1 INFO-VAX	Sat, 11 Nov 2000	Volume 2000 : Issue 631       Contents:F Re: %EDT-F-INTERERR, Internal software error; contact a Compaq support! Re: *.*;-0: undocumented feature? ! Re: *.*;-0: undocumented feature? , ? Tape Device owned by dead process on Alpha0 Re: ? Tape Device owned by dead process on Alpha! Re: Alpha Hardware Experts...Help  Re: Cluster Problem  Re: COBOL F$EDIT Re: COBOL F$EDIT Re: COBOL F$EDIT Re: Compiler /tune switches. Re: DecWindows1 Re: Disk write-back cache - how do I turn it off? * Re: Enabling DECwindows on a VAX 4000/300. Re: GAWK for VMS? , Re: login prompts inhibited with logins=0 ??, Re: login prompts inhibited with logins=0 ?? Malloc + debug question * Re: No doubts here (Was: Re: Galaxy doubt) SPEC_int92? SPECrate_fp95? etc.  Re: sqlpre70 get sqlbugchk.dmp
 Re: Time Zone 
 Re: Time Zone   F ----------------------------------------------------------------------  % Date: Sat, 11 Nov 2000 13:49:49 -0500 2 From: "Richard B. Gilbert" <DRAGON@compuserve.com>O Subject: Re: %EDT-F-INTERERR, Internal software error; contact a Compaq support 7 Message-ID: <200011111350_MC2-BA75-2C2E@compuserve.com>   G         I believe that SOS was never native VMS code;  I believe it ran D under the RSX emulator.  It was my first VMS editor.  (When I had an el-cheapo brand-X terminal).  =     " Message text written by Alan GreigJ >>Someone once told me that SOS originally stood for Son Of Standalone. I= s  that	 >correct?   E Son Of Stopgap it said in the sources. There used to be a VMS version F of SOS. Wonder if that still exists anywhere for possible inclusion onC the Freeware CD. There are times on a dumb terminal or with poor VT A emulation where I know I could do something quicker with SOS than < trying to use EDT in line mode. Suppose there's always teco.  D The biggest annoyance with SOS/EDIT was that the editors were almostD identical (common source which then diverged) but that many commandsD had changed letters For example on TOPS-10 to save a program withoutD line numbers and then invoke the last compile class command you gaveE the command *gsb (go strip nobackup). On TOPS-20 the same command was 1 *gun (go unsequence no backup). Just plain daft.<    ------------------------------  % Date: Sat, 11 Nov 2000 08:02:05 +0100 2 From: martin@radiogaga.harz.de (Martin Vorlaender)* Subject: Re: *.*;-0: undocumented feature?; Message-ID: <3a0ceeed.524144494f47414741@radiogaga.harz.de>   $ Jim Agnew (agnew@hsc.vcu.edu) wrote:I : there was a thread for a while a while back about tryign to detect file H : version numbers "too close" to the 32K limit.. Anyone keep the command : file that found those???  6 As any system manager worth his money should have DFU:  , $ MCR DFU SEARCH device /VERSION=MINIMUM=nnn  % And it's fast, too (uses INDEXF.SYS).    cu,    Martin --J One OS to rule them all       | Martin Vorlaender  |  VMS & WNT programmer7 One OS to find them           | work: mv@pdv-systeme.de N 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: Sat, 11 Nov 2000 09:54:16 -0500 , From: Howard S Shubs <hshubs@mindspring.com>* Subject: Re: *.*;-0: undocumented feature?> Message-ID: <hshubs-869E8B.09541611112000@news.mindspring.com>  < In article <3a0ceeed.524144494f47414741@radiogaga.harz.de>,  martin@radiogaga.harz.de wrote:   7 >As any system manager worth his money should have DFU:   % Wow, is -that- ever a rash statement!  --   Howard S ShubsD "Run in circles, scream and shout!"  "I hope you have good backups!"   ------------------------------  % Date: Sat, 11 Nov 2000 02:01:40 -0500 ( From: Robert Boyd <rlboyd@bellsouth.net>5 Subject: ? Tape Device owned by dead process on Alpha - Message-ID: <3A0CEEC4.5083F990@bellsouth.net>   N There is a tool available to help clear things up when a tape drive is left inP an allocated state when a process terminates without getting a proper rundown to! clear the UCB/etc on OpenVMS VAX.   P I'm wondering if anyone knows of such a tool for doing the same thing on Alpha? P We have a system running VMS V7.1 that occasionally winds up with the tape driveI allocated to a non-existant process.  Any pointers would be appreciated.  L Supposedly we shouldn't be having this problem any more, but .... it happens anyway.    Bob Boyd rlb14162@glaxowellcome.com   ------------------------------  # Date: Sat, 11 Nov 2000 12:29:00 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) 9 Subject: Re: ? Tape Device owned by dead process on Alpha 0 Message-ID: <009F2F39.861D4DCF@SendSpamHere.ORG>  X In article <3A0CEEC4.5083F990@bellsouth.net>, Robert Boyd <rlboyd@bellsouth.net> writes:O >There is a tool available to help clear things up when a tape drive is left in Q >an allocated state when a process terminates without getting a proper rundown to " >clear the UCB/etc on OpenVMS VAX. > Q >I'm wondering if anyone knows of such a tool for doing the same thing on Alpha?  Q >We have a system running VMS V7.1 that occasionally winds up with the tape drive J >allocated to a non-existant process.  Any pointers would be appreciated. M >Supposedly we shouldn't be having this problem any more, but .... it happens 	 >anyway.    I Watch out!  This dead process might be elected posthumously to the senate  in Missouri.  ;)     > 	 >Bob Boyd  >rlb14162@glaxowellcome.com   ) I think you might like to take a look at:   8 http://www2.wku.edu/scripts/fileserv/vileserv.com?CLRREF     --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM              O city, n., 1. a place where trees are cut down and streets are named after them.    ------------------------------  % Date: Sat, 11 Nov 2000 13:49:51 -0500 2 From: "Richard B. Gilbert" <DRAGON@compuserve.com>* Subject: Re: Alpha Hardware Experts...Help7 Message-ID: <200011111350_MC2-BA75-2C2F@compuserve.com>   J         ISTR from another message in this thread, that you were looking f= orJ 16Gb of disk space.  One 18Gb SCSI disk meets that requirement.  These ar= e J readily available and not even terribly expensive.  You could put four of=  J them in the DS20 if you were so minded.   Need a little more?   I believe=  ) that 36 or 50 Gb disks are available now.   B         Please specify your requirements a little more completely.   Message text written by fooguyJ >Sadly, I can't make Compaq put more than four drive bays in a DS20, or 1=  
 in the DS10L.  <    ------------------------------  # Date: Sat, 11 Nov 2000 07:07:23 GMT 1 From: CSABA  HARANGOZO   <csabah@zipworld.com.au>  Subject: Re: Cluster Problem8 Message-ID: <Le6P5.143$aQ6.22005@nostril.pacific.net.au>  + Phillip Helbig <helbig@astro.rug.nl> wrote: N >> Also, a device specifically set unavailable with the SET DEVICE/NOAVAILABLE >> dev.    > OK.   J >> > > I'll have to admit that I'm not really sure what causes a device toM >> > > fail the "avl" test.  I couldn't find any devices which exist but fail J >> > > "avl" on the systems I tried.  I did find that if a host leaves the >> >  J >> >         A device mounted privately, maybe ? It exists, but may not beB >> >         available. Tape drives in use could be good examples.+ >> >         I haven't checked this though.    > Nope.   > 	You are right, Phillip. After I posted my message, I realised> 	that here we are talking about availability for the _system_,C 	and not for a user. It would be useful to be around a VMS machine,  	but alas...     > $ wso f$getjpi("","PID")
 > 20200FC6 > $ sh dev disk$floppy  R > Device                  Device           Error    Volume         Free  Trans MntR >  Name                   Status           Count     Label        Blocks Count Cnt0 > GLADIA$DVA0:            Online               0 > $ allocate disk$floppy' > %DCL-I-ALLOC, _GLADIA$DVA0: allocated  > $ sh dev disk$floppy  R > Device                  Device           Error    Volume         Free  Trans MntR >  Name                   Status           Count     Label        Blocks Count Cnt0 > GLADIA$DVA0:            Online alloc         0% > $ wso f$getdvi("disk$floppy","avl")  > TRUE   > in another DECterm:    > $ $ wso f$getjpi("","PID")
 > 202010C4' > $ $ wso f$getdvi("disk$floppy","avl")  > TRUE > $   H > If ALLOCATE does not make it unavailable to another process, then this > goes against my intuition.    C 	I think a status of "Unavailable" or "HostUnavailable" and similar > 	cases when a system can not access a device, then the lexical> 	returns a FALSE. And of course, as it was written above, with9 	SET DEV/[NO]AVAILABLE device   ( needs dismount first ).    			Cheers,		Csaba   I    ---------------------------------------------------------------------- E    * Csaba I. Harangozo     |    'To err is human', said the hedgehog E    * csabah@zipworld.com.au |           as he dismounted a wirebrush. I    ---------------------------------------------------------------------- ;    EARTH::AUSTRALIA:[SYDNEY]HARANGOZO.CSABA;1, delete? [N]:    ------------------------------  # Date: Sat, 11 Nov 2000 12:22:47 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)  Subject: Re: COBOL F$EDIT 0 Message-ID: <009F2F38.A7A5294E@SendSpamHere.ORG>   In article <1137A4A23A51D311B2D600105A1D5213026FDFDC@seantexch.unitedad.com>, Terry Marosites <TMarosites@unitedad.com> writes:  >Hi all, > M >   I want to an f$edit from COBOL if I call lib$do_command I don't come back  >to my program  M LIB$DO_COMMAND will essentially queue the command and it will not be executed + until the image had rundown and terminated.   J > if I do a lib$spawn my symbol is set in the spawned process. does anyone3 >know an easy method of doing a f$edit from COBOL.    K What do you want to do?  Do you want to manipulate a DCL symbol within your B program or do you just want F$edit-like functions in your program?  K There are STR$ RTLs which can do, quite easily too, some of the things that L the F$edit will/can do:  STR$TRIM, STR$UPCASE; or with combinations of these. routines you can accomplish similar results.    L David Dachtera suggests using the BASIC run-time support for the EDIT$ oper-M ator (VAX: BAS$EDIT, Alpha: DBASIC$EDIT).  The format, at least for the Alpha L version, is DBASIC$EDIT(dest-string by descriptor, src-string by descriptor,L and an edit mask by value).  You'd need to explore the BASIC EDIT$ operator K to learn the significance of the various mask bits.  This would likely work L but I'm sure would be pooh-poohed by a number of folks hare as an unsupport-
 ed hack.    K I'd suggest you look into the STR$ routines.  I don't do COBOL so I'm gues- K sing that it does not have very good string manipulation features itself if $ you are asking for access to F$edit.     --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM              O city, n., 1. a place where trees are cut down and streets are named after them.    ------------------------------  % Date: Sat, 11 Nov 2000 08:03:50 -0500 2 From: norm lastovica <norman.lastovica@oracle.com> Subject: Re: COBOL F$EDIT * Message-ID: <3A0D43B6.C9A4CFE0@oracle.com>  > it is no problem at all to call bas$edit from other languages.< the first parameter is the character string to be edited (it> will be changed in place; if any of the edit "functions" that ? you request could change the length of the string, I'd suggest  > using a dynamic descriptor) and the second parameter is a bit = mask of the functions to perform.  according to online help,   the valid functions are:    1%    Trim parity bits "  2%    Discard all spaces and tabs<  4%    Discard characters: CR, LF, FF, ESC, RUBOUT, and NULL&  8%    Discard leading spaces and tabs*  16%   Reduce spaces and tabs to one space%  32%   Convert lowercase to uppercase    64%   Convert [ to ( and ] to )'  128%  Discard trailing spaces and tabs ,  256%  Do not alter characters inside quotes  & and the bit definitions would then be:    trim_parity_bit = 0  disc_sp_tab_bit = 1  disc_junk_bit   = 2  disc_lead_bit   = 3  compact_bit     = 4  upper_case_bit  = 5  parens_bit      = 6  disc_trail_bit  = 7  ignor_quote_bit = 8     "David J. Dachtera" wrote: >  > Terry Marosites wrote: > >  > > Hi all,  > > P > >    I want to an f$edit from COBOL if I call lib$do_command I don't come back > > to my program M > >  if I do a lib$spawn my symbol is set in the spawned process. does anyone 5 > > know an easy method of doing a f$edit from COBOL.  > B > I have memories from long ago of being able to call the BAS$EDITD > function from BASIC run-time library from a non-BASIC 3GL. PerhapsG > someone else will chime in with the specifics. I think I read it here  > first a while back.  > E > I don't have BASIC installed on my hobbyist Alpha just yet, and the D > MicroVAX 3100 does not run just now. Otherwise I could compile the9 > following /MACHINE_CODE/LIST and post the machine code:  >  >         MAP (SAMPLE) & >         STRING >         TEST_STRING     = 32%  > ( >         A$ = EDIT$( TEST_STRING, 16% ) >  >         END PROGRAM  > G > A MACRO coder could then divine the calling sequence to make BAS$EDIT J > work in another language. Basically (no pun intended), you pass a stringE > by descriptor, a word integer by reference (I believe), and another I > string by descriptor to receive the output of the function. The integer D > is a bitmap of functions to perform: trim leading/trailing spaces,H > upcase/locase/compress/collapse the string, string control characters,E > etc. Trouble is, I don't have the doc.'s or a compiler available. I G > don't know the sequence of arguments or the bits for the functions by 	 > memory.  > E > Now that I think about it, I think I saw that back in the 80's in a G > DIBOL program. I may have the source on an old TK50, if I can spin up % > the old MicroVAX (it has a TZ30)...  >  > -- > David J. Dachtera  > dba DJE Systems  > http://www.djesys.com/ > < > Unofficial Affordable OpenVMS Home Page and Message Board:! > http://www.djesys.com/vms/soho/  > H > This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings > is to be expected. > B > Feel free to exercise your rights of free speech and expression. > H > However, attacks against individual posters, or groups of posters, are > strongly discouraged.    --  > norman lastovica / oracle rdb engineering / usa / 610.696.4685   ------------------------------  % Date: Sat, 11 Nov 2000 08:28:39 -0500 2 From: norm lastovica <norman.lastovica@oracle.com> Subject: Re: COBOL F$EDIT * Message-ID: <3A0D4987.11E22167@oracle.com>  5 whoops - I lied - there are 3 parameters to bas$edit:   ? 	destination_string_descriptor - string descriptor by reference : 	source_string_descriptor - string descriptor by reference 	edit_mask - word by value  6 So only the destination descriptor needs to be dynamic9 if it has a different length than the source.  Of course, 6 the same string descriptor can be used for both source and destination.  3 Here's an example of how a call might look from C):   ) 	BAS$EDIT (&dststr, &srcstr, 16 + 32 + 4)   = All of this stuff *used* to be documented back in VMS version < 3 in the (as I recall) "language support routines" RTL book.@ I regret the day just this past year when mine hit the dumpster.   	fwiw    norm lastovica wrote:  > @ > it is no problem at all to call bas$edit from other languages.> > the first parameter is the character string to be edited (it? > will be changed in place; if any of the edit "functions" that @ > you request could change the length of the string, I'd suggest? > using a dynamic descriptor) and the second parameter is a bit > > mask of the functions to perform.  according to online help, > the valid functions are: >  >  1%    Trim parity bits $ >  2%    Discard all spaces and tabs> >  4%    Discard characters: CR, LF, FF, ESC, RUBOUT, and NULL( >  8%    Discard leading spaces and tabs, >  16%   Reduce spaces and tabs to one space' >  32%   Convert lowercase to uppercase " >  64%   Convert [ to ( and ] to )) >  128%  Discard trailing spaces and tabs . >  256%  Do not alter characters inside quotes > ( > and the bit definitions would then be: >  >  trim_parity_bit = 0 >  disc_sp_tab_bit = 1 >  disc_junk_bit   = 2 >  disc_lead_bit   = 3 >  compact_bit     = 4 >  upper_case_bit  = 5 >  parens_bit      = 6 >  disc_trail_bit  = 7 >  ignor_quote_bit = 8 >  > "David J. Dachtera" wrote: > >  > > Terry Marosites wrote: > > >a
 > > > Hi all,  > > > R > > >    I want to an f$edit from COBOL if I call lib$do_command I don't come back > > > to my programeO > > >  if I do a lib$spawn my symbol is set in the spawned process. does anyonep7 > > > know an easy method of doing a f$edit from COBOL.  > > D > > I have memories from long ago of being able to call the BAS$EDITF > > function from BASIC run-time library from a non-BASIC 3GL. PerhapsI > > someone else will chime in with the specifics. I think I read it hereu > > first a while back.  > >)G > > I don't have BASIC installed on my hobbyist Alpha just yet, and theiF > > MicroVAX 3100 does not run just now. Otherwise I could compile the; > > following /MACHINE_CODE/LIST and post the machine code:  > >  > >         MAP (SAMPLE) & > >         STRING! > >         TEST_STRING     = 32%  > >n* > >         A$ = EDIT$( TEST_STRING, 16% ) > >s > >         END PROGRAM- > >2I > > A MACRO coder could then divine the calling sequence to make BAS$EDITeL > > work in another language. Basically (no pun intended), you pass a stringG > > by descriptor, a word integer by reference (I believe), and anotherbK > > string by descriptor to receive the output of the function. The integernF > > is a bitmap of functions to perform: trim leading/trailing spaces,J > > upcase/locase/compress/collapse the string, string control characters,G > > etc. Trouble is, I don't have the doc.'s or a compiler available. IbI > > don't know the sequence of arguments or the bits for the functions byc > > memory.t > > G > > Now that I think about it, I think I saw that back in the 80's in asI > > DIBOL program. I may have the source on an old TK50, if I can spin upo' > > the old MicroVAX (it has a TZ30)...n > >m > > -- > > David J. Dachterax > > dba DJE Systemsa > > http://www.djesys.com/ > >s> > > Unofficial Affordable OpenVMS Home Page and Message Board:# > > http://www.djesys.com/vms/soho/P > >tJ > > This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings > > is to be expected. > >-D > > Feel free to exercise your rights of free speech and expression. > >(J > > However, attacks against individual posters, or groups of posters, are > > strongly discouraged.1 >  > --@ > norman lastovica / oracle rdb engineering / usa / 610.696.4685   -- f> norman lastovica / oracle rdb engineering / usa / 610.696.4685   ------------------------------  % Date: Sat, 11 Nov 2000 08:15:25 -0500 2 From: norm lastovica <norman.lastovica@oracle.com>% Subject: Re: Compiler /tune switches.H* Message-ID: <3A0D466C.D6A588D4@oracle.com>   Dan Sugalski wrote:) > 1 > At 08:08 PM 11/10/00 +0000, Hoff Hoffman wrote: A > >   The /OPTIMIZE=TUNE should generally be left set to GENERIC.  > M > Is there much benefit to doing otherwise? The help info's rather... sparse.wJ > :) If a /TUNE=EV6 gets me a boost of a few percent I'm all for the extra > compile time.e  4 	It depends very much on the application and what it> does.  If you know for certain that you're going to be running> on an EV6 system and you've got the very latest compilers (and@ thus the latest GEM backend), you might as well test and compare$ with and without /TUN=EV6+/ARCH=EV6.8 	It appears unlikely that compiling this way and running+ on an EV6 would make the performance worse.    > - >                                         DanS > K > --------------------------------------"it's like this"--------------------4 > Dan Sugalski                          even samuraiA > dan@sidhe.org                         have teddy bears and even:= >                                       teddy bears get drunk0   -- 0> norman lastovica / oracle rdb engineering / usa / 610.696.4685   ------------------------------  % Date: Sat, 11 Nov 2000 09:23:42 +0100 2 From: martin@radiogaga.harz.de (Martin Vorlaender) Subject: Re: DecWindowsV; Message-ID: <3a0d020e.524144494f47414741@radiogaga.harz.de>g   ryanfleming@my-deja.com wrote:= : I have two Alpha 800s running OS V7.1-2 and UCX V4.2 ECO 3. > : I am able to use exceed to launch a remote decwindow on a PCF : I AM NOT able to launch a remote decwindow from one PC to the other.  ? Why would you want to launch an app from one eXceed to another?e - Sorry, couldn't resist.-  H : I have gone into the style manager under security and have two entries : on each alpha: :     TCPIP x.x.x.x *- :     TCPIP * *?/ : Where x.x.x.x is the other Alphas IP address.s  + The second entry should probably be enough.   < : I also added the SYS$MANAGER:DECW$PRIVATE_SERVER_SETUP.COM5 : I made sure the TCPIP was mentioned as a transport.  :pF : I attempt to launch the window by connecting to Alpha B using telnet6 : (from Alpha A's decwindow) and typing the following:8 :      $ SET DISPLAY/CREATE/NODE=x.x.x.x/TRANSPORT=TCPIP :      $ CREATE/TERMINAL/DETACHo : I get the following reult:2 :      %DECW-E-CANT_OPEN_DISPL, Can't open displayH : The above two commands works just fine for exceed.  Would anyone be so+ : kind as to show me what I am doing wrong?o  C What do the SYS$MANAGER:DECW$SERVER_0_ERROR.LOGs say? Another idea: G Try running DECW$EXAMPLES:ICO.EXE. Its error messages are usually a bitl more detailed.   cu,s   Martin --J One OS to rule them all       | Martin Vorlaender  |  VMS & WNT programmer7 One OS to find them           | work: mv@pdv-systeme.devN 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: Sat, 11 Nov 2000 08:09:28 -0500-2 From: norm lastovica <norman.lastovica@oracle.com>: Subject: Re: Disk write-back cache - how do I turn it off?* Message-ID: <3A0D4508.9CA21041@oracle.com>  : and to further muddy the water, another very large effect 8 of mounting a disk with an explicit /NOCACHE is that it 9 also disables VIOC (vms's read cache).  while it has beens; a while since there were problems with it, it could consumen- CPU time that might be used for other things.o   	fwiw        Michael Shield wrote:s >  > Ok,  > M > Having been through all this with  big O, it is clear that the situation is H > still not clearly understood,  so here goes, make of it what you will. > H > There are three separate areas which Oracle get messed up, which isn'tL > helped by the way VMS displays its drive info. Starting at the disk level. > K > The standard modern scsi drive contains an amount of internal memory thatlL > can be used to cache data as its transfered to the drive itself. It CAN beN > turned on (default is off I believe) by clever programs (undocumented optionM > in the SCU utility in TRU64) but there is no way to do so using VMS without N > writing a program yourself. Thus this is not a factor in VMS/Oracle systems. > L > The only area where WBC is in use is if the option has been enabled on theF > specific drive on a hsk/z/g controller. Certain cache policies mixedE > together with cache battery failures could in the past have lead toxL > problems, but the main difficulty is getting Compaq and Oracle to come outL > and say for definate that a particular solution (hsz70 with mirrored cache > for instance) was supported. > K > The one that gets them every time is the mount option. Every disk, exceptyL > the system disk, is mounted by default with an option that makes it appearH > to have WBC enabled. This usually creates a wild goose chase trying toL > figure out which of the previous two scenarios it is reffering to, when inA > fact it is only ever used by applications that make use of XQP+ , > funnctionallity (pathworks 5 and onwards). > L > Thus mount disks writethrough at the operating system makes not one jot ofD > difference, and after having done so, your oracle bug will remain. > ; > And in answer to your question, connect to the HSZ and goN >  > Set DXXX nowriteback > H > I live in hope of the day when enough VMS people get to talk to Oracle4 > support and explain this to them once and for all. > 
 > Mike Shield  > / > At home, on holiday, not speaking for Compaq.)3 > <Steve.Spires@yellowpages.co.uk> wrote in message + > news:00256992.0061BBD2.00@quegw01.btyp...  > > cc:l > > bcc:J > > Contact:   Tel: 3063  -  IS - Infrastructure, 1st Floor, Bridge Street > Plazav > >l1 > > Disk write-back cache - how do I turn it off?  > >m > >n > > Hi all,e > > L > > We are having some problems here with file corruptions related to Oracle > files.K > > Oracle have suggested that we turn off write-back caching on the disks.  > >-I > > I have turned off write-back cache on the controllers (HSZ70s) but amC > unsure asgM > > to how I mount the disk - or set the volume - so that write-back cache is  > not 2 > > enabled as shown if you do a SHOW DEVICE/FULL; > >-N > > Volume Status:  subject to mount verification, write-back caching enabled. > >AJ > > I had a look throught the MOUNT and SET VOLUME help, but couldn't find
 > anythingM > > - not to say it isn't there somewhere - so if someone could help, I would  > > appreciate it. > > L > > Does the above output segment from SHOW DEVICE mean just that write-back > cache0M > > is enabled, or does it mean that it is both enabled and in use? Or have Id > gotrJ > > the wrong end of the stick because I'm now too tired to grasp things I > should > > know...? > >  > > Steve Spires > > VMS System Manager > > BT/Yellow Pages  > >t > >   > > [Information] -- PostMaster:H > > This transmission is intended solely for the addressee(s) and may beK > > confidential. If you are not the named addressee, or if the message has  > beenF > > addressed to you in error, you must not read, disclose, reproduce, > distribute or  > > use this transmission. > >uL > > Delivery of this message to any person other than the named addressee is > notoL > > intended in any way to waive confidentiality.  If you have received thisJ > > transmission in error please contact the sender or delete the message. > >e > > Thank you. > >  > >  > >i   -- o> norman lastovica / oracle rdb engineering / usa / 610.696.4685   ------------------------------   Date: 11 Nov 2000 15:04:14 GMT& From: Cthulhu <cthulhu@kadath.deep.it>3 Subject: Re: Enabling DECwindows on a VAX 4000/300.e( Message-ID: <8ujn5e$q7$1@kadath.deep.it>  + Peter LANGSTOEGER <eplan@kapsch.net> wrote:i   > $ SET DEFAULT DECW$EXAMPLES: > $ RUN DOGS   Doh!   I feel so stupid...r   	retruingly, 	   Cthulhu  -- '  G        Ph'nglui mglw'nafh Cthulhu http://www.rlyeh.it wgah'nagl fhtgan! # 		        <cthulhu at rlyeh dot it>s   ------------------------------  % Date: Sat, 11 Nov 2000 18:44:06 +0010.% From: paddy.o'brien@zzz.tg.nsw.gov.aut Subject: Re: GAWK for VMS?5 Message-ID: <01JWF9XJLRUQ006DNN@tgmail.tg.nsw.gov.au>a  M Try re-compilation of the source.  Not tried, but should work with the error - provided below.    Regards, Paddy   Paddy O'Brien, Transmission Development,o
 TransGrid, PO Box A1000, Sydney South,  NSW 2000, Australia-   Tel:   +61 2 9284-3063 Fax:   +61 2 9284-3050& Email: paddy.o'brien@zzz.tg.nsw.gov.au  M Either "\'" or "\s" (to escape the apostrophe) seems to work for most people, ; but that little whizz-bang apostrophe gives me little spam.e       Bernd wrote: >l >johnson@umtc.de wrote:s >> tJ >> Can anybody tell me how to go about getting awk for a VMS machine which >> doesn't have a C compiler?r >> t. >http://www.montagar.com/freeware/GAWK-2_15_6/     $ r gawk.exe. %DCL-W-ACTIMAGE, error activating image MTHRTLJ -CLI-E-IMGNAME, image file UMTC$DKA0:[SYS0.SYSCOMMON.][SYSLIB]UVMTHRTL.EXE: -SYSTEM-F-SHRIDMISMAT, ident mismatch with shareable image $    ------------------------------  # Date: Sat, 11 Nov 2000 12:33:57 GMT 8 From: Veli =?iso-8859-1?Q?K=F6rkk=F6?= <korkko@decus.fi>5 Subject: Re: login prompts inhibited with logins=0 ?? ' Message-ID: <3A0C4901.498A3AD@decus.fi>d  9 And since I am (relatively speaking) wet-eared newcomer, 33 having started with VMS/MicroVMS V4.5, I apparently_6 may not even have had a chance to experience such nice5 feature. But unfortunately after fair amount of years 5 and VMS versions it is slightly difficult to remember$8 all those ancient things and what feature came with what version.  ; However, every now and then one hits an older VMS version, a5 and somehow magically all those years ago experiencedd; (painful ?) happenings come back to ones mind, sort of dejag vu.b   _veli    Uwe Zessin wrote:  > * > In article <3A09A1BF.9C74D648@decus.fi>,= >   Veli =?iso-8859-1?Q?K=F6rkk=F6?= <korkko@decus.fi> wrote:c5 > > For many years and many, many versions, I've only  > > done > >a > >       $ set logins/inter=0 > >h7 > > and it will do the trick. Also, I have usually someo
 > > code likec > >c  > >       $ if somecond .nes. "" > >       $ then- > >       $  startup$interactive_logins :== 0& > >       $  exit  > >       $ endifN > > : > > in SYSTARTUP_VMS.COM and this will nicely disable user6 > > logins but I have always been able to login due to< > > possessing OPER privilege. Of course my memory with this1 > > issue goes to back to versions like V5 or so.t > > 	 > > _velip > A > Ok, then I should have said 'many, many years' and 'many, many,  > many versions' ;-) > B > It was on VAX/VMS V4.1 or V4.2 and for obvious reasons I haven't > tried it again.n >  > Uwe Zessin >  > ...l > > Uwe Zessin wrote: 1 > > > In article <C7GiK1D1e5B1@eisner.decus.org>,bB > > >   Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) wrote:E > > > > In article <3A085833.9736D2CC@ohio.edu>, "Richard D. Piccard"o  > > > <piccard@ohio.edu> writes:E > > > > > There is an echo in brain that VMS once upon a time behavedhH > > > > > differently if LOGINS were left at zero, in a scenario such asH > > > > > you describe, as opposed to having been briefly above zero andF > > > > > then set back to zero.  I found that objectionable, since itG > > > > > meant having, however briefly, a time when the non-zero valueV > > > > > was in effect. > > > > F > > > > I believe the first enabling started the Job Controller at its, > > > > task of creating LOGINOUT processes. > > >,D > > > Correct. I found that out the hard way on my first VMS upgrade5 > > > many years ago. I wanted to be 'clever' and setdD > > > 'STARTUP$INTERACTIVE_LOGINS' to 0 to keep any users off. Oops. > > > I > > > Later I used 'SET LOGINS/INTERACTIVE=1', 'SET LOGINS/INTERACTIVE=0'aE > > > in SYSTARTUP so I could log in (with OPER privilege, of course)aB > > > on a local terminal port while the system was still booting. > > >i > > > -- > > > Uwe Zessin > ( > Sent via Deja.com http://www.deja.com/ > Before you buy.g   ------------------------------  # Date: Sat, 11 Nov 2000 17:52:40 GMTn% From: Uwe Zessin <zessin@my-deja.com>f5 Subject: Re: login prompts inhibited with logins=0 ?? ) Message-ID: <8uk116$9st$1@nnrp1.deja.com>r  ' In article <3A0C4901.498A3AD@decus.fi>, ;   Veli =?iso-8859-1?Q?K=F6rkk=F6?= <korkko@decus.fi> wrote:s: > And since I am (relatively speaking) wet-eared newcomer,5 > having started with VMS/MicroVMS V4.5, I apparently)8 > may not even have had a chance to experience such nice
 > feature.  7 'nice feature'? Ugh. I was glad when I was able to pulli4 myself out of that mess - well, I _did_ have the old  system disk still available. ;-)  6 > But unfortunately after fair amount of years and VMS9 > versions it is slightly difficult to remember all thoset9 > ancient things and what feature came with what version.I   Agreed.   < > However, every now and then one hits an older VMS version,7 > and somehow magically all those years ago experiencede= > (painful ?) happenings come back to ones mind, sort of dejat > vu.2  < Yes. I have been to a customer this week to check his system9 for an upgrade. It's a mixed-architecture / mixed-versions4 system (V5.5-2 / V6.2-1H3), and I was tempted to use' '$ DIRECTORY/PAGE' on the VAX system...e   > _veli6   --
 Uwe Zessin    & Sent via Deja.com http://www.deja.com/ Before you buy.2   ------------------------------  % Date: Sat, 11 Nov 2000 04:01:06 -0400o- From: JF Mezei <jfmezei.spamnot@videotron.ca>s  Subject: Malloc + debug question, Message-ID: <3A0CFCC1.12DFE90D@videotron.ca>   DECC, VMS 7.2 (vax)_  K I have a bug which propped up on a program. It reproduces at the same place8% when processing a batch of messages.    D early in the message, I malloc a 21 byte string to hold a file name.G then, as I process the data for that message, I malloc various strings.   L In that processing, malloc at one point gives me an address that is equal toH that of the first string that was malloced. The program then effectively' overwrites the data of the first field.n  K Are there known problems where malloc  will reuse space that wasn't freed ?a  J If not, is there a way for me to set, in debug, a breakpoint if a specificK address/value is passed as an argument to any routine ? (eg: where would myaJ programm pass the address of the first string to the "free" routine, which+ would result in malloc reusing that space).m  E While I am at it, the DECCRTL now includes the "strdup" routine which L duplicates a string into newly allocated space (returning its address). Does1 anyone know if strdup is compatible with "free" ?,   eg:   / 	myvar = strdup("Delicious chocolate cookies");o
 	free(myvar);d   ------------------------------  % Date: Sat, 11 Nov 2000 13:49:53 -0500 2 From: "Richard B. Gilbert" <DRAGON@compuserve.com>3 Subject: Re: No doubts here (Was: Re: Galaxy doubt))7 Message-ID: <200011111350_MC2-BA75-2C30@compuserve.com>u  <         It may be the best but it's not the most economical!  J         What do you want from a home/small business system?  A WYSIWYG wo= rd6 processor?  A spreadsheet?  A small/simple database? =  J Quicken/Quickbooks/Money?  Games?  You can get everything I've mentioned = onH a PC for under $2,000.   Hardware, licenses, media, and documentation! =    A         Now price out a VMS workstation with licenses, media, and J documentation.  Add WordPerfect for VMS.  Add Lotus 1-2-3 for VMS.  Assum= eh? that someone was willing to update the last VMS versions to thecJ functionality of recent PC versions!   What do you pay?  $10K? $15K?  Is = it worth it?   In most cases, not!e  D         What sort of volume would you need to make it economical?  AH million units per year?  What kind of investment would it take to get to9 that point?  What would the return on that investment be?f  J         Maybe if DEC had made some cluefull moves in the early 1980s. . .=  .   It's too late now!  & Message text written by Paul Repacholi > = Different markets at the moment. VMS has few apps to even get ? started in that area. The sad theng is, VMS is *THE BEST* small.C bussiness/Home system base. It's *KID* proof! And manger proofable.5 But, no apps, no start...<   ------------------------------  % Date: Sat, 11 Nov 2000 17:22:17 +0100 0 From: Didier Morandi <Didier.Morandi@Easynet.fr>( Subject: SPEC_int92? SPECrate_fp95? etc.) Message-ID: <3A0D7239.94D5C55@Easynet.fr>O  = In http://www5.compaq.com/alphaserver/archive/comp/nov95.htmlt  V I see SPEC_int92, SPECrate_int92 and in the VMS 20 years presentation, SPECint_rate95.  * What are the differences, and the meaning?  G I remember that SPECmarks superceded the VUPS, then spec int (integer?)aE and fp (floating point?) came. Where are these units defined, please?i7 And are they still evaluating nr of CPU operations/sec?l   Thanks,h D.   ------------------------------  % Date: Sat, 11 Nov 2000 07:55:40 -0500t2 From: norm lastovica <norman.lastovica@oracle.com>' Subject: Re: sqlpre70 get sqlbugchk.dmpl* Message-ID: <3A0D41CC.FAAA5B0E@oracle.com>  
 chan fei -  2 	Please contact oracle rdb support and report this9 problem formally if it can be reproduced with the currenta: supported version of rdb (v7.0.5).  we'd very much like to9 make sure that the bug is corrected.  support will likely ; need the database definition and a copy of the source code.    	thanksm( 	norm lastovica / oracle rdb engineering   CHAN Fei wrote:  > C > I found that the problem occurs only for the non-ansi style code, = > after I've changed to ANSI format, this problem was solved.:D > I think sqlpre70 got serious prolbem in pre-compile non-ansi style > sql-cobol programs.d > $ > CHAN Fei (f1chan@hkucs.org) wrote:
 > : Dear all, E > :       I got serious problem when pre-compile an embeded-SQL COBOLiM > : program, the problem was compiled okay all the way, until the source codet8 > : grows over 1280 lines(the space lines are inclusive) >  > : errors:  >  > : $ scob == "$sql$pre /cobol"g > : $ scob rnpextrntivb.scod3 > : %SQL-I-BUGCHKDMP, generating bugcheck dump file % > : DKA200:[D07CHINESE]SQLBUGCHK.DMP;  > ' > : and messages in the dump file are :s > 9 > : This is a AlphaServer 800 5/333 running VMS  V7.1-1H2a+ > : Current time is 10-NOV-2000 14:28:42.87PF > : Bugcheck Dump Identification is "QZYQZ8MXQ3HBAAAE2U75GU6VAAUKAAAA" > : T > : ================================================================================ > :           Stack DumpT > : ================================================================================ > T > : --------------------------------------------------------------------------------B > : ***** Exception at 006D81AC : SQL$$CREATE_SQL_FIELD + 00000844? > : %SYSTEM-F-ACCVIO, access violation, reason mask=00, virtualh > : address=0000000000D_ > & > : ..... (error codes in hex deleted) > N > : I've increase Page fault quota, Working set quota/extends, but seems theseG > : values are not violated. Anyone could provide hints on the problem?M > : many thanks.   -- S> norman lastovica / oracle rdb engineering / usa / 610.696.4685   ------------------------------  % Date: Sat, 11 Nov 2000 19:37:24 +0010t% From: paddy.o'brien@zzz.tg.nsw.gov.aud Subject: Re: Time Zone5 Message-ID: <01JWFBSMWAVM0068XX@tgmail.tg.nsw.gov.au>9   Bill Gunshannon wrote:3 >In article <3a014cd2.2060383875@news.newsguy.com>,s) > agreig@my-deja.com (Alan Greig) writes:  >|>gJ >|>                                              but the Spring US date isJ >|> out of synch with almost everywhere else. I think that's the right way
 >|> round. >bE >What do you mean the US date is out of synch??  We invented DaylighteH >Savings Time.  It's the rest of the world that's out of sync!!  :-) :-)    Q There was some comment earlier as to who had "invented" it.  Someone gave a case oK for Ben Franklin, I had always thought that it was UK in one of the wars.  aB Perhaps there is a distinction betwixt inventing and implementing.  O Since in most parts of the world some areas (States - including some US states  O as in Australia, local countries) have it or not, either those who have it are mO out of sync, or those who don't are. (Yep, I saw your smilies, and my comments c are full of invisible ones :-)  L My read of Alan was not that parts had it or not, but the timing of when it N occurred.  In comments earlier by myself and J-F Mezei, it was noted that the M Eastern side of Australia (NSW and Victoria -- not Queensland who never have oL daylight saving -- it fades the curtains) went rather silly by shifting the Q normal change date by two months because of the Olympics.  Still not sure of the h
 rationale.  O I know at the time a step-daughter in South Australia (which did not change at bP the abnormal time) had to completely change her work hours.  The head office in G Sydney expected her to be at work at 9:00 Sydney time -- 7:30 in South  ? Australia.  There is normally only a half-hour time difference.   Q It's a mess, and since we are not at war world wide (when I believe it was first tM implemented) why do we still need it?  I also have vague memories of hearing  : that cows had to be "educated" to give milk earlier/later.   Regards, Paddy   Paddy O'Brien, Transmission Development, 
 TransGrid, PO Box A1000, Sydney South,  NSW 2000, Australial   Tel:   +61 2 9284-3063 Fax:   +61 2 9284-3050& Email: paddy.o'brien@zzz.tg.nsw.gov.au  M Either "\'" or "\s" (to escape the apostrophe) seems to work for most people,i; but that little whizz-bang apostrophe gives me little spam.t   ------------------------------  % Date: Sat, 11 Nov 2000 12:43:34 -0500 & From: Jerome Fine <jhfine@idirect.com> Subject: Re: Time Zone* Message-ID: <3A0D8546.91436AD@idirect.com>   >Hoff Hoffman wrote:  j > >In article <39FF8A2F.E44231FD@earthlink.net>, "David J. Dachtera" <djesys.nospam@earthlink.net> writes:- > >re: How to reset for daylight savings timec > >See HELP SUBMIT /AFTER3  > >..and do this in a batch job:( > >In autumn:     $ SET TIME="-01:00:00"J >   That very likely won't have the expected results in all cases, and theL >   SET TIME itself will be rejected if DTSS or other time-keeping mechanismI >   is enabled, and would definitely not be the approach I would suggest.tJ >   Please see the OpenVMS FAQ for directions on the daylight savings timeJ >   switch-over.  (Suggestions for updates to the time-related informationI >   in the FAQ -- or most anything else, for that matter -- are welcome.)a   Jerome Fine replies:  N I understand that every so often, there is also a Leap Second required.  WhileI it occurs, on average, less than once a year, which way would be the best . to set the clock back by just that ONE second.  N In addition, what is the current state of the code for keeping track of dates?M I understand that the subroutines are written in 32-bit C and will need to be0K modified - or better still - changed to 64-bit C.  What is the cut-off date H for this to be done?  And will it be done for VMS on the VAX or just forG VMS on Alpha since the switch to a 64-bit C on Alpha will probably taket place within a few years?a   Sincerely yours,   Jerome Fine    ------------------------------   End of INFO-VAX 2000.631 ************************