN games, setting up the networks on the fly, I've  seenC > this many times. You might also want to check you've got the file  sharing 7 > loaded by trying to share out one of the PC's drives.  >  > Shane  > 9 > casinoop2@aol.com (CasinoOp2) on 11/08/2000 03:57:56 PM  >  > To:   Info-VAX@Mvb.Saic.Com  > cc:  > . > Subject:  Accessing a OpenVMS server from NT > D > Does anyone know what is required to access an OpenVMS file server from	 > Windows @ > NT? What I want to do is map a network drive to a specific VMS
 directory. > WhenD > I try to search the network (Network Neighborhood/ Entire Network/ Netware  > orE > Compatible Network) all I get is "Unable to browse the network. The  network C > path was not found.". What am I not thinking of? What is meant by  "networkE > path" anyway? I have installed Pathworks and the server is properly  > providing E > the licenses for pwxxwinat07.01, and pwlmxxxca07.02. I can login to  the  > serverE > with Reflection X or PowerTerm 525. I can transfer files using ftp.  But for  > the G > life of me I cannot map a drive. There must be someone out there that  can  > point  > me in the right direction. >  > Clark Calkins, engineer  > Schafer Corporation  > ccalkins@schafercorp.com >  > D You didn't list any version info so I'll make some assumptions.  TheE client licenses you did mention are for Pathworks Advanced Server 6.0 E and Advanced Server 7.2.  These versions of Pathworks Advanced Server A do not support Novell's IPX/SPX transport, only DECnet, tcpip and G NetBUI.  It appears that your NT system may only be configured to use a G Novell IPX/SPX transport.  This is probably why you can't "browse" your  VMS server.   D You need to configure both the client and the server to use a commonG transport.  Which one(s) you can/need to use, I can't tell you based on  the few details you gave.  --	 Bill Hall % Distributed Systems Technical Support $ International Truck and Engine Corp.    & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------   Date: 8 Nov 2000 11:53:56 -0500 9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) * Subject: Re: Alpha Hardware Experts...Help+ Message-ID: <6eh0SMCyIEkv@eisner.decus.org>   N In article <8ubv51$q40$1@nnrp1.deja.com>, fooguy <jweisen@my-deja.com> writes:  G > But now comes the troubling part...shopping for Alpha Hardware. We're H > looking at a DS20, but as we know the trend in hardware (Alpha, Intel,# > Sparc...) is to detached storage.   @ You should be guided by your own needs, rather than "the trend".   ------------------------------  % Date: Tue, 07 Nov 2000 21:13:20 -0600 7 From: "David J. Dachtera" <djesys.nospam@earthlink.net> 6 Subject: Re: Backup written on TZ87 to be read on TZ86- Message-ID: <3A08C4D0.75AE65DE@earthlink.net>    Phillip Helbig wrote:  > = > In article <3A0848D1.2CDB58EE@y12.doe.gov>, "Dale A. Marcy"  > <dqm@y12.doe.gov> writes:  > Q > >      I have a user that has to transfer files via tape from one VAX system to O > > another VAX system on a non regular basis.  He had to do it today, the last R > > time was approximately a month ago, but it is not at a specific interval.  TheO > > problem is that the system that is writing the tape is a VAXstation 4000-60 O > > running VMS V7.1 using a TZ87 drive, so when the tape is mounted on the VAX O > > 4000-500A also running VMS V7.1, but using a TZ86 drive, it cannot read the N > > TZ87 format.  I can and have each time forced the TZ87 to emulate the TZ86Q > > format, but I would like to automate this so that it does not need to be done O > > each time.  The user is using a Backup/Rewind command to create the tape on R > > the writing node.  Is there a qualifier that can be used on the backup commandM > > to force the TZ87 to write at the TZ86 density?  I looked at the /Density R > > qualifier, but I thought it was for the old reel to reel tapes and if it wouldM > > work for this, I did not know the bits per inch to specify to emulate the L > > TZ86.  Any help would be appreciated.  Hopefully, I have provided enoughP > > information here.  I do not have the exact backup command, but could find itG > > out if needed.  I know that the backup command does not specify any  > > /Media_Format. > I > A while back, I asked but never got a (to me) satisfactory answer about F > a similar problem regarding TLZ (DAT) devices.  That is, what is theH > relation between device name, length of tapes, compression, DDS level,D > /MEDIA_FORMAT=COMPACTION, /DENSITY, how much will fit on the tape,- > compatibility between different drives etc.   = Well, on the MOUNT command, the /DENSITY qualifier is largely @ meaningless on all but the newest OpenVMS and all but the newest@ versions of OVMS and DLT drives, with the noteworty exception of	 9-tracks.   D /MEDIA_FORMAT=COMPACTION is, of course, only valid for those devices( which support hardware data compression.  E How much will fit on the tape depends on the integrity of the medium, H the blocksize, the number of datasets (and the tape marks which separate them) and other factors.    H When hardware compaction is employed, this is further complicated by theD nature (compressibility) of the data being recorded: .ZIP, .GZ, etc.H files will not compress much at all. RMS indexed files where the data isA stored on disk in a compressed form (data/index compression) will E compress a bit more (DCX is not very efficient), but still not a lot.   B How much you get on a tape is how much you get on a tape. It's notH easily predictable with any degree of accuracy. Believe me - I've tried.F Experience with a specific body of data is usually the best indicator.  " Sorry - that's just the way it is.   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/   F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.    ------------------------------  % Date: Thu, 09 Nov 2000 07:31:23 +0800 4 From: Dave Sneddon - bigpond <dbsneddon@bigpond.com>6 Subject: Re: Backup written on TZ87 to be read on TZ86+ Message-ID: <3A09E24B.538BD2B8@bigpond.com>    Phillip Helbig wrote:  > = > In article <3A0848D1.2CDB58EE@y12.doe.gov>, "Dale A. Marcy"  > <dqm@y12.doe.gov> writes:  > 
 	[...snip...]    I > A while back, I asked but never got a (to me) satisfactory answer about F > a similar problem regarding TLZ (DAT) devices.  That is, what is theH > relation between device name, length of tapes, compression, DDS level,D > /MEDIA_FORMAT=COMPACTION, /DENSITY, how much will fit on the tape,- > compatibility between different drives etc.   E The following is from a card "Care and Maintenance of your DAT drive" 6 (EK-SM2TE-RC.A01) that comes with most new DAT drives.  ( device		usable tapes		capacity (my info, 					i think it is correct)    TLZ04		60m			1.3GB TLZ06/TKZ6L	60m,90m			2GB/4GB ! TLZ07/TLZ7L	60m,90m,120m		4GB/8GB ! TLZ09/TLZ9L	60m,90m,120m		4GB/8GB ' TLZ10/TLZ1L	60m,90m,120m,125m	12GB/24GB  					nocompaction/compaction 60m and 90m are DDS-1  120m are DDS-2 125m are DDS-3   --   Regards, Dave. I ------------------------------------------------------------------------- I David B Sneddon (dbs)  OpenVMS Systems Programmer   dbsneddon@bigpond.com I DBS software at ...   http://www.users.bigpond.com/dbsneddon/software.htm I "Life is what happens to you while you're busy making other plans" Lennon    ------------------------------  $ Date: Wed, 8 Nov 2000 21:07:05 +0100" From: "Hans Vlems" <hvlems@iae.nl>' Subject: Re: Bug in PIPE: please check. ( Message-ID: <8ucb60$ddk$1@news.IAEhv.nl>  7 Confirmed that behaviour on a VAX/VMS 7.2 (no patches).    $ pipe ( dir & ) $ pipe ( show time & ) $ pipe ( show time & ) $ pipe ( show time & ) $ pipe ( show time & ) $ pipe ( show time & ) $ pipe ( show time & )    8-NOV-2000 19:58:00 $ pipe ( show time & )    8-NOV-2000 19:58:01 $ pipe ( dir & ) $ pipe ( dir & ) $ pipe ( dir & ) $ pipe ( dir & ) $ pipe ( dir & ) $ pipe ( dir & ) $ pipe ( dir & ) $   
 Hans Vlems  9 Gotfryd Smolik, VMS lists heeft geschreven in bericht ...  > 8 > Have found a definite PIPE inconsistency; regarding to >HELP the syntax:  >  >$ PIPE (command &)  > ? > even, if may be problematic say if not required extra spaces:  >  >$ PIPE ( command & )  > ! > is correct and suposed to work.  > 9 > The example will not work for me for any command, where < >uses user-mode program (at least haven't see any time work)< >and works randomly with supervisor-mode command and command >procedures, as in examples: > # >$ pipe ( dir & ) ! Will never work  > , >$ pipe ( show time & ) ! Will sometime work >   8-NOV-2000 11:21:05  >$ pipe ( show time & )  >$ pipe ( show time & )  >   8-NOV-2000 11:21:06  > ? > The "probability of work" is not depended of the time between  >PIPE commands.  > ? >BTW: The expected bahaviour is usefull for "command shortcut", : > when allows parallel run of "in the process" and "out of9 > the process" command in one (the examples are minimised  > for check !).  > B > System: AXP VMS 7.2-1, not patched, but have check all suspected- >READMEs for the word PIPE and haven't found.  >  > Any comments ?? A > Can anyone confirm on differrent VMS version the bug, or inform ( >that with some patches PIPE will work ? >  > Regards - Gotfryd  >  >-- F >=====================================================================G >$ ON F$ERROR("LANGUAGE","ENGLISH","IN_MESSAGE").GT.F$ERROR("NORMAL") -  > THEN EXCUSE/OBJECT=ME / >$!                        GS@stanpol.zabrze.pl F >===================================================================== >  >  >    ------------------------------   Date: 8 NOV 2000 19:07:53 GMT + From: Dave Greenwood <greenwoodde@ornl.gov>  Subject: Re: Cluster Problem1 Message-ID: <8NOV00.19075364@feda34.fed.ornl.gov>   + helbig@astro.rug.nl (Phillip Helbig) wrote: B > In article <8NOV00.15382223@feda34.fed.ornl.gov>, Dave Greenwood! > <greenwoodde@ornl.gov> writes:   >   0 > >  $	if .not. f$getdvi(dev,"exists") then exit >   	 > Why not  >   F > $	if .not. f$getdvi(dev,"exists") .or. f$getdvi(dev,"mnt") then exit >   B > in case it has already been mounted (e.g. in SYPAGSWPFILES.COM)? >  i; > What is the difference between f$getdvi(dev,"exists") andn2 > f$getdvi(dev,"avl")?  Which should be used here?  C I'll have to admit that I'm not really sure what causes a device totF fail the "avl" test.  I couldn't find any devices which exist but failC "avl" on the systems I tried.  I did find that if a host leaves the G cluster its disks still exist and are "avl" even though the host itselfl is down.   So a better sequence might bel     if f$getdvi(dev,"exists")n   then     if f$getdvi(dev,"avl")     then#       if f$getdvi(dev,"host_avail")i
       then         ...   C You need to check "exists" first, because the "avl" check will givelE a NOSUCHDEV error if the device doesn't exist.  Which is also why you.E don't want to combine the "exists" and "mnt" clauses in a single "if"r statement as you show above.  E The other problem with the "mnt" test is that your system device willeE be mounted locally (and therefore will pass the "mnt" test) but won'tC be mounted remotely.  9 > Does "mnt" only refer to mounted on the node from whichp  > f$getdvi(dev,"mnt") is issued?  > Yes - "mnt" only refers to the mount status on the local node.  @ Btw, another issue you might want to think about is what happens? if node A crashes and leaves the cluster for a long enough timeqD that the other nodes put its disks in mount verify timeout.  I think> you have to dismount the disks on all other nodes in order forC mount/cluster to remount the disks.  I don't know of a way for nodep@ A to detect that other nodes have A's disks in MntVerifyTimeout.  E It's also possible that if you're re-executing your mount script thatiC your local node may have disks in MntVerifyTimeout for some reason. 6 Here's how I check for MntVerify and MntVerifyTimeout:     $ mntvfy = %x4000n   $ valid  = %x0800a   $ sts = f$getdvi(vol,"sts")sE   $ if (sts .and. mntvfy) .eq. mntvfy then ... ! Disk is in MntVerifyrL   $ if (sts .and. valid) .ne. valid then ...   ! Disk is in MntVerifyTimeout   Dave --------------9 Dave Greenwood                Email: Greenwoodde@ORNL.GOVnH Oak Ridge National Lab        %STD-W-DISCLAIMER, I only speak for myself   ------------------------------   Date: 8 Nov 2000 19:19:00 GMT-2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: Cluster Problem6 Message-ID: <8uc8v4$71i$3@mailint03.im.hou.compaq.com>  [ In article <8u9n1i$a5u$4@info.service.rug.nl>, helbig@astro.rug.nl (Phillip Helbig) writes:i5 :...Imagine a cluster where all disks are physically rE :connected only to one node.  Unless one does some fancy footwork to  G :synchronise the booting of several machines, when a given node boots, iG :some disks will be either already mounted on other nodes which booted -G :earlier, or will not be available if connected to nodes which haven't oI :yet booted.  Either way, it seems that each machine, when it boots, has 2I :to MOUNT the disks somehow.  Of course, if one waits until all machines qA :are up, then one MOUNT/CLUSTER will do it, but things should be n7 :automatic and work whetever machines are currently up.y     SYS$EXAMPLES:MSCPMOUNT.COM  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Wed, 08 Nov 2000 15:13:46 -0500a  From: norm.raphael@jamesbury.com Subject: Re: Cluster Problem4 Message-ID: <C2256991.006E6FA1.00@jklh21.valmet.com>  7 hoffman@xdelta.zko.dec.nospam on 11/08/2000 02:19:00 PMr  / Please respond to hoffman@xdelta.zko.dec.nospamc   To:   Info-VAX@mvb.saic.come cc:o Subject:  Re: Cluster Probleme          L /In article <8u9n1i$a5u$4@info.service.rug.nl>, helbig@astro.rug.nl (Phillip Helbig) writes: 5 /:...Imagine a cluster where all disks are physically0E /:connected only to one node.  Unless one does some fancy footwork tosG /:synchronise the booting of several machines, when a given node boots,9G /:some disks will be either already mounted on other nodes which bootedtG /:earlier, or will not be available if connected to nodes which haven'teI /:yet booted.  Either way, it seems that each machine, when it boots, has I /:to MOUNT the disks somehow.  Of course, if one waits until all machines A /:are up, then one MOUNT/CLUSTER will do it, but things should ben8 /:automatic and work whetever machines are currently up. /  /  SYS$EXAMPLES:MSCPMOUNT.COM:  K The trick in this example is to create a detached process that checks everyeI 15 minutes for new disks to mount.  It is otherwise difficult to make onelH node in a cluster aware that another node may start to serve one or moreG disks that do not necessarily want to be mounted to the entire cluster.i  L Provision should also be made for disks that go mountverifytimeout and whose host then becomes available.w   / O / --------------------------- pure personal opinion --------------------------- M /   Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com  /e   ------------------------------   Date: 8 Nov 2000 19:51:29 GMTe* From: helbig@astro.rug.nl (Phillip Helbig) Subject: Re: Cluster Problem. Message-ID: <8ucas1$5d8$1@info.service.rug.nl>  @ In article <8NOV00.19075364@feda34.fed.ornl.gov>, Dave Greenwood <greenwoodde@ornl.gov> writes: g  - > helbig@astro.rug.nl (Phillip Helbig) wrote:pD > > In article <8NOV00.15382223@feda34.fed.ornl.gov>, Dave Greenwood# > > <greenwoodde@ornl.gov> writes: n > >  r2 > > >  $	if .not. f$getdvi(dev,"exists") then exit > >  m > > Why not- > >  cH > > $	if .not. f$getdvi(dev,"exists") .or. f$getdvi(dev,"mnt") then exit > >  @D > > in case it has already been mounted (e.g. in SYPAGSWPFILES.COM)? > >  o= > > What is the difference between f$getdvi(dev,"exists") and 4 > > f$getdvi(dev,"avl")?  Which should be used here? > E > I'll have to admit that I'm not really sure what causes a device tos > fail the "avl" test.    < Perhaps if it is somehow occupied with another command.  In  SYPAGSWPFILES.COM I have     $ LOOP1: $   ON WARNING THEN GOTO LOOP1 $   WAIT 0000 00:00:00.50o% $   READY = F$GETDVI("DKA300:","AVL")3* $   IF READY .EQS. "FALSE" THEN GOTO LOOP1  ; which I think was suggested in the documentation somewhere.   2 > I couldn't find any devices which exist but failE > "avl" on the systems I tried.  I did find that if a host leaves thenI > cluster its disks still exist and are "avl" even though the host itselfe
 > is down.  E Interesting.  Since my VAXstation won't power up now (see my post in l4 comp.sys.dec), I can't test that at the moment.  :-(   > So a better sequence might be  >  >   if f$getdvi(dev,"exists")  >   then >     if f$getdvi(dev,"avl")
 >     then% >       if f$getdvi(dev,"host_avail")d >       then
 >         ...w   Right!  E > You need to check "exists" first, because the "avl" check will give G > a NOSUCHDEV error if the device doesn't exist.  Which is also why youiG > don't want to combine the "exists" and "mnt" clauses in a single "if"  > statement as you show above.  & Yes, been there, made that mistake....  G > The other problem with the "mnt" test is that your system device will G > be mounted locally (and therefore will pass the "mnt" test) but won't  > be mounted remotely. >k; > > Does "mnt" only refer to mounted on the node from whicho" > > f$getdvi(dev,"mnt") is issued? > @ > Yes - "mnt" only refers to the mount status on the local node.  I Right.  So this would be a problem only when the local system disk is to II be mounted remotely on other nodes.  Hmmmm.  I like the "mnt" test since rG it avoids the fatal error if the device has already been mounted.  The nA system disk will always be mounted locally when the MOUNT.COM is s	 executed.   G Perhaps there is no reason to mount a system disk on another node.  On r. the other hand, perhaps this is useful.  Hmmm.  C A related question---when in the startup sequence is SYS$SYSDEVICE - defined?  I I guess one would need some extra code for the system disk, assuming one n wants other nodes to mount it.  R   ------------------------------   Date: 8 NOV 2000 15:38:22 GMTd+ From: Dave Greenwood <greenwoodde@ornl.gov>d Subject: Re: Cluster Problem1 Message-ID: <8NOV00.15382223@feda34.fed.ornl.gov>a  + helbig@astro.rug.nl (Phillip Helbig) wrote:kH > In article <G3oF1E.BIE@world.std.com>, moroney@world.std.spaamtrap.com > (Michael Moroney) writes:  >  o0 > > helbig@astro.rug.nl (Phillip Helbig) writes: > > M > > >Let me back up a bit.  Imagine a cluster where all disks are physically 0I > > >connected only to one node.  Unless one does some fancy footwork to BK > > >synchronise the booting of several machines, when a given node boots, hK > > >some disks will be either already mounted on other nodes which booted .K > > >earlier, or will not be available if connected to nodes which haven't  M > > >yet booted.  Either way, it seems that each machine, when it boots, has  M > > >to MOUNT the disks somehow.  Of course, if one waits until all machines wE > > >are up, then one MOUNT/CLUSTER will do it, but things should be e; > > >automatic and work whetever machines are currently up.a > > E > > In this case the best thing to do is to do a MOUNT/CLUSTER on thepH > > server when it comes up, and a MOUNT/SYSTEM on the clients when they > > come up. >  n > OK.... >   L > > The MOUNT/CLUSTER will mount the disk on all nodes in the cluster at theI > > time when the server makes its disks available.  The MOUNT /SYSTEM iseG > > for nodes that enter the cluster later than the server and want to tJ > > access the disk. Be sure that there is error handling here if the diskG > > is not available (server offline)  Redundant mounts are effectivelyy > > ignored. >  /C > This means, however, that one has to have different MOUNT.COM or lH > whatever files for clients and servers.  What about having each node, I > when it boots, execute the same .COM file?  In that case, MOUNT/SYSTEM pH > or MOUNT/CLUSTER would have, in the end, the same effect.  The former K > would have less overhead per individual command, but the latter might do aK > more with the first command so that later ones from other nodes would be  
 > ignored. [snip]  J I haven't followed this discussion very carefully so this may be somethingE of a non sequitor, but I'm going to throw in my $.02 anyway.  I use aoH single "MOUNT.COM" for all systems in my cluster.  And I like to be ableG to add/subtract a disk with minimal editing so I make a subroutine callc for each disk, eg:  $ $!call mnt  Device(s)		Label(s)	Name $! $ call mnt  FEDA01$DRB1		USERA $ call mnt  FEDA01$DRB2		USERB  J The "Name" parameter, if present, specifies the logical name for the disk.: If not present then the label is used as the logical name.  B The "mnt" subroutine handles all the /cluster vs /system nonsense. Something like (untested):    $ mnt: subroutine  $      dev = p1
  $	label = p2e"  $	if p3 .nes. "" then lognam = p3"  $	if p3 .eqs. "" then lognam = p2I  $! Can't mount a device if it doesn't exist (probably host isn't up yet)a,  $	if .not. f$getdvi(dev,"exists") then exit3  $! If we host the device then mount it clusterwidem:  $	if f$getdvi(dev,"host_name") .eqs. f$getsyi("nodename")  $	then	qual = "/cluster"f  $	else	qual = "/system"  $	endif'  $	mount '/qual' 'dev' 'label' 'lognam'c  $	exitt  $	endsubroutine   Hth, Dave --------------9 Dave Greenwood                Email: Greenwoodde@ORNL.GOVeH Oak Ridge National Lab        %STD-W-DISCLAIMER, I only speak for myself   ------------------------------   Date: 8 Nov 2000 16:53:13 GMTt* From: helbig@astro.rug.nl (Phillip Helbig) Subject: Re: Cluster Problem. Message-ID: <8uc0dp$1hh$2@info.service.rug.nl>  @ In article <8NOV00.15382223@feda34.fed.ornl.gov>, Dave Greenwood <greenwoodde@ornl.gov> writes: w  - > helbig@astro.rug.nl (Phillip Helbig) wrote:eJ > > In article <G3oF1E.BIE@world.std.com>, moroney@world.std.spaamtrap.com > > (Michael Moroney) writes:  > >   2 > > > helbig@astro.rug.nl (Phillip Helbig) writes: > > > O > > > >Let me back up a bit.  Imagine a cluster where all disks are physically  K > > > >connected only to one node.  Unless one does some fancy footwork to nM > > > >synchronise the booting of several machines, when a given node boots, iM > > > >some disks will be either already mounted on other nodes which booted rM > > > >earlier, or will not be available if connected to nodes which haven't sO > > > >yet booted.  Either way, it seems that each machine, when it boots, has dO > > > >to MOUNT the disks somehow.  Of course, if one waits until all machines gG > > > >are up, then one MOUNT/CLUSTER will do it, but things should be Y= > > > >automatic and work whetever machines are currently up.r > > > G > > > In this case the best thing to do is to do a MOUNT/CLUSTER on theTJ > > > server when it comes up, and a MOUNT/SYSTEM on the clients when they > > > come up. > >  (
 > > OK.... > >  iN > > > The MOUNT/CLUSTER will mount the disk on all nodes in the cluster at theK > > > time when the server makes its disks available.  The MOUNT /SYSTEM is	I > > > for nodes that enter the cluster later than the server and want to  L > > > access the disk. Be sure that there is error handling here if the diskI > > > is not available (server offline)  Redundant mounts are effectivelyd > > > ignored. > >  eE > > This means, however, that one has to have different MOUNT.COM or wJ > > whatever files for clients and servers.  What about having each node, K > > when it boots, execute the same .COM file?  In that case, MOUNT/SYSTEM  J > > or MOUNT/CLUSTER would have, in the end, the same effect.  The former M > > would have less overhead per individual command, but the latter might do nM > > more with the first command so that later ones from other nodes would be & > > ignored. > [snip] > L > I haven't followed this discussion very carefully so this may be somethingG > of a non sequitor, but I'm going to throw in my $.02 anyway.  I use a(J > single "MOUNT.COM" for all systems in my cluster.  And I like to be ableI > to add/subtract a disk with minimal editing so I make a subroutine calli > for each disk, eg: > & > $!call mnt  Device(s)		Label(s)	Name > $!  > $ call mnt  FEDA01$DRB1		USERA  > $ call mnt  FEDA01$DRB2		USERB > L > The "Name" parameter, if present, specifies the logical name for the disk.< > If not present then the label is used as the logical name. > D > The "mnt" subroutine handles all the /cluster vs /system nonsense. > Something like (untested): >  >  $ mnt: subroutine >  $      dev = p1 >  $	label = p2n$ >  $	if p3 .nes. "" then lognam = p3$ >  $	if p3 .eqs. "" then lognam = p2K >  $! Can't mount a device if it doesn't exist (probably host isn't up yet)$. >  $	if .not. f$getdvi(dev,"exists") then exit5 >  $! If we host the device then mount it clusterwide < >  $	if f$getdvi(dev,"host_name") .eqs. f$getsyi("nodename") >  $	then	qual = "/cluster"  >  $	else	qual = "/system"
 >  $	endif) >  $	mount '/qual' 'dev' 'label' 'lognam' 	 >  $	exitd >  $	endsubroutine  G I think this is the elegant solution I am looking for.  One could have dE ONE file to mount common disks (obviously, disks mounted only on one XG node should be handled separately; one could, of course, have these in wF the file and then check the nodename before mounting etc), split into H the calls for individual disks and the subroutine as above.  Yes, makes H sense---if a node hosts a disk, it should mount it /cluster, which will I mount it on all nodes which are up.  If not, it mounts it /system, since "H it will already have been mounted on all other nodes which are up.  The @ test only mounts it if it is visible.  If, for some reason, one I dismounted some disks on one node, one could execute this .com and mount -2 them all again---multiple mounts would be ignored.  # Sounds good.  Still some questions.o  G Would it be better to check if a disk is mounted then mount it if not, fB or just mount away and trust the command will be ignored if it is : already mounted.  (What are the differences in overheads.)  E As in the example above, mount/cluster is nice since it mounts it on dI nodes which are already up, and normally all this has to be done just at eC boot time.  However, could one ALWAYS use mount/cluster?  It would sG result in some extra overhead, and is not necessary, but would it hurt?i   ------------------------------   Date: 8 Nov 2000 17:47:40 GMT?* From: helbig@astro.rug.nl (Phillip Helbig) Subject: Re: Cluster Problem. Message-ID: <8uc3js$32e$2@info.service.rug.nl>  @ In article <8NOV00.15382223@feda34.fed.ornl.gov>, Dave Greenwood <greenwoodde@ornl.gov> writes:    . >  $	if .not. f$getdvi(dev,"exists") then exit   Why notl  D $	if .not. f$getdvi(dev,"exists") .or. f$getdvi(dev,"mnt") then exit  @ in case it has already been mounted (e.g. in SYPAGSWPFILES.COM)?  9 What is the difference between f$getdvi(dev,"exists") andn0 f$getdvi(dev,"avl")?  Which should be used here?  7 Does "mnt" only refer to mounted on the node from whicht f$getdvi(dev,"mnt") is issued?   ------------------------------  # Date: Thu, 09 Nov 2000 06:12:44 GMT 1 From: CSABA  HARANGOZO   <csabah@zipworld.com.au>  Subject: Re: Cluster Problem6 Message-ID: <wfrO5.27$aQ6.8238@nostril.pacific.net.au>  , Dave Greenwood <greenwoodde@ornl.gov> wrote:- > helbig@astro.rug.nl (Phillip Helbig) wrote: C >> In article <8NOV00.15382223@feda34.fed.ornl.gov>, Dave Greenwoods" >> <greenwoodde@ornl.gov> writes:  >>  1 >> >  $	if .not. f$getdvi(dev,"exists") then exity >>  
 >> Why not >>  G >> $	if .not. f$getdvi(dev,"exists") .or. f$getdvi(dev,"mnt") then exit  >>  C >> in case it has already been mounted (e.g. in SYPAGSWPFILES.COM)?I >>  < >> What is the difference between f$getdvi(dev,"exists") and3 >> f$getdvi(dev,"avl")?  Which should be used here?   E > I'll have to admit that I'm not really sure what causes a device toiH > fail the "avl" test.  I couldn't find any devices which exist but failE > "avl" on the systems I tried.  I did find that if a host leaves the   > 	A device mounted privately, maybe ? It exists, but may not be6 	available. Tape drives in use could be good examples. 	I haven't checked this though.  						Csaba     I > cluster its disks still exist and are "avl" even though the host itselfi
 > is down.  
 	[...snip...]V     > Dave > --------------; > Dave Greenwood                Email: Greenwoodde@ORNL.GOV.J > Oak Ridge National Lab        %STD-W-DISCLAIMER, I only speak for myself    I    ----------------------------------------------------------------------oE    * Csaba I. Harangozo     |    'To err is human', said the hedgehogmE    * csabah@zipworld.com.au |           as he dismounted a wirebrush.<I    ----------------------------------------------------------------------t;    EARTH::AUSTRALIA:[SYDNEY]HARANGOZO.CSABA;1, delete? [N]:    ------------------------------  " Date: Wed, 8 Nov 2000 22:43:20 GMT/ From: "Richard L. Dyson" <rick-dyson@uiowa.edu>n% Subject: Re: Curious printing probleml) Message-ID: <3A0982A8.6EBED6A0@uiowa.edu>h   Phil Howell wrote: > M > <abirkett@my-deja.com> wrote in message news:8ubtf5$og9$1@nnrp1.deja.com...h > > Hi,t > >s
 > > VMS 7.1-2i > > UCX 4.2 eco 4h > > L > > We are trying to send a postscript job to a Canon printer over IP and itJ > > doesn't work. Setup as below. (names changed to protect the innocent).  @ 	What Canon printer are you using?  We once demo'd one and foundG that it's EFI Fiery board did not do LPD correctly...  We decided to go-E with someone else.  Is this a high end like a 500 or higher or one ofo the# lower desktop models?-   Rick -- -H Richard L. Dyson                                    rick-dyson@uiowa.eduH  _   _      _____                http://www-pi.physics.uiowa.edu/~dyson/H | | | |    |_   _|   Systems Analyst                     O: 319/335-1879H | | | | of   | |     The University of Iowa            FAX: 319/335-17536 | \_/ |     _| |_    Department of Physics & Astronomy-  \___/     |_____|   Iowa City, IA 52242-1479s   ------------------------------  # Date: Wed, 08 Nov 2000 16:02:50 GMTi From: abirkett@my-deja.com! Subject: Curious printing problemt) Message-ID: <8ubtf5$og9$1@nnrp1.deja.com>t   Hi,s  	 VMS 7.1-2s
 UCX 4.2 eco 4f  H We are trying to send a postscript job to a Canon printer over IP and itF doesn't work. Setup as below. (names changed to protect the innocent).  F Printer queue PRINTER_1, idle, on node::"printer_1:9100", mounted form DEFAULTh>   /BASE_PRIORITY=4 /BLOCK_LIMIT=(1,"") /DEFAULT=(FORM=DEFAULT)G   /LIBRARY=HP_DEVCTL Lowercase /OWNER=[SYSTEM] /PROCESSOR=UCX$TELNETSYMTA   /PROTECTION=(S:M,O:D,G:R,W:RS) /RETAIN=ERROR /SCHEDULE=(NOSIZE)e   $ sho queue/full PRINTER_2= Server queue PRINTER_2, idle, on node::, mounted form DEFAULTm?   /BASE_PRIORITY=4 /DEFAULT=(FEED,FORM=DEFAULT) /OWNER=[SYSTEM]sE   /PROCESSOR=UCX$LPD_SMB /PROTECTION=(S:M,O:D,G:R,W:RS) /RETAIN=ERRORi    D PRINTER_1 is a HP Laserjet and its setup works fine - a standard UCX9 queue. The other is a Canon and set up as an LPD queue inu UCX$LPD_PRINTCAP as;   PRINTER_2|printer_2:\k1         :lf=/SYS$SPECIFIC/UCX_LPD/PRINTER_2.LOG:\b         :lp=PRINTER_2:\          :rm=printer_2:\-D         :rp=print:\                          <- this was recommended,         :sd=/SYS$SPECIFIC/UCX_LPD/PRINTER_2:  C The real problem is that when you send postscript data to PRINTER_2 E contained in a file that has RAT:FTN set the file doesn't print. When:F you convert it to being standard carriage control type format it does.  @ Has anybody seen this, and if so, how is it cured. The output isA generated by the application and so changing it is not an option.l   Thanks,p   Aden --D Intelligence is being able to listen to the William Tell overture...5                ...without thinking of the Lone Ranger$   --D Intelligence is being able to listen to the William Tell overture...7                  ...without thinking of the Lone Rangert    & Sent via Deja.com http://www.deja.com/ Before you buy.P   ------------------------------  # Date: Wed, 08 Nov 2000 21:04:28 GMT,% From: Kevin Iliff <kevin@digikey.com>? Subject: DECforms...+ Message-ID: <3A09BFF5.559CBD32@digikey.com>    Hello,  G Are there any DECforms users, or former DECforms users, monitoring thismF newsgroup?  Our company is moving from VTs to PCs, so we're looking toE replace DECforms with something more PC compatible.  Does anyone have B any insights as to what our options might be?  Does anyone know of( another newsgroup that might be helpful?   Thanks,c   Kevin Iliffw Senior Software Engineer Digi-Key Corporation kevin.iliff@digikey.comt   ------------------------------  % Date: Wed, 08 Nov 2000 18:56:47 -0400i- From: JF Mezei <jfmezei.spamnot@videotron.ca>o Subject: Re: DECforms..., Message-ID: <3A09DA20.B05EEBAD@videotron.ca>   Kevin Iliff wrote: >  > Hello, > I > Are there any DECforms users, or former DECforms users, monitoring this H > newsgroup?  Our company is moving from VTs to PCs, so we're looking to5 > replace DECforms with something more PC compatible.   G Haven't used DECforms, but you might want to search the compaq site for M deforms info. I beleive that they now have a web front end that allows you tox use decforms through the web.    ------------------------------  # Date: Wed, 08 Nov 2000 13:55:57 GMT - From: "Compusync" <hsappleton@sprintmail.com>,: Subject: FA: Reflection4 For Win DEC VT340/300 ReGIS/SixelF Message-ID: <NXcO5.31717$Pw6.2453990@newsread1.prod.itd.earthlink.net>  @ http://cgi.ebay.com/aw-cgi/eBayISAPI.dll?ViewItem&item=491691846G Auction Ends on:                 Thursday, Nov 16, 2000 at 18:22:03 PSTd   ------------------------------   Date: 8 Nov 2000 13:38:42 PSTtT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)I Subject: Re: Fatal:  Insufficient virtual memory to continue compilation. 3 Message-ID: <u9z9CScGKOVI@mccdev.slac.stanford.edu>d  * In article <3A09236C.498981D3@gtech.com>, D     	Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com> writes: > Sapper wrote:rG >> Anyone know how to fix this problem.  Check virtual mem. and set it.a# >> I'm compiling a program in BASIC  >> a? >> Fatal:  Insufficient virtual memory to continue compilation.e0 >> %LIB-F-INSVIRMEM, insufficient virtual memory >> eN >> Here is my SH MEM. Notice the bottom says The Virtual I/O Cache is DISABLED >> on this node. > > > Virutal I/O cache is disk cache and has *NOTHING* to do with > insufficient virtual memory. > A > Solution: increase SYSUAF PGFLQUOTA of the username. Check thatgA > SYSGEN VIRTUALPAGECNT are >= the new value. Check that you haver > big enough page-files.  H         I also  answered  this...on  vmsnet.alpha.   Too  bad  it wasn'tH     cross-posted.  :-(  My reason for following-up here is that, if  youH     take a look at the original posting you'll find  the  system  is  anH     Alpha,  not  a VAX.  Therefore VIRTUALPAGECNT is irrelavent as well.H     He also has 512MB of physical  memory,  so it shouldn't be a problem;     to boost PGFLQUO to multi-100K with no adverse affects.i           -Ken -- aM  Kenneth H. Fairfield            |  Internet: Fairfield@SLC.Slac.Stanford.Edus:  SLAC, 2575 Sand Hill Rd, MS 46  |  Voice:    650-926-2924:  Menlo Park, CA  94025           |  FAX:      650-926-3515N  -----------------------------------------------------------------------------B  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------  % Date: Wed, 08 Nov 2000 19:01:12 -0400e- From: JF Mezei <jfmezei.spamnot@videotron.ca>eI Subject: Re: Fatal:  Insufficient virtual memory to continue compilation.E, Message-ID: <3A09DB29.DA405BC3@videotron.ca>  9 "Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515" wrote:eJ >     Alpha,  not  a VAX.  Therefore VIRTUALPAGECNT is irrelavent as well.J >     He also has 512MB of physical  memory,  so it shouldn't be a problem= >     to boost PGFLQUO to multi-100K with no adverse affects.R  M What if the page file is 50k blocks, the PGFLQUO is 200k, and the WSEXTENT is 
 only 16k ?  N If the maximum allowed working set is small, it will force the process to pageM to the page file, right ? At this point, you face the risk of using more pagee file than the actual page file.T   ------------------------------   Date: 8 Nov 2000 16:55:01 PSTaT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)I Subject: Re: Fatal:  Insufficient virtual memory to continue compilation.i3 Message-ID: <vfPKaP+1ZgLv@mccdev.slac.stanford.edu>l  \ In article <3A09DB29.DA405BC3@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:; > "Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515" wrote:lK >>     Alpha,  not  a VAX.  Therefore VIRTUALPAGECNT is irrelavent as well.uK >>     He also has 512MB of physical  memory,  so it shouldn't be a problem > >>     to boost PGFLQUO to multi-100K with no adverse affects. > O > What if the page file is 50k blocks, the PGFLQUO is 200k, and the WSEXTENT ise > only 16k ? > P > If the maximum allowed working set is small, it will force the process to pageO > to the page file, right ? At this point, you face the risk of using more page ! > file than the actual page file..  H         Actually, I don't _think_  so...   As  I  said before, our usersH     have  PGFLQUO  of  400000,  but WSEXTENT is more  like  16K  (that'sH     certainly true for my account).  But we managed to run for the firstH     three years without _ever_ paging to disk (2 GB of  physical  memoryH     can  be _really_ nice :-).  Basically, the paged-out pages never getH     written to disk (pagefile)  unless  there  is  a  demand on physicalH     pages.   This  is not to say you shouldn't have a  reasonably  sizedH     page file and/or swap  file.   But,  for  example,  VMS  does  _not_*     require a page file in order to run...  H         I'll leave the details  of  paging  and  when pages actually getH     written  to  disk to someone else.  I'm merely reporting what  we've     experienced.           -Ken -- .M  Kenneth H. Fairfield            |  Internet: Fairfield@SLC.Slac.Stanford.Edu :  SLAC, 2575 Sand Hill Rd, MS 46  |  Voice:    650-926-2924:  Menlo Park, CA  94025           |  FAX:      650-926-3515N  -----------------------------------------------------------------------------B  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------  # Date: Wed, 08 Nov 2000 17:50:18 GMTf From: carlhirsch@my-deja.com, Subject: Fibre Channel Encapsulation/Gateway) Message-ID: <8uc3om$ue2$1@nnrp1.deja.com>i  E I work for a company that runs VMS clusters for its core application.r@ One project we're working on is an attempt to extend a VMS FibreF Channel SAN across a 40-mile run of fiber to a disaster recovery site.D One limitation we're running into is that all but the most expensive. solutions don't support Fibre Channel signals.  ? What I'm wondering is this - is anyone in this forum aware of adE technology that allows Fibre Channel to ride over IP or Ethernet? I'mtA not aware of any gateways of this sort, but I figured I'd ask thew	 question.s  C Of course, since it's VMS latency is an issue as well. So something(: hardware-based would be preferable to a software solution.   thanks,l -carl hirsch network engineer    & Sent via Deja.com http://www.deja.com/ Before you buy.T   ------------------------------   Date: 8 Nov 2000 21:39:09 GMT * From: helbig@astro.rug.nl (Phillip Helbig)2 Subject: IP-Verbindung von zu Hause in Deutschland. Message-ID: <8uch5t$7au$1@info.service.rug.nl>  G Inzwischen denke ich, dass ich wei, wie ich alles mache, was ich dazu eF brauche usw.  (Bei der TCPIP-Konfigurierung, hat der ISDN-Router dann I eine IP-Adresse, die ich angebe?  Wer kann erklren, warum ich 3 weniger eE NUTZBARE IP-Adressen habe, als insgesamt?)  Allerdings kenne ich nur CH zwei Firmen, die so etwas anbieten (mit statischen IP-Adressen, dial-in B dial-out falls billiger als echte Standleitung usw.).  Diese sind F Lnecom in Lneburg und Komtel in Flensburg.  Die Preise und auch die ; Herangehensweisen sind unterschiedlich, also lohnt es sich  G wahrscheinlich, noch mehr Angebote zu haben---die Frage ist nur, woher?b   ------------------------------  " Date: Wed, 8 Nov 2000 13:48:36 GMT- From: "Richard D. Piccard" <piccard@ohio.edu>b5 Subject: Re: login prompts inhibited with logins=0 ??o( Message-ID: <3A0959B1.C2E82FFF@ohio.edu>  U Larry, thanks for the confirmation.  Hoff or others, is this still the case in modernA versions of VMS?#                                 RDPr     Larry Kilgallen wrote:  Z > In article <3A085833.9736D2CC@ohio.edu>, "Richard D. Piccard" <piccard@ohio.edu> writes:K > > There is an echo in brain that VMS once upon a time behaved differently G > > if LOGINS were left at zero, in a scenario such as you describe, as K > > opposed to having been briefly above zero and then set back to zero.  IhL > > found that objectionable, since it meant having, however briefly, a time* > > when the non-zero value was in effect. >sE > I believe the first enabling started the Job Controller at its task ! > of creating LOGINOUT processes.e   --B ==================================================================B Dick Piccard                           Academic Technology ManagerB piccard@ohio.edu                                 Computer ServicesB http://oak.cats.ohiou.edu/~piccard/                Ohio University   ------------------------------  # Date: Wed, 08 Nov 2000 16:01:26 GMTl% From: Uwe Zessin <zessin@my-deja.com>w5 Subject: Re: login prompts inhibited with logins=0 ??e) Message-ID: <8ubtcg$of4$1@nnrp1.deja.com>d  + In article <C7GiK1D1e5B1@eisner.decus.org>, <   Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) wrote:? > In article <3A085833.9736D2CC@ohio.edu>, "Richard D. Piccard"s <piccard@ohio.edu> writes:? > > There is an echo in brain that VMS once upon a time behaved B > > differently if LOGINS were left at zero, in a scenario such asB > > you describe, as opposed to having been briefly above zero and@ > > then set back to zero.  I found that objectionable, since itA > > meant having, however briefly, a time when the non-zero valuea > > was in effect. >oE > I believe the first enabling started the Job Controller at its taskr! > of creating LOGINOUT processes.a  C Correct. I found that out the hard way on my first VMS upgrade manycG years ago. I wanted to be 'clever' and set 'STARTUP$INTERACTIVE_LOGINS'o! to 0 to keep any users off. Oops.   C Later I used 'SET LOGINS/INTERACTIVE=1', 'SET LOGINS/INTERACTIVE=0'd? in SYSTARTUP so I could log in (with OPER privilege, of course)d< on a local terminal port while the system was still booting.   --
 Uwe Zessin    & Sent via Deja.com http://www.deja.com/ Before you buy.t   ------------------------------  $ Date: Wed, 8 Nov 2000 21:01:11 +0100" From: "Hans Vlems" <hvlems@iae.nl>1 Subject: Re: NCP commands in the 7.2 Cluster doc? ( Message-ID: <8ucaqv$cnp$1@news.IAEhv.nl>  F Agreed, phase V is a RPITA on systems that rate below, say, 5 VUPS andH have less than 32 MB main memory. That said, the information you can get, out of ncl is a lot more than NCP ever gave.G Provided you can remember where you put the paper bits with the correctr invocations ;-)o  
 Hans Vlems  H JF Mezei heeft geschreven in bericht <3A08CFD6.D1F603D5@videotron.ca>... >"David J. Dachtera" wrote: I >> Upgrading from Phase-V to Phase-IV is no small task, but the rewards -oJ >> in sleep-filled nights, stable systems and connections, etc. - are well) >> worth the nominally monumental effort.o >@L >I spent more time downgrading to phase 5 from 4 than I did upgrading from 5 toJ >4. I had everything running on the phase 5 stuff. But for my machines, it wasrD >overkill and took way too many resources, and issued too many opcom	 messages.t >>L >Also, installing it on my all mightly microvax II was impossible because itL >requires 80k blocks to install, whereas Phase 4 does the same with far less! >system resources and disk space.e   ------------------------------  % Date: Tue, 07 Nov 2000 21:17:20 -0600 7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>h1 Subject: Re: NCP commands in the 7.2 Cluster doc?o- Message-ID: <3A08C5C0.893FAB80@earthlink.net>    Didier Morandi wrote:i > 	 > in page  > I > www.openvms.compaq.com:8000/72final/4477/4477pro_005.html#decnet_configu > B > of the OpenVMS Cluster Systems dated january 1999, there are NCP  > commands instead of NCL. Why ?  F Speaking strictly for myself, I do not recommend Phase-V, unless there; is something specific in it that you or your company needs.   F Upgrading from Phase-V to Phase-IV is no small task, but the rewards -G in sleep-filled nights, stable systems and connections, etc. - are welly& worth the nominally monumental effort.   ...IMHO.   -- - David J. Dachtera- dba DJE SystemsI http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/w  F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.u   ------------------------------  " Date: Wed, 8 Nov 2000 16:52:39 GMT) From: Jack Haskell <jhaskell@hotmail.com>: Subject: Re: OpenVMS on 5300+ Message-ID: <3A0984D7.1E632A1D@hotmail.com>a  Y I agree.  I traced the links provided by the seller and the Compaq link shows it to be as ^ Hoffman describes.  The seller seemed to be confused by the fact that (as you state) it was an Alpha chip.c  ] It's a good thing for me that I asked and Hoffman replied because I was seriously consideringF2 purchasing the system so that I could run OpenVMS.   -jack    Tim Llewellyn wrote:   > Jack Haskell wrote:  >h > > Hoff Hoffman wrote:o > >e] > > > In article <3A06EAD4.B012F808@hotmail.com>, Jack Haskell <jhaskell@hotmail.com> writes:oK > > > :I searched the net for information on whether or not the alphaservero$ > > > :5300 will run OpenVMS or not. > > > L > > >   There is no such thing as the AlphaServer 5300 series, I will assumeC > > >   that this is a reference to the Digital Server 5300 series.s > >t] > > OK. I guess that the seller got it wrong. He listed it as an Alpha Server. It was sellinge+ > > on ebay for 800.00 with no takers. See:e > >- >a] > My take on what Hoff is trying to say, an alpha server is a server with an Alpha inside it. ^ > An AlphaServer is a particular range of alpha based servers  qualified to run VMS and Tru64.Q > The Digital Server 5300 is a server with an alpha chip designed to run NT only.r >e > --8 > Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project2 > MedAS at the BBC, Whiteladies Road, Bristol, UK.C > Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.uk] >RC > I speak for myself only and my views in no way represent those ofR > MedAS or the BBC.    ------------------------------   Date: 8 Nov 2000 19:59:21 GMT:2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: OpenVMS on 53006 Message-ID: <8ucbap$7hq$1@mailint03.im.hou.compaq.com>  Y In article <3A094585.3DECF648@bbc.co.uk>, Tim Llewellyn <tim.llewellyn@bbc.co.uk> writes:d :X :u :Jack Haskell wrote: :  :> Hoff Hoffman wrote: :>\ :> > In article <3A06EAD4.B012F808@hotmail.com>, Jack Haskell <jhaskell@hotmail.com> writes:J :> > :I searched the net for information on whether or not the alphaserver# :> > :5300 will run OpenVMS or not.s :> >K :> >   There is no such thing as the AlphaServer 5300 series, I will assumeiB :> >   that this is a reference to the Digital Server 5300 series. :>N :> OK. I guess that the seller got it wrong. He listed it as an Alpha Server.  :dL :My take on what Hoff is trying to say, an alpha server is a server with an M :Alpha inside it. An AlphaServer is a particular range of alpha based serverssN :qualified to run VMS and Tru64.  The Digital Server 5300 is a server with an $ :alpha chip designed to run NT only.  G   Correct.  There is no "5300" used on any "AlphaServer" series member.nG   The only recent use of the "5300" I can recall was with the "DIGITAL hE   Server 5300" series, a platform that is targeted for and supported .F   only for Microsoft Windows NT.  The DIGITAL Server 5300 does contain   an Alpha microprocessor.  J   I am aware of various hacks that can potentially allow at least some of I   the systems targeted for use with Windows NT to boot OpenVMS, and I am rG   also aware of some folks that have had problems.  Some of these AlpharI   systems targeting Windows NT have platform variants that are supported  I   by OpenVMS, and some do not.  Other than the hobbyist-type licenses, I eL   am not aware of any OpenVMS licenses that would permit OpenVMS operations G   on these.  (eg: I am not aware of any licenses that would permit the f/   commercial use of these "converted" systems.)   G   All this reminds me of some text I need to write up, but I digress...i  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Wed, 08 Nov 2000 07:35:33 +0100n- From: "I.A. Saez" <i.a.saez.scheihing@tue.nl>cF Subject: Re: Oracle/VMS performance problems with Multithreaded Server& Message-ID: <3A08F435.5B417C7F@tue.nl>   Jan Vorbrueggen wrote: > 2 > "I.A. Saez" <"i.a.saez.<secret>"@tue.nl> writes: > L > > That MTS does not work under Unix is not such a problem. Proces creationI > > (for dedicated processes) is cheap under Unix. Process creation underc# > > VMS costs a lot more (seconds!)n > O > This is mostly folklore. Actual process creation is remarkbly similar in UnixaN > and VMS. What differs is that the created process by default receives a copyP > of the original process's address space in Unix; however, this isn't used thatN > often (most fork() calls are followed by an exec() call). What tends to makeJ > the perception of process creation as slow is the additional overhead ofP > setting up the environment of DCL (symbols, logical names) when SPAWNing (when  ) Thus process creation on VMS is slower:-)a  O > will that improvement to the DCL SPAWN code finally be released, I wonder!?);aO > however, all that can be turned off. And on a current Alpha, the situation ismL > quite different fron a VAX-11/725 with regard to the resources consumed by > a $CREPRC call.i > 
 >         Jany   ------------------------------   Date: 8 Nov 2000 19:09:25 GMTa2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: PERL on VMS6 Message-ID: <8uc8d5$71i$1@mailint03.im.hou.compaq.com>  ` In article <3A0709D7.CB0B4A57@boeing.com>, Lewis McDonald <lewis.d.mcdonald2@boeing.com> writes:E :Is there a web site written after 1996 (or 1999) that is specific toEI :PERL on OpenVMS other than the Compaq site? How do I find out if PERL is I :working on OpenVMS? I found the executable. Thanks for any help on this!   D   I have a Perl 5.5.3 kit (PCSI-based) going onto the next Freeware.  C   A Perl kit will likely be part of the OpenVMS DII COE environmentnE   for OpenVMS, as well.  (Please see the OpenVMS FAQ for pointers to l   DII COE information.)   B   Please also see the OpenVMS FAQ for pointers to Perl information   and to a Perl mailing list.e  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Wed, 08 Nov 2000 02:43:03 GMT 5 From: njc@CC47532-A.ewndsr1.nj.home.com (Neil Cherry)-) Subject: Re: Porting from Unix to OpenVMS > Message-ID: <slrn90hff2.b25.njc@CC47532-A.ewndsr1.nj.home.com>  : On Tue, 7 Nov 2000 08:08:52 -0600, John E. Malmberg wrote:  > >http://eisner.decus.org/DECUServe/DECnotes/VMS/3064/SDIR.HTML  9 That is so cool, can I post that as part of my LHA pages?g  4 BTW: what language is that being used in (CM11.BLI):  ; http://eisner.decus.org/DECUServe/DECnotes/VMS/3064.12.HTML    --  H Linux Home Automation           Neil Cherry             ncherry@home.netC http://members.home.net/ncherry                         (Text only)FB http://meltingpot.fortunecity.com/lightsey/52           (Graphics)0 http://linuxha.sourceforge.net/				(SourceForge)   ------------------------------  $ Date: Tue, 7 Nov 2000 21:34:06 -0600) From: "John E. Malmberg" <wb8tyw@qsl.net>i) Subject: Re: Porting from Unix to OpenVMSk/ Message-ID: <t0hialrf5guh38@corp.supernews.com>i  B "Neil Cherry" <njc@CC47532-A.ewndsr1.nj.home.com> wrote in message8 news:slrn90hff2.b25.njc@CC47532-A.ewndsr1.nj.home.com...< > On Tue, 7 Nov 2000 08:08:52 -0600, John E. Malmberg wrote: > @ > >http://eisner.decus.org/DECUServe/DECnotes/VMS/3064/SDIR.HTML >c; > That is so cool, can I post that as part of my LHA pages?e  > Decuserve is for information sharing, so I do not see why not.  6 > BTW: what language is that being used in (CM11.BLI): > = > http://eisner.decus.org/DECUServe/DECnotes/VMS/3064.12.HTML   H It is BLISS, a programming language available for OpenVMS and is on it's freeware CD-ROM.  D Also available on-line from a link at http://www.openvms.compaq.com/ -Johnf wb8tyw@qsl.network   ------------------------------  + Date: Wed, 08 Nov 2000 20:27:49 +0100 (CET)w: From: "Gotfryd Smolik, VMS lists" <gotfryd@stanpol.com.pl>( Subject: Re: RECALL question/suggestion.I Message-ID: <Pine.LNX.4.21.0011082015420.3799-100000@irys.stanpol.com.pl>   # On Wed, 1 Nov 2000, JF Mezei wrote:.  + +Is there a way to RECALL with a wildcard ?G +U9 +If no, I would suggest the following improvement to VMS:  +e +RECALL/all commandc +nI +It would display all commands that begin with "command" that were issuedgL +recently. (Currently, without /ALL, you can only retreive the most recently) +entered command that matches "command".)t  ;  Och, yes, we must know that my "have find bug in PIPE" hasl some reason -;)i  ?  Two commandprocedures for test included: the one with "mailboxf= only" *can* hang (check the info at end), the second requires 8 temporary file (but the files deleted as fast as can !).:  The PIPE resolution disallows of course the RECAL/SAVE in6 "one shot", but with 1 subcommand (shortcut) after the selection may be acceptable.  N +For instance, if I have 5 or 6 TYPE commands with long/complex file names, itL +would be nice to be able to recall all of them and then recall the specific +one I want.    Regards - Gotfryd? PS: will warn, that can be unavailable up to end of next week !-  E ===================================================================== = ---------------- RECSEAM.COM - mailbox only ------------------1 $ ver='f$verify(0)' ! XI.2000, (C) Gotfryd Smolik 1 $! RECALL/SEARCH, freeware with unlimited use -:)l $! Check warning at end. $ stat=1 $ if p1.eqs."#" then goto exec $ if p1.eqs."^" then goto save/ $! define&run PIPE symbol for Recall/search -:)j' $ rr*ecall:==pipe close/nolog rec$out -e#   ; open/read/write rec$out MPA0: -4   ; recall/out=rec$out: -s%   ; @'f$environment("PROCEDURE")' "#"' $ if p1 then goto exit $ type sys$input RR*ECALL ==> RECALL/SEARCH RR/HELP describes syntax8  The definition .COM can be run with "TRUE" parameter to>  prevent this message or "HELP" to see full HELP for run-time.   $hlp:e $ if p1.nes."" $ then $   type sys$input4  Empty paramaters "" are dissallowed and unsupported1  Allowed qualifier must be first in command line,S  separated with spaces:s     /HELP       - this helpo%     /RUN, where is by default: /QUERYs     /EXACT      - as in SEARCH;     /BEGIN      - first param. from start of line; normally-6                   the command line is searched as text=     /SAVE       - to recall buffer. Not supported with /RUN !n<                    /SAVE *REQUIRES* to run after the command7                   one additional pseudocommand: RRS*AVE,:                   (saves the selected lines) or RRF*FLUSH.;                    /SAVE *REQUIRES* also CREMBX.COM commande=                   procedure be available thru symbol or path.     In next versions:=     /PRECISE    - as RECALL does (not a "free search" but thed'                   command must match !)d;  and "compressed form" (like /RUN/BEGIN without space) willu
  be included.e   $ endife $ goto exitr   $exec: $ close/nolog rec$savr- $ if f$extract(0,2,p2).eqs."/H" then goto hlpr; $ if f$trnlnm("rec$OUT","LNM$PROCESS","EXECUTIVE").eqs."" -n+         then exit %x3808A       ! Hm... -;)  $ on control then goto CTL_Y $ on error then goto err $ write rec$out ""9 $! Here you can do with the RECALL contens what you will;g) $! EXCEPT - the EOF signal is not here :(s: $! You must close the channel when empty record is found !; $! Supose, we will allow run with 1 positional qualifier :]d  $! Supose /SEARCH is default -;)% $! Temporary only works as /MATCH=ANDi3 $!-------------------------------------------------o
 $ Q_run  :=NO 
 $ Q_query:=NOc
 $ Q_exact:=NOw
 $ Q_begin:=NOr
 $ Q_save :=NO  $ pn=2  ! first parametert> $ chkq="if f$extract(0,f$len(p'pn'),qual).eqs.p'pn' then pipe" $Qual:2 $ if f$extract(0,1,p'pn').nes."/" then goto noQual3 $ if f$len(p'pn').lt.2 then stat=%x380BA ; goto end  $ qual:=/RUN2 $ 'chkq' Q_run:=YES     ; Q_query:=YES ; goto Qnxt
 $ qual:=/SAVE  $ 'chkq' goto setSave  $ qual:=/EXACT# $ 'chkq' Q_exact:=YES   ; goto Qnxta $ qual:=/BEGIN# $ 'chkq' Q_begin:=YES   ; goto QnxtO $ qual:=/QUERY3 $ if f$len(p'pn').lt.4 then stat=%x380BA ; goto end=# $ 'chkq' Q_query:=YES   ; goto Qnxt  $ qual:=/NOEXACT# $ 'chkq' Q_exact:=NO    ; goto Qnxti $ qual:=/NOQUERY# $ 'chkq' Q_query:=NO    ; goto Qnxt  $ stat=%x3880A
 $ goto end $Qnxt:	 $ pn=pn+1 . $ if pn.gt.6 then pipe stat=%x3880A ; goto end $ goto qual-  	 $setSave:, $ close/nolog rec$mbxM $ crembx/out=NLA0: rec$mbx/reade
 $ tmp=$statusc
 $ if .not.tmpo $ thenF $   write sys$error "--> Using /SAVE requires CREMBX.COM be available" $   stat=tmp( $   if tmp.eqs.%X38090 then stat=%x3220A $   goto end $ endifa $ open/write rec$sav MPA0:3 $ spawn/nolog/nological/nosymbol/nowait/inp=NLA0: -.,         @'f$environment("PROCEDURE")' "^"  -(         'f$getdvi("REC$SAV","DEVNAM")' -(         'f$getdvi("REC$MBX","DEVNAM")' &; $ RRS*AVE :==pipe RECALL/INPUT=REC$MBX: ; close REC$MBX ; - B         delete/symbol/global RRFLUSH ; delete/symbol/global RRSAVE3 $ RRF*LUSH:==pipe TYPE REC$MBX: ; close REC$MBX ; -eB         delete/symbol/global RRFLUSH ; delete/symbol/global RRSAVE
 $ Q_save:=YESo $ Q_query:=YES $ goto Qnxti   $noQual:3 $ if p'pn'.eqs."" then pipe stat=%x38812 ; goto endT	 $ pm=pn+1a $chkp: $ if p'pm'.eqs."" then goto runy3 $ if .not.Q_exact then p'pm'=f$edit(p'pm',"UPCASE")o	 $ pm=pm+1' $ if pm.le.8 then goto chkp_ $run:o	 $nxtRead:e $ on error then goto err# $ read rec$out line/end=end/err=errE $ text=lineI1 $ if .not.Q_exact then text=f$edit(line,"UPCASE")( $ if line.eqs."" then goto end $ p=pn  ! Parameter number $nxtSea: $ if q_begin.and.(p.eq.pn) $ then1 $   if f$locate(p'p',text).ne.0 then goto nxtRead  $ else+ $   if text.eqs.text-p'p' then goto nxtRead  $ endif, $ p=p+10 $ if p.ge.pm then goto found
 $ goto nxtSeai   $found:j $ write sys$output linet $ prompt="Next,End [Next]: "& $ if q_save then prompt="Save,"+prompt& $ if q_run  then prompt="Run," +prompt $ if q_query $ then# $   prompt="Select one of: "+prompty1 $   read sys$command ask/prompt=&prompt/error=endl/ $   ask=f$extract(0,1,f$edit(ask+"N","UPCASE"))c$ $   if ask.eqs."N" then goto nxtRead  $   if ask.eqs."E" then goto end1 $   if q_run .and.(ask.eqs."R") then gosub runCmdb1 $   if q_save.and.(ask.eqs."S") then gosub savCmd  $ else $   if q_run  then gosub runCmdh $   if q_save then gosub savCmdk $ endif  $ goto nxtRead   $runCmd: $ on severe then continuen$ $ define/nolog sys$input sys$command 'line' $ deassign sys$input $ on error then goto err $ return   $savCmd: $ write rec$sav line $ return   $CTL_Y:o $ stat=%x30610
 $ goto end   $err:  $ stat=$status $end:m
 $ set noon $ close/nolog rec$out  $ if q_savem $ then $   create sys$outputaF --> /SAVE detected, you *must* do RRS*SAVE or RRF*FLUSH in next line ! $   write rec$sav "" $   close/nolog rec$savl $ endifn $ if .not.stat $ then6 $   write sys$output "--> Check the syntax with /HELP"# $   write sys$output f$messge(stat)-9 $   if f$trnlnm("SYS$OUTPUT").nes.f$trnlnm("SYS$ERROR") -J+         then write sys$error f$messge(stat)u $ endif: $exit:+ $ exit stat.or.%x10000000 + 0*f$verify(ver)3   $save: $ on error then goto ecp $ open/read  inp 'p2'  $ open/write out 'p3'w $ptCp: $ read  inp line $ if line.nes."" $ then $   write out line
 $   goto ptCpe $ endift $eCp:e
 $ set noon $ close inpe $ close out , $ exit	! - Do NOT cut the next part, please:/  ----------------- WARNING --------------------e9  The code suposes, that default p-mailbox (MPAnn: device)a5 size on your node is enought big to save whole RECALLs buffer output.5  If the assumption is wrong - the code can hang. Thatd3 problem can be overriden with second MP -:), but...o  = ====================== END OF FILE  =========================   = ---------------- RECSEAF.COM - temp. files  ----------------- 1 $ ver='f$verify(0)' ! XI.2000, (C) Gotfryd Smolik/1 $! RECALL/SEARCH, freeware with unlimited use -:)p $ stat=1 $ if p1.eqs."#" then goto exec $ if p1.eqs."^" then goto save/ $! define&run PIPE symbol for Recall/search -:) ' $ rr*ecall:==pipe close/nolog rec$inp -r7   ; recall/out=SYS$SCRATCH:REC'f$getjpi(0,"PID")'.TMP --%   ; @'f$environment("PROCEDURE")' "#"e $ if p1 then goto exit $ type sys$input RR*ECALL ==> RECALL/SEARCH RR/HELP describes syntax8  The definition .COM can be run with "TRUE" parameter to>  prevent this message or "HELP" to see full HELP for run-time.   $hlp:  $ if p1.nes."" $ then $   type sys$input4  Empty paramaters "" are dissallowed and unsupported1  Allowed qualifier must be first in command line,e  separated with spaces:      /HELP       - this helpl%     /RUN, where is by default: /QUERYo     /EXACT      - as in SEARCH;     /BEGIN      - first param. from start of line; normally 6                   the command line is searched as text=     /SAVE       - to recall buffer. Not supported with /RUN !t<                    /SAVE *REQUIRES* to run after the command7                   one additional pseudocommand: RRS*AVE.:                   (saves the selected lines) or RRF*FLUSH.    In next versions:=     /PRECISE    - as RECALL does (not a "free search" but the0'                   command must match !)e;  and "compressed form" (like /RUN/BEGIN without space) will 
  be included.n   $ endif0 $ goto exitt   $exec:- $ if f$extract(0,2,p2).eqs."/H" then goto hlpt: $ open/read rec$out SYS$SCRATCH:REC'f$getjpi(0,"PID")'.TMP1 $ delete SYS$SCRATCH:REC'f$getjpi(0,"PID")'.TMP;*a $ on control then goto CTL_Y $ on error then goto err9 $! Here you can do with the RECALL contens what you will;r; $! Supose, we will allow run with 1 positional qualifier :]r  $! Supose /SEARCH is default -;)% $! Temporary only works as /MATCH=AND 3 $!-------------------------------------------------l
 $ Q_run  :=NOR
 $ Q_query:=NO 
 $ Q_exact:=NO 
 $ Q_begin:=NOw
 $ Q_save :=NOo $ pn=2  ! first parameter > $ chkq="if f$extract(0,f$len(p'pn'),qual).eqs.p'pn' then pipe" $Qual:2 $ if f$extract(0,1,p'pn').nes."/" then goto noQual3 $ if f$len(p'pn').lt.2 then stat=%x380BA ; goto end@ $ qual:=/RUN2 $ 'chkq' Q_run:=YES     ; Q_query:=YES ; goto Qnxt
 $ qual:=/SAVEw $ 'chkq' goto setSavec $ qual:=/EXACT# $ 'chkq' Q_exact:=YES   ; goto Qnxt> $ qual:=/BEGIN# $ 'chkq' Q_begin:=YES   ; goto Qnxt0 $ qual:=/QUERY3 $ if f$len(p'pn').lt.4 then stat=%x380BA ; goto end># $ 'chkq' Q_query:=YES   ; goto Qnxto $ qual:=/NOEXACT# $ 'chkq' Q_exact:=NO    ; goto Qnxtp $ qual:=/NOQUERY# $ 'chkq' Q_query:=NO    ; goto Qnxtt $ stat=%x3880A
 $ goto end $Qnxt:	 $ pn=pn+1>. $ if pn.gt.6 then pipe stat=%x3880A ; goto end $ goto qual.  	 $setSave:I $ open/write/share=write -6         rec$sav SYS$SCRATCH:REC'f$getjpi(0,"PID")'.TMP $ open/read/share=write -s6         rec$inp SYS$SCRATCH:REC'f$getjpi(0,"PID")'.TMP@ $ RRS*AVE:==pipe recall/input=rec$inp: ; close/nolog rec$inp ; -C          delete/symbol/global RRFLUSH ; delete/symbol/global RRSAVE ) $ RRF*LUSH:==pipe close/nolog rec$inp ; -lC          delete/symbol/global RRFLUSH ; delete/symbol/global RRSAVE 
 $ Q_save:=YES  $ Q_query:=YES $ goto Qnxt    $noQual:3 $ if p'pn'.eqs."" then pipe stat=%x38812 ; goto enda	 $ pm=pn+1i $chkp: $ if p'pm'.eqs."" then goto runr3 $ if .not.Q_exact then p'pm'=f$edit(p'pm',"UPCASE")c	 $ pm=pm+1n $ if pm.le.8 then goto chkp  $run:e	 $nxtRead:  $ on error then goto err# $ read rec$out line/end=end/err=errm $ text=liner1 $ if .not.Q_exact then text=f$edit(line,"UPCASE")  $ if line.eqs."" then goto end $ p=pn  ! Parameter number $nxtSea: $ if q_begin.and.(p.eq.pn) $ then1 $   if f$locate(p'p',text).ne.0 then goto nxtReadV $ else+ $   if text.eqs.text-p'p' then goto nxtReadc $ endife $ p=p+1t $ if p.ge.pm then goto found
 $ goto nxtSeam   $found:o $ write sys$output line( $ prompt="Next,End [Next]: "& $ if q_save then prompt="Save,"+prompt& $ if q_run  then prompt="Run," +prompt $ if q_query $ then# $   prompt="Select one of: "+promptt1 $   read sys$command ask/prompt=&prompt/error=end / $   ask=f$extract(0,1,f$edit(ask+"N","UPCASE"))H$ $   if ask.eqs."N" then goto nxtRead  $   if ask.eqs."E" then goto end1 $   if q_run .and.(ask.eqs."R") then gosub runCmd 1 $   if q_save.and.(ask.eqs."S") then gosub savCmde $ else $   if q_run  then gosub runCmd/ $   if q_save then gosub savCmdt $ endif  $ goto nxtRead   $runCmd: $ on severe then continueo$ $ define/nolog sys$input sys$command 'line' $ deassign sys$input $ on error then goto err $ return   $savCmd: $ write rec$sav line $ return   $CTL_Y:  $ stat=%x30610
 $ goto end   $err:d $ stat=$status $end:o
 $ set noon $ close/nolog rec$out( $ if q_save  $ then $   close/nolog rec$sava $   create sys$outpute: --> /SAVE detected, do RRS*SAVE to really save the lines ! $ endif  $ if .not.stat $ then6 $   write sys$output "--> Check the syntax with /HELP"# $   write sys$output f$messge(stat) 9 $   if f$trnlnm("SYS$OUTPUT").nes.f$trnlnm("SYS$ERROR") - +         then write sys$error f$messge(stat)s $ endifn $exit:+ $ exit stat.or.%x10000000 + 0*f$verify(ver)t8 ======================= END OF FILE ====================   ------------------------------  + Date: Wed, 08 Nov 2000 20:31:16 +0100 (CET) : From: "Gotfryd Smolik, VMS lists" <gotfryd@stanpol.com.pl>( Subject: RE: RECALL question/suggestion.I Message-ID: <Pine.LNX.4.21.0011082028220.3799-100000@irys.stanpol.com.pl>i  ' On Wed, 1 Nov 2000, John Vottero wrote:   L +I agree.  I would also like to have a /WILDCARD qualifier so you could say: + " +$ RECALL/ALL/WILDCARD *MYPROGRAM*  A  Hm... Will disagree: the command interpreter can already found, - that wildcard applies !04  Will you say "not at end of command ? (ok, agree !)*  Check the long message in answe to JF -;)A BTW: can you answer *after* citate (better readable) ? Thanks -:)     Regards - Gotfryd   -- PE ===================================================================== F $ ON F$ERROR("LANGUAGE","ENGLISH","IN_MESSAGE").GT.F$ERROR("NORMAL") - 		THEN EXCUSE/OBJECT=ME.. $!                        GS@stanpol.zabrze.plE =====================================================================n   ------------------------------  + Date: Wed, 08 Nov 2000 20:46:42 +0100 (CET) : From: "Gotfryd Smolik, VMS lists" <gotfryd@stanpol.com.pl>( Subject: Re: RECALL question/suggestion.I Message-ID: <Pine.LNX.4.21.0011082032200.3799-100000@irys.stanpol.com.pl>-  * On Thu, 2 Nov 2000, Alan E. Feldman wrote: [...] E +I suggest RECALL/OUTPUT/PREPEND_DOLLAR_SIGNS or RECALL/OUTPUT/DOLLAR-D +which would prepend "$ " to each line in the output file. One could> +even add /BLANKS=n to put n spaces between $ and the command.  #  But what with proper command ? -;)f9  DCL command always starts with "$", in some case the "$"0> can be omitted; in interactive mode the case has place always,1 and 99.98% command really put without "$". But...t  - [interactive mode, the first "$" is prompt !]g& $ x=1	! Improper command: "$" misses ! $ sh sym/all-   X = 1   Hex = 00000001  Octal = 00000000001n $ $x=2	! Proper command -:)  $ sh sym/all-   X = 2   Hex = 00000002  Octal = 00000000002e8 $ $$x=3	! Also proper: the command set the symbol "$x" ! $ sh sym/all.   $X = 3   Hex = 00000003  Octal = 00000000003-   X = 2   Hex = 00000002  Octal = 00000000002s    -:)  ;  Serious - I think, that you will "include" the RECALL texta9 to b.ex. commandprocedure: but then (within editor) this " is not a problem !    Regards - Gotfryd   -- xE ===================================================================== F $ ON F$ERROR("LANGUAGE","ENGLISH","IN_MESSAGE").GT.F$ERROR("NORMAL") - 		THEN EXCUSE/OBJECT=MEw. $!                        GS@stanpol.zabrze.plE =====================================================================S   ------------------------------  # Date: Wed, 08 Nov 2000 21:04:29 GMTa* From: Alan E. Feldman <alan48@my-deja.com>( Subject: Re: RECALL question/suggestion.) Message-ID: <8ucf4n$9h0$1@nnrp1.deja.com>u  - In article <Pine.LNX.4.21.0011082032200.3799-/ 100000@irys.stanpol.com.pl>,=   "Gotfryd Smolik, VMS lists" <gotfryd@stanpol.com.pl> wrote:n, > On Thu, 2 Nov 2000, Alan E. Feldman wrote: > [...]oG > +I suggest RECALL/OUTPUT/PREPEND_DOLLAR_SIGNS or RECALL/OUTPUT/DOLLARnF > +which would prepend "$ " to each line in the output file. One could@ > +even add /BLANKS=n to put n spaces between $ and the command. >c% >  But what with proper command ? -;)g; >  DCL command always starts with "$", in some case the "$"0@ > can be omitted; in interactive mode the case has place always,3 > and 99.98% command really put without "$". But...  > / > [interactive mode, the first "$" is prompt !] ( > $ x=1	! Improper command: "$" misses ! > $ sh sym/all/ >   X = 1   Hex = 00000001  Octal = 00000000001  > $ $x=2	! Proper command -:)/ > $ sh sym/all/ >   X = 2   Hex = 00000002  Octal = 00000000002a: > $ $$x=3	! Also proper: the command set the symbol "$x" ! > $ sh sym/all0 >   $X = 3   Hex = 00000003  Octal = 00000000003/ >   X = 2   Hex = 00000002  Octal = 00000000002n >  >  -:) >(= >  Serious - I think, that you will "include" the RECALL textu: > to b.ex. commandprocedure: but then (within editor) this > is not a problem ! >R  G I understand that without prepending a dollar sign to each command, onetD can run RECALL/INPUT to place long commands in the recall buffer forD easy recall. Of course, one could just define symbols for them also.  F But I thought it would be nice -- it doesn't have to be the default --G if you could have RECALL prepend the dollar signs for you. Yes, some or B many commands work without it, but it is not good coding practice!  8 With this option, you could have the following scenario:  A Hey, the last nine commands I just ran would make a great command/
 procedure!  * $ RECALL /O=new-exciting.com /P /L[INES]=9  C And you'd have it! Yeah, you could get by without it, but you couldn+ also get by without the RECALL command! :-)a  F P.S. You could also make .com the default file type! I mean, the wholeE point of the RECALL command is to save typing, is it not? That's alsoP' why I typed such short qualifier names.    -- Disclaimer: JMHO Alan E. Feldman    w:R  alan48  &-)        afeldman  &-) dellnet.com        gfigroup.com     & Sent via Deja.com http://www.deja.com/ Before you buy.P   ------------------------------  # Date: Wed, 08 Nov 2000 16:16:03 GMTp- From: goathunter@goatley.com (Hunter Goatley)d( Subject: Re: remote disconnect on telnet/ Message-ID: <3a097afa.3852810@swen.process.com>   4 On Tue, 7 Nov 2000 15:30:05 +0200, "Phillip du Ploy"! <itbpjdp@puknet.puk.ac.za> wrote:   M >I am a sysadmin on a Dec ES40 (ev67 ) . Is has 2 network cards in it, namely H >de500 (one is used for decnet) .  It runs OpenVMS 7.2 -1 . We are usingL >multinet 4.3 for our IP stack.  We get a problem of "axp remote disconnect"M >on the telnet client side.  This happens when you are typing (never when you < >are inactive), but the session will stay active on the axp. >/A Sounds like the client may be interpreting something incorrectly.r> Without more details, it's hard to say.  A TCPDUMP showing the, problem would probably show what's going on.  = Have you opened a support call with us (support@process.com)?p  M >We also use the LPR/LPD function to print to remote network printers via IP. I >I am thinking of replacing Multinet with UCX .  Any commends on this ( IH >also want to implement SSH ). >c3 MultiNet V4.3 includes an SSH client and server....h   Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/t: goathunter@goatley.com      http://www.goatley.com/hunter/   ------------------------------  % Date: Tue, 07 Nov 2000 20:36:52 -0600s7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>d Subject: Re: SHOW PROCESSw- Message-ID: <3A08BC44.F5A146DE@earthlink.net>    Arne Vajhj wrote: >  > Phillip Helbig wrote:/G > > Why does $  SHOW PROCESS <name> work only if the process is running L > > under the same User Identifier as the process which does $  SHOW PROCESSI > > (note: not the same User:), but $  SHOW PROCESS/ID=<process-ID> worksh > > from any process?f > >aL > > If I want to see a process with known name but unknown ID from a processJ > > with another user identifier, do I first have to get the ID (with SHOWJ > > USER/FULL, or is there a better way?) and then run $  SHOW PROCESS/ID? > @ > You are aware of that process-names are implicit prefixed with
 > UIC group ?=   That's right. Check this out:e  " DJAS01::DDACHTERA$ sh proc/subproc  G  7-NOV-2000 21:34:29.48   User: DDACHTERA        Process ID:   0000005Fo>                           Node: DJAS01           Process name: "DDACHTERA"A    There is 1 process in this job:      DDACHTERA (*)e DJAS01::DDACHTERA$ spawn+ %DCL-S-SPAWNED, process DDACHTERA_1 spawned$= %DCL-S-ATTACHED, terminal now attached to process DDACHTERA_1o! DJAS01::DDACHTERA$ set uic [20,1] * DJAS01::DDACHTERA$ set proc/name=ddachtera" DJAS01::DDACHTERA$ sh proc/subproc  G  7-NOV-2000 21:35:24.00   User: DDACHTERA        Process ID:   00000060A>                           Node: DJAS01           Process name: "DDACHTERA"m  # There are 2 processes in this job: t     DDACHTERAD     DDACHTERA (*)   G Although, if you logout at this point, the master process in the job is  hung.E   --   David J. Dachteras dba DJE SystemsU http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/F  F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.m   ------------------------------  % Date: Thu, 09 Nov 2000 01:04:04 -0400o- From: JF Mezei <jfmezei.spamnot@videotron.ca> ) Subject: Re: Starting VMS install with cdd, Message-ID: <3A0A303B.A82EFDA1@videotron.ca>   Dan Cauley wrote:eH > operating system on the drive), What must I do in order to accsess the > CD-ROM with OpenVMS on it   9 1- to format the disk drive: (not needed but recommended)  at the >>> TEST 75  ; SCSI-A, enter the SCSI ID of the drive, answer 1 at RUSURE.     2- once your CDROM is installed: >>> SHOW DEVICES  M If your CD rom is properly installed, you'll see your hard disk drive as wellq as your cdrom drive.   To boot from cd:  > >>> BOOT DKAxxx  (where the XXX is the DKA number for your CD.  I You will then be brought to a standalone backup facility. If you have thet9 hobyist kit, there are instructions on how to install it.a  @ something line BACKUP/INIT/IMAGE DKA300:VAXVMS072.B/SAVE DKA200:  L This will put a minimum VMS onto your hard drive, and you can then boot fromK the hard drive, and the rest of the installation will then continue and the H rest of the CD will be restored. The backup command will initialise your system disk.   ------------------------------  % Date: Tue, 07 Nov 2000 21:20:28 -0600t7 From: "David J. Dachtera" <djesys.nospam@earthlink.net> 8 Subject: Re: Suggestion for DCL scripts which open files- Message-ID: <3A08C67C.8A97283D@earthlink.net>   8 "Ed James, 410-295-1919, ed.james@telecomsys.com" wrote: > A > "BAZLEY, Sebastian" <Sebastian.BAZLEY@sema.co.uk> wrote in ? on " > Tue, 7 Nov 2000 17:07:54 -0000 : > L > > Suggestion: the following idiom is useful when accessing files from DCL: > >e0 > > $ close/nolog CONFIG ! clean up if necessary  > > $ open/read   CONFIG xxx.yyy > > $ read CONFIG line > > ...  > > $ close CONFIG > >vO > > The close/nolog does no harm if the file is not currently open, but is veryCI > > useful if the previous invocation of a command file has left the fileo > > open...i >  > This also works: > 6 >   $ if f$trnlnm("CONFIG") .nes. "" then close CONFIG  , Yes - it is:  & DJAS01::DDACHTERA$ open lgi lgicmd.com  DJAS01::DDACHTERA$ sh log/fu lgiH    "LGI" [super,confine] = "_DJAS01$DKA0" [terminal] (LNM$PROCESS_TABLE)  D ...but does not guarantee that the channel is open. In your example,; "CONFIG" may be a valid logical name for some other reason..   -- = David J. Dachtera= dba DJE Systems= http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/,  F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.S   ------------------------------  % Date: Wed, 08 Nov 2000 03:52:12 +0100e2 From: martin@radiogaga.harz.de (Martin Vorlaender)8 Subject: Re: Suggestion for DCL scripts which open files; Message-ID: <3a08bfdc.524144494f47414741@radiogaga.harz.de>e  L Ed James, 410-295-1919, ed.james@telecomsys.com (jamese@beast.dtsw.army.mil) wrote:: : "BAZLEY, Sebastian" <Sebastian.BAZLEY@sema.co.uk> wrote:L : > Suggestion: the following idiom is useful when accessing files from DCL: : > 0 : > $ close/nolog CONFIG ! clean up if necessary  : > $ open/read   CONFIG xxx.yyy : > $ read CONFIG line : > ...  : > $ close CONFIG : > J : > The close/nolog does no harm if the file is not currently open, but isI : > very useful if the previous invocation of a command file has left the  : > file open... :R : This also works: :m6 :   $ if f$trnlnm("CONFIG") .nes. "" then close CONFIG  F To be precise, it should be f$trnlnm("CONFIG","LNM$PROCESS"), in orderK to not inadvertently catch other logicals (common problem with file logical0 names like TEMP).    cu,    Martin --J One OS to rule them all       | Martin Vorlaender  |  VMS & WNT programmer7 One OS to find them           | work: mv@pdv-systeme.de2N One OS to bring them all      |       http://www.pdv-systeme.de/users/martinv/> And in the Darkness bind them.| home: martin@radiogaga.harz.de   ------------------------------  % Date: Wed, 08 Nov 2000 01:12:22 -0500t2 From: rdeininger@mindspring.com (Robert Deininger)' Subject: Re: Sun "uptime" belly laugherrL Message-ID: <rdeininger-0811000112230001@user-2ive688.dialup.mindspring.com>  R In article <tvag0tcsqmahhlfubes16fek9m9ph9vipg@4ax.com>, jlsuexxxz@home.com wrote:      F > Why are you doing this on the production cluster?  What's the matter* > with workstations?  Development systems? > F > In any case, I sincerely doubt that the entire cost and value of theH > production servers are based on your utilization of the system.  I.E.,B > you're just a small drop in the bucket as far as the business isG > concerned, they didn't purchase the servers for your benefit.  That's>G > not to say that you're completely unimportant, but you should be able @ > to recognize that you're pretty far down in the priority scale) > compared to keeping the business alive.     D The "production servers" ARE workstations.  The main activities are D programming and number crunching.  If these activities stop, nothing6 of value is produced, and the whole group will vanish.  E And BTW, I am the main system manager, so I get to decide when a nodeEC shuts down (barring network failure).  So yes, we are talking about C inconvenience, not disaster.  But I was only responding to the "no =; application down time" claim, which is not true in general.o   -- L Robert Deininger rdeininger@mindspring.comH   ------------------------------   Date: 8 Nov 2000 13:06:09 PSTLT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)4 Subject: Re: Transfering users via FTP & email lost.3 Message-ID: <LNa1pgp4Q0kQ@mccdev.slac.stanford.edu>t  B In article <20001107043225.25832.qmail@ww156.netaddress.usa.net>, 7     	Michael Angello <techwebsite@netscape.net> writes:n  H > I am using OVMS AXP V7.2 on a AS800, and I am migrating some user dataG > from other OVMS system to this one via ftp, because in the old systemoL > has no installed the network software feature needed to apply the COPY/RCPG > command which could may transfer all user data (including mail files)x& > with all of the atributes preserved.  D         Does COPY/RCP preserve attributes?  RCP by itself is pretty )     poor, but perhaps VMS-to-VMS is OK...E  F >                                      So, I tried to transfer of thatH > info. via ftp, but when I try to read a user's email, the system shows > a message like:c  H         FTP does _not_ preserve  VMS  file  attributes for indexed filesH     (except  for  possibly  cooperating VMS-to-VMS transfers  using  theH     _same_ vendor's IP stack...you never said which IP stack and versionH     you have installed on the two  systems).   MAIL.MAI  is  an  indexedH     file and will _not_ TP correctly.  COPY/RCP would likely have ruined     it as well.   H         The solution is to use  BACKUP  to  create an (on-disk) saveset,H     then FTP the saveset in BINARY mode, and finally restore the saveset     on the target system.            -Ken -- eM  Kenneth H. Fairfield            |  Internet: Fairfield@SLC.Slac.Stanford.Eduq:  SLAC, 2575 Sand Hill Rd, MS 46  |  Voice:    650-926-2924:  Menlo Park, CA  94025           |  FAX:      650-926-3515N  -----------------------------------------------------------------------------B  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------  # Date: Thu, 09 Nov 2000 01:29:07 GMTr From: trdorr@my-deja.com Subject: Re: Use P1 in PIPE0) Message-ID: <8ucukt$n25$1@nnrp1.deja.com>=   Juse to follow up on this. The correct syntax is-. $ PIPE SHOW SYSTEM | SEARCH SYS$INPUT ""'P1'""   Works like a charm!!  ' In article <3A020C96.7D90@adldata.com>,p   sol@adldata.com wrote:& > If the p1 is in quotes. it should be/ > $ PIPE SHOW SYSTEM | SEARCH SYS$INPUT "''p1'"n >.
 > sol gongolaW >  > trdorr@my-deja.com wrote:e > > 7 > > I want to search with a different string each time.r' > > Is it possible to use P1 with PIPE?D > >* > > TYPE SHOWSYS.COM > > $!FILE: SHOWSYS.COM 0 > > $ PIPE SHOW SYSTEM | SEARCH SYS$INPUT "'p1'" > >  > > Thanks,  > > Toma > > * > > Sent via Deja.com http://www.deja.com/ > > Before you buy.F >)    & Sent via Deja.com http://www.deja.com/ Before you buy.k   ------------------------------  % Date: Wed, 08 Nov 2000 11:14:04 -0600=/ From: Chris Scheers <chris@applied-synergy.com>A* Subject: Re: VAXstation 4000-90 Error CodeO Message-ID: <42EB9E790731ACD8.62BB26F08EC0BBBC.4E8FE70B498A7E4B@lp.airnews.net>.   "Zane H. Healy" wrote: > N > I've had a VAXstation 4000-90 without any RAM for a couple years now, and asN > of yesterday I've got a VAXstation 4000-60 with enough 16MB SIMMs that I can > test out the -90.m > L > I pulled 4 16MB SIMMs and installed them as per the Service Manual for theN > -90.  Upon turning the -90 on it gives me a error code of "F2h".  Looking atL > the list of error codes in the service manual I don't see this one listed.! > Does anyone know what it means?  > M > I'm assuming the -90 is dead, and I suspected that when I got it, but wouldw > like to verify.n > K > Is the PowerSupply or anything else from the -90 usable as spares for thei > -60?  G VS4000-60 memory is not necessarily compatible with a -90.  (AFAIK, -90,* memory will always work in a -60, though.)  tG -----------------------------------------------------------------------c$ Chris Scheers, Applied Synergy, Inc.  C Voice: 817-237-3360            Internet: chris@applied-synergy.com :   Fax: 817-237-3074    ------------------------------  % Date: Wed, 08 Nov 2000 11:04:50 -0800a@ From: "Russell E. Owen" <owen@astroNOJNK.washington.edu.invalid>3 Subject: VMS Fortran question: entering hex numbers 2 Message-ID: <8uc84j$fj6k$1@nntp6.u.washington.edu>  F The VMS Fortran help say that a hex constant is written as Z'd...' or F Z"d...d". I suspect they are referring to constants within the source C code, because it does NOT work for free-format reading values into  	 integers.R  G Is there some way to specify a hex value when free-format reading into /H an integer? (We have a several configuration files that would be easier H to read if their values were expressed in hex or even binary instead of 	 decimal.)-  
 -- Russell   ------------------------------  % Date: Wed, 08 Nov 2000 16:37:09 -0500 , From: Steve Lionel <Steve.Lionel@compaq.com>7 Subject: Re: VMS Fortran question: entering hex numberse8 Message-ID: <oohj0t8d0psfpg5snrltam19kkilsdonh2@4ax.com>  5 On Wed, 08 Nov 2000 11:04:50 -0800, "Russell E. Owen">/ <owen@astroNOJNK.washington.edu.invalid> wrote:r  G >The VMS Fortran help say that a hex constant is written as Z'd...' or UG >Z"d...d". I suspect they are referring to constants within the source .D >code, because it does NOT work for free-format reading values into 
 >integers. >BH >Is there some way to specify a hex value when free-format reading into I >an integer? (We have a several configuration files that would be easier pI >to read if their values were expressed in hex or even binary instead of i
 >decimal.)  @ No, there is not currently support for list-directed input usingE hexadecimal notation.  You can use formatted input with the Z format,  however.    - Steve Lionel (mailto:Steve.Lionel@compaq.com)  Fortran Engineering & Compaq Computer Corporation, Nashua NH  6 Compaq Fortran web site: http://www.compaq.com/fortran   ------------------------------   Date: 8 Nov 2000 14:06:54 PST0T From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)7 Subject: Re: VMS Fortran question: entering hex numbers 3 Message-ID: <3u2SqNipWH8i@mccdev.slac.stanford.edu>n  3 In article <8uc84j$fj6k$1@nntp6.u.washington.edu>, cG     	"Russell E. Owen" <owen@astroNOJNK.washington.edu.invalid> writes:uH > The VMS Fortran help say that a hex constant is written as Z'd...' or H > Z"d...d". I suspect they are referring to constants within the source E > code, because it does NOT work for free-format reading values into o > integers.c  H         A list-directed (what you  call  "free  format") read (or write)H     has  _many_  restrictions,  one of which is that you  can't  read  aH     hexadecimal format string into an integer variable without a  formatH     specification,  i.e.,  can't  be  done  with the "*" (list-directed)H     format.  Note als to that  the  Z"..."  and Z'...' formats are newlyH     (this decade!) standardized forms in Fortran 90/95.  They were neverH     part  of the F77 although VAX Fortran supported the '...'X style  as     well as IBM's Z'...' style.   H         In addition, if you _do_ want  to read a hexadecimal string ontoH     an  integer variable, you need to omit the "z" prefix and quotes  or     apostrophes.  I > Is there some way to specify a hex value when free-format reading into 0J > an integer? (We have a several configuration files that would be easier J > to read if their values were expressed in hex or even binary instead of  > decimal.)0  H         This sort  of  question  comes  up  on  comp.lang.fortran fairlyH     frequently.   The short answer is you can't use list-directed  inputH     for the problem at hand.  Is there a  reason  you  are  avoiding  an;     explicit format?  That will usually solve your problem.x  H         In the case that the external file already has the z'...' forms,H     you can always read the  whole  line into a character variable, thenH     "parse" character variable for the fields you want to read, creatingH     appropriate  formats (with correct field widths) as you go, then  doH     an  internal  read  of  the  character  variable  into  you  numericH     variables (that you would have otherwise done with the list-directedH     read).    This  would  allow  you  to  strip  the  z's,  quotes  andH     apostrophes before reading  the  hex  string  (with  a Zw.n format).2     Been there, done that, not all that difficult.           -Ken -- rM  Kenneth H. Fairfield            |  Internet: Fairfield@SLC.Slac.Stanford.Edur:  SLAC, 2575 Sand Hill Rd, MS 46  |  Voice:    650-926-2924:  Menlo Park, CA  94025           |  FAX:      650-926-3515N  -----------------------------------------------------------------------------B  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------  % Date: Wed, 08 Nov 2000 15:59:15 -0400a- From: JF Mezei <jfmezei.spamnot@videotron.ca>n0 Subject: Re: VMS VMAIL, SMTP transports, headers, Message-ID: <3A09B08F.41694C1D@videotron.ca>   Hunter Goatley wrote:SF > There was no To: line, which surprises me a little, but technically,E > it's because PMDF wasn't told by VMS Mail to process a To: address.aF > PMDF was only told about the CC address by VMS Mail, and that's what
 > it reports.t  J When you type a transport%address at the TO or CC prompts in VMS mail, theM transport is called with the same function code (value 2) whether the addresshQ is entered in the TO or CC fields. (At least on character cell VMSmail, vax 7.2).s  N Therefore, the transport can't really tell if the address it is being asked to handle is a TO or CC.m  K However, if the transport scans the "text" list of TOs and CCs, it can then1G determine if a specific address was present as part of a TO or CC list.A   ------------------------------  # Date: Wed, 08 Nov 2000 22:24:29 GMT$- From: goathunter@goatley.com (Hunter Goatley)a0 Subject: Re: VMS VMAIL, SMTP transports, headers0 Message-ID: <3a09d25c.26223317@swen.process.com>  , On Wed, 08 Nov 2000 15:59:15 -0400, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote::   >Hunter Goatley wrote:G >> There was no To: line, which surprises me a little, but technically,uF >> it's because PMDF wasn't told by VMS Mail to process a To: address.G >> PMDF was only told about the CC address by VMS Mail, and that's what  >> it reports. >aK >When you type a transport%address at the TO or CC prompts in VMS mail, the N >transport is called with the same function code (value 2) whether the addressR >is entered in the TO or CC fields. (At least on character cell VMSmail, vax 7.2). >eO >Therefore, the transport can't really tell if the address it is being asked tor >handle is a TO or CC. >-L >However, if the transport scans the "text" list of TOs and CCs, it can thenH >determine if a specific address was present as part of a TO or CC list.  > Correct; that's what PMDF, MX, MultiNet, et al, do to get that info....   Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/p: goathunter@goatley.com      http://www.goatley.com/hunter/   ------------------------------   Date: 8 Nov 2000 13:53:22 -0500e2 From: malmberg@eisner.decus.org (John E. Malmberg)2 Subject: Re: Where's C's PRV$M_READALL & _GRPPRV ?+ Message-ID: <Xo1hkhmuiuma@eisner.decus.org>   . In article <3A096D60.43345936@CCAgroup.co.uk>,4 Chris Sharman <Chris.Sharman@CCAgroup.co.uk> writes:  6 > Trying to build/install John Malmberg's FRONTPORT on% > our Alpha (DECC V6.2-008, VMS 7.1).  > ; > It fails, complaining about PRV$M_READALL & PRV$M_GRPPRV.e& > Other PRV$M_whatever symbols are OK. > < > $ lib/tex/extr=prvdef sys$share:sys$starlet_c/out=prvdef.h > 5 > gives me a module which is indeed missing the masks & > for the bits in the second longword. > 6 > The library was created Jun-98, last updated Jun-99.+ > The PRVDEF module was inserted 25-Nov-96.- > 4 > I've installed patches ALPACRT08_071 & CCAE02_062.   Thank you for the feedback.a  6 Apparently the symbols were not added to starlet until OpenVMS 7.2.  : If you look in the module that the compiler is complaining: about, you will see that there is alternative code for the: VAX architecture that does not expect to be able to handle
 64 bit masks.n    : My VAX is running 7.1, and my Alpha Multia is running 7.2.9 The VAX did not show the symbols, and since the DEC C 6.0:9 compiler for VAX did not admit to being able to handle 64 : bit masks, I wrote a separate version of the code for VAX.  8 I made the assumption that the missing symbols were only# defined for the ALPHA architecture.n    ; I will make a note of this for a potential future update tod< FRONTPORT to put in a #define PRV$M_READALL and PRV$M_GRPPRV  if they are not already defined.    : I would also warn you that the FPORT_DL* rountines may not9 compile and link on OpenVMS Alpha either.  If so, for now 2 just modify them to compile clean and ignore them.8 The next update (that may be a while) will be to rewrite8 the fport_dl*() routines to use LIB$FIND_IMAGE_SYMBOL so9 that they will work with versions of OpenVMS earlier thant 7.2.    0 SAMBA 2.0.6 does not use the FPORT_DL* routines.    2 The FRONTPORT kits were also missing a header file8 HISTORY.H.  It is only an empty file.  A replacement for5 it is at http://eisner.decus.org/~malmberg/frontport/p  1 The HISTORY.H is not needed for building or usingp8 the FRONTPORT library.  It is needed for building SAMBA.  < A bug has just turned up in the SAMBA 2.0.6 UN*X source code7 that a patch for has been posted on the SAMBA-TECHNICAL05 or SAMBA-VMS mailing list, that did not show up in myV5 testing on ALPHA.  It should be in their archives at c) http://www.samba.org for non-subscribers.9  : The bug may show up with TCP/IP versions other than Compaq
 TCPIP 5.0.  5 It showed up on the VAX version, which even with thatS6 fixed is still suffering from an apparent data overrun4 problem that I am currently assuming is a problem in
 lib_xfer.mar.   ; I do not know if the found SAMBA bug only showed up becausep: of the buffer overrun problem in the VAX version, or if it' could show up with other TCP/IP stacks.o    8 At this time I can not troubleshoot it further as I must< shutdown my home VMS systems for a few months in preparation for relocation.i   -John. wb8tyw@qsl.network" malmberg@eisner.decus.organization   ------------------------------   Date: 8 Nov 2000 19:17:05 GMTm2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman), Subject: Re: Where's the Y2K compliance page6 Message-ID: <8uc8rh$71i$2@mailint03.im.hou.compaq.com>  F In article <8u9a5c$jvf$1@nnrp1.deja.com>, abirkett@my-deja.com writes:  F :In particular DEC C 6.0 and DEC Pascal 5.1-11 (now prefixed Compaq ofF :course) He also requested macro-64 but that's on the freeware CD. TheF :client will be using Alpha VMS 7.2-1 by the way. I'll probably end up' :putting the latest versions on anyway.   C   I am not aware of any particular Y2K problems in any of the cited)6   products, nor with any current versions of products.  F   I am aware of reasons to run Compaq C V6.2 over V6.0, not the least B   of which are 21264 code generation support and the detection of    uninitialized variables.  I   I do not know that Macro64 was ever formally investigated.  That said, nI   I would also avoid the use of Macro64 in general.  (I will assume thereiI   is a reason for considering the Macro64 assembler here, and that there n-   is no confusion with the Macro32 compiler.)r  F   The supported version product list -- the software rollout (swroll) I   pages --  is probably the best resource here.  Not the (old) Y2K lists.t4   Search for swroll in the OpenVMS FAQ for pointers.  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   Date: 8 Nov 2000 20:33:37 GMTs2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman), Subject: Re: Where's the Y2K compliance page6 Message-ID: <8ucdb1$80m$1@mailint03.im.hou.compaq.com>  Z In article <0$mvbei+BgnI@eisner.decus.org>, koehler@eisner.decus.org (Bob Koehler) writes:J :  The Feds required year 2000 day 366 testing as a part of Y2K testing.  @ :  Why retire info related to something that hasn't happend yet?  I   Last I looked, 29-Feb-2000 did already happen.  Of course, 2038 hasn't  H   happened yet, and that may well be as much (or more) "fun" as was Y2K.  C   The information that was posted on the Y2K pages is now outdated.g  J   The software rollout report (search the FAQ for "swroll") is likely now J   the best resource for getting (current) version and support information.  K   Static web pages are quite obviously trivially easy to maintain.  Upkeep nJ   of the search engines and the inevitable updates to the current-version I   information are not free -- this data gets stale, and stale informationp3   can generate support questions and support calls.a  I   Archives of the Y2K information have been preserved within Compaq, for  I   those folks that are moving to antique OpenVMS versions and/or layered     product versions.-  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Wed, 08 Nov 2000 21:10:23 GMTt+ From: John Santos <john.santos@verizon.net> , Subject: Re: Where's the Y2K compliance page> Message-ID: <MPG.14738b3b34d5300b989693@news.bellatlantic.net>  7 In article <8ucdb1$80m$1@mailint03.im.hou.compaq.com>, e% hoffman@xdelta.zko.dec.nospam says...i > \ > In article <0$mvbei+BgnI@eisner.decus.org>, koehler@eisner.decus.org (Bob Koehler) writes:L > :  The Feds required year 2000 day 366 testing as a part of Y2K testing.  B > :  Why retire info related to something that hasn't happend yet? > K >   Last I looked, 29-Feb-2000 did already happen.  Of course, 2038 hasn't eJ >   happened yet, and that may well be as much (or more) "fun" as was Y2K.  F Yes, but Year 2000 day 365 and Year 2000 day 366 haven't happened yet.  E (I can't imagine why this year would be different than any other leap-A year, but one patch release of Java on Sun got 29-Feb-2000 wrong,oE so anything is possible.  I think it was in 1.8, the previous release0I worked okay, and by the time we discovered the problem, there was anothere$ version out that worked okay again.)  H Please hold out for another couple of months before deleting everything!  E >   The information that was posted on the Y2K pages is now outdated.C > L >   The software rollout report (search the FAQ for "swroll") is likely now L >   the best resource for getting (current) version and support information. > M >   Static web pages are quite obviously trivially easy to maintain.  Upkeep gL >   of the search engines and the inevitable updates to the current-version K >   information are not free -- this data gets stale, and stale information15 >   can generate support questions and support calls.j > K >   Archives of the Y2K information have been preserved within Compaq, for /K >   those folks that are moving to antique OpenVMS versions and/or layered e >   product versions.  > P >  --------------------------- pure personal opinion ---------------------------N >    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com  I I think a static page or set of pages would serve anyone's purposes just 9G fine.  If the version you are investigating isn't listed and is higher  J than what is listed, I think you could safely assume it was released post E Y2K and so not an issue.  No need to keep updating the list with new uH releases, unless a specific Y2K problem is found with a post-Y2K releaseG of something.  But 99.99% of the time, just saying "Vx.y was current at G Y2K and was tested and found not to have any problems.  Later versions i@ are post-Y2K." would be true and would never need to be updated.   -- e John Santos    ------------------------------   Date: 8 Nov 2000 08:19:38 -0500e, From: koehler@eisner.decus.org (Bob Koehler), Subject: Re: Where's the Y2K compliance page+ Message-ID: <0$mvbei+BgnI@eisner.decus.org>)  k In article <8u939v$g4p$2@mailint03.im.hou.compaq.com>, hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) writes:  > G >   Uh, why?  (Do you know something about 1-Jan-2001 that I don't? :-)  > G >   Do you have a specific Y2K-related problem or Y2K-related question?0 >1J >   Much of the Y2K information has been taken off-line and the retirementM >   process will continue, as the information is no longer of any particular $H >   relevence or value, and there is also a cost to maintain the search M >   engines and related tools.  eg: I removed the Y2K section of the OpenVMS aI >   FAQ (with advanced notice) in a recent issue, as I'd figured that fewi@ >   (or no) readers would be interested in half-year-old "news". >   I   The Feds required year 2000 day 366 testing as a part of Y2K testing.  i?   Why retire info related to something that hasn't happend yet?e  F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences Corporation = NASA GSFC Flight Software       | Federal Sector, Civil GroupSE                                 | please remove ".aspm" when replyingn   ------------------------------   Date: 8 Nov 2000 17:44:11 GMTR* From: helbig@astro.rug.nl (Phillip Helbig)2 Subject: Re: www.northernlight.com using Solaris ?. Message-ID: <8uc3db$32e$1@info.service.rug.nl>  
 In articleA <OF8D98C718.F45159E2-ON03256991.0066820D@ep-bc.petrobras.com.br>,b, fabio_compaq@ep-bc.petrobras.com.br writes:   L > I checked  at     http:// www.netcraft.com  and sounds like Northern Light > isF > using a Solaris server  as http server ?  It=B4s not OpenVMS powered > anymore???  . This should be in the FAQ: The BACKEND is VMS.   ------------------------------   End of INFO-VAX 2000.626 ************************