1 INFO-VAX	Fri, 03 Nov 2006	Volume 2006 : Issue 604       Contents:! Re: Another Vax/Alpha/VMS triumph . Re: Debugger GUI only working on some machines. Re: Don Jenkins, VP BCS, has left the building+ Re: DS10L hanging problem, tracking it down + Re: DS10L hanging problem, tracking it down + Re: DS10L hanging problem, tracking it down + Re: DS10L hanging problem, tracking it down + Re: DS10L hanging problem, tracking it down + Re: DS10L hanging problem, tracking it down + Re: DS10L hanging problem, tracking it down + Re: DS10L hanging problem, tracking it down + Re: DS10L hanging problem, tracking it down 3 Re: Itanium model numbers {re F$getsyi("HW_MODEL")} 3 Re: Itanium model numbers {re F$getsyi("HW_MODEL")} . Re: Login Break-in LGI parameters (Resolution) More COV Self-Help right here!4 Re: Observations on Samba for VMS (CIFS Kit from HP) Re: PIPE question  Re: PIPE question / Re: Rdb Network Security - Multinet 5.2 Webinar 4 Re: Rich Marcello retiring to perform volunteer work4 Re: Rich Marcello retiring to perform volunteer work4 Re: Rich Marcello retiring to perform volunteer work4 Re: Rich Marcello retiring to perform volunteer work4 Re: Rich Marcello retiring to perform volunteer work4 Re: Rich Marcello retiring to perform volunteer work4 Re: Rich Marcello retiring to perform volunteer work$ Strange CDE overnight disapearance !( The future of Internet Socket Protection Re: Time change questions !  Re: Time change questions !  Re: Time change questions !  Re: Time change questions !  Re: Time change questions !  Re: Time change questions !  Re: Time change questions !  Re: Time change questions !  Re: Time change questions !  Re: Toggle Operator Console  Re: Toggle Operator Console   F ----------------------------------------------------------------------   Date: 2 Nov 2006 15:48:53 -0600 % From: frey@encompasserve.org (Sharon) * Subject: Re: Another Vax/Alpha/VMS triumph3 Message-ID: <RAF6Zmrplxct@eisner.encompasserve.org>   N In article <87mz7924e7.fsf@k9.prep.synonet.com>, prep@prep.synonet.com writes:) > frey@encompasserve.org (Sharon) writes:  > @ >> 	Um, nope, actually it was TOTAL downtime to the entire groupC >> of 911 dispatchers.  The dispatching system is what ties all the > >> other components together; phones, radios, etc. like a hub.A >> Although it takes advantage of VMS clustering to provide quick E >> failover, there's no way to do a phased switchout of the hardware.  >> Ain't it great?  :-)  >  > That in Seatle?   7 	Nope!  A very large customer in the state of New York.   	  - Sharon " "Gravity...  is a harsh mistress!"   ------------------------------   Date: 2 Nov 2006 14:48:39 -0800  From: amelia_airhead@yahoo.com7 Subject: Re: Debugger GUI only working on some machines C Message-ID: <1162507719.117821.152100@b28g2000cwb.googlegroups.com>   F   There's a very minor difference in the Buffered I/O byte count quota between A the two systems but the rest of the process quotas are identical.      Amelia   Paul Sture wrote: D > In article <1160766107.644813.27810@h48g2000cwc.googlegroups.com>," >  amelia_airhead@yahoo.com wrote: >  > > Paul Sture wrote:  > N > > > Could this be a quota problem? Can you check the user quotas against the. > > > system where you _can_ run the debugger? > > >  > > 	 > > Paul,  > > ( > >   There's no quota on either system. > >  > E > Sorry, I meant process quotas in the User Authorization File (UAF).  > C > Do you see any differences in the output of the following command < > between the system that does work and the ones that don't? >  > $ SHOW PROCESS /QUOTA  >  > --   > Paul Sture   ------------------------------  % Date: Fri, 03 Nov 2006 01:18:45 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> 7 Subject: Re: Don Jenkins, VP BCS, has left the building , Message-ID: <454ADF3A.156E56A4@teksavvy.com>   Ken Farmer wrote:  > K > Word has it Don Jenkins left HP today.  Don't know if he retired or what.    Don Jenkins E Vice President, Marketing, Business Critical Servers  Hewlett-Packard ? Company  Don Jenkins is responsible for the marketing of HPs G business-critical server products,  including HP Integrity, HP 9000 and H HP Alpha servers as well as the HP-UX, Tru64  UNIX and OpenVMS operating environments.   E Prior to this position, Jenkins was director of Operating Environment 8 Product Marketing  for Business Critical Systems at HP.   D During his 20-plus years with Compaq and Digital Equipment Corp., heA held a variety  of executive and technology management positions, @ including vice president of High Performance Systems Product andD Solutions Marketing. He also held the position of vice  president ofD Product Management and Marketing for Compaq's UNIX Systems Business C Unit, and he was responsible for Strategic Business Development for 3 Compaqs High  Performance Server Business Unit.    ? Jenkins also served as director of Sales, Marketing and Product C Management for  Intuitive Technology Corporation, a startup company E providing client/server and  Internet-based software products for the A process manufacturing market, and as chief  executive officer and = president of Integrated Systems Assemblies, a venture-backed  A multichip module company.  Jenkins holds a bachelor of science in F chemical engineering and an MBA from the  University of Massachusetts.   ------------------------------   Date: 2 Nov 2006 13:09:05 -0800 ( From: "Rich Jordan" <jordan@ccs4vms.com>4 Subject: Re: DS10L hanging problem, tracking it downC Message-ID: <1162501745.019593.118380@b28g2000cwb.googlegroups.com>    JF Mezei wrote:  > Rich Jordan wrote:I > > minutes.  It doesn't tend to change once the system has been up about I > > 15 minutes, and has never shown a change immediately prior to a hang. > > > I could tighten up the loop to get more frequent readings. > K > If the temperature stabilises well before a failure, then the temperature ' > at the sensor may not be the problem.  > L > However, the temperature elsewhere in the system may be an issue. Prior toK > my modifications, a lot of hot air from the CPU bled over to the PCI card * > side and that heated up that whole area. > L > If your current temperature is 40=B0 (Are you in Coober Pedy ? :-), then = evenK > the front side of the DS10 would be hot since air coming in to "cool" the L > disk drives would already be hot. (Or were to talking about 40 USA degree= s ?) > 6 > I take it the error log shows no hint of a problem ?  G System temp reported in centigrade; most recent runs (yesterday) got up D to 36C reported.  Its cooler upstairs even with the furnace running.F Earlier when the problem started I was seeing 44C logged; the computer& room stays warm even with A/C running.  G Outside temp is expected to be in the 40s Fahrenheit when I do the open E box testing.  Depending on ambient in the house, and especially if it A ends up warmer outside, I might also turn on the A/C and duct air E directly to the DS10 inlets.  This is the last shot, may as well do a < thorough job of it before spending real money to replace it.   Rich   ------------------------------   Date: 2 Nov 2006 15:00:01 -0800 ( From: "Rich Jordan" <jordan@ccs4vms.com>4 Subject: Re: DS10L hanging problem, tracking it downB Message-ID: <1162508401.693582.102640@m7g2000cwm.googlegroups.com>   JF Mezei wrote:  > Rich Jordan wrote:E > > ends up warmer outside, I might also turn on the A/C and duct air I > > directly to the DS10 inlets.  This is the last shot, may as well do a @ > > thorough job of it before spending real money to replace it. > J > If it is an not-obvious problem like a failed fan etc, then the odds areH > that one or more chips are getting near their death and any warming upE > causes flakieness. If the only way to keep the system running is to L > constantly pour liquid nitrogen on the unit, it may satisfy your hobbyists7 > curiosity, but it wouldn't result in a viable system.  > I > Note: Once my old mac failed to boot. Not even some sad mac icon on the J > display. Pretty scary. Turns out that removing the CPU daughterboard andL > putting it back in fixed the problem. No idea what would have caused it toL > fail between the time I turned it off and the time I turned it back on theK > next day. Random glitch caused by the alignement of Venus with regards to  > Alpha Centuri. > K > So sometimes, there are things that can be done to "permanently" fix some . > glitch. But other times, it is a lost cause.  D A powercycle is by far the most common time for hardware failures toA occur.  Could be thermal, could be a capacitor or other component E failing when the output voltage upsurges, etc.  Machines left on tend D to last longest.  unfortunately this DS10L was left on all the time, but failed anyway.  F Not a fan, and all removable parts/cables/connectors have been removedF and reseated.  The cold test will tell the rest, although even if thatD passes its still time to replace the box.  Its just nice to know the cause when things break.   Rich   ------------------------------  $ Date: Thu, 2 Nov 2006 20:02:41 -0500/ From: "William Webb" <william.w.webb@gmail.com> 4 Subject: Re: DS10L hanging problem, tracking it downI Message-ID: <8660a3a10611021702w41817457k83215bece110f047@mail.gmail.com>   F On 16 Oct 2006 08:15:39 -0700, Rich Jordan <jordan@ccs4vms.com> wrote: > My DS10L is misbehaving. > > > DS10L 6/466, 1GB memory (compatible, not Compaq), KZPBA SCSIF > controller, Quantum Viking II 4.5GB UWSE drive, attached BA364 tower$ > with two UWSE disks and one CDROM. > % > OpenVMS V8.2, patched up to 6/2006.  > F > The system ran completely reliably  for months.  About 3 weeks ago II > found it not responding, though I could still get into the RMC from the A > console.  Rebooted, ran about a week then hung again.  Each run G > interval got shorter until now it will sometimes stay up less than 30 F > minutes after a long powerdown; a simple powercycle will usually not0 > get all the way through a boot before hanging. > G > A temp monitor run when VMS is up shows system temp at 26C after long C > powerdown, and peaking at 40C during mid day when the room is the F > warmest.  The system has plenty of open ventilation space, and has a< > small desk fan blowing at its front panel (unchanged since > installation). > H > There are no error messages, no crash, no log entries.  It just stops.I > RMC is still accessible, but performing a reset has only worked once to C > reboot; usually a power cycle is required to even get to the SRM. H > Front panel HALT, console BREAK and console CTRL-P do nothing when itsF > in this state (CTRL-P definitely halts the system when its operating > normally). > B > I had one test where I was in SRM, not having booted yet, and itC > stopped responding, so I'm pretty sure its system hardware, not a E > problem with the SCSI controller or drives.  Further testing needed   > there since it was a one-shot. >  > Work done so far: , > - ran without external drives (still hung)I > - pulled and reseated SCSI card and memory (both risers and DIMMs) with  > static protection H > - removed heat sink to verify (after seeing pics of overheated units);H > the CPU and heat sink looked fine but the grafoil pad was wrinkled andH > twisted around the studs a bit.  I pulled it, cleaned the surfaces and- > used a high quality thermal grease instead. I > - verified all fans are operating, and the little flexible air dams are  > properly located.  > H > The system is still hanging after a short period of operations.  Still > no errors or log messages. > H > Next test will be to pull the SCSI controller and just play with it atI > SRM level to see what happens.  The POST does pass, a memory test seems H > to pass.  No memory errors were logged at VMS level.  I'll run as many6 > tests as possible at the SRM level as soon as I can. > I > This is a hobby system, so no support contract.  Thanks for any advice.  >  > Rich >  >   E If you're anywhere in the vicinity of southeastern Pennsylvania, drop  me an email    WWWebb   ------------------------------   Date: 2 Nov 2006 17:05:25 -0800  From: bob.birch@gmail.com 4 Subject: Re: DS10L hanging problem, tracking it downB Message-ID: <1162515925.490239.12590@m73g2000cwd.googlegroups.com>   Rich Jordan wrote: > JF Mezei wrote:  > > Rich Jordan wrote:G > > > ends up warmer outside, I might also turn on the A/C and duct air K > > > directly to the DS10 inlets.  This is the last shot, may as well do a B > > > thorough job of it before spending real money to replace it. > > L > > If it is an not-obvious problem like a failed fan etc, then the odds areJ > > that one or more chips are getting near their death and any warming upG > > causes flakieness. If the only way to keep the system running is to N > > constantly pour liquid nitrogen on the unit, it may satisfy your hobbyists9 > > curiosity, but it wouldn't result in a viable system.  > > K > > Note: Once my old mac failed to boot. Not even some sad mac icon on the L > > display. Pretty scary. Turns out that removing the CPU daughterboard andN > > putting it back in fixed the problem. No idea what would have caused it toN > > fail between the time I turned it off and the time I turned it back on theM > > next day. Random glitch caused by the alignement of Venus with regards to  > > Alpha Centuri. > > M > > So sometimes, there are things that can be done to "permanently" fix some 0 > > glitch. But other times, it is a lost cause. > F > A powercycle is by far the most common time for hardware failures toC > occur.  Could be thermal, could be a capacitor or other component G > failing when the output voltage upsurges, etc.  Machines left on tend F > to last longest.  unfortunately this DS10L was left on all the time, > but failed anyway. > H > Not a fan, and all removable parts/cables/connectors have been removedH > and reseated.  The cold test will tell the rest, although even if thatF > passes its still time to replace the box.  Its just nice to know the > cause when things break.  3 One 1960/70's era troubleshooting technique mite be 3 usable here, depends on the layout of the CPU board  and the size of your gonads ;-) 6 If it's a heat related problem sometime you can touch,: with your finger, the surface of the plastic, ceramic chip6 package, NOT THE LEADS, to see if it is really getting5 hot. Sometimes you will see other chips with the same 6 part# or in the say area and you can judiciously touch4 one, than the other and compare the temp. A low cost tester. 4 We use to put the boards on extenders and touch each8 chip with the tip of the nose (most sensitive) to detect3 chips, components running hot. The customer used to 6 think you lost it, but it used to work as a technique. Mite work here. 6 Ditto for shock sensitive components, where we wud tap7 lightly with the plastic end of a screwdriver to see if 5 we cud induce an error. Tap judiciously, lighly, than , heavier at different parts of the box/board.5 Try anything, if the alternative is swapping the box.    >  > Rich   ------------------------------  $ Date: Thu, 2 Nov 2006 20:15:33 -0500/ From: "William Webb" <william.w.webb@gmail.com> 4 Subject: Re: DS10L hanging problem, tracking it downI Message-ID: <8660a3a10611021715m72fc6da3jff76cf4abf10f1ff@mail.gmail.com>   N On 2 Nov 2006 17:05:25 -0800, bob.birch@gmail.com <bob.birch@gmail.com> wrote: > Rich Jordan wrote: > > JF Mezei wrote:  > > > Rich Jordan wrote:I > > > > ends up warmer outside, I might also turn on the A/C and duct air M > > > > directly to the DS10 inlets.  This is the last shot, may as well do a D > > > > thorough job of it before spending real money to replace it. > > > N > > > If it is an not-obvious problem like a failed fan etc, then the odds areL > > > that one or more chips are getting near their death and any warming upI > > > causes flakieness. If the only way to keep the system running is to P > > > constantly pour liquid nitrogen on the unit, it may satisfy your hobbyists; > > > curiosity, but it wouldn't result in a viable system.  > > > M > > > Note: Once my old mac failed to boot. Not even some sad mac icon on the N > > > display. Pretty scary. Turns out that removing the CPU daughterboard andP > > > putting it back in fixed the problem. No idea what would have caused it toP > > > fail between the time I turned it off and the time I turned it back on theO > > > next day. Random glitch caused by the alignement of Venus with regards to  > > > Alpha Centuri. > > > O > > > So sometimes, there are things that can be done to "permanently" fix some 2 > > > glitch. But other times, it is a lost cause. > > H > > A powercycle is by far the most common time for hardware failures toE > > occur.  Could be thermal, could be a capacitor or other component I > > failing when the output voltage upsurges, etc.  Machines left on tend H > > to last longest.  unfortunately this DS10L was left on all the time, > > but failed anyway. > > J > > Not a fan, and all removable parts/cables/connectors have been removedJ > > and reseated.  The cold test will tell the rest, although even if thatH > > passes its still time to replace the box.  Its just nice to know the > > cause when things break. > 5 > One 1960/70's era troubleshooting technique mite be 5 > usable here, depends on the layout of the CPU board ! > and the size of your gonads ;-) 8 > If it's a heat related problem sometime you can touch,< > with your finger, the surface of the plastic, ceramic chip8 > package, NOT THE LEADS, to see if it is really getting7 > hot. Sometimes you will see other chips with the same 8 > part# or in the say area and you can judiciously touch6 > one, than the other and compare the temp. A low cost	 > tester. 6 > We use to put the boards on extenders and touch each: > chip with the tip of the nose (most sensitive) to detect5 > chips, components running hot. The customer used to 8 > think you lost it, but it used to work as a technique. > Mite work here. 8 > Ditto for shock sensitive components, where we wud tap9 > lightly with the plastic end of a screwdriver to see if 7 > we cud induce an error. Tap judiciously, lighly, than . > heavier at different parts of the box/board.7 > Try anything, if the alternative is swapping the box.  >  > >  > > Rich >  >   = I trust you didn't use this method with flyback transformers.   2 If you did, please post a picture of you new "do".   WWWebb WWWebb   ------------------------------  % Date: Thu, 02 Nov 2006 21:01:27 -0500 ' From: Dave Froble <davef@tsoft-inc.com> 4 Subject: Re: DS10L hanging problem, tracking it down9 Message-ID: <H8ednduK9eetPtfYnZ2dnUVZ_o2dnZ2d@libcom.com>    Rich Jordan wrote:  A > When VMS is running there's a job recording system temp every 5 
 > minutes.  G How are you getting the temperature while VMS is running?  I've looked  0 for a lexical function and didn't find anything.   --  4 David Froble                       Tel: 724-529-0450> Dave Froble Enterprises, Inc.      E-Mail: davef@tsoft-inc.com DFE Ultralights, Inc.  170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------   Date: 2 Nov 2006 18:30:47 -0800  From: bob.birch@gmail.com 4 Subject: Re: DS10L hanging problem, tracking it downC Message-ID: <1162521047.277417.234960@b28g2000cwb.googlegroups.com>    William Webb wrote: P > On 2 Nov 2006 17:05:25 -0800, bob.birch@gmail.com <bob.birch@gmail.com> wrote: > > Rich Jordan wrote: > > > JF Mezei wrote:  > > > > Rich Jordan wrote:K > > > > > ends up warmer outside, I might also turn on the A/C and duct air O > > > > > directly to the DS10 inlets.  This is the last shot, may as well do a F > > > > > thorough job of it before spending real money to replace it. > > > > P > > > > If it is an not-obvious problem like a failed fan etc, then the odds areN > > > > that one or more chips are getting near their death and any warming upK > > > > causes flakieness. If the only way to keep the system running is to R > > > > constantly pour liquid nitrogen on the unit, it may satisfy your hobbyists= > > > > curiosity, but it wouldn't result in a viable system.  > > > > O > > > > Note: Once my old mac failed to boot. Not even some sad mac icon on the P > > > > display. Pretty scary. Turns out that removing the CPU daughterboard andR > > > > putting it back in fixed the problem. No idea what would have caused it toR > > > > fail between the time I turned it off and the time I turned it back on theQ > > > > next day. Random glitch caused by the alignement of Venus with regards to  > > > > Alpha Centuri. > > > > Q > > > > So sometimes, there are things that can be done to "permanently" fix some 4 > > > > glitch. But other times, it is a lost cause. > > > J > > > A powercycle is by far the most common time for hardware failures toG > > > occur.  Could be thermal, could be a capacitor or other component K > > > failing when the output voltage upsurges, etc.  Machines left on tend J > > > to last longest.  unfortunately this DS10L was left on all the time, > > > but failed anyway. > > > L > > > Not a fan, and all removable parts/cables/connectors have been removedL > > > and reseated.  The cold test will tell the rest, although even if thatJ > > > passes its still time to replace the box.  Its just nice to know the > > > cause when things break. > > 7 > > One 1960/70's era troubleshooting technique mite be 7 > > usable here, depends on the layout of the CPU board # > > and the size of your gonads ;-) : > > If it's a heat related problem sometime you can touch,> > > with your finger, the surface of the plastic, ceramic chip: > > package, NOT THE LEADS, to see if it is really getting9 > > hot. Sometimes you will see other chips with the same : > > part# or in the say area and you can judiciously touch8 > > one, than the other and compare the temp. A low cost > > tester. 8 > > We use to put the boards on extenders and touch each< > > chip with the tip of the nose (most sensitive) to detect7 > > chips, components running hot. The customer used to : > > think you lost it, but it used to work as a technique. > > Mite work here. : > > Ditto for shock sensitive components, where we wud tap; > > lightly with the plastic end of a screwdriver to see if 9 > > we cud induce an error. Tap judiciously, lighly, than 0 > > heavier at different parts of the box/board.9 > > Try anything, if the alternative is swapping the box.  > >  > > > 
 > > > Rich > >  > >  > ? > I trust you didn't use this method with flyback transformers.  > 4 > If you did, please post a picture of you new "do".  9 Low voltage, low current was key. Flybacks were generally 4 cheap and readily available, not "nose friendly" ;-)   >  > WWWebb > WWWebb   ------------------------------   Date: 2 Nov 2006 20:24:00 -0800  From: bob.birch@gmail.com 4 Subject: Re: DS10L hanging problem, tracking it downB Message-ID: <1162527839.886547.79600@f16g2000cwb.googlegroups.com>   Jeff Campbell wrote: > bob.birch@gmail.com wrote: > > Rich Jordan wrote: > >> JF Mezei wrote: > >>> Rich Jordan wrote:H > >>>> ends up warmer outside, I might also turn on the A/C and duct airL > >>>> directly to the DS10 inlets.  This is the last shot, may as well do aC > >>>> thorough job of it before spending real money to replace it. N > >>> If it is an not-obvious problem like a failed fan etc, then the odds areL > >>> that one or more chips are getting near their death and any warming upI > >>> causes flakieness. If the only way to keep the system running is to P > >>> constantly pour liquid nitrogen on the unit, it may satisfy your hobbyists; > >>> curiosity, but it wouldn't result in a viable system.  > >>> M > >>> Note: Once my old mac failed to boot. Not even some sad mac icon on the N > >>> display. Pretty scary. Turns out that removing the CPU daughterboard andP > >>> putting it back in fixed the problem. No idea what would have caused it toP > >>> fail between the time I turned it off and the time I turned it back on theO > >>> next day. Random glitch caused by the alignement of Venus with regards to  > >>> Alpha Centuri. > >>> O > >>> So sometimes, there are things that can be done to "permanently" fix some 2 > >>> glitch. But other times, it is a lost cause.I > >> A powercycle is by far the most common time for hardware failures to F > >> occur.  Could be thermal, could be a capacitor or other componentJ > >> failing when the output voltage upsurges, etc.  Machines left on tendI > >> to last longest.  unfortunately this DS10L was left on all the time,  > >> but failed anyway.  > >>K > >> Not a fan, and all removable parts/cables/connectors have been removed K > >> and reseated.  The cold test will tell the rest, although even if that I > >> passes its still time to replace the box.  Its just nice to know the  > >> cause when things break.  > > 7 > > One 1960/70's era troubleshooting technique mite be 7 > > usable here, depends on the layout of the CPU board # > > and the size of your gonads ;-) : > > If it's a heat related problem sometime you can touch,> > > with your finger, the surface of the plastic, ceramic chip: > > package, NOT THE LEADS, to see if it is really getting9 > > hot. Sometimes you will see other chips with the same : > > part# or in the say area and you can judiciously touch8 > > one, than the other and compare the temp. A low cost > > tester.  > + > Ah, yes! The old 5 second rule!  8-)  8-)   + Sometimes these youngsters need some senior , guidance on fixing tough problems and a good3 bit of it is attitude, creating a physic connection  to become one with the machine.   + Even though you have no spares, training or - test equipment or help (ideal conditions) you 1 can still fix it by creating the right enviroment 
 and attitude.   . I used to watch the tech's during the midnight2 shift at Ampex, on the left coast, tweeking a room0 full of AVR-2 video recorders with "Deep Throat"3 as the alignment tape and Etta James blaring in the 1 background " I got my Mo-Jo working...but it just / won't work on you....going down to Louisana..." + Man that's the enviroment you need on tough . problems, no swapping out, you fix everything.$ But then again that was the 60's....  + Well sorry got to go, here at the home it's 0 Lights out at 1130PM after the milk and cookies.     > 8 > > We use to put the boards on extenders and touch each< > > chip with the tip of the nose (most sensitive) to detect7 > > chips, components running hot. The customer used to : > > think you lost it, but it used to work as a technique. > > Mite work here. : > > Ditto for shock sensitive components, where we wud tap; > > lightly with the plastic end of a screwdriver to see if 9 > > we cud induce an error. Tap judiciously, lighly, than 0 > > heavier at different parts of the box/board.9 > > Try anything, if the alternative is swapping the box.  > > 	 > >> Rich  > >  > S > ----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==---- U > http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups M > ----= East and West-Coast Server Farms - Total Privacy via Encryption =----    ------------------------------  $ Date: Thu, 2 Nov 2006 23:48:46 -0500/ From: "William Webb" <william.w.webb@gmail.com> 4 Subject: Re: DS10L hanging problem, tracking it downH Message-ID: <8660a3a10611022048rfd63286m8227a4aa74afbe40@mail.gmail.com>  ( ------=_Part_20412_1165220.1162529326653; Content-Type: text/plain; charset=ISO-8859-1; format=flowed  Content-Transfer-Encoding: 7bit  Content-Disposition: inline   G On 2 Nov 2006 20:24:00 -0800, bob.birch@gmail.com <bob.birch@gmail.com>  wrote: >  > Jeff Campbell wrote: > > bob.birch@gmail.com wrote: > > > Rich Jordan wrote: > > >> JF Mezei wrote: > > >>> Rich Jordan wrote:J > > >>>> ends up warmer outside, I might also turn on the A/C and duct airL > > >>>> directly to the DS10 inlets.  This is the last shot, may as well do > a E > > >>>> thorough job of it before spending real money to replace it. L > > >>> If it is an not-obvious problem like a failed fan etc, then the odds > are K > > >>> that one or more chips are getting near their death and any warming  > upK > > >>> causes flakieness. If the only way to keep the system running is to H > > >>> constantly pour liquid nitrogen on the unit, it may satisfy your > hobbyists = > > >>> curiosity, but it wouldn't result in a viable system.  > > >>> K > > >>> Note: Once my old mac failed to boot. Not even some sad mac icon on  > the L > > >>> display. Pretty scary. Turns out that removing the CPU daughterboard > and L > > >>> putting it back in fixed the problem. No idea what would have caused > it to K > > >>> fail between the time I turned it off and the time I turned it back  > on theF > > >>> next day. Random glitch caused by the alignement of Venus with > regards to > > >>> Alpha Centuri. > > >>> L > > >>> So sometimes, there are things that can be done to "permanently" fix > some4 > > >>> glitch. But other times, it is a lost cause.K > > >> A powercycle is by far the most common time for hardware failures to H > > >> occur.  Could be thermal, could be a capacitor or other componentL > > >> failing when the output voltage upsurges, etc.  Machines left on tendK > > >> to last longest.  unfortunately this DS10L was left on all the time,  > > >> but failed anyway.  > > >>E > > >> Not a fan, and all removable parts/cables/connectors have been 	 > removed H > > >> and reseated.  The cold test will tell the rest, although even if > thatK > > >> passes its still time to replace the box.  Its just nice to know the  > > >> cause when things break.  > > > 9 > > > One 1960/70's era troubleshooting technique mite be 9 > > > usable here, depends on the layout of the CPU board % > > > and the size of your gonads ;-) < > > > If it's a heat related problem sometime you can touch,@ > > > with your finger, the surface of the plastic, ceramic chip< > > > package, NOT THE LEADS, to see if it is really getting; > > > hot. Sometimes you will see other chips with the same < > > > part# or in the say area and you can judiciously touch: > > > one, than the other and compare the temp. A low cost
 > > > tester.  > > - > > Ah, yes! The old 5 second rule!  8-)  8-)  > - > Sometimes these youngsters need some senior . > guidance on fixing tough problems and a good5 > bit of it is attitude, creating a physic connection ! > to become one with the machine.  > - > Even though you have no spares, training or / > test equipment or help (ideal conditions) you 3 > can still fix it by creating the right enviroment  > and attitude.  > 0 > I used to watch the tech's during the midnight4 > shift at Ampex, on the left coast, tweeking a room2 > full of AVR-2 video recorders with "Deep Throat"5 > as the alignment tape and Etta James blaring in the 3 > background " I got my Mo-Jo working...but it just 1 > won't work on you....going down to Louisana..." - > Man that's the enviroment you need on tough 0 > problems, no swapping out, you fix everything.& > But then again that was the 60's.... > - > Well sorry got to go, here at the home it's 2 > Lights out at 1130PM after the milk and cookies. >  >  > > : > > > We use to put the boards on extenders and touch each> > > > chip with the tip of the nose (most sensitive) to detect9 > > > chips, components running hot. The customer used to < > > > think you lost it, but it used to work as a technique. > > > Mite work here. < > > > Ditto for shock sensitive components, where we wud tap= > > > lightly with the plastic end of a screwdriver to see if ; > > > we cud induce an error. Tap judiciously, lighly, than 2 > > > heavier at different parts of the box/board.; > > > Try anything, if the alternative is swapping the box.  > > >  > > >> Rich  > > >  > > J > > ----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet > News==----L > > http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ > NewsgroupsI > > ----= East and West-Coast Server Farms - Total Privacy via Encryption  > =----  >  >  Good tech support story-  D The site where I cut my VMS teeth had an LP27 line printer- it beganK throwing error 38s, I believe- something like band gate board sensor out of J range- and there was no discernable pattern to when it would throw errors.  I The Bell Atlantic guy adjusted the sensors, scoped boards, swapped boards J out-- it got to the point where they started coming in in pairs to work on this beast-   L And just at the point where they were going to throw in the towel and get meH a "new" printer, one of the field service guys slammed the gate a little hard- and it threw an error.  H This led to them carefully testing the wiring harness that went from theE frame of the printer around the hinge of the gate to the gate itself-   L It turns out there was a break in one of the wires in the harness, sometimes* it was closed, and sometimes it opened up.  = But those guys just about went crazy trying to figure it out.    WWWebb  ( ------=_Part_20412_1165220.1162529326653+ Content-Type: text/html; charset=ISO-8859-1  Content-Transfer-Encoding: 7bit  Content-Disposition: inline    <br><br><div><span class="gmail_quote">On 2 Nov 2006 20:24:00 -0800, <b class="gmail_sendername"><a href="mailto:bob.birch@gmail.com">bob.birch@gmail.com</a></b> &lt;<a href="mailto:bob.birch@gmail.com">bob.birch@gmail.com </a>&gt; wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Jeff Campbell wrote:<br>&gt; <a href="mailto:bob.birch@gmail.com">bob.birch@gmail.com8</a> wrote:<br>&gt; &gt; Rich Jordan wrote:<br>&gt; &gt;&gt; JF Mezei wrote:<br>&gt; &gt;&gt;&gt; Rich Jordan wrote:<br>&gt; &gt;&gt;&gt;&gt; ends up warmer outside, I might also turn on the A/C and duct air<br>&gt; &gt;&gt;&gt;&gt; directly to the DS10 inlets.&nbsp;&nbsp;This is the last shot, may as well do a<br>&gt; &gt;&gt;&gt;&gt; thorough job of it before spending real money to replace it.<br>&gt; &gt;&gt;&gt; If it is an not-obvious problem like a failed fan etc, then the odds are<br>&gt; &gt;&gt;&gt; that one or more chips are getting near their death and any warming up<br>&gt; &gt;&gt;&gt; causes flakieness. If the only way to keep the system running is to<br>&gt; &gt;&gt;&gt; constantly pour liquid nitrogen on the unit, it may satisfy your hobbyists<br>&gt; &gt;&gt;&gt; curiosity, but it wouldn't result in a viable system.0<br>&gt; &gt;&gt;&gt;<br>&gt; &gt;&gt;&gt; Note: Once my old mac failed to boot. Not even some sad mac icon on the<br>&gt; &gt;&gt;&gt; display. Pretty scary. Turns out that removing the CPU daughterboard and<br>&gt; &gt;&gt;&gt; putting it back in fixed the problem. No idea what would have caused it to <br>&gt; &gt;&gt;&gt; fail between the time I turned it off and the time I turned it back on the<br>&gt; &gt;&gt;&gt; next day. Random glitch caused by the alignement of Venus with regards to<br>&gt; &gt;&gt;&gt; Alpha Centuri. <br>&gt; &gt;&gt;&gt;<br>&gt; &gt;&gt;&gt; So sometimes, there are things that can be done to &quot;permanently&quot; fix some<br>&gt; &gt;&gt;&gt; glitch. But other times, it is a lost cause.<br>&gt; &gt;&gt; A powercycle is by far the most common time for hardware failures to<br>&gt; &gt;&gt; occur.&nbsp;&nbsp;Could be thermal, could be a capacitor or other component<br>&gt; &gt;&gt; failing when the output voltage upsurges, etc.&nbsp;&nbsp;Machines left on tend<br>&gt; &gt;&gt; to last longest.&nbsp;&nbsp;unfortunately this DS10L was left on all the time, <br>&gt; &gt;&gt; but failed anyway.<br>&gt; &gt;&gt;<br>&gt; &gt;&gt; Not a fan, and all removable parts/cables/connectors have been removed<br>&gt; &gt;&gt; and reseated.&nbsp;&nbsp;The cold test will tell the rest, although even if that <br>&gt; &gt;&gt; passes its still time to replace the box.&nbsp;&nbsp;Its just nice to know the<br>&gt; &gt;&gt; cause when things break.<br>&gt; &gt;<br>&gt; &gt; One 1960/70's era troubleshooting technique mite be<br>&gt; &gt; usable here, depends on the layout of the CPU board  <br>&gt; &gt; and the size of your gonads ;-)<br>&gt; &gt; If it's a heat related problem sometime you can touch,<br>&gt; &gt; with your finger, the surface of the plastic, ceramic chip<br>&gt; &gt; package, NOT THE LEADS, to see if it is really getting  <br>&gt; &gt; hot. Sometimes you will see other chips with the same<br>&gt; &gt; part# or in the say area and you can judiciously touch<br>&gt; &gt; one, than the other and compare the temp. A low cost<br>&gt; &gt; tester.<br>&gt;<br>&gt; Ah, yes! The old 5 second rule!&nbsp;&nbsp;8-)&nbsp;&nbsp;8-)<br><br>Sometimes these youngsters need some senior<br>guidance on fixing tough problems and a good<br>bit of it is attitude, creating a physic connection<br>to become one with the machine.  <br><br>Even though you have no spares, training or<br>test equipment or help (ideal conditions) you<br>can still fix it by creating the right enviroment<br>and attitude.<br><br>I used to watch the tech's during the midnight <br>shift at Ampex, on the left coast, tweeking a room<br>full of AVR-2 video recorders with &quot;Deep Throat&quot;<br>as the alignment tape and Etta James blaring in the<br>background &quot; I got my Mo-Jo working...but it just  <br>won't work on you....going down to Louisana...&quot;<br>Man that's the enviroment you need on tough<br>problems, no swapping out, you fix everything.<br>But then again that was the 60's....<br><br>Well sorry got to go, here at the home it's
<br>Lights out at 1130PM after the milk and cookies.<br><br><br>&gt;<br>&gt; &gt; We use to put the boards on extenders and touch each<br>&gt; &gt; chip with the tip of the nose (most sensitive) to detect<br>&gt; &gt; chips, components running hot. The customer used to  <br>&gt; &gt; think you lost it, but it used to work as a technique.<br>&gt; &gt; Mite work here.<br>&gt; &gt; Ditto for shock sensitive components, where we wud tap<br>&gt; &gt; lightly with the plastic end of a screwdriver to see if <br>&gt; &gt; we cud induce an error. Tap judiciously, lighly, than<br>&gt; &gt; heavier at different parts of the box/board.<br>&gt; &gt; Try anything, if the alternative is swapping the box.<br>&gt; &gt;<br>&gt; &gt;&gt; Rich  <br>&gt; &gt;<br>&gt;<br>&gt; ----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----<br>&gt; <a href="http://www.newsfeeds.com">http://www.newsfeeds.com</a> The #1 Newsgroup Service in the World! 120,000+ Newsgroupso<br>&gt; ----= East and West-Coast Server Farms - Total Privacy via Encryption =----<br><br></blockquote></div><br>Good tech support story- <br><br>The site where I cut my VMS teeth had an LP27 line printer- it began throwing error 38s, I believe- something like band gate board sensor out of range- and there was no discernable pattern to when it would throw errors. l<br><br>The Bell Atlantic guy adjusted the sensors, scoped boards, swapped boards out-- it got to the point where they started coming in in pairs to work on this beast-<br><br>And just at the point where they were going to throw in the towel and get me a &quot;new&quot; printer, one of the field service guys slammed the gate a little hard- and it threw an error.<br><br>This led to them carefully testing the wiring harness that went from the frame of the printer around the hinge of the gate to the gate itself- <br><br>It turns out there was a break in one of the wires in the harness, sometimes it was closed, and sometimes it opened up.&nbsp;  g <br><br>But those guys just about went crazy trying to figure it out.<br><br>WWWebb<br clear="all"><br>   * ------=_Part_20412_1165220.1162529326653--   ------------------------------  % Date: Thu, 02 Nov 2006 14:08:34 -0500 8 From: Stephen Hoffman <Hoff@HoffmanLabs-RemoveThis-.Org>< Subject: Re: Itanium model numbers {re F$getsyi("HW_MODEL")}) Message-ID: <eidfnj$1kr7$1@pyrite.mv.net>     VAXman- @SendSpamHere.ORG wrote:A > Will all Itanium boxes return 4096 when requesting the HW_MODEL  > as per F$getsyi("HW_MODEL")? > A > I've noted the EXE$GW_CPUTYPE is always zero on Itaniums I have @ > been on so I'm assuming the $GETSYI just feigns this valeu for > all Itaniums.   F    Integrity servers don't have a hardware model code stored anywhere G I'm aware of (and I looked for it back during the IA-64 port), as that  F is simply not something that is available within the standard Itanium H platform.  Identifying the particulars of the box during bootstrap (for B baseline I/O and system configuration) obviously required various C changes, as did the licensing -- this is related to the per-core-,  H per-processor- and per-socket-related changes that arose within OpenVMS C I64.  There is a model text string in the platform, but there's no  I standard format for the name stored in same; it's a text string intended   for display.   ------------------------------  # Date: Thu, 02 Nov 2006 19:51:57 GMT " From:   VAXman-  @SendSpamHere.ORG< Subject: Re: Itanium model numbers {re F$getsyi("HW_MODEL")}0 Message-ID: <00A5E216.E87B2FF3@SendSpamHere.ORG>  d In article <eidfnj$1kr7$1@pyrite.mv.net>, Stephen Hoffman <Hoff@HoffmanLabs-RemoveThis-.Org> writes: >  > ! >VAXman- @SendSpamHere.ORG wrote: B >> Will all Itanium boxes return 4096 when requesting the HW_MODEL >> as per F$getsyi("HW_MODEL")?  >>  B >> I've noted the EXE$GW_CPUTYPE is always zero on Itaniums I haveA >> been on so I'm assuming the $GETSYI just feigns this valeu for  >> all Itaniums. > G >   Integrity servers don't have a hardware model code stored anywhere  H >I'm aware of (and I looked for it back during the IA-64 port), as that G >is simply not something that is available within the standard Itanium  I >platform.  Identifying the particulars of the box during bootstrap (for  C >baseline I/O and system configuration) obviously required various  D >changes, as did the licensing -- this is related to the per-core-, I >per-processor- and per-socket-related changes that arose within OpenVMS  D >I64.  There is a model text string in the platform, but there's no J >standard format for the name stored in same; it's a text string intended 
 >for display.   B Are you saying that my "HP rx2600  (1.50GHz/6.0MB)" returned from   , $ WRITE SYS$OUTPUT "''F$getsyi("HW_NAME")'"   B could change from release to release of VMS?  I know the amount of8 memory show could but that's not dependent upon OS rev.    --  K VAXman- A Bored Certified VMS Kernel Mode Hacker   VAXman(at)TMESIS(dot)COM              5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------  $ Date: Thu, 2 Nov 2006 15:41:34 -0500 From: norm.raphael@metso.com7 Subject: Re: Login Break-in LGI parameters (Resolution) Q Message-ID: <OFC98CB614.B0D933EE-ON8525721A.00711B1B-8525721A.0071AB5C@metso.com>   ; With the help of others (TNX), this is how it really works.  ==  F Users get LGI_RET_LIM chances to login successfully before the current connection drops.   E Users get LGI_RETRY_TMO after entering a Username to enter a Password $ before the current connection drops.  H Users get LGI_BRK_LIM login failures before being blocked as an INTRUDERK (The failure count is logged in the INTRUSION entry even though lockout has  yet to occur).  J Once the break-in limit is reached,  LGI_HID_TIM is multiplied by a randomG value between 1 and 1.5 and that value is added to the time of the next D failure and stored in the INTRUSION Expiration value and the type isJ changed from SUSPECT to INTRUDER.  The source is prevented from login even> with the correct Username and Password until this time passes.  I Monitoring of login failure then continues until the INTRUSION Expiration K time is reached.  For each subsequent invalid login, the INTRUSION count is H incremented and the INTRUSION Expiration time is recomputed as above andI stored as the new INTRUSION Expiration time. A valid login attempt during K this lockout period only increments the INTRUSION count; it does not change K the INTRUSION Expiration Time.  Once the Expiration time has passed with no ? more invalid login attempts, the INTRUSION record is discarded.   J That means that from that source, a user cannot achieve a successful loginE until (LGI_HID_TIM * [1-1.5]) from the time of the last invalid login : attempt, a time which can be seen in the INTRUSION record.  7 norm.raphael@metso.com wrote on 11/01/2006 01:10:00 PM:    >  > F > peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) wrote on 11/01/2006 > 12:13:46 PM: > A > > In article <OF80E79240.FB9FD36D-ON85257219.0054EE54-85257219. 7 > > 0055D0F3@metso.com>, norm.raphael@metso.com writes: C > > >Can someone please expain in simple English what happens here.  > > I > > Use the default values of the mentioned SYSGEN parameters and you see  > > your arguments vanish... > > H > > Someone did introduce some illogic into your parameters. Change them back > ;-)  > D > I don't see that at all.  Break timeout = hidetimeout = 5 mins for default.K > So why have hidetimeout?  Hit the limit, wait 5, fail again, wait 5 more. K > The hidetimeout is now expired, and unless you tack it onto the end after  > all J > failure timeouts are done, in which case it's additive, it does nothing. > F > If I am misconstruing, please show how the defaults would work for 7 > failures.  > J > P.S.  The SOX boys want the time to regain valid login to be 30 minutes. > How would you do that? >  > >  > > -- > > Peter "EPLAN" LANGSTOEGER ) > > Network and OpenVMS system specialist   > > E-mail  peter@langstoeger.atJ > > A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist >    ------------------------------  $ Date: Fri, 3 Nov 2006 06:37:16 +08003 From: "Richard Maher" <maher_rj@hotspamnotmail.com> ' Subject: More COV Self-Help right here! 1 Message-ID: <eidrne$1fj$1@news-02.connect.com.au>   
 Hi Elliot,  D > I think it is sad that someone like Rich could not or chose not toG > survive at hp. It means that hp has sunk much further from the hp way & > than I would have otherwise thought. > F > I felt the same way about DEC and Ken Olsen when the Palmer tendency > got control.   Yeah me too! I need a hug.   Regards Richard Maher   3 "Elliott Roper" <nospam@yrl.co.uk> wrote in message + news:021120062207560759%nospam@yrl.co.uk... A > In article <f159$454a5c94$cef8887a$4837@TEKSAVVY.COM>, JF Mezei ' > <jfmezei.spamnot@teksavvy.com> wrote:  >  > > Elliott Roper wrote: > > > It is still there. > > > 161.114.21.165C > > > h20325.www2.hp.com.  7200  IN CNAME activeanswers.compaq.com.  > > >  > > 0 > > Thanks. You must have it on your DNS caches. > >  > > ! > > $ nslookup h20325.www2.hp.com  > > Server:  VELO.vaxination.ca  > > Address:  10.0.0.11  > > F > > *** VELO.vaxination.ca can't find h20325.www2.hp.com: Non-existent host/domain  > >  > >  > > $ nslookup 161.114.21.165  > > Server:  VELO.vaxination.ca  > > Address:  10.0.0.11  > > $ > > Name:    vzaqswebtest.cce.hp.com > > Address:  161.114.21.165 > >  > >  > > however,G > >  >http://161.114.21.165/blogs/marcello/archive/2006/09/19/1630.html  > >  > > Does work.8 > That's why I put the numeric address in there for you. > > H > > The contents do not surprise me.  Such "peace and love" mentality is neatL > > on the surface, but it does not really foster change because to push forI > > changes, you often end up creating conflicts against those who refuse  change.  > > J > > His peace and love mentality hasn't gotten VMS anywhere within HP. AndL > > while he may have taken the right decisions for his own carreer ( he hasF > > survived 2 mergers and gone up in the ranks during this time), his decisionJ > > to be a quiet sheep that doesn't disturb the shit has also resulted in/ > > nobody fighting for VMS' success within HP.  > > F > > You'd think that at the very least, in his position, he would haveF > > sucessfully argued for a secons renaissance for VMS (eg: marketing budget) I > > since the first time showed that it worked well and gerenated serious  growth.  > G > You can't reach that conclusion without further evidence. He may have : > succeeded spectacularly in avoiding an earlier disaster. > F > I only had one personal contact with him, a long phone call long agoH > when DEC was in its death throes and I wanted the UK management to notI > welsh on a deal I thought I had for 300 low spec alpha VMS workstations F > that were very important to my business. He was OK then. I was a bitI > disappointed with the result, but I did not know how bad things were at  > DEC at the time. > D > I think it is sad that someone like Rich could not or chose not toG > survive at hp. It means that hp has sunk much further from the hp way & > than I would have otherwise thought. > F > I felt the same way about DEC and Ken Olsen when the Palmer tendency > got control. >  > --  3 > To de-mung my e-mail address:- fsnospam$elliott$$ E > PGP Fingerprint: 1A96 3CF7 637F 896B C810  E199 7E5C A9E4 8E59 E248    ------------------------------   Date: 2 Nov 2006 19:50:57 GMT  From: healyzh@aracnet.com = Subject: Re: Observations on Samba for VMS (CIFS Kit from HP) , Message-ID: <eidi7101tnj@enews3.newsguy.com>  0 Paul Sture <paul.sture.nospam@hispeed.ch> wrote:H > I didn't notice any nasties such as overwriting the current versions,  > but an area to check out.   E OK, this is a big one, as I believe that Samba 2 does that just that.   4 > The following is independent of CIFS or Pathworks:  H > I can't say without looking it up whether DCPS supports that printer,  > but OS X certainly does.  K Oddly enough it looks like the 4MP is supported, and just about every model H of the 5 series, except the 5P/5MP.  I suspect I can just tell it that IA have a 4MP, or that I have a Postscript printer, but will have to C experiment.  The first thing I have to do though is switch from the E Ethertalk to Localtalk to a one of the several printservers I've been I collecting over the past 6+ years with the intention of doing this.  Yes, L I've been threatening to make this change for that long, like I said, what I have right now works just fine.   J > An alternative solution is, as long as the Mac can successfully talk to I > your HP, switch on printer sharing on the Mac side, then set up an LPD  G > queue on VMS, aimed at it. Yep, I can print Postscript from VMS even  ' > though my printer doesn't support it.   G This is an option, but it is not one I really want to use.  I went this K route with the new PC I have, as it doesn't seem to want to print via Samba E to the VMS box.  It's a real pain, as the Mac isn't up 24x7.  This is G actually helping to incourage me to move off of Appletalk for printing.   I > In what ways weren't they working well enough for you? I did manage to  K > get at NFS on VMS from the command line on the Mac, but never the Finder  @ > or application programs. I just got used to using FTP instead.  F There was that problem with NFS, and IIRC, with Samba 2, it overwritesJ files when I try to edit them from the Mac.  To make matters worse, (againJ IIRC) each save, deletes the previous version, so the more saves, the more versions you trash.    		Zane   ------------------------------  % Date: Thu, 02 Nov 2006 20:07:07 -0600 3 From: David J Dachtera <djesys.no@spam.comcast.net>  Subject: Re: PIPE question0 Message-ID: <454AA44B.AD512903@spam.comcast.net>   "Bart.Zorn@gmail.com" wrote: > E > In a pipeline segment, SYS$INPUT and SYS$PIPE have, by default, the 
 > same value. > > The explicit definition of SYS$INPUT is therefore redundant!  D ...except in a DCL procedure where SYS$INPUT points to the procedureC and SYS$PIPE points to SYS$OUTPUT of the previous PIPEline segment.    --   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: Thu, 02 Nov 2006 20:07:44 -0600 3 From: David J Dachtera <djesys.no@spam.comcast.net>  Subject: Re: PIPE question/ Message-ID: <454AA470.569B586@spam.comcast.net>    briggs@encompasserve.org wrote:  > h > In article <45495EE8.4A53E066@spam.comcast.net>, David J Dachtera <djesys.no@spam.comcast.net> writes:! > > norm.raphael@metso.com wrote:  > > $ PIPE -< > >       (WRITE SYS$OUTPUT D ; WRITE SYS$OUTPUT "EXIT") | -6 > >       (DEFINE/USER SYS$INPUT SYS$PIPE: ; DATEPROG) > E > It seems rather silly to define SYS$INPUT as SYS$PIPE in a pipeline 4 > since SYS$INPUT is already SYS$PIPE in a pipeline. >  > $ d = "asdf"= > $ pipe ( write sys$output d ; write sys$output "EXIT" ) | - ; > _$     ( show logical sys$pipe ; show logical sys$input ) 8 >    "SYS$PIPE" = "_EISNER$MPA1745:" (LNM$PROCESS_TABLE)9 >    "SYS$INPUT" = "_EISNER$MPA1745:" (LNM$PROCESS_TABLE)   D ...except in a DCL procedure where SYS$INPUT points to the procedureC and SYS$PIPE points to SYS$OUTPUT of the previous PIPEline segment.    --   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: Thu, 02 Nov 2006 14:48:11 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> 8 Subject: Re: Rdb Network Security - Multinet 5.2 Webinar8 Message-ID: <1e260$454a4b7a$cef8887a$25161@TEKSAVVY.COM>   Richard Maher wrote:  # > It's on tomorrow, so sign-up now!  > % > http://www.process.com/webinar.html   I What is a "webinar" ????? Unless they provide the technology platform to  J host their conference BEFORE you sign up, then why bother signing up only I to find out tomorrow when you try to access some link that it requires a  & specific version of windows software ?  K In other words, why block your own time for some conference when you don't  ( even know if you will be able to watch ?   ------------------------------  % Date: Thu, 02 Nov 2006 14:50:26 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> = Subject: Re: Rich Marcello retiring to perform volunteer work 8 Message-ID: <4360b$454a4c00$cef8887a$25161@TEKSAVVY.COM>   Alan Greig wrote: ) > Rich's most recent Blog is interesting  G > http://h20325.www2.hp.com/blogs/marcello/archive/2006/09/19/1630.html     K I am unable to resolve that host name... Can you provide a hint on how one  ; can navigate to his blog ? Or has it been removed already ?   , Could you provide the jist of what he said ?   ------------------------------  % Date: Thu, 02 Nov 2006 20:39:01 +0000 & From: Elliott Roper <nospam@yrl.co.uk>= Subject: Re: Rich Marcello retiring to perform volunteer work 1 Message-ID: <021120062039010642%nospam@yrl.co.uk>   A In article <4360b$454a4c00$cef8887a$25161@TEKSAVVY.COM>, JF Mezei % <jfmezei.spamnot@teksavvy.com> wrote:    > Alan Greig wrote: + > > Rich's most recent Blog is interesting  I > > http://h20325.www2.hp.com/blogs/marcello/archive/2006/09/19/1630.html  >  > M > I am unable to resolve that host name... Can you provide a hint on how one  = > can navigate to his blog ? Or has it been removed already ?  It is still there. 161.114.21.165= h20325.www2.hp.com.  7200  IN CNAME activeanswers.compaq.com.   G A thoughtful piece on how to not escalate violence in your personal and  business life.   --  1 To de-mung my e-mail address:- fsnospam$elliott$$ C PGP Fingerprint: 1A96 3CF7 637F 896B C810  E199 7E5C A9E4 8E59 E248    ------------------------------  % Date: Thu, 02 Nov 2006 16:02:15 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> = Subject: Re: Rich Marcello retiring to perform volunteer work 6 Message-ID: <f159$454a5c94$cef8887a$4837@TEKSAVVY.COM>   Elliott Roper wrote: > It is still there. > 161.114.21.165? > h20325.www2.hp.com.  7200  IN CNAME activeanswers.compaq.com.  >   , Thanks. You must have it on your DNS caches.     $ nslookup h20325.www2.hp.com  Server:  VELO.vaxination.ca  Address:  10.0.0.11   N *** VELO.vaxination.ca can't find h20325.www2.hp.com: Non-existent host/domain     $ nslookup 161.114.21.165  Server:  VELO.vaxination.ca  Address:  10.0.0.11     Name:    vzaqswebtest.cce.hp.com Address:  161.114.21.165     however,C  >http://161.114.21.165/blogs/marcello/archive/2006/09/19/1630.html   
 Does work.  J The contents do not surprise me.  Such "peace and love" mentality is neat I on the surface, but it does not really foster change because to push for  M changes, you often end up creating conflicts against those who refuse change.   G His peace and love mentality hasn't gotten VMS anywhere within HP. And  I while he may have taken the right decisions for his own carreer ( he has  L survived 2 mergers and gone up in the ranks during this time), his decision G to be a quiet sheep that doesn't disturb the shit has also resulted in  + nobody fighting for VMS' success within HP.   C You'd think that at the very least, in his position, he would have  K sucessfully argued for a secons renaissance for VMS (eg: marketing budget)  M since the first time showed that it worked well and gerenated serious growth.    ------------------------------  % Date: Thu, 02 Nov 2006 22:07:56 +0000 & From: Elliott Roper <nospam@yrl.co.uk>= Subject: Re: Rich Marcello retiring to perform volunteer work 1 Message-ID: <021120062207560759%nospam@yrl.co.uk>   ? In article <f159$454a5c94$cef8887a$4837@TEKSAVVY.COM>, JF Mezei % <jfmezei.spamnot@teksavvy.com> wrote:    > Elliott Roper wrote: > > It is still there. > > 161.114.21.165A > > h20325.www2.hp.com.  7200  IN CNAME activeanswers.compaq.com.  > >  > . > Thanks. You must have it on your DNS caches. >  >  > $ nslookup h20325.www2.hp.com  > Server:  VELO.vaxination.ca  > Address:  10.0.0.11  > P > *** VELO.vaxination.ca can't find h20325.www2.hp.com: Non-existent host/domain >  >  > $ nslookup 161.114.21.165  > Server:  VELO.vaxination.ca  > Address:  10.0.0.11  > " > Name:    vzaqswebtest.cce.hp.com > Address:  161.114.21.165 >  > 
 > however,E >  >http://161.114.21.165/blogs/marcello/archive/2006/09/19/1630.html  >  > Does work.6 That's why I put the numeric address in there for you. > L > The contents do not surprise me.  Such "peace and love" mentality is neat K > on the surface, but it does not really foster change because to push for  O > changes, you often end up creating conflicts against those who refuse change.  > I > His peace and love mentality hasn't gotten VMS anywhere within HP. And  K > while he may have taken the right decisions for his own carreer ( he has  N > survived 2 mergers and gone up in the ranks during this time), his decision I > to be a quiet sheep that doesn't disturb the shit has also resulted in  - > nobody fighting for VMS' success within HP.  > E > You'd think that at the very least, in his position, he would have  M > sucessfully argued for a secons renaissance for VMS (eg: marketing budget)  O > since the first time showed that it worked well and gerenated serious growth.   E You can't reach that conclusion without further evidence. He may have 8 succeeded spectacularly in avoiding an earlier disaster.  D I only had one personal contact with him, a long phone call long agoF when DEC was in its death throes and I wanted the UK management to notG welsh on a deal I thought I had for 300 low spec alpha VMS workstations D that were very important to my business. He was OK then. I was a bitG disappointed with the result, but I did not know how bad things were at  DEC at the time.  B I think it is sad that someone like Rich could not or chose not toE survive at hp. It means that hp has sunk much further from the hp way $ than I would have otherwise thought.  D I felt the same way about DEC and Ken Olsen when the Palmer tendency got control.   --  1 To de-mung my e-mail address:- fsnospam$elliott$$ C PGP Fingerprint: 1A96 3CF7 637F 896B C810  E199 7E5C A9E4 8E59 E248    ------------------------------  % Date: Thu, 02 Nov 2006 17:33:49 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> = Subject: Re: Rich Marcello retiring to perform volunteer work 7 Message-ID: <2f147$454a724c$cef8887a$8099@TEKSAVVY.COM>    Elliott Roper wrote:  D > I think it is sad that someone like Rich could not or chose not toG > survive at hp. It means that hp has sunk much further from the hp way & > than I would have otherwise thought.  L Well, I view it as a success that he managed to not only stay so long after K the merger, but also manager to get promoted within that time. My guess is  E that he was a "yes man" to Carly in order to keep his job (and avoid  B conflicts) and that under Hurd, this was not seen as a good thing.  8 If Hurd dislikes "yes man"s , then this is a good thing.  F But despite all that, Marcello is to be credited for saving VMS under 
 Curly/Compaq.    ------------------------------  % Date: Thu, 02 Nov 2006 23:24:52 -0500 ( From: Bill Todd <billtodd@metrocast.net>= Subject: Re: Rich Marcello retiring to perform volunteer work G Message-ID: <uLKdnfSYi6IIWdfYnZ2dnUVZ_v2dnZ2d@metrocastcablevision.com>    Dave Froble wrote: > Bill Todd wrote: > D >> With nothing but Montvale (late next year, though last I knew it E >> wasn't promising much beyond what Montecito already offers, which  G >> barely beats today's x86-64 systems) to carry that wallowing vessel  F >> forward for the next three years, POWER5+ already far ahead of it, I >> POWER6 appearing to be on-track to double that performance next year,  G >> POWER6+ scheduled for 2008, and x86-64 servers apparently likely to  G >> get EV7-style on-chip memory and glueless routing facilities before  H >> Itanic acquires them with Tukwila, his confident statement that "the G >> alliance agreement we have with them [Intel] goes through 2011" may  > >> not have seemed quite as reassuring as it did in September. > 1 > I'm guessing that you're discussing only Intel.   F That's indeed the specific flavor of x86-64 that I was thinking of in H the last item I listed above, but you are of course correct in pointing I out that I did not limit the observation appropriately (and that Opteron  F has had such EV7-style features for about 3.5 years now, highlighting 1 even more Intel's tardiness in introducing them).    ...   J > Don't know how Intel got Conroe to be so fast in comparison to Opteron, & > at least in single core performance,  C Well, being a full process generation ahead helps a lot, you know:  G Intel parts of very similar design were only clocking a little above 2  F GHz in the 90 nm. process size that Opteron still has to make do with.  H Aside from that, Intel recovered remarkably well from its NetBurst (GHz F uber alles!) fiasco:  the Israeli-designed parts that were originally G optimized for low-power (e.g., notebook) applications turned out to be  F very well-suited to more general use at higher power and clock rates, F and saved Intel's ass big-time.  AMD's K8 design, while well-balanced C and well-executed, is in terms of its core largely a 32-bit Athlon  G derivative:  it's not too surprising that a much newer design (without  : the NetBust handicaps) can out-perform it clock-for-clock.  %   but I'm hoping AMD has an answer to  > this.   I Getting to 65 nm. and (IIRC) strained silicon should help (AMD's process  F partner IBM seems to be in fat city with its own 65 nm. process, with F second-silicon POWER6s already running at 4 - 5 GHz - though a lot of I that clock-rate increase reportedly comes from circuit-tweaking):  while  I all I've heard so far about 65 nm. K8 clock rates is minor increases (to  G 3.2 or perhaps 3.4 GHz), the accompanying power figures are rumored to  F be really good, leaving them considerable headroom on that score (or, F alternatively, the ability to run quad-core parts considerably faster * than Intel's can due to heat limitations).  D    The memory controller and glueless routing will keep AMD ahead in; > SMP systems only until Intel gets the same on their CPUs.   H A double-edged sword, since the earlier Intel gets that for its own x86 H processors, the more it will hurt its Itanic prospects (given Tukwila's I delay).  Then again, if Intel has actually already thrown in that towel,   it won't hold it back at all.   G In any event, AMD should remain very competitive in quad-socket-and-up  C SMP systems for the next couple of years (next year their 8-socket  E configurations should become far more generally useful, which brings  E them up to 32-core configurations - not shabby at all), and may come  E back to rough equality in smaller systems if their 'K8L' core design  D turns out well.  In 2008, AMD supposedly goes to glueless 32-socket F capability, just in time to be competitive with Intel's glueless x86s D (one early rumor was that Intel might not scale gluelessly above 16 H sockets, to leave room for its big OEMs to create their own chipsets to F differentiate their large-system products as they currently must with C Itanic, but I'm not sure I buy that strategy as making much sense:  E there's just not that much niche left above 16 quad-core sockets...).   H I actually don't mind at all seeing Intel bounce back with a product as G worthy as its new x86s are - mostly, I hope that both companies thrive  G and keep each other honest, so that fiascoes like Itanic aren't likely  	 to recur.   H An interesting tidbit that I stumbled upon today after having forgotten I about it is that IBM is reportedly designing its (2009?) POWER7 products  I to plug into (presumably high-end) commodity AMD motherboards (reprising  C an old EV6 strategy that never bore fruit:  IIRC that's one of the  D reasons that DEC licensed AMD to use the EV6 bus for Athlon).  This H might be interpreted to suggest that IBM has a *great* deal of faith in  AMD's SMP directions...    - bill   ------------------------------  % Date: Thu, 02 Nov 2006 23:42:23 -0500 ( From: Bill Todd <billtodd@metrocast.net>= Subject: Re: Rich Marcello retiring to perform volunteer work G Message-ID: <feednWrrPtAtVdfYnZ2dnUVZ_rOdnZ2d@metrocastcablevision.com>    Bill Todd wrote:   ...   J > I actually don't mind at all seeing Intel bounce back with a product as I > worthy as its new x86s are - mostly, I hope that both companies thrive  I > and keep each other honest, so that fiascoes like Itanic aren't likely   > to recur.   E Damn - had one more (hardly original, but worth repeating) editorial   comment to make in that vein:   H Can you imagine where we'd be today if Intel hadn't gotten so entangled F in pursuing 'strategy' (rather than just sticking to its knitting and I giving customers what they needed) over the past decade-plus, first with  F Itanic and more recently adding NetBust to the mix?  The mind boggles E (though then again, AMD would now be toast, and Intel would be being  E tempted to be 'strategic' now, with no one left to keep them honest).   H Can you imagine where we'd be today if DEC hadn't made similar mistakes G two decades ago, and instead continued a real horse-race that had kept   IBM on its toes as well?  G I don't believe that we can place *all* the blame for both fiascoes on  H management:  even engineers can be overcome by delusions of grandeur to I the point where they lose sight of the fact that they're really supposed  I to be providing a *service*, not playing Masters of the Universe.  But I  G do believe that we can blame management for fanning such flames rather  : than reining them in and directing them more productively.   - bill   ------------------------------  % Date: Thu, 02 Nov 2006 16:11:19 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> - Subject: Strange CDE overnight disapearance ! 7 Message-ID: <acd70$454a5eb3$cef8887a$5312@TEKSAVVY.COM>   J Woke up this morning to find that the CDE desktop (front panel and window C manager) had disapeared. Other windows remained without any window  
 decorations !   K Thankfully, I had a old Session manager window up, and from there was able  H to start the old MWM window manager and am now back in business with my < session essentially converted from CDE to the old desktop !.  G One notable difference: I still have a CDE file manager window and the  J later is still fully functional and I am able to launch  .COMs located in  that directory !!!!    780021A DTSESSION       LEF  27800224 JFMEZEI$CDE002  LEF 27800226 DECW$TE_0226    COM 27800227 _FTA50:         LEF 2780022A JFMEZEI         LEF 27800131 DECW$SERVER_0   HIB 27800246 _FTA56:         CUR 2780025E VUE$JFMEZEI_2   LEF 27800266 VUE$JFMEZEI_1   LEF 27800267 VUE$JFMEZEI_5   LEF 2780026F DECW$MWM        COM 27800274 JFMEZEI$CDE005  HIB  M So I guess it was just the combi-front panel and window manager that crashed.   L the [.dt]errorlog.; file contains no relevant warnings about this occurance ! that happened while I was asleep.    Os this a common occurance ?  A Does anyone know the DCL command needed to restart the CDE front   panel/window manager ?  J Oh, and by the way,  starting Mozilla interactively on the same node that H runs the display generates a LOT more OPCOM security violation attempts G than when I ran it as a detached process targetted at a different node.    ------------------------------  $ Date: Fri, 3 Nov 2006 06:27:15 +08003 From: "Richard Maher" <maher_rj@hotspamnotmail.com> 1 Subject: The future of Internet Socket Protection 0 Message-ID: <eidr4k$lr$1@news-02.connect.com.au>   Hi Ian,   < Yeah, I can't make it either :-( looks like we're both mugs!  L In the comming days I'll look forward do yet another of your informative and( perceptive distillations on openvms.org.  L Anyway, if in fact not giving the talk, I'm sure Bob Gezelter will be makingI more cameo appearances than "Where's Wally Now?". That should make up for  the rest of us :-)   Cheers Richard Maher  G PS. Can someone please show JF how to use a phone in the next couple of A hours? I'm going out on a limb here, but there is certainly every J possibility that there will be a number to simply dial if you just want to
 listen in?  + "Ian Miller" <ijm@uk2.net> wrote in message = news:1162471238.574615.170490@h48g2000cwc.googlegroups.com... G > Unfourtunally 8pm GMT on a Friday I have other things to do. I expect G > the webinar presentation will appear on their web site sometime after & > the event and I can look at it then. >    ------------------------------  % Date: Fri, 03 Nov 2006 00:41:58 +0800  From: prep@prep.synonet.com $ Subject: Re: Time change questions !0 Message-ID: <87irhx242x.fsf@k9.prep.synonet.com>  + "Neil Rieck" <n.rieck@sympatico.ca> writes:   A > Also be aware of the fact that if you are using NTP to maintain A > clock synchronization, that you may also need to tweak your NTP B > settings. For example, in TCPware you need to set WAYTOOBIG to aE > value smaller than 3600 so that VMS and TCPware don't try to adjust $ > the system clock at the same time.  ? NO, NO, No!! NTP is totally UT. The ONLY `time change' it knows & or care about are the UT leap seconds.  4 It does not care a flying F*** about Summer TIme....   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  $ Date: Thu, 2 Nov 2006 14:35:55 -0500) From: "Neil Rieck" <n.rieck@sympatico.ca> $ Subject: Re: Time change questions !< Message-ID: <454a4812$0$14841$9a6e19ea@news.newshosting.com>  ) <prep@prep.synonet.com> wrote in message  * news:87irhx242x.fsf@k9.prep.synonet.com...- > "Neil Rieck" <n.rieck@sympatico.ca> writes:  > B >> Also be aware of the fact that if you are using NTP to maintainB >> clock synchronization, that you may also need to tweak your NTPC >> settings. For example, in TCPware you need to set WAYTOOBIG to a F >> value smaller than 3600 so that VMS and TCPware don't try to adjust% >> the system clock at the same time.  > A > NO, NO, No!! NTP is totally UT. The ONLY `time change' it knows ( > or care about are the UT leap seconds. > > > Paul Repacholi                               1 Crescent Rd.,9 > +61 (08) 9257-1001                           Kalamunda. A >                                             West Australia 6076   M You are correct about NTP, but wrong in how it interfaces with OpenVMS. I've  A had incorrect settings in the past which have done the following:   I 1) enabled in sysgen, OpenVMS changes the system clock to move in/out of   daylight savings  I 2) a fraction of a second later, NTP (via the stack) notices the time is  0 wrong (by over an hour) and so adjusts it again.  L This is one reason why you must set WAYTOOBIG so that the two modules don't M fight each other. The other alternative is to disable time changes in sysgen  = and only have the stack (TCPware in my case) change the time.   
 Neil Rieck Kitchener/Waterloo/Cambridge,  Ontario, Canada.! http://www3.sympatico.ca/n.rieck/    ------------------------------    Date: 02 Nov 2006 14:56:20 -05003 From: Rich Alderson <news@alderson.users.panix.com> $ Subject: Re: Time change questions !. Message-ID: <mddhcxh1v2z.fsf@panix5.panix.com>  = koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:   > > In article <mdd8xiu4vv1.fsf@panix5.panix.com>, Rich Alderson) > <news@alderson.users.panix.com> writes:     M >> It's too bad that ideas rarely got imported into VMS from the 36-bit side,  >> in this case from Tops-20.   L >> On Tops-20, system time is expressed in days and fractions of a day sinceN >> 11/17/1858 0:00:00 UTC, and time-of-day displays use a simple set of system3 >> calls to present the local timezone to the user.   E >    And in TOPS-20 the FORTRAN libary couldn't handle times near the C >    begining of the year if you ran your system on GMT.  So we ran H >    our system on GMT, but lied to it and claimed we were running it onH >    local time.  (Nobody ever actually looked at the timezone setting.)  H >    I hope this was eventually fixed, but I don't recall ever receivingF >    anything from DEC other than "that's a known bug"; and I reviewed$ >    all the patches they announced.  L That's news to me, but then I didn't do very much Fortran programming on theO -20.  Still, you would think that I'd have heard something about it in the last 	 30 years.    --  L Rich Alderson                                       | /"\ ASCII ribbon     |L news@alderson.users.panix.com                       | \ / campaign against |L "You get what anybody gets. You get a lifetime."    |  x  HTML mail and    |L                          --Death, of the Endless    | / \ postings         |   ------------------------------  % Date: Thu, 02 Nov 2006 15:23:05 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> $ Subject: Re: Time change questions !8 Message-ID: <b4949$454a53a7$cef8887a$28768@TEKSAVVY.COM>   Stephen Hoffman wrote:J >   The time value is directly retrieved through most system services and D > most any of the existing low-level system calls, and directly set J > through, and there are timezone offsets applied via various of the more D > recent UTC-related calls and via libraries such as the C standard  > library.      K If VMS engineers (I was going to say "You guys", but alas, you've been set  L free from the VMS engineering habitat), were to change SYS$GETTIM and other I   official systems services that access the system time to automatically  E add some delta time taken from a logical name, how much would break ?   G Applications that cheat and fetch the quadword from memory would get a  F GMT/UTC time. They wouldn't fail. Applications and DCL would properly ( convert the system time to "local" time.  K You could then code SYS$GETUTC_TIM which would just return the system time  ! without any localised conversion.   I DCL would need an update for all the /AFTER switches to properly convert   the supplied value into GMT.  G And you might then be able to allow a job table logical to define that  K job's time offset to system time. (and if the job logical doesn't exist, a   system logical would be used.   L And for users who cannot cnhange, then they can still set their system time ;   to their local time zone and declare an offset of 00:00 .    ------------------------------  $ Date: Thu, 2 Nov 2006 17:16:13 -0500) From: "Neil Rieck" <n.rieck@sympatico.ca> $ Subject: Re: Time change questions !< Message-ID: <454a6da2$0$14804$9a6e19ea@news.newshosting.com>  ; "JF Mezei" <jfmezei.spamnot@teksavvy.com> wrote in message  2 news:b4949$454a53a7$cef8887a$28768@TEKSAVVY.COM... > Stephen Hoffman wrote:K >>   The time value is directly retrieved through most system services and  E >> most any of the existing low-level system calls, and directly set  K >> through, and there are timezone offsets applied via various of the more  E >> recent UTC-related calls and via libraries such as the C standard   >> library.  >  > M > If VMS engineers (I was going to say "You guys", but alas, you've been set  H > free from the VMS engineering habitat), were to change SYS$GETTIM and A > other official systems services that access the system time to  H > automatically add some delta time taken from a logical name, how much  > would break ?  > I > Applications that cheat and fetch the quadword from memory would get a  H > GMT/UTC time. They wouldn't fail. Applications and DCL would properly * > convert the system time to "local" time. >   M As I said in a previous post, this would be a big job and the modality would  K need to be enabled / disabled as required. If I were doing this, using per  L process commands similar to $SET PROC/PARSE_STYLE and $SET PROC/CASE_LOOKUP 0 would make sense. Then we would be able to do a K SUBMIT/AFTER=time-with-time-zone and be assured of it running only once on   the morning of time transition.    > M > You could then code SYS$GETUTC_TIM which would just return the system time  # > without any localised conversion.  >   L How about just use $BINTIM and $ASCTIM the way they were meant and then use E SYS$ASCUTC, $BINUTC, $GETUTC, $NUMUTC, $TIMCON where they are needed.    > K > DCL would need an update for all the /AFTER switches to properly convert   > the supplied value into GMT. >   K Before swicthes too. Don't forget SET TIME, SHOW TIME and probably quite a  	 few more.   I > And you might then be able to allow a job table logical to define that  M > job's time offset to system time. (and if the job logical doesn't exist, a   > system logical would be used.  > I > And for users who cannot cnhange, then they can still set their system  @ > time to their local time zone and declare an offset of 00:00 .  C Like I already mentioned, SET PROC would default to the old way of  
 operation.  
 Neil Rieck Kitchener/Waterloo/Cambridge,  Ontario, Canada." http://www3.sympatico.ca/n.rieck/    ------------------------------  % Date: Thu, 02 Nov 2006 17:23:00 -0500 3 From: "Richard B. Gilbert" <rgilbert88@comcast.net> $ Subject: Re: Time change questions !: Message-ID: <ksudnRwzdexY8tfYnZ2dnUVZ_rSdnZ2d@comcast.com>   Neil Rieck wrote:   + > <prep@prep.synonet.com> wrote in message  , > news:87irhx242x.fsf@k9.prep.synonet.com... > - >>"Neil Rieck" <n.rieck@sympatico.ca> writes:  >> >>B >>>Also be aware of the fact that if you are using NTP to maintainB >>>clock synchronization, that you may also need to tweak your NTPC >>>settings. For example, in TCPware you need to set WAYTOOBIG to a F >>>value smaller than 3600 so that VMS and TCPware don't try to adjust% >>>the system clock at the same time.  >>A >>NO, NO, No!! NTP is totally UT. The ONLY `time change' it knows ( >>or care about are the UT leap seconds. >>> >>Paul Repacholi                               1 Crescent Rd.,9 >>+61 (08) 9257-1001                           Kalamunda. A >>                                            West Australia 6076  >  > O > You are correct about NTP, but wrong in how it interfaces with OpenVMS. I've  C > had incorrect settings in the past which have done the following:  > K > 1) enabled in sysgen, OpenVMS changes the system clock to move in/out of   > daylight savings > K > 2) a fraction of a second later, NTP (via the stack) notices the time is  2 > wrong (by over an hour) and so adjusts it again. > N > This is one reason why you must set WAYTOOBIG so that the two modules don't O > fight each other. The other alternative is to disable time changes in sysgen  ? > and only have the stack (TCPware in my case) change the time.  >  > Neil Rieck > Kitchener/Waterloo/Cambridge,  > Ontario, Canada.# > http://www3.sympatico.ca/n.rieck/  >  >  > F I don't seem to have this problem with the UCX version of NTP.  I do, B however, use the UTC$CONFIGURE_TDF.COM;1 procedure to change from ) standard to daylight time and back again.    ------------------------------  $ Date: Thu, 2 Nov 2006 17:25:28 -0500) From: "Neil Rieck" <n.rieck@sympatico.ca> $ Subject: Re: Time change questions !< Message-ID: <454a6fcd$0$14820$9a6e19ea@news.newshosting.com>  F "Stephen Hoffman" <Hoff@HoffmanLabs-RemoveThis-.Org> wrote in message # news:eiddqe$1k5s$1@pyrite.mv.net...  > J >   OpenVMS stores a quadword offset from 17-Nov-1858, and does not store L > the system time as or in UTC.  The time value is whatever you choose. (It 5 > could obviously be UTC, if that's what you choose.)  > J >   The time value is directly retrieved through most system services and M > most any of the existing low-level system calls, and directly set through,  H > and there are timezone offsets applied via various of the more recent K > UTC-related calls and via libraries such as the C standard library.  (In  J > some ways, this scheme is the reverse of the way Unix is implemented in I > this area.)  On Unix, the low-level is UTC and the various conversions   > happen higher up.  >    Well stated.   > M >   The classic off-by-one-hour is associated with direct access to SET TIME  M > and/or a failure to use the provided procedures.  (It would have been nice  G > to consolidate all this into some core commands rather than into the  M > current collection of interfaces and tools, and to have had the UTC values  J > attached to all the objects and embedded in all the structures, but the I > chances of implementing that in any reasonable fashion basically ended   > back in the 1970s.)  > H >   In general, I've learned over the years that the more I learn about J > time, the less I really understand.  Most of what I've gleaned over the M > years around time and time management (and that is specifically related to  M > OpenVMS) is included in the FAQ, and I've made a deliberate effort to stay  M > out of trying to define this stuff.  If you want to learn more (um, less?)  L > around how this time stuff works, wander over to the NIST/NBS site and to L > the other resources that are referenced within the OpenVMS FAQ.  And then J > once you (think you) understand the terrestrial implementation of time, - > there's a whole 'nother layer with physics.  >   L So true. Last night I attended a Perimeter Institute lecture on the history L of Time and Motion. (what is motion without reference to another object? Is L the reference object in motion with respect to something else? what is time F without referenced to another time?) By the time the lectured ended I K doubted that I could drive my car from the parking and respected those who    could put a skate board on Mars.   :-)   
 Neil Rieck Kitchener/Waterloo/Cambridge,  Ontario, Canada." http://www3.sympatico.ca/n.rieck/    ------------------------------  % Date: Thu, 02 Nov 2006 21:33:48 -0500 ' From: Dave Froble <davef@tsoft-inc.com> $ Subject: Re: Time change questions !9 Message-ID: <nJGdnefjU5taN9fYnZ2dnUVZ_sOdnZ2d@libcom.com>    Neil Rieck wrote:   M > In large networks (which sometimes cross time zones) a person can log into  N > any UNIX work station and access his own work files in the remote location. M > In most cases you are working from a storage area network (so the clock is  M > not the issue) but you may optionally copy files to your local machine and  J > then later copy them back. For this reason it was decided that all UNIX K > machines in a network would be running the same internal clock (UTC) but  I > that each user would SEE their own local time as defined by a timezone  O > parameter in their profile. It is not a big leap to see what can happen when  ) > networks are connected to the internet.  > I > So whether people are all logging into a single system or a network of  N > distributed systems, it might make sense to set the clock of all systems in I > the world to UTC (if possible). Where it is not possible or practical,  G > you've got to provide pseudo timezone support for the UNIX languages  G > (C/C++/Java/Javascript). BTW, the pseudo method is what OpenVMS uses.   A Well, yeah, all true and great 20-20 hindsight.  Reality is much  G tougher.  You face the question, "Do we 'do it right' (if there is any  B such thing) and break a bunch of existing applications, or, do we I maintain backward compatibility and never 'get it right'?"  I don't have  @ an answer.  I think the answer is different for different users.  E At one time, not breaking things was a high priority.  Since killing  A Alpha, I'm guessing users would accept anything without too much   complaining.  E Hey, Bob, what's the 'right thing' with respect to timekeeping for a   spacecraft orbiting Mars?    --  4 David Froble                       Tel: 724-529-0450> Dave Froble Enterprises, Inc.      E-Mail: davef@tsoft-inc.com DFE Ultralights, Inc.  170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------  % Date: Thu, 02 Nov 2006 21:37:22 -0500 ' From: Dave Froble <davef@tsoft-inc.com> $ Subject: Re: Time change questions !9 Message-ID: <nJGdnebjU5sGNtfYnZ2dnUVZ_sOdnZ2d@libcom.com>    JF Mezei wrote:  > Stephen Hoffman wrote:G >>   The time value is directly retrieved through most system services  I >> and most any of the existing low-level system calls, and directly set  F >> through, and there are timezone offsets applied via various of the J >> more recent UTC-related calls and via libraries such as the C standard  >> library.  >  > I > If VMS engineers (I was going to say "You guys", but alas, you've been  H > set free from the VMS engineering habitat), were to change SYS$GETTIM F > and other  official systems services that access the system time to H > automatically add some delta time taken from a logical name, how much  > would break ?  > I > Applications that cheat and fetch the quadword from memory would get a  H > GMT/UTC time. They wouldn't fail. Applications and DCL would properly * > convert the system time to "local" time. > H > You could then code SYS$GETUTC_TIM which would just return the system ( > time without any localised conversion. > K > DCL would need an update for all the /AFTER switches to properly convert   > the supplied value into GMT. > I > And you might then be able to allow a job table logical to define that  K > job's time offset to system time. (and if the job logical doesn't exist,  ! > a system logical would be used.  > I > And for users who cannot cnhange, then they can still set their system  A > time  to their local time zone and declare an offset of 00:00 .   H I think it's already been mentioned, but apparently you didn't read it, D or ignored it, what about procedures that look directly at the time @ quadword and don't use the system services and library routines?   Can you say "broken app"?    --  4 David Froble                       Tel: 724-529-0450> Dave Froble Enterprises, Inc.      E-Mail: davef@tsoft-inc.com DFE Ultralights, Inc.  170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------  $ Date: Thu, 2 Nov 2006 12:42:37 -0700, From: David D Miller <ddmiller@raytheon.com>$ Subject: Re: Toggle Operator ConsoleV Message-ID: <OF3A47729D.017671AE-ON0725721A.006C0C05-0725721A.006C367A@mck.us.ray.com>   Folks:  J I know this is bad form, but I researched my problem (in the DEC/Compaq/HPH docs) and answered my questions.  Gosh, maybe the 3rd edition of _System* Management Guide_ will do a better job! :)   dave.   ? David D Miller/RWS/Raytheon/US wrote on 11/02/2006 08:49:42 AM:    > Folks  > # > I'd like to return to this topic:  > - > > > Control-F2 for CDE/DECwindows/DW-Motif-  > B > is the correct keystroke to toggle, provided (as someone pointed* > out) the proper DECW logical is enabled. > > > However, CTRL-F2 doesn't immediately toggle from DW-Motif toD > Operator, but I think it toggles the next time an operator messageB > is posted.  Is this the correct behavior?  I'd rather it toggledD > immediately so I can see the last message written to the operator./ > BTW, it toggles back to DW-Motif immediately.  > D > On that subject, not all messages posted on the screen are writtenE > to OPERATOR.LOG.  Nor do the ones that are in OPERATOR.LOG have the A > detail that the operator display has.  How do I control what is  > written to OPERATOR.LOG? >  > dave.    ------------------------------  % Date: Thu, 02 Nov 2006 12:54:09 -0700 & From: Jim Mehlhop <Mehlhop@parsec.com>$ Subject: Re: Toggle Operator Console' Message-ID: <454A4CE1.50002@parsec.com>    David D Miller wrote:  > Folks  > # > I'd like to return to this topic:  >  > * >>>Control-F2 for CDE/DECwindows/DW-Motif- >  > K > is the correct keystroke to toggle, provided (as someone pointed out) the ! > proper DECW logical is enabled.  > L > However, CTRL-F2 doesn't immediately toggle from DW-Motif to Operator, butJ > I think it toggles the next time an operator message is posted.  Is thisK > the correct behavior?  I'd rather it toggled immediately so I can see the I > last message written to the operator.  BTW, it toggles back to DW-Motif  > immediately. > G > On that subject, not all messages posted on the screen are written to I > OPERATOR.LOG.  Nor do the ones that are in OPERATOR.LOG have the detail E > that the operator display has.  How do I control what is written to  > OPERATOR.LOG?   ( Check the OPC logicals in sylogicals.com     >  > dave.  >    ------------------------------   End of INFO-VAX 2006.604 ************************