1 INFO-VAX	Sun, 30 Apr 2006	Volume 2006 : Issue 238       Contents: Re: Capturing console activity Re: Capturing console activity Re: Capturing console activity8 Re: DCL versus Unix CLIs, was: Re: File output like Unix8 Re: DCL versus Unix CLIs, was: Re: File output like Unix8 Re: DCL versus Unix CLIs, was: Re: File output like Unix Re: Debugging a crashing editor / Re: Help Required With VAX 4000-500 Diagnostics / Re: Help Required With VAX 4000-500 Diagnostics / Re: Help Required With VAX 4000-500 Diagnostics P Re: Now this is (a little) insult (Was Re: Quorum, locks and application questio) Re: Various comments (decus presentation) ) Re: Various comments (decus presentation)   F ----------------------------------------------------------------------  % Date: Sat, 29 Apr 2006 22:00:25 -0500 @ From: "David J. Dachtera" <djesys.nospam@NeOaSrPtAhMlNiOnWk.net>' Subject: Re: Capturing console activity 6 Message-ID: <44542849.589A2DAD@NeOaSrPtAhMlNiOnWk.net>   Thomas Wirt wrote: >  > Wilm Boerhout wrote:7 > > glen.thompson@gmail.com wrote on 28-4-2006 19:45...  > > - > >> System: OpenVMS 7.2 on a Vaxstation 4000  > >>L > >> Our system console is a Decwriter printing terminal because the log wasL > >> required many years ago by our internal audit department.  We'd like toK > >> get rid of the Decwriter and replace it with one of our spare VTs.  Is I > >> there anyway to automatically capture all the activity on the system L > >> console?  I'd like an electronic version of what is currently on paper,8 > >> including both operator input and output displayed. > >  > > - > > Would the auditing dept. settle for this:  > > K > > Connect a PC to the console through a serial cable (COM port); start up > > > a terminal emulator that allows session logging to a file. > > D > > Several Terminal Emulators will do this, among them the good old > > Pathworks VT320 TE.  > G > This is a great use for an old laptop that is too slow to do anything A > else.  We do this on one of our 4 Alphas and move it as needed.   H Another re-use of gear is to get something like IOnetwork's (Now Digi's)F Edgeport/8 device (USB to multiple serial port adapter) to log data onF multiple consoles. Use the right version of Windows (XP-something thatH allows remote console) on an old PC with at least USB-2 and you can even@ have remote access to them! (Just not like PCM which allowed one> read/write and one or more read-only accesses simultaneously.)   --   David J Dachtera dba DJE Systems  http://www.djesys.com/  & Unofficial OpenVMS Marketing Home Page! http://www.djesys.com/vms/market/   ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/   ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/   ------------------------------  % Date: Sat, 29 Apr 2006 22:03:14 -0500 @ From: "David J. Dachtera" <djesys.nospam@NeOaSrPtAhMlNiOnWk.net>' Subject: Re: Capturing console activity 6 Message-ID: <445428F2.611F3F84@NeOaSrPtAhMlNiOnWk.net>   Rob Brown wrote: > 4 > On Fri, 28 Apr 2006 glen.thompson@gmail.com wrote: > G > > Our system console is a Decwriter printing terminal because the log G > > was required many years ago by our internal audit department.  We'd C > > like to get rid of the Decwriter and replace it with one of our @ > > spare VTs.  Is there anyway to automatically capture all theF > > activity on the system console?  I'd like an electronic version ofH > > what is currently on paper, including both operator input and output > > displayed. > B > I asked a similar question way back in 1999 in the DECUServe VMS? > conference, topic 2988 (telnet://eisner.encompasserve.org/ or H > http://eisner.encompasserve.org).  More or less, the answer was "write  > something to do what you want. > E > Since you already have a VMS system, it seems to me that everything F > that gets sent there (except system crashes) already gets written to > SYS$MANAGER:OPERATOR.LOG.   D Well, not quite everything. I recently discovered that network loginD failures one of my machines was not being recorded in the OPCOM log,G though they did appear on OPA0: Still haven't quite untangled that one.   G Also, complaints from VMS like page/swap file related issues only go to  OPA0: and not to OPCOM.   - > So you must be interested in also recording - > interactive sessions on the system console.  > H > One solution, which comes to me now but didn't 7 years ago, is to turnH > on "autoprint" mode on the VT and run a cable from the VT printer port@ > back to the host, where it gets read and saved to a file via a* > detached COPY LTAx: LOGFILE.LOG command.  	 Good one!    --   David J Dachtera dba DJE Systems  http://www.djesys.com/  & Unofficial OpenVMS Marketing Home Page! http://www.djesys.com/vms/market/   ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/   ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/   ------------------------------  % Date: Sun, 30 Apr 2006 00:10:16 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> ' Subject: Re: Capturing console activity , Message-ID: <445438A3.93619A2F@teksavvy.com>   "David J. Dachtera" wrote:F > Well, not quite everything. I recently discovered that network loginF > failures one of my machines was not being recorded in the OPCOM log,I > though they did appear on OPA0: Still haven't quite untangled that one.   
 SHOW LOG OPC*   * These are set/documented in SYLOGICALS.COM  H And yes, you are right that not all of the operator console is logged byD opcom. The nastiest messages (system crash, system boot etc) are not logged in OPCOM.   ------------------------------    Date: 29 Apr 2006 11:16:28 -0700$ From: "AEF" <spamsink2001@yahoo.com>A Subject: Re: DCL versus Unix CLIs, was: Re: File output like Unix B Message-ID: <1146334588.644455.22310@e56g2000cwe.googlegroups.com>   FredK wrote:4 > "Rob Brown" <mylastname@gmcl.com> wrote in messageA > news:Pine.LNX.4.61.0604281211340.19111@localhost.localdomain...  > D > > > Why include Unix?  In Unix, no filename (or extension) has anyI > > > implicit meaning.  Applications may impart meaning, but the OS just  > > > don't care.  > > 0 > > I think that the same is true in VMS or RSX. > >  > 
 >     .DIR     Well, not by itself!  
     .DIR;1  " And the directory bit must be set.  E But is this really more basic than .COM and SYS$SYSTEM:.EXE to VMS? I C guess in a sense it is in that directory files *have* to be of this @ form to function as directories. Aside from this they are reallyB similar in that RMS assumes that for [DIR1.DIR2.DIR3] there existsC appropriate files ending with .DIR;1 to look for and @ assumes that G @BLAH means there is a file ending with .COM to look for (and similarly B for RUN). One could go on and on with all these subtle variations!    I > > (MS) Windows, on the other hand, takes extension-based file typing to G > > another level (higher or lower, depending on your opinion).  Simply H > > invoking FOO.DOC runs MSWORD (for example) without you invoking Word > > at all.  > >  >  >     As will CDE    AEF    ------------------------------  % Date: Sat, 29 Apr 2006 15:39:00 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> A Subject: Re: DCL versus Unix CLIs, was: Re: File output like Unix , Message-ID: <4453C0CE.482C924A@teksavvy.com>  G Actually, VMS and DCL itself do not base processing based on file type.    When you type @CHOCOLATE  G , then it is the @ "command" which fills any missing file specification 6 with .COM.  But you can do @CHOCOLATE.TXT if you want.  @ When you do CC chocolate, it is the CC command which tags a ".C"1 extension if it isn't specified in the file name.   F When you say PRINT chocolate , it is the PRINT command which puts in aC .LIS default file extension if you didn't specify a file extension.   D So basically, the SY$PARSE system service has the uncanny ability toH complete a file spec with application defined defaults if those part are not specified.    A When you look at DOS however, you could type in a filename at the H prompt, and DOS would automatically look up what files are available and< then either invoke the "batch file" processor if there was a= "chocolate.bat" file, the executable processor if there was a T chocolate.exe or another processor if there was a special image file "chocolate.com"  E VMS itself doesn't seem to have any such "based on the file type I do H something differently".  It in fact has fairly explicit processing basedE on how a command is invoked. (@, RUN, foreign image, DCL command etc)    ------------------------------  % Date: Sat, 29 Apr 2006 21:53:20 -0500 @ From: "David J. Dachtera" <djesys.nospam@NeOaSrPtAhMlNiOnWk.net>A Subject: Re: DCL versus Unix CLIs, was: Re: File output like Unix 6 Message-ID: <445426A0.C03686B2@NeOaSrPtAhMlNiOnWk.net>   "Richard B. Gilbert" wrote:  >  > AEF wrote: > > Simon Clubley wrote: > > c > >>In article <1145989404.330424.164290@g10g2000cwb.googlegroups.com>, bob@instantwhip.com writes:  > >>D > >>>why do users want convuluted unix commands like pipe grep slurp? > >>>gulp ... when they use common english makes sense commands  > >>>like COPY on vms? > >>>  > >>< > >>Pipe is a DCL command. Unix uses redirection characters. > >>K > >>Those Unix commands are generally more powerful than the VMS equivalent = > >>and naming commands using English words only goes so far.  > >>L > >>For example, what English verbs and qualifiers would you use to describeK > >>the functionality that ghostscript offers and by the time that you have Q > >>finished have you really gained anything over the current approach (as people 2 > >>still have to learn what your qualifiers do) ? > >>+ > >>Two examples of more powerful commands: J > >>      grep (the Unix version of search) can do full regular expression > >>      pattern matching.  > >  > > C > > Can it do something like $ SEARCH FILE.TYP WORD1,WORD2,... with E > > /MATCH=(one of OR, NOR, AND, NAND, ...) Maybe it can ... I'm just ( > > asking but I think the answer is no. > >  > 8 > You are right about that.  But, OTOH, can SEARCH find:7 > 1. a record that "begins with" a string or a pattern? 5 > 2. a record that "ends with" a string or a pattern?   F For 1 & 2, see HELP SEARCH /KEY, with the caveat that "ends with" will# be limited to fixed-length records.   J > 3. find a record containing one or more spaces followed by exactly threeJ > decimal digits, then one upper case alphabetic character and then one or > more spaces?????  . Convoluted, but possible though not practical.   --   David J Dachtera dba DJE Systems  http://www.djesys.com/  & Unofficial OpenVMS Marketing Home Page! http://www.djesys.com/vms/market/   ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/   ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/   ------------------------------  # Date: Sat, 29 Apr 2006 18:47:25 GMT , From: Hoff Hoffman <hoff-remove-this@hp.com>( Subject: Re: Debugging a crashing editor2 Message-ID: <1xO4g.6988$B42.1655@news.cpqcorp.net>   Sidney Cammeresi wrote:   ! > 		Vim: Caught deadly signal BUS  > 		Vim: Finished.  G    Those are vim-generated errors, too, and not OpenVMS errors.  (They  F may well be OpenVMS signals which were translated into C signals, and I that vim has then captured and reported.  But in that format, the errors  5 are initially arising from vim and not from OpenVMS.)   A    There were some questions posted in an earlier reply, and I'd  I definitely look to the particular vim you were using.  There's a copy of  G   vim available at <http://www.hp.com/go/openvms/freeware/>, and there  9 are likely a number of other updates and versions around.   F    As for debugging this, you can compile and link the vim image with F debugging -- the compilation generally sees /DEBUG/NOOPTIMIZE and the E linker generally sees /DEBUG added to the respective build commands.  F When you then invoke the resulting image, you'll be in the DECwindows E debugger interface, or in the character cell interface, depending on  H your current environment.  (To disable the DECwindows debugger and fall B back to the character cell interface, you can use the DCL command I <DEFINE[/NOLOG][/JOB] DBG$DECW$DISPLAY "  "> during your login.com login  E command procedure, or at any time before you start up an image built  G with debug.)  HELP, EXAMINE, STEP and a large number of other commands  I are available within the debugger, and a GO command will generally start  
 up the image.   G    The programming concepts manual and the debugging manual are a good  C start here, if you are looking to learn how the various pieces fit  B together.  These and most other OpenVMS manuals and documents are H available at <http://www.hp.com/go/openvms/doc/>.   The OpenVMS FAQ has H information on common C programming bugs, and pointers to lists of same.   ------------------------------  # Date: Sun, 30 Apr 2006 03:34:59 GMT 0 From: John Santos <john.santos@post.harvard.edu>8 Subject: Re: Help Required With VAX 4000-500 Diagnostics> Message-ID: <MPG.1ebdfa6bc614e009989710@news.bellatlantic.net>  F In article <e2ptl7$fcg$1@sun-cc204.lut.ac.uk>, g.j.parker@lboro.ac.uk  says...  > JF Mezei wrote:  > >...D > > Note: if you did not terminate the DSSI bus, it is possible thatI > > overnight, the DSSI just burned itself out. The DSSI controllers were H > > quite picky about this.  You could remove terminators for very shortH > > periods, but not for long periods (more than a minute for instance). > % > Very useful info there, JF, thanks.  > E > I don't like the sound of 'burned itself out', is this a game over  H > situation for the CPU module? As far as I know I didn't leave the bus E > unterminated but did leave the machine running with the drive cage  J > containing two RF36s uninstalled. My knowledge of DSSI is based only on D > what I know of SCSI so may not have a great deal of mileage in it. > K > Bernhard Wulf also commented, off-ng, that the CPU card has a small fuse  J > on it that supplies termination power to the DSSI bus and this may have 8 > been damaged. I'm going to check this out later today.  E Urgh!!  (Due to a news problem, I missed a couple of days' posts this E week, now going back and catching up...  A little behind, but, hey, I C distributed the last of my Christmas presents earlier this evening. - Still got one coming to me, apparently... :-)   E Anyway, my "new" VAX 4000-300 says one of the DSSI adapters is broken E and may have a blown fuse.  It says this 50 times every boot and then C shuts up.  Is the fuse easy to replace?  Is blowing it a symptom of E running without enough terminators?  How many do I need?  There are 2 D DSSI buses, currently no DSSI devices, but there appear to be 3 DSSIF connectors, two on the CPU and one on the panel at the left which also@ has a SCSI connector.  I could only scavenge 2 DSSI terminators.  I > My machine also has a KZQSA SCSI card in it (currently hosting a CDROM  K > and a DAT drive). If the DSSI controller turns out to be toast can I pop  I > some external SCSI hard drives on this chain and still use the machine?   C My understanding is this controller is unsupported for anything but F tapes and CDROMs, but it does work with some disks.  I also understandC it is quite slow, but maybe that is in comparison with modern SCSI, B and it maybe quite a bit faster than DSSI.  It would be limited by@ the Qbus as well.  If this is a hobbyist system, I would try it.B SCSI disks are much easier to find than DSSI, and are available inE much larger sizes and with faster seek and rotational latencies, even 6 if you can't take advantage of greater transfer rates.   >  > Gary >    --   John   ------------------------------  % Date: Sun, 30 Apr 2006 00:21:48 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> 8 Subject: Re: Help Required With VAX 4000-500 Diagnostics, Message-ID: <44543B56.7AAFD1FC@teksavvy.com>   John Santos wrote:G > Anyway, my "new" VAX 4000-300 says one of the DSSI adapters is broken G > and may have a blown fuse.  It says this 50 times every boot and then  > shuts up.   G On the -500/-600, the terminator looks like  resistor that is "exposed" H on the back of the console panel (eg: open the panel, look at the bottom right of the back.  E Not sure about the -300. On my -200, that terminator is not visible.    F > DSSI buses, currently no DSSI devices, but there appear to be 3 DSSIH > connectors, two on the CPU and one on the panel at the left which also > has a SCSI connector.   E One of the DSSI busses is not terminated at either end. So you have 2 A plugs for it. One on the left side of the cabinet, and one on the  console panel.    H This is so that you could have   [term]-[HOST]----[HOST]---[HOST]-[term]C So it needs 2 terminators to function if you have no other cabinets  connected to it.    F The other DSSI is terminated at one end, unterminated at the other. So$ it needs one terminator to function.  E > it is quite slow, but maybe that is in comparison with modern SCSI, D > and it maybe quite a bit faster than DSSI.  It would be limited by > the Qbus as well.   & DSSI is considerably faster than QBUS.   ------------------------------  # Date: Sun, 30 Apr 2006 05:14:56 GMT 0 From: John Santos <john.santos@post.harvard.edu>8 Subject: Re: Help Required With VAX 4000-500 Diagnostics> Message-ID: <MPG.1ebe11dddc0476f7989712@news.bellatlantic.net>  - In article <44543B56.7AAFD1FC@teksavvy.com>,  $ jfmezei.spamnot@teksavvy.com says... > John Santos wrote:I > > Anyway, my "new" VAX 4000-300 says one of the DSSI adapters is broken I > > and may have a blown fuse.  It says this 50 times every boot and then 
 > > shuts up.  > I > On the -500/-600, the terminator looks like  resistor that is "exposed" J > on the back of the console panel (eg: open the panel, look at the bottom > right of the back. > G > Not sure about the -300. On my -200, that terminator is not visible.   > H > > DSSI buses, currently no DSSI devices, but there appear to be 3 DSSIJ > > connectors, two on the CPU and one on the panel at the left which also > > has a SCSI connector.  > G > One of the DSSI busses is not terminated at either end. So you have 2 C > plugs for it. One on the left side of the cabinet, and one on the  > console panel. >  > J > This is so that you could have   [term]-[HOST]----[HOST]---[HOST]-[term]E > So it needs 2 terminators to function if you have no other cabinets  > connected to it. >  > H > The other DSSI is terminated at one end, unterminated at the other. So& > it needs one terminator to function. >   & Sounds like I need another terminator.  G > > it is quite slow, but maybe that is in comparison with modern SCSI, F > > and it maybe quite a bit faster than DSSI.  It would be limited by > > the Qbus as well.  > ( > DSSI is considerably faster than QBUS. >   C Yup, and the built-in DSSI buses (on the CPU card) don't go through A the Qbus, but straight to memory.  (The memory isn't on the Qbus, C either.  PMI or something similar on the MVII through the 4000-200, E and something else (at least, it's mechanically different and doesn't F use the across-the-front ribbon cables) on VAX 4000-300 and up.  Maybe6 the MVI was pure Qbus memory, like most Qbus PDP-11s?)  < But what I meant was that the KFQSA would be limited by QbusC bandwidth, unlike the DSSI buses (unless you were using a Qbus DSSI ? card), so even if the SCSI disks were intrinsically faster, the B system might be slower than with DSSI.  But if your application is? seek-bound, it might be faster on the faster SCSI disks despite ) their being throttled by the slower Qbus.      --   John   ------------------------------  # Date: Sat, 29 Apr 2006 18:33:58 GMT , From: Hoff Hoffman <hoff-remove-this@hp.com>Y Subject: Re: Now this is (a little) insult (Was Re: Quorum, locks and application questio 2 Message-ID: <qkO4g.6987$XT1.3833@news.cpqcorp.net>  C    There is no obvious evidence of any references to a freeze lock  I within the OpenVMS SYS system facility, so I can't say what Rdb might be  I doing with it, nor how Rdb might have implemented its locking.    Having  G just searched the whole of it, I see only a few references to "freeze"  > any where in the OpenVMS SYS system facility, and none appear G particularly relevant to your question.  (I've been using Rdb for eons  B now, and can't recall ever having specifically seen this, either.)  E    There are certainly various texts available that describe locking  D schemes and algorithms, if that is of interest here.  Knuth, et al. E DECdtm and RTR documentation is also available, and Rdb is certainly  E able to integrate into a distributed transactional environment.  (As  F I've previously commented, I have usually found it more economical to G use existing middleware than to write it fresh -- though that approach  F is certainly far less educational and sometimes also rather less fun, B and the up-front acquisition costs can be tough(er) to justify in H certain situations and in some organizational administrations.)  But as I to what Rdb is doing here or as to what portions of or what sequences of  D the OpenVMS distributed lock manager might be in use here, I simply  don't know.    ------------------------------  # Date: Sun, 30 Apr 2006 03:56:14 GMT 0 From: John Santos <john.santos@post.harvard.edu>2 Subject: Re: Various comments (decus presentation)> Message-ID: <MPG.1ebdff6ed741a346989711@news.bellatlantic.net>  : In article <zW24g.1701$1V4.137891@news20.bellglobal.com>, $ newsonly@weaverconsulting.ca says...= > "JF Mezei" <jfmezei.spamnot@teksavvy.com> wrote in message  ( > news:44500D37.BC7B5357@teksavvy.com... > >...D > > (Note, I had told the organisers to bring duct tape to cover my 
 > > mouth,- > > and they didn't. So it is their fault :-)  > F > That gave me a great idea, but I checked the website of the place I F > got the flashlights made and they do not do custom printing on duct  > tape. :-(  >    Check out Lee Valley. G   <http://www.leevalley.com/wood/page.aspx?c=1&p=32180&cat=1,110,43466>   E Great Canadian company.  They have real duct tape.  (Not cheap.) They > also have Gaffer's tape.  The owner does usenet - hangs out onA rec.woodworking.  Don't know if they customize it or if they were * the official supplier to the Possum Lodge.     > >...E > > When a speaker does insist that people ask questions to make the   > > session * > > "interactive", why not ask questions ? > I > I never said there was anything wrong with asking questions, just that  C > I noticed the person trying to count how many questions you were  G > asking gave up when the person sitting next to him refused to remove  $ > his shoes and socks to help. :):)   A If he counted in binary, he should have been able to hit 1023 (1  B Kiloquestion) just on his fingers.  1 Gigaquestion on his fingers,E toes and his neighbor's fingers.  Or did JF exceed that?  :-) :-) :-)    --   John   ------------------------------  % Date: Sun, 30 Apr 2006 00:37:44 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> 2 Subject: Re: Various comments (decus presentation), Message-ID: <44543F11.FC4C67AB@teksavvy.com>   John Santos wrote:G > Great Canadian company.  They have real duct tape.  (Not cheap.) They @ > also have Gaffer's tape.  The owner does usenet - hangs out onC > rec.woodworking.  Don't know if they customize it or if they were , > the official supplier to the Possum Lodge.  E 3M duct tape was the original/official supplier to Possum Lodge which G unfortunatly closed a few weeks ago, closing a long chapter of canadian H history. Canadian Tire has not sold 3M real duct tape in years. But someD other hardware stores do have it. The cheap stuff called "duck tape"@ made is mexico is useless unless you want just temporary stuff.     B > If he counted in binary, he should have been able to hit 1023 (1D > Kiloquestion) just on his fingers.  1 Gigaquestion on his fingers,G > toes and his neighbor's fingers.  Or did JF exceed that?  :-) :-) :-)    I can count to 99 on my hands.    ( right thumb = 5, other right fingers = 1( left thumb = 50, other left fingers = 10     1 = 1 right finger down. 4 = 4 right fingers down- 5 = right thumb down, other right fingers up. 1 9 = right thumb down, 4 other right fingers down. ) 10 = 1 left finger down, right fingers up      so, 74 =    4 left thumb down, 2 left fingers down  (50 + 10 + 10)4 right thumb up, 4 right fingers down (1 + 1 + 1 + 1)  G Never considered using the fingers in binary mode.  Not quite that good A at converting from decimal to binary in my head for any arbritary  number.    ------------------------------   End of INFO-VAX 2006.238 ************************