1 INFO-VAX	Sat, 11 Nov 2000	Volume 2000 : Issue 630       Contents:! RE: *.*;-0: undocumented feature? , Re: A DEC Saga book currently being written?$ Re: ABS/MDMS Tape rotation questions% Re: Adding a SCSI disk to a PWS 433au 8 Re: Backup written on TZ87 to be read on TZ86 - resolved8 Re: Backup written on TZ87 to be read on TZ86 - resolved Re: Backup: Restore of SubDirs
 Re: BALSETCNT 
 Re: BALSETCNT  Re: Broken TK50  COBOL F$EDIT Re: COBOL F$EDIT Re: COBOL F$EDIT Re: Compiler /tune switches. Re: Compiler /tune switches.! Re: DECnet 5 -> 4 routing problem ! Re: DECnet 5 -> 4 routing problem @ Re: Fatal:  Insufficient virtual memory to continue compilation.@ Re: Fatal:  Insufficient virtual memory to continue compilation. FTP beyond a firewall  Re: FTP beyond a firewall 2 Re: I have made some questions, without any answer9 Re: Is there some way to make MAIL> SET TRANSPORT useful?  Re: More fuel on the fire  Re: Nested DCL symbol names  OpenVMS technical reference kit  Pathworks Advanced Server error  RE: RECALL question/suggestion.  Re: sqlpre70 get sqlbugchk.dmp Re: Technology of US elections Re: Technology of US elections! Re: VAXstation 4000-90 Error Code ! Re: VAXstation 4000-90 Error Code ! Re: VAXstation 4000-90 Error Code ! Re: VAXstation 4000-90 Error Code  Re: VMS, BSD, and FORTRAN  Re: VS4000-60 Parts Needed, Re: We Interrupt the US Electoral Follies..., Re: We Interrupt the US Electoral Follies..., Re: We Interrupt the US Electoral Follies..., Re: We Interrupt the US Electoral Follies...# Re: Where's the Y2K compliance page   F ----------------------------------------------------------------------  % Date: Fri, 10 Nov 2000 09:55:00 -0600 > From: "Dryjanski, Tomasz" <Tomasz.Dryjanski@intl.fritolay.com>* Subject: RE: *.*;-0: undocumented feature?- Message-ID: <0G3U00MS61K6VN@mx.west.saic.com>   3      In article <0G3S00NW1BCOQ3@mx.east.saic.com>,  I          "Dryjanski, Tomasz" <Tomasz.Dryjanski@intl.fritolay.com> writes: <      >>      I observed, that *.*;-0 is equivalent to *.*;0,      >G      >No, that is not  the  case.   x.y;0  gives the _highest_ version  C      >number, x.y;-0 gives the _lowest_ version number of the file.       >(...) F      >If you can show an  example  where  what I've described is _not_J      >the  case, post the evidence here, along with VMS version: it  would      >be a bug in DCL.                     -Ken              OK, this was my mistake. ,      Thanks for the answer - now it's clear.             T. Dryjanski    ------------------------------  % Date: Fri, 10 Nov 2000 20:30:03 +0100 0 From: Didier Morandi <Didier.Morandi@Easynet.fr>5 Subject: Re: A DEC Saga book currently being written? * Message-ID: <3A0C4CBC.26510BD5@Easynet.fr>   "Thomas H. Pauli" wrote: >  > You might want to look at  > 8 > http://www.openvms.compaq.com/openvms/20th/vmsbook.pdf   Yep.   I printed it already.   G Fantastic pictures of Ben Schreiber, lovely stories from Andy Goldstein D and his snow storm, and so many other names I know, even Jay Nichols3 (page 4). This guy speaks French better than me :-)   6 I have to spend a pretty big time to read that book...  H But my question was more about an independant author, like Terry Shannon2 for example, or Mr Monosson (does he still exist?)   Thanks anyway. D.   ------------------------------  % Date: Sat, 11 Nov 2000 01:50:28 -0500 ( From: Robert Boyd <rlboyd@bellsouth.net>- Subject: Re: ABS/MDMS Tape rotation questions - Message-ID: <3A0CEC28.776FE8C0@bellsouth.net>    Dave,   L This is a process that requires either manual management or some DCL code toP manage this part.  You could have a post function procedure set the offsite dateP on the tape(s) that the current save request just wrote to.  Or you could have aL completely separate cleanup job that runs after all of the backups that willM modify the offsite date on all of the tapes.  Then as part of the weekly full L backup, run another step that moves all of the tapes that have offsite dates$ that are in the correct time window.  P If you are working with V3.0B be sure to get the latest builds from Compaq.  TheK ABS kit is up to build 57 and MDMS is up to 271 at least.  You can see what @ version you have by doing ABS SHOW VERSION and MDMS SHOW SERVER.  H There are LOTS of hints/kinks to be had about ABS/MDMS depending on your@ specific requirements for managing the backup streams and media.   Bob Boyd rlb14162@glaxowellcome.com   Dave Harrold wrote:  > G > I am in the process of installing and configuring ABS and MDMS V3.0B. 2 > We are using a TL895 library with 4 TZ89 drives. > D > I am having a difficult time trying to figure out and plan for theE > tape rotation.  What I would like to have is a Full backup once per F > week and incrementals the other 6 days.  After that cycle is done, ID > want all of the tapes used in that rotation to be ejected from theH > library to be sent off-site.  I can't seem to have this setting of theE > tape in MDMS to be OFFSITE be automatically.  Do I have to manually 5 > set the offsite date?  If so, how are you doing it?    ------------------------------  % Date: Fri, 10 Nov 2000 16:12:36 -0500 2 From: rdeininger@mindspring.com (Robert Deininger). Subject: Re: Adding a SCSI disk to a PWS 433auL Message-ID: <rdeininger-1011001612370001@user-2ive7o5.dialup.mindspring.com>  k In article <3a0bd387.4932001@news.rz.uni-duesseldorf.de>, Wolfgang.Poll@hhunospam.uni-duesseldorf.de wrote:      > R > Error Log Report Generator                                      Version V6.1          C >                                        ANSI APPROVED VERSION = 3.     H This looks kind of familiar from our problem with a larger quantum disk. It was fixed by VMS 7.1-2.   Again, I make to promises.   --   Robert Deininger rdeininger@mindspring.com    ------------------------------  % Date: Fri, 10 Nov 2000 15:25:41 -0500 ' From: "Dale A. Marcy" <dqm@y12.doe.gov> A Subject: Re: Backup written on TZ87 to be read on TZ86 - resolved * Message-ID: <3A0C59C5.E46D969@y12.doe.gov>   Group,  J      It looks like I will be able to resolve this problem by adding a TZ87H drive to the target system, so both systems will be using the same styleM drive.  Thanks to the various people who have provided input on this problem.   
 Dale A. Marcy . Science Applications International Corporation   "Dale A. Marcy" wrote: >  > Group, > O >      I have a user that has to transfer files via tape from one VAX system to M > another VAX system on a non regular basis.  He had to do it today, the last P > time was approximately a month ago, but it is not at a specific interval.  TheM > problem is that the system that is writing the tape is a VAXstation 4000-60 M > running VMS V7.1 using a TZ87 drive, so when the tape is mounted on the VAX M > 4000-500A also running VMS V7.1, but using a TZ86 drive, it cannot read the L > TZ87 format.  I can and have each time forced the TZ87 to emulate the TZ86O > format, but I would like to automate this so that it does not need to be done M > each time.  The user is using a Backup/Rewind command to create the tape on P > the writing node.  Is there a qualifier that can be used on the backup commandK > to force the TZ87 to write at the TZ86 density?  I looked at the /Density P > qualifier, but I thought it was for the old reel to reel tapes and if it wouldK > work for this, I did not know the bits per inch to specify to emulate the J > TZ86.  Any help would be appreciated.  Hopefully, I have provided enoughN > information here.  I do not have the exact backup command, but could find itE > out if needed.  I know that the backup command does not specify any  > /Media_Format. > 	 > Thanks,  >  > Dale A. Marcy 0 > Science Applications International Corporation   ------------------------------  % Date: Fri, 10 Nov 2000 17:06:00 -0500 5 From: John Johnstone <jj_usenet@onay-amspay.mail.com> A Subject: Re: Backup written on TZ87 to be read on TZ86 - resolved 3 Message-ID: <3A0C2AF8.A09514B@onay-amspay.mail.com>   J I have seen the exact same problem with the TZ87 to TZ86 issue.  It was soJ long ago that I'd forgotten about it.  My work-around was to INIT the tapeF in 6GB format either by doing it on a TZ86 or by using the front panelH density override on the TZ87 and then creating a dummy backup saveset onF the tape.  Backup, running on the TZ87, would then properly append theH saveset as a second one on the tape without initializing it first.  ThisJ thing was real frustrating when I had to deal with it but this was about 5
 years ago.     Dale A. Marcy wrote: >  > Group, > L >      It looks like I will be able to resolve this problem by adding a TZ87J > drive to the target system, so both systems will be using the same styleO > drive.  Thanks to the various people who have provided input on this problem.  >  > Dale A. Marcy 0 > Science Applications International Corporation >    ------------------------------  # Date: Fri, 10 Nov 2000 23:45:37 GMT * From: Alan E. Feldman <alan48@my-deja.com>' Subject: Re: Backup: Restore of SubDirs ) Message-ID: <8ui1b0$so9$1@nnrp1.deja.com>   . In article <8ugb46$sbs$2@info.service.rug.nl>,   helbig@astro.rug.nl wrote:; > In article <8ufr83$3j6$1@nnrp1.deja.com>, Alan E. Feldman  > <alan48@my-deja.com> writes: > < > > > > Why is it a problem that the directory dates change? > > > ? > > > DIR [.IMPORTANT_PROJECTS...]*.DIR/SIN=<some crucial date>  > > > B > > > Perhaps when I start a project, I create a directory for it.
 > > Probably, C > > > I put some files in it soon after, but due to PURGE only much  newer  > > onesH > > > might be around.  Thus, the only way to keep track is by DIRECTORY > > > dates. > > @ > > You can jot the dates down in a notebook or type them into a	 database,  > > if it's that important.  > A > Sure, but the whole point is to avoid redundant work.  Going to F > extremes, why do we need file dates at all---we can just jot down in a C > notebook or type into a database when they were created, revised,  backed > up,....  :-|  0 Hey, RELAX!!! I SAID RELAX!!! Hmmmm..., sorry...  E Well, keeping track of project dates is doable and is much, much less F work than keeping track of all dates of all files. Yeah, that's pretty7 extreme. But it boils down to a question of priorities.    > G > > Old files can be copied to new directories in many other cases. I'd  > > consider that quite normal.  > < > Yes, but not what one expects when a whole tree is copied. >   G At first, yes. However, using the current time for the creation of said F directories is technically correct. While the files are actually beingG copied, the directories are not. They are created on the target disk if C need be, and then the files are added one by one with the .dir file ( being updated with each additional file.  C And what if the directories already exist on the target disk? Which E date would you prefer? The original date of the .dir on the source or  on the target?  @ BUT!!!, aside from that situation, you do have a good point. TheG original creation date of the directory is more useful to the user. And D using such would be more in spirit with the lnm philosophy of device: independence. Well, I'm afraid were stuck with the current implementation.   G Also, I will admit that I too once looked at .DIR creation dates in the F hopes of seeing the original creation dates only to find that they areG what they are. But I did not remember that when I asked my question; it $ was many years ago when it happened.   Moral: Keep a notebook.    -- Disclaimer: JMHO Alan E. Feldman  &-)/ w: afeldman@gfigroup.BorisYeltsinCooksCajun.com 5 h: alan48@dellnet.YouCantBelieveEverythingYouRead.com     & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------  # Date: Fri, 10 Nov 2000 18:46:28 GMT 1 From: "Mark D. Jilson" <jilly@clarityconnect.com>  Subject: Re: BALSETCNT2 Message-ID: <3A0C43DD.1D1F9119@clarityconnect.com>  @ Most likely AUTOGEN is setting VIRTUALPAGECNT to the new size ofF physical memory and thus you are exceeding the size of S0 space.  LookE at your old parameters by finding the VAXVMSSYS.OLD file generated by , the initial AUTOGEN done to size this system $ SET DEFAULT SYS$SYSTEM
 $  MCR SYSGEN  USE VAXVMSSYS.OLD;n  SHOW VIRTUALPAGECNT 
 USE ACTIVE SHOW VIRTUALPAGECNT  EXIT  E If the value in the Current column from the 1st SHOW is less than the F value in the 2nd SHOW then hard set VIRTUALPAGECNT in MODPARAMS.DAT to& the lower value and run AUTOGEN again.  E AUTOGEN only knows about one application, ANALYZE/CRASH, thus AUTOGEN E sizes VIRTUALPAGECNT to be able to analyze a full memory dump.  Since F your system used to run with the (probably) lower VIRTUALPAGECNT value? it still should and the only problems you might have will be in  analyzing a full memory dump.    operagost@my-deja.com wrote: > G > We recently upped the RAM on our VAX4000-705a running OpenVMS 5.5-2H4 E > from 448 to 512 MB. Autogen (not running setparams) now reports the  > following: > > > ** WARNING ** - The user-modified value for BALSETCNT of 198J >         very likely would have caused system virtual address space to beF >         exceeded rendering your system unbootable. You should reviewF >         records pertaining to BALSETCNT, VIRTUALPAGECNT and WSMAX inJ >         MODPARAMS.DAT; these primarily affect the size of balance slots., >         BALSETCNT is being reduced to 182. > G > I'm hesitant to adjust this value, because it might have been set for  > Oracle 7 (MODPARAMS.DAT):  >  > Min_Balsetcnt           = 198 C > Min_GBLPages            = 200000                ! For ORACLE V7.0 H > Min_GBLSections         = 500                   ! For ORACLE V7.0 (min > 400) > J > Why is it that with more RAM I can now only have fewer working sets? How > should I proceed?  > ( > Sent via Deja.com http://www.deja.com/ > Before you buy.    --  D Jilly	- Working from Home in the Chemung River Valley - Lockwood, NY0 	- jilly@clarityconnect.com			- Brett Bodine fan. 	- Mark.Jilson@Compaq.com			- since 1975 or so, 	- http://www.jilly.baka.com               -   ------------------------------  % Date: Fri, 10 Nov 2000 14:20:39 -0500 0 From: David Beatty <David.Beatty.NOSPAM@sas.com> Subject: Re: BALSETCNT2 Message-ID: <ZkkMOj8J7I7WdlG8DMiNLRM1a55C@4ax.com>   I've got some questions:  8     1.  Did any other SYSGEN parameters change recently?6     2.  What are the values for WSMAX, VIRTUALPAGECNT,          and MAXPROCESSCNT? 1     3.  What does the SHOW MEMORY command display           on your system?,     4.  Did the Oracle SGA increase in size?  3 Also, check on MCR SYSGEN HELP PARAMETERS BALSETCNT > for an explanation of BALSETCNT.  Each memory resident process requires one balance set slot.  > On Fri, 10 Nov 2000 18:17:46 GMT, operagost@my-deja.com wrote:  F >We recently upped the RAM on our VAX4000-705a running OpenVMS 5.5-2H4D >from 448 to 512 MB. Autogen (not running setparams) now reports the >following:  > = >** WARNING ** - The user-modified value for BALSETCNT of 198 I >        very likely would have caused system virtual address space to be E >        exceeded rendering your system unbootable. You should review E >        records pertaining to BALSETCNT, VIRTUALPAGECNT and WSMAX in I >        MODPARAMS.DAT; these primarily affect the size of balance slots. + >        BALSETCNT is being reduced to 182.  > F >I'm hesitant to adjust this value, because it might have been set for >Oracle 7 (MODPARAMS.DAT): >  >Min_Balsetcnt           = 198B >Min_GBLPages            = 200000                ! For ORACLE V7.0G >Min_GBLSections         = 500                   ! For ORACLE V7.0 (min  >400)  > I >Why is it that with more RAM I can now only have fewer working sets? How  >should I proceed? >  > ' >Sent via Deja.com http://www.deja.com/  >Before you buy.   ------------------------------  % Date: Fri, 10 Nov 2000 18:30:21 -0500   From: John Santos <JOHN@egh.com> Subject: Re: Broken TK505 Message-ID: <1001110181326.2092C-100000@Ives.egh.com>   + On Thu, 9 Nov 2000, Robert Deininger wrote:   m > In article <MPG.1474de0592a4f025989695@news.bellatlantic.net>, John Santos <john.santos@verizon.net> wrote:  > # > > I think I got it working again!    May have spoken too soon... :-(    > > & > > Thanks to everyone who helped out.   [...]s  N > Yes.  The LED pushes into one hole, and the photodiode pushes into the otherL > (the one at the top of the tach assembly).  I actually don't know which isK > which, since the only one I've taken apart didn't have power at the time.o8 > Could you see light from the LED, or was it invisible? >   F I had to pull power from the drive before removing it from the cabinetD (power cable isn't long enough), so I don't know.  It was amber, andB looked like an LED, but I'm not sure what a photodiode looks like.  L > The photodiode "looks" at the LED through the spinning "screen" inside theH > tach (attached to the shaft), so it alternately sees light and dark as > the shaft turns. >  tD > > I'm not sure if this was the original problem or if I knocked it4 > > out while trying to fix the original stuck tape. > M > Likely the original problem.  With a bad tach, the drive would have no ideadO > whether the tape loaded successfully, or how much tape it had wound.  I thinkrM > it just gives up at some point.  If it tried to rewind and unload, it wouldhJ > risk breaking the tape.  So it refuses to power the little solenoid that7 > unlocks the door latch, and lets a human sort it out.A  J Hmm...  Current problem is it seems to be having major problems unloading.E Once I get a tape loaded, and mounted, it seems to read okay (haven'tcB tried writing), but trying to unload the tape, it rewinds to aboutI beginning of tape, seems to wobble a bit, and then starts fast forwarding I through most of the tape.  Eventually it stops, rewinds again to or close-E to the BOT and then does it again.  One time while I was watching it,0E it seemed to go all the way to BOT, I heard a THUNK like the solenoidoD had tripped, the red button light went out, and green light came on,D and I tried to lift the handle and discovered it was fast-forwardingF again!  Ugh!  Then the red and green lights started flashing again, itC went through its dance a few more times, repeated the THUNK and theeC green on, red off business.  I let it sit for about 10 seconds, andsC looked inside, and it really seemed unloaded this time.  Lifted thesC handle, sure enough, the tape was rewound and entirely in the cart,mA leader in place, etc.  So I tried to load another tape.  Took it, D went to load point (I think), but I couldn't mount it.  Said "VolumeA invalid" in VMS.  (I was able to read the previously-mounted tapee@ fine.)  I did a manual unload, it started its dance, I let it be@ for an hour and when I came back, it was unloaded.  That's where I left it for now.  B Maybe a problem with the BOT sensing?  There seems to be a slot in@ the part of the leader that is permantly attached to the take-upA reel that should be sensed by the same LED/photodiode that senses A the hole at EOT, maybe that isn't working right?  I think the EOTpB hole is probably ignored when reading.  When it reaches the end of@ a track, the read head probably senses this and initiates a head= reposition to the next track?  (This is the way 9-track tapessC worked.  The EOT was only monitored on writes, and caused a specialg@ write completion status.  The program was supposed to then writeB end-of-volume labels (past the EOT), rewind the tape, and start onG the next vol.  When reading the tape, you ignore the EOT, but when you -@ saw the end-of-volume labels and tape marks, you switched to theA next reel.  It would make sense for the TK50 to handle the end of I track the same way, but then if there was a problem with the end-of-trackcA sensor, you would only notice while writing or unloading.  I likea+ this theory... ;-)  I will check this next.    -- r John Santoss Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Fri, 10 Nov 2000 15:30:02 -0800t/ From: Terry Marosites <TMarosites@unitedad.com>t Subject: COBOL F$EDIToM Message-ID: <1137A4A23A51D311B2D600105A1D5213026FDFDC@seantexch.unitedad.com>o   Hi all,t  L    I want to an f$edit from COBOL if I call lib$do_command I don't come back
 to my programtI  if I do a lib$spawn my symbol is set in the spawned process. does anyoneo2 know an easy method of doing a f$edit from COBOL.    Thanks Terry    ------------------------------  % Date: Fri, 10 Nov 2000 18:52:15 -0400o- From: JF Mezei <jfmezei.spamnot@videotron.ca>g Subject: Re: COBOL F$EDITe, Message-ID: <3A0C7C11.F9E43D7C@videotron.ca>   Terry Marosites wrote:N >    I want to an f$edit from COBOL if I call lib$do_command I don't come back > to my programB    T I beleive that most of the functionality of F$edit is provided by the STR$ routines.  @ HELP RTL STR$ will give you a listing of the available routines.  M From cobol, you can call these routines, and pass strings as arguments . It'sl/ been a while, but you can do something such as    P CALL "STR$UPCASE" USING destination-str by descriptor, source-str by descriptor.   ------------------------------  % Date: Fri, 10 Nov 2000 21:24:39 -0600t7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>  Subject: Re: COBOL F$EDITo- Message-ID: <3A0CBBF7.E704A8E9@earthlink.net>-   Terry Marosites wrote: > 	 > Hi all,2 > N >    I want to an f$edit from COBOL if I call lib$do_command I don't come back > to my programPK >  if I do a lib$spawn my symbol is set in the spawned process. does anyoner3 > know an easy method of doing a f$edit from COBOL.d  @ I have memories from long ago of being able to call the BAS$EDITB function from BASIC run-time library from a non-BASIC 3GL. PerhapsE someone else will chime in with the specifics. I think I read it hereA first a while back.=  C I don't have BASIC installed on my hobbyist Alpha just yet, and theaB MicroVAX 3100 does not run just now. Otherwise I could compile the7 following /MACHINE_CODE/LIST and post the machine code:    	MAP (SAMPLE) &- 	STRING- 	TEST_STRING	= 32%   	A$ = EDIT$( TEST_STRING, 16% )    	END PROGRAM  E A MACRO coder could then divine the calling sequence to make BAS$EDITsH work in another language. Basically (no pun intended), you pass a stringC by descriptor, a word integer by reference (I believe), and anothervG string by descriptor to receive the output of the function. The integervB is a bitmap of functions to perform: trim leading/trailing spaces,F upcase/locase/compress/collapse the string, string control characters,C etc. Trouble is, I don't have the doc.'s or a compiler available. IaE don't know the sequence of arguments or the bits for the functions by  memory.   C Now that I think about it, I think I saw that back in the 80's in aoE DIBOL program. I may have the source on an old TK50, if I can spin upb# the old MicroVAX (it has a TZ30)...g   -- p David J. Dachtera0 dba DJE Systems  http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/8  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: 10 Nov 2000 20:08:16 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)% Subject: Re: Compiler /tune switches./6 Message-ID: <8uhkjg$5gt$1@mailint03.im.hou.compaq.com>  o In article <rjq66lwvxu3.fsf@morwong.ucc.gu.uwa.edu.au>, Paul Repacholi <prep@morwong.ucc.gu.uwa.edu.au> writes: E :The C compiler has EV6 and EV6_2 as options. What is the difference?h  '   Um, the latter is not documented? :-)a  B :Is it possible to use MM ( EV6 )etc without getting the byte/word! :load/stores ( EV56 ) generated? n     Not that I am aware of.h     What are you up to?c  @   My preference for most compilations is /ARCH=EV56, when I needA   the performance (and when I am not targeting pre-EV56 systems).   =   The /OPTIMIZE=TUNE should generally be left set to GENERIC.c  @   Note that recent versions of GEM will generate AMASK-protected"   code sequences for use in loops.  & :If not, can this be a wish list item.  ?   I don't know enough about the GEM internals to know what the b=   level of difficulty would be for this suggestion.  (It doesh5   run contrary to the current mechanisms, obviously.)e  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Fri, 10 Nov 2000 15:59:13 -0500u" From: Dan Sugalski <dan@sidhe.org>% Subject: Re: Compiler /tune switches.a; Message-ID: <5.0.0.25.0.20001110155434.01a75dd0@24.8.96.48>w  / At 08:08 PM 11/10/00 +0000, Hoff Hoffman wrote:7? >   The /OPTIMIZE=TUNE should generally be left set to GENERIC.   L Is there much benefit to doing otherwise? The help info's rather... sparse. I :) If a /TUNE=EV6 gets me a boost of a few percent I'm all for the extra r
 compile time.    					Dan  I --------------------------------------"it's like this"-------------------d2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and evena;                                       teddy bears get drunkg   ------------------------------  % Date: Fri, 10 Nov 2000 17:08:55 -0500   From: John Santos <JOHN@egh.com>* Subject: Re: DECnet 5 -> 4 routing problem5 Message-ID: <1001110164014.2092A-100000@Ives.egh.com>S  * On Wed, 8 Nov 2000, antonio.carlini wrote:   > John Santos wrote: > > = > > Yes.  Local nodes are all area 1.x  There are currently 2 8 > > remote areas connected through the DR250: 33.x where: > > the problematic node lives, and 4.x, which works fine. > 2 > Just out of curiosity what is the node number of. > the DR250 that connects to VAX 6620 (33.11)?  6 Local DR250 is 1.17.  There is no remote DR250 at that7 site.  Serial line from DR250 (lease line) connected to 5 a TX port on the 6620 (async DECnet.)  The connectionl& between the 6620 and 6630 is ethernet.  - > I cannot remember the relevant NCP commands,1 > but you should be able to find out whether this 2 > DR250 sees the VAX as an area router. (I know it- > works in segregated mode but I'm curious asu1 > to how this could fail ... and someone's nickedl# > my DR250 so I cannot try it out!). >   
 On local VAX,   	 $ mcr ncp , NCP>set exec node rtr user foo password pass NCP>sho adj node stat     / Node Volatile Status as of 10-NOV-2000 16:47:27l   Executor node = 1.17 (RTR)   State                    = onp, Physical address         = AA-00-04-00-11-04    L     Node        State      Active  Delay  Type           Cost  Hops  Circuit                             Links  L  1.5 (ALPHA1)   reachable       0      2  nonrouting IV     1     1  ETHER-1( Next node to destination =  1.5 (ALPHA1)L  1.16 (VAX)     reachable       1      2  area IV           1     1  ETHER-1& Next node to destination =  1.16 (VAX)L  1.23 (ALPHA2)  reachable       0         nonrouting IV     1     1  ETHER-1) Next node to destination =  1.23 (ALPHA2)mM  2.4 (RMDRTR)                                                        SERIAL-2i( Next node to destination =  2.4 (RMDRTR)M 33.11 (RVAX1)                                                        SERIAL-5n( Next node to destination = 33.11 (RVAX1)  K Alpha1 & 2 are the local Alphas.  Alpha1 is the one I changed to segregatedr* mode.  Alpha2 is still in integrated mode.  & VAX1 is the local VAX (Phase 5 router)  A RMDRTR is the remote DR250 at the site that works.  (The only VMSeC node at that site is a VAX 6650 VMS V7.1 DECnet Ph 5., address 2.3.eG There used to be another VAX at that site -- 11/785!, clustered, sharedp? system disk, same software, address 2.2 that also worked fine.)r  H RVAX1 is the 6620, VMS V7.1 DECnet Ph 4, is clustered with RVAX2, sharedB system disk, same software, address 33.12.  It is connections fromG ALPHA1 and ALPHA2 to RVAX2 that don't work.  (ALPHA1 to RVAX2 now worksRF since changing to segregated mode.)  BTW, there used to be (until lastB November) a pair of PDP-11's running RSTS/E V10.1 at the same site: (33.13 and 33.14) that showed the same behaviour as RVAX1.  D I think this conforms with what you are thinking about, but I posted# the whole thing to avoid confusion.e   -- y John SantosE Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Fri, 10 Nov 2000 17:34:43 -0500v  From: John Santos <JOHN@egh.com>* Subject: Re: DECnet 5 -> 4 routing problem5 Message-ID: <1001110170925.2092B-100000@Ives.egh.com>o  * On Wed, 8 Nov 2000, antonio.carlini wrote:   [...]y  - > I cannot remember the relevant NCP commands 1 > but you should be able to find out whether thise2 > DR250 sees the VAX as an area router. (I know it- > works in segregated mode but I'm curious ase1 > to how this could fail ... and someone's nickeda# > my DR250 so I cannot try it out!)r  4 P.S.  About DR250's...  Since upgrading our MOP boot3 systems to V7.1/DECnet 5, downline loads of DR250's 2 no longer work!  (At least not without significant manual intervention.)   5 Downloading everything else (Decserver 100, 200, 700,e6 550, MUXserver 300, X25 routers (now defunct, I forget4 the model numbers, DEMSA-xx? maybe) works fine.  The4 only boxes that have problems are DR250's.  This has3 happened with 3 different DR250's at 3 sites.  (OneN8 site had field service swap in 2 DR250's before deciding it wasn't a hardware problem.)  5 Here's what happens:  The DR250 appears to load okay, 4 but immediately crashes and does an upline dump back7 to the load host.  If you don't catch it (middle of the 7 night) you soon use up all the free space on the system  disk of the load host.  4 Watching on a terminal, you see it init, then try an4 ISO8802 load.  If it can't find a load host, it then7 tries an Ethernet load, followed by a Serial line load.w6 It then waits a short while and tries again.  (This is7 what you see happen if you don't have a load host or if-6 you unplug the ethernet port before it has time to try5 a load.)  DECnet seems to respond to the ISO8802 load % request, but must do something wrong.	  # Here's the trick to get it to load:.  < 1) rename the load image file from RT2rtrname.SYS to .XXX on> the load host.  (It is much easier to do this if you only have one load host configured.)    ( 2) reply/enable=network on the load host  > 3) type ahead a command to rename back the load image to .SYS, but don't type the <CR>.   4) power-up the DR250.  ; 5) After it finishes self-test, it will say it is trying and7 [ISO8802] download.  There will be a bunch (6-8?) OPCOMn9 messages about the load host seeing the download request,e/ but not being able to find the load image file.E  : 6) The DR250 will give up on ISO8802 and try an [ETHERNET]	 download.r  ; 7) Hit the <CR> on terminal, restoring the load image file.0  8 8) with any luck, you'll do it while the DR250 is trying9 Ethernet and it will work.  If it doesn't, the DR250 will5: try serial, wait a bit and go back to ISO8802.  While it's8 waiting, rename the file back to .XXX and try again.  If: it crashes, you will see OPCOM messages about dumping; try again.  : It is possible to do this blind (i.e. a remote DR250 while8 dialed into the remote load host on a dial-up line), but6 it usually takes 3-4 tries to guess when to rename the load image file back to .SYS.A  : So, there is obviously some difference between ISO8802 and7 Ethernet downloads, but I couldn't find this documenteds9 anywhere.  I also tried using LANCP instead of DECNet forf9 MOP handling, with the same results.  (I could have sworn 8 it worked once when I tried to force LANCP to use V3 MOP3 only, but I was never able to replicate this.)  The 5 difference between MOP V3 and MOP V4 isn't documentedi8 anywhere, but LANCP HELP says /V3 is for nodes that have9 a problem with MOP V4.  Is this the same thing as ISO8802h: vs. Ethernet, or does that have to do with Ethernet packet- format, an entirely different protocol level?   5 P.S.  We ended up putting all the DR250's on UPS's too minimize rebooting.    -- 0 John Santos1 Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------   Date: 10 Nov 2000 15:45:53 PSTT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)I Subject: Re: Fatal:  Insufficient virtual memory to continue compilation.t3 Message-ID: <obJ7JBRNv+BR@mccdev.slac.stanford.edu>E  3 In article <3A0C1E0E.24A6EBC8@clarityconnect.com>, t7     "Mark D. Jilson" <jilly@clarityconnect.com> writes: A > I'd bet that your UAF WSEXTENT entries were being overridden bynE > PQL_MWSEXTENT which gets set to WSMAX which gets set to the size ofL > physical memory.  6         Yep, mea culpa.  Sorry for the misinformation.  H         Our PQL_MWSEXTENT on the AS4100's is  150000.  And we _do_ page,H     just  not very much, and since each system has four page files sizedH     to the max allowed, the percent usage in those files has been  quiteH     small.   What  I  should  have said was that with 2 GB of memory, itD     took three years before we ever _swapped_ on these systems.  :-)  H         I'm glad Mark posted  this  since  a  colleague  pointed out theH     error  to  me yesterday and I hadn't yet had a chance to look up  my1     erroneous missive and issue the correction...e           -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: Fri, 10 Nov 2000 19:03:07 -0400u- From: JF Mezei <jfmezei.spamnot@videotron.ca>tI Subject: Re: Fatal:  Insufficient virtual memory to continue compilation. , Message-ID: <3A0C7E9D.191BEBB8@videotron.ca>  9 "Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515" wrote:gJ >         Our PQL_MWSEXTENT on the AS4100's is  150000.  And we _do_ page,J >     just  not very much, and since each system has four page files sizedJ >     to the max allowed, the percent usage in those files has been  quiteJ >     small.   What  I  should  have said was that with 2 GB of memory, itF >     took three years before we ever _swapped_ on these systems.  :-)  M I compiled KERMIT once, on a Microvax 3100 with 16 megs.  A few days into thesK compile :-) it got to a file that was just too challenging for the compiler N :-)  I had to increase the PGFLQUO for that process and made damened sure that) my physical page files were ample enough.e  M Not everyone has forgotten what paging and swapping is all about. When I do aeL backup on my all mighty microvax II, about 45 of the 50 or so processes thatH live there get swapped out to give Backup room to get a 16k working set.  , Not everyone has the luxury of a 2gig RAM...   ------------------------------  % Date: Fri, 10 Nov 2000 14:04:49 -0500t2 From: "Akshaya Gupta" <akshaya@nortelnetworks.com> Subject: FTP beyond a firewall0 Message-ID: <8uhh8g$nok$1@bcrkh13.ca.nortel.com>  H I need to FTP files from VMS to a UNIX box.  This UNIX box is behind theI firewall and needs another authentication of username and password.  I am I able to connect to this unix box interactively.  But, when I am trying tonK put things in a DCL script, it fails.  As FTP common on VMS side has optionlL of one /USERNAME and one /PASSWORD attributes only.  Is there any work aroudH for doing the FTP across a firewall.  I am sure some one must have tried this before.   Regards,	 ~ Akshayav   ------------------------------  # Date: Fri, 10 Nov 2000 22:30:04 GMTn+ From: Craig A. Berry <calepine@my-deja.com>e" Subject: Re: FTP beyond a firewall) Message-ID: <8uhst4$p5v$1@nnrp1.deja.com>r  0 In article <8uhh8g$nok$1@bcrkh13.ca.nortel.com>,5   "Akshaya Gupta" <akshaya@nortelnetworks.com> wrote:l? > I need to FTP files from VMS to a UNIX box.  This UNIX box isk
 behind the; > firewall and needs another authentication of username and- password.  I am-A > able to connect to this unix box interactively.  But, when I am0	 trying toeB > put things in a DCL script, it fails.  As FTP common on VMS side
 has option9 > of one /USERNAME and one /PASSWORD attributes only.  Isc there any work aroud? > for doing the FTP across a firewall.  I am sure some one mustc
 have tried > this before.  A Assuming the firewall is set up to function as a proxy, somethingnC like the following syntax might work, where ruser and rpass are ther; logins you need for the remote host and puser and ppass are0% what you need for the proxy/firewall:e  < $ ftp/username="ruser@remotehostaddress"/password="rpass"  - proxyaddress"puser ppass"::n    & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------  % Date: Fri, 10 Nov 2000 16:13:09 -0500e2 From: norm lastovica <norman.lastovica@oracle.com>; Subject: Re: I have made some questions, without any answer * Message-ID: <3A0C64E5.454CC5AE@oracle.com>   Rudolf Wingert wrote:r >  > Hello, > D > days ago I have made some questions without any answer. What couldH > be happened. Is there nobody with an answer, or are my emails blocked?" > The subject of the emails where: > ! >         1. Change of IP subnet:0E >            I have to change our IP adresses from class C to class bn0 >            net and have asked for an easy way.  5 	sounds like a good time to use your support contract % and call compaq vms support directly.$  A >         2. Network PostScript printer and non PostScript files:rE >            I have asked for a convert to PostSrict mechanisem using ' >            a standard queue symbiont.i > / >         3. OpenVMS 7.1-2 and cluster forming: H >            We have problems to reboot a bootserver under OpenVMS 7.1-2< >            and I have asked for a patch or other solution.  5 	sounds like a good time to use your support contract % and call compaq vms support directly.o   > 3 >         4. Good news (ironical) from Sun service:n; >            This was only a report about good Sun service.C > E > Hope that there is anybody, who can give an answer to the questionsT > one to three.  >   > TIA and regards Rudolf Wingert   -- O> norman lastovica / oracle rdb engineering / usa / 610.696.4685   ------------------------------  % Date: Fri, 10 Nov 2000 16:57:09 -0400p- From: JF Mezei <jfmezei.spamnot@videotron.ca>rB Subject: Re: Is there some way to make MAIL> SET TRANSPORT useful?, Message-ID: <3A0C611F.7E7BA162@videotron.ca>   Christoph Gartmann wrote:rN > The above looks all right to me. But the SMTP% will only be used if you sendM > a mail from within VMS-Mail. And it will not pass through PMDF in this caser0 > (unless you told Multinet somewhere to do so).N > Your forwarding address is not related to the default transport. That is, ifJ > you receive mail this mail will be directly forwared to PMDF via the IN% > transport.  N Consider that SET TRANSPORT does not have a help entry (and hence undocumentedL ?).  I also had, under some circumstances, some problems with some automatedH emails sent to my VMSmail address. The forwarded address would receive aI non-delivery notification for the original message (and that non-deliveryeH notification contained the original message). So I still got the messageI anyways, but it was in the form of "Sorry, I couldn't deliver to you, thee following message" ! ! !    I While I suspect it was a MRGATE problem, it nevertheless seemed to make aaM difference to your forwarding address under some circumstances when you had a  default transport defined.  N But more importantly, consider the V7.2 behaviour of assuming any address thatF contains an @ goes out with SMTP% as transport, disregarding any otherL transport information. Also, consider the differences between a message thatI is sent interactively on the same node, and one which goes through decnetgN where the message is actually delivered by another process on the target node.   ------------------------------  % Date: Fri, 10 Nov 2000 17:03:13 -0400s- From: JF Mezei <jfmezei.spamnot@videotron.ca>d" Subject: Re: More fuel on the fire, Message-ID: <3A0C628A.12E1B69F@videotron.ca>   Jim Agnew wrote: >  > hhmm.. interesting... won't those guys EVER learn??? everybody from Clinton to Firestone, Intel, all tried to cover up like blue  L Didn't Clinton try to UNCOVER instead of cover his stuff ? I though that theM mere presence of the pizza delivery girl would cause Clinton to uncover his 2o! nodes and hard ethernet segment ?b   ------------------------------  # Date: Fri, 10 Nov 2000 23:31:16 GMT-* From: Alan E. Feldman <alan48@my-deja.com>$ Subject: Re: Nested DCL symbol names) Message-ID: <8ui0g4$s6h$1@nnrp1.deja.com>-  ) In article <8uh390$23a$1@nnrp1.deja.com>,u$   Fatz <fatz_nyc@my-deja.com> wrote: >j > [...stuff omitted...]wG > (2) CALL doesn't seem to do more than 1 level symbol substitution for F > parameters to a subroutine, whereas it does for the subroutine name. > [...more stuff omitted...]    F You can do 2 levels, but perhaps not with prefixes being added at each step. Example:   dcl> TYPE MAKE.COM $   ZERO == "ONE"f $   ONE == "TWO" $   TWO == "THREE" $   THREE == "FOUR"D $   FOUR == "FIVE" dcl>
 dcl> @MAKE dcl> dcl> TYPE P.COM- $ SH SYM /LOC/ALL  dcl>" dcl> @P '&ZERO '&ONE '&TWO '&THREE   P1 = "TWO"   P2 = "THREE"
   P3 = "FOUR" 
   P4 = "FIVE"a	   P5 = ""a	   P6 = "" 	   P7 = "" 	   P8 = ""  dcl>  C You must use uppercase values for the symbols, however, because DCL G does not upcase anything after ampersand substitution. The CALL commandD works the same.i  A What's happening here is that the apostrophe is forcing ampersand ; substitution during phase 1. If the result of the ampersandr? substitution is a symbol spelled with no lowercase letters, therE apostrophe will perform substitution for that, yielding two levels of > substitution in one line ***in phase one***, whereas ampersandD substitution normally happens at or near the end of phase 2. And forE some reason, phase 2 ampersand subst. just doesn't happen for P1 thru2 P8 of the @ and CALL commands.  D You can also use '& on the right side of the :=[=] operator, and youC can use ''& inside quoted strings. And you can use &'& to get threea< levels of substitution in situations where phase 2 ampersand substitution is valid.  @ Also, '& is iterative the same way ' is. However, ZERO = "'ONE'"C followed by value = '&ZERO' will not work because the ' attempts toaD translate 'ONE', which is not a symbol. But with ZERO = "'&ONE'" andC the values given above for TWO and up, then TYPE 'ZERO will exhibits> iterative substitution, yielding TYPE THREE, for example. IOW,D the "intermediate" symbols values of '& translation must not containC any subst. operators. But if the final result of '& subst. contains / either ' or '&, iterative subst. will continue!s  9 Now making this work with prefixes, that's another issue.L    ! >  [...yet more stuff omitted...]u@ > (3) There are always alternative ways of doing things and it's probablyC > extremely rare, if it happens at all, that people are forced intoR having3 > to perform this kind of substitution on one line.e  ? Agreed. And those who write their DCL code in lower case shouldd% definitely avoid & whenever possible.4  F > Having said all that - any comments on (2)?  And why does & not workE > when there's a prefix?  It appears there's something wrong with DCL  > because this works:  > 
 > $ test& = 5R > $ sho sym test&o3 >   TEST& = 5   Hex = 00000005  Octal = 00000000005c >AD > That's probably why you can't substitute with prefixes because DCL isn't @ > recognising the & as a special character when it appears later# > than position 0 in a symbol name.a >oE > According to the docs, & shouldn't be a valid character in a symbolh name? > and to my mind, docs or no docs, it's a special character: it!	 shouldn't,@ > be allowed in symbol, file, logical or any other kind of name.  G Well, it is. And since it's not documented as such, I would consider it2D [the use of special characters in symbol names] also as unsupported.E Smolik symbols (those whose names contain lowercase letters) are alsooF unsupported (but fun!) and can be created using the ampersand! To wit:   dcl> set verify  dcl> dcl> type smolik.com
 $ A = "Hello"t
 $ &A = "HAH!"3 $ sh sym H"elloh dcl>
 dcl> @smolik;h
 $ A = "Hello".
 $ &A = "HAH!"s $ sh sym H"elloa   Hello = "HAH!" dcl>  E Very clever! (These Smolik symbols were first brought to my attentionE@ in a post by Gotfryd Smolik.) This works because & works *after*2 command evaluation and *before* command execution.  E (Logical names containing special characters is a whole nuther story!r
 Don't do it!)   G ***********************************************************************0G Also, the VMS user's manual warns you not to use the ampersand operatorPA for symbol subst. unless it is required to translate your symbolsR
 correctly.G ***********************************************************************r  G There is, however, a fun, yet safe, undocumented use for the ampersand:    dcl> TYPE COMMENT-DEMO.COM $    SET VERIFY. $    ZERO = "ONE"u $!   ZERO = '&ZERO'o $    SET NOVERIFYn dcl> dcl> @COMMENT-DEMO;l $    ZERO = "ONE"a $!   ZERO = 'ONE'i $    SET NOVERIFYr dcl>  G As you can see, you can translate symbols embedded in comments when SETvD VERIFY is in effect. To do this you ***MUST*** spell the symbol nameD using uppercase letters or it won't work. ($!'&zero' will not work.)  E There is also a use for the ampersand in PIPE commands (see the FAQ).d< There may be still yet another use, but I forget what it is.  ! All examples were run on VMS 6.1.p -- Disclaimer: JMHO Alan E. Feldman  &-)/ w: afeldman@gfigroup.BorisYeltsinCooksCajun.com 5 h: alan48@dellnet.YouCantBelieveEverythingYouRead.com     & Sent via Deja.com http://www.deja.com/ Before you buy.i   ------------------------------  # Date: Fri, 10 Nov 2000 22:47:57 GMT & From: "Marco Shaw" <marco@nbnet.nb.ca>( Subject: OpenVMS technical reference kit7 Message-ID: <xW_O5.2735$rm4.60651@sodalite.nbnet.nb.ca>h  K Looking for someone who may want to sell their 'OpenVMS technical references kit'.-  6 Email me directly, by simply replying to this message.   Marco    ------------------------------  # Date: Sat, 11 Nov 2000 02:42:54 GMTi From: dankel1963@my-deja.com( Subject: Pathworks Advanced Server error) Message-ID: <8uibne$4nm$1@nnrp1.deja.com>   0 I get an error setting permissions on a share in/ pathworks advanced server.  I am trying to givep/ DALKEM access to his root directory on VMS.  Hea0 is a member of the group Domain Users only.  Any suggestions?  ( NTSERVER\\ALPHA1> mod share dalkem/perm=
 (dalkem=full)h( %PWRK-E-ERRMODSHAREPERM, error modifying permissions for share "DALKEM"- -LM-E-NOTALLMAPPED, one or more user or groupl names were not found    & Sent via Deja.com http://www.deja.com/ Before you buy.i   ------------------------------   Date: 10 Nov 2000 21:36:26 GMT* From: helbig@astro.rug.nl (Phillip Helbig)( Subject: RE: RECALL question/suggestion.. Message-ID: <8uhpoq$l6i$1@info.service.rug.nl>  H In article <8uhal4$907$1@nnrp1.deja.com>, jgessling@my-deja.com writes:   D > Yes, I've noticed that too.  That's why I proposed something like: >  > $ set proc/pipe=all  > or > $ set proc/pipe=explicit > = > To make all dcl commands have an implied pipe (all) or makelB > you specify it (explicit).  Wouldn't make much (any) performanceE > impact and would save a lot of typing.  Like when I start a commandeH > and the realize continuing it with a pipe would be nice and then doing= > the "<crtl>A <ctrl>H pipe" dance to add it to the begining.n  
 How about    $ set prompt="$ PIPE "  9 which might help if you are fast with cut and paste.  :-)e  I One reason I like $ as the prompt is that one can quickly copy the WHOLE N/ LINE with the mouse and it will be valid input!e   ------------------------------   Date: 11 Nov 2000 02:03:23 GMT! From: f1chan@hkucs.org (CHAN Fei)s' Subject: Re: sqlpre70 get sqlbugchk.dmp * Message-ID: <8ui9db$loc$1@www.csis.hku.hk>  A I found that the problem occurs only for the non-ansi style code,m; after I've changed to ANSI format, this problem was solved.uB I think sqlpre70 got serious prolbem in pre-compile non-ansi style sql-cobol programs.       " CHAN Fei (f1chan@hkucs.org) wrote: : Dear all,-> : 	I got serious problem when pre-compile an embeded-SQL COBOLK : program, the problem was compiled okay all the way, until the source code)6 : grows over 1280 lines(the space lines are inclusive)  	 : errors:a   : $ scob == "$sql$pre /cobol"@ : $ scob rnpextrntivb.sco 1 : %SQL-I-BUGCHKDMP, generating bugcheck dump file,# : DKA200:[D07CHINESE]SQLBUGCHK.DMP;t          & : and messages in the dump file are :   7 : This is a AlphaServer 800 5/333 running VMS  V7.1-1H2h) : Current time is 10-NOV-2000 14:28:42.87 D : Bugcheck Dump Identification is "QZYQZ8MXQ3HBAAAE2U75GU6VAAUKAAAA" :  nR : ================================================================================ :           Stack DumpR : ================================================================================  R : --------------------------------------------------------------------------------@ : ***** Exception at 006D81AC : SQL$$CREATE_SQL_FIELD + 00000844= : %SYSTEM-F-ACCVIO, access violation, reason mask=00, virtualt : address=0000000000D   $ : ..... (error codes in hex deleted)    L : I've increase Page fault quota, Working set quota/extends, but seems theseE : values are not violated. Anyone could provide hints on the problem?  : many thanks.   ------------------------------   Date: 10 Nov 2000 23:50:18 GMT2 From: mathog@seqaxp.bio.caltech.edu (David Mathog)' Subject: Re: Technology of US elections , Message-ID: <8ui1jq$fkq@gap.cco.caltech.edu>  \ In article <3A0B8581.753F833F@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:O >From the coverage of the florida elections, I got quick glances at what looked N >like the old punched cards, and there was mention of voters punching holes to >make their selections.n  G In my district we have punch cards and we use a sort of little nail to fJ knock out one of the prepunched holes.  The card is a little lighter than I the punch cards I recall from the days of old. The problem is that if you,G don't knock the little tab all the way out it may read either closed ordH open.  Hence the "hand count" they are doing in many places in Florida. G Seems people can reliably tell if the tab is in or out but the machinesaJ cannot.  On the other hand, it's hard to believe that they don't make someB other sort of error to offset the gain in accuracy - for instance,7 accidentally scoring the same card twice or not at all.c  O >What sort of technology is actually used ? And is this used nationwide or just  >in Florida.  D It varies from place to place.  And thanks to the electoral college K nonsense votes in different states have slightly different values.  To makexJ it even more ridiculous in this election the difference in Florida betweenF the candidates' totals are well within the error levels of the variousJ technologies and the voters themselves.  (Ie, folks pushing the wrong holeG by accident.)   The founding fathers got a lot of things right but theyuM wrote the constitution before statistics were well understood - and it shows.   J I think the best solution would be to utilize another 18th century problemC solving methodology - the two candidates could engage in a good oldaK fashioned duel.  The winner gets to be president.  And with any luck they'dn both be crack shots! c   David Mathog mathog@seqaxp.bio.caltech.eduo? Manager, sequence analysis facility, biology division, Caltech h   ------------------------------  % Date: Fri, 10 Nov 2000 17:36:45 -0800e! From: Shane.F.Smith@Healthnet.come' Subject: Re: Technology of US elections-D Message-ID: <OF5CE6598F.ED24DDBE-ON88256994.0008C48C@foundation.com>  K Paraphrasing Douglas Adams badly from memory, "Anyone sufficiently vain andhF arrogant to campaign to get themselves elected, is the last person who should be elected."p   Shanea          F mathog@seqaxp.bio.caltech.edu (David Mathog) on 11/10/2000 03:50:18 PM  / Please respond to mathog@seqaxp.bio.caltech.eduo   To:   Info-VAX@Mvb.Saic.Com  cc:   ( Subject:  Re: Technology of US elections    5 In article <3A0B8581.753F833F@videotron.ca>, JF Mezei & <jfmezei.spamnot@videotron.ca> writes:H >From the coverage of the florida elections, I got quick glances at what lookedK >like the old punched cards, and there was mention of voters punching holesm to >make their selections.i  F In my district we have punch cards and we use a sort of little nail toI knock out one of the prepunched holes.  The card is a little lighter than I the punch cards I recall from the days of old. The problem is that if youtG don't knock the little tab all the way out it may read either closed orrG open.  Hence the "hand count" they are doing in many places in Florida..G Seems people can reliably tell if the tab is in or out but the machines-J cannot.  On the other hand, it's hard to believe that they don't make someB other sort of error to offset the gain in accuracy - for instance,7 accidentally scoring the same card twice or not at all.>  J >What sort of technology is actually used ? And is this used nationwide or just >in Florida.  C It varies from place to place.  And thanks to the electoral college K nonsense votes in different states have slightly different values.  To makegJ it even more ridiculous in this election the difference in Florida betweenF the candidates' totals are well within the error levels of the variousJ technologies and the voters themselves.  (Ie, folks pushing the wrong holeG by accident.)   The founding fathers got a lot of things right but theytF wrote the constitution before statistics were well understood - and it shows.  J I think the best solution would be to utilize another 18th century problemC solving methodology - the two candidates could engage in a good oldvK fashioned duel.  The winner gets to be president.  And with any luck they'dl both be crack shots!   David Mathog mathog@seqaxp.bio.caltech.edu > Manager, sequence analysis facility, biology division, Caltech   ------------------------------  % Date: Fri, 10 Nov 2000 13:55:50 -0600o/ From: Chris Scheers <chris@applied-synergy.com>y* Subject: Re: VAXstation 4000-90 Error CodeO Message-ID: <89D1F0EF52388285.1516F121DB1C1A44.D39DF0CD66D8E30C@lp.airnews.net>/   Robert Deininger wrote:s > j > In article <8udvpc$1qh$1@porthos.nl.uu.net>, "Maarten van Breemen" <Maarten.van.Breemen@notox.nl> wrote: > ; > > VS400-60 memory is not compatible with VS400-90 memory.a > >I5 > > Model 60 can take MS44-DA (2x16) or MS44-BA (2x4)25 > > Model 90 can take MS44-DC (4x16) or MS44-BC (4x4)v >  > I don't know exactly what these numbers refer to.  Maybe they are order numbers for _kits_?  I have moved SIMMS between the two systems without problems.  Clearpoint, who make memory for these systems, claim they are the same SIMMS. > W > The only difference is that the -60 takes 2 at a time, and the -90 takes 4 at a time.d    H I have not personally encountered this, but (supposedly) early VS4000-60; memory used slower chips and may not work in the VS4000-90.r  H Later VS4000-60 memory seems to interchange without problems.  (No, I do/ not know what constitutes "early" and "later".)e  8 AFAIK, VS4000-90 memory will always work in a VS4000-60.  G -----------------------------------------------------------------------l$ Chris Scheers, Applied Synergy, Inc.  C Voice: 817-237-3360            Internet: chris@applied-synergy.com g   Fax: 817-237-3074    ------------------------------  % Date: Fri, 10 Nov 2000 16:09:50 -0500n2 From: rdeininger@mindspring.com (Robert Deininger)* Subject: Re: VAXstation 4000-90 Error CodeL Message-ID: <rdeininger-1011001609500001@user-2ive7o5.dialup.mindspring.com>  k In article <35IO5.7920$Qz2.323684@typhoon.aracnet.com>, "Zane H. Healy" <healyzh@shell1.aracnet.com> wrote:n  J > > The SCSI ID of the controller can be changed at the console. IIRC the G > > default was 6, but it can be anything.  Do you have an ID conflict?  > L > If I did, then I'd be seeing wierdness when I do the SHOW DEVICE, wouldn't > I?  F I would think so.  It will tell you the ID of the controller, however.  H The other post that suggested checking the boot flags sounds good to me.? Looking in the wrong system root would certainly cause trouble.    -- h Robert Deininger rdeininger@mindspring.com    ------------------------------  # Date: Sat, 11 Nov 2000 02:51:59 GMTy2 From: "Zane H. Healy" <healyzh@shell1.aracnet.com>* Subject: Re: VAXstation 4000-90 Error Code7 Message-ID: <jv2P5.7988$Qz2.337718@typhoon.aracnet.com>d  ( John E. Malmberg <wb8tyw@qsl.net> wrote:G > Please post the exact Version of OpenVMS that you have on the CD-ROM.   F > Some models VAX and VAXstations require an -Hn kit (Hardware supportH > release) as they came out after the original operating system release.  I Well, it's V7.1, so I'm hoping that's not the problem :^)  The Hard DriveA6 has the same version on it (fresh install on the -60).   			Zane    ------------------------------  % Date: Fri, 10 Nov 2000 23:46:17 -0600.) From: "John E. Malmberg" <wb8tyw@qsl.net>v* Subject: Re: VAXstation 4000-90 Error Code/ Message-ID: <t0pn5t5ktokvb1@corp.supernews.com>   L Since you can verify that the CDROM and the hardrive work on another system,I I will also make the assumption that you have tried them one at a time on/L the bus, and that you have checked the cable connections and the terminator.  J If you can boot the system from the network, that will get you some better) diagnostics as has been sugested already.   J I would also inspect the internal SCSI cable and the connectors carefully.K It is possible that it is damaged.  Or you could have a bad SCSI controllert on your system.e  I If you can use it as an LAVC satellite though, it still can be useful fore you.   -Johnu   ------------------------------   Date: 11 Nov 2000 06:49:23 GMT- From: djweath@attglobal.net (Dave Weatherall)w" Subject: Re: VMS, BSD, and FORTRAN5 Message-ID: <DTiotGxQ0bj6-pn2-dnMeKTOfQnIR@localhost>n  7 On Tue, 7 Nov 2000 02:28:23, "Craig D. Lansing (root)" e <lansing@beldar.com> wrote:b   > ...  Is anyoneH > familiar with whether or not sockets will work on a VMS platform?  CanD > this be written in FORTRAN?  I doubt that we can get the vendor toH > change his mind, but I am mostly just curious about whether or not the > capability exists.  D It can be done in FORTRAN. We've done it,  both with QIO and Socket C calls. Pointers to Pointers can be a pain. C does that better than tF FORTRAN of course. In FORTRAN you have a little more overhead managingF POINTERS as a result but once you've got your library in place its OK.E Looking back at it, IIRC I still prefer the QIO method as it is more EE suitable for asynchronous activity. And yes you can have a QIO based oF TCP/IP on VMS talk to Socket based TCP/IP (We've done it on Tru64 and 
 Solaris).    Cheers - Dave.   ------------------------------  % Date: Fri, 10 Nov 2000 14:01:42 -0600 / From: Chris Scheers <chris@applied-synergy.com> # Subject: Re: VS4000-60 Parts NeededaO Message-ID: <DA0CA6974BE2F616.85EADB6BEB5916B3.CC678C85AB29AADD@lp.airnews.net>n   Dan O'Reilly wrote:r > I > I have a 4000-60 that needs the bank of switches that contains the halt G > button, the headphone connector, etc., that lives under the flip-downeF > panel at the right front of the system.  It appears to connect via aG > 50-pin cable to the motherboard.  I think I can make this system workeH > if somebody has parts from a dead system they would be willing to part > with.     D FWIW: The part you are looking for is actually a board.  It is aboutH 6"x6" with switches and lights at one end, a connector at the other, and; a couple of chips, components, and a speaker in the middle.t  B It is labelled "PV2 LIGHTS + SWITCHES".  It seems to have the partC number 54-20367, but I don't find a match for that in Businesslink.d   (No, you can't have mine!)  
 Good luck!  G -----------------------------------------------------------------------f$ Chris Scheers, Applied Synergy, Inc.  C Voice: 817-237-3360            Internet: chris@applied-synergy.com c   Fax: 817-237-3074    ------------------------------    Date: 10 Nov 2000 14:50:15 -05001 From: kaplow_r@eisner.decus.org.mars (Bob Kaplow)c5 Subject: Re: We Interrupt the US Electoral Follies...v+ Message-ID: <rAvMD+8rjyks@eisner.decus.org>l  C Speaking of electoral follies, today is the ballot deadline for thelI Encompass bylaws and ammendments. And just to cause havoc, it's a federaleL mail holiday! I assume someone will be posting results in less time and withJ less suspicion, threats of lawsuits, and demands for another election than we're now seeing in Florida.   ------------------------------  % Date: Fri, 10 Nov 2000 15:15:53 -0500 0 From: David Beatty <David.Beatty.NOSPAM@sas.com>5 Subject: Re: We Interrupt the US Electoral Follies...s2 Message-ID: <OlcMOtLZpDV1N908CN7nWMp0P2EY@4ax.com>  D Actually, the Post Offices should be open today and closed tomorrow.? The administrative offices are closed today (as are all Federal*5 offices), since they are normally closed on Saturday.   B On 10 Nov 2000 14:50:15 -0500, kaplow_r@eisner.decus.org.mars (Bob Kaplow) wrote:  D >Speaking of electoral follies, today is the ballot deadline for theJ >Encompass bylaws and ammendments. And just to cause havoc, it's a federalM >mail holiday! I assume someone will be posting results in less time and withEK >less suspicion, threats of lawsuits, and demands for another election thann >we're now seeing in Florida.    ------------------------------  # Date: Fri, 10 Nov 2000 21:30:41 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)r5 Subject: Re: We Interrupt the US Electoral Follies... 0 Message-ID: <009F2EBC.07B80D21@SendSpamHere.ORG>  _ In article <rAvMD+8rjyks@eisner.decus.org>, kaplow_r@eisner.decus.org.mars (Bob Kaplow) writes:gD >Speaking of electoral follies, today is the ballot deadline for theJ >Encompass bylaws and ammendments. And just to cause havoc, it's a federalM >mail holiday! I assume someone will be posting results in less time and withe  ^^^^^^^^^^^^ ????  H I had mail delivered today.  Perhaps somebody forgot to tell my carrier.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMr            iO city, n., 1. a place where trees are cut down and streets are named after them.0   ------------------------------  % Date: Fri, 10 Nov 2000 16:55:30 -0600v1 From: "Dave Gudewicz" <david.gudewicz@abbott.com> 5 Subject: Re: We Interrupt the US Electoral Follies...a8 Message-ID: <8uhu78$6uv$1@fizban.fizban.pprd.abbott.com>  K I happened to go home for lunch today.  While I was there, the mailman camenK by and delivered the mail.  And I'm a veteran!  Wonder if that matters? ;-)u   Dave...-  > "Bob Kaplow" <kaplow_r@eisner.decus.org.mars> wrote in message% news:rAvMD+8rjyks@eisner.decus.org...oE > Speaking of electoral follies, today is the ballot deadline for theoK > Encompass bylaws and ammendments. And just to cause havoc, it's a federal1I > mail holiday! I assume someone will be posting results in less time and  withL > less suspicion, threats of lawsuits, and demands for another election than > we're now seeing in Florida.   ------------------------------  % Date: Fri, 10 Nov 2000 19:08:49 -0500   From: John Santos <JOHN@egh.com>, Subject: Re: Where's the Y2K compliance page5 Message-ID: <1001110184634.2092D-100000@Ives.egh.com>p  & On Wed, 8 Nov 2000, John Santos wrote:  9 > In article <8ucdb1$80m$1@mailint03.im.hou.compaq.com>, u' > hoffman@xdelta.zko.dec.nospam says...r > > ^ > > In article <0$mvbei+BgnI@eisner.decus.org>, koehler@eisner.decus.org (Bob Koehler) writes:N > > :  The Feds required year 2000 day 366 testing as a part of Y2K testing.  D > > :  Why retire info related to something that hasn't happend yet? > > M > >   Last I looked, 29-Feb-2000 did already happen.  Of course, 2038 hasn't wL > >   happened yet, and that may well be as much (or more) "fun" as was Y2K. > H > Yes, but Year 2000 day 365 and Year 2000 day 366 haven't happened yet. > G > (I can't imagine why this year would be different than any other leap C > year, but one patch release of Java on Sun got 29-Feb-2000 wrong, G > so anything is possible.  I think it was in 1.8, the previous releasep   Whoops:  1.1.8, not 1.8...  K > worked okay, and by the time we discovered the problem, there was anotherg& > version out that worked okay again.)  C But I did a little more checking.  It was okay in 1.1.6, apparentlywC okay in the initial releases of 1.1.8, got broken simultaneously in8A 1.1.8H and in 1.2 (released at the same time, I think), was still A broken in 1.2.1, but got fixed in 1.2.2 and in a later release of A 1.1.8.  This info is from an e-mail from a co-worker who did sometE testing and verification on his own and also got some info from Sun'sSJ web site.  As far as I know, this is fixed in all current Java variants.    F The point here is not to trash Sun about it, it's just that we aren't $ completely out of the Y2K woods yet.   -- = John Santos= Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------   End of INFO-VAX 2000.630 ************************