1 INFO-VAX	Fri, 08 Apr 2005	Volume 2005 : Issue 195       Contents:> Re: A blast from the past - PDP history, and a bit of TECO too> Re: A blast from the past - PDP history, and a bit of TECO too Re: Broken vax4000 Re: Broken vax4000 Re: Broken vax4000 Re: Broken vax4000 Re: Broken vax4000 Re: Broken vax4000 Re: Broken vax4000 Re: Broken vax4000 Re: Broken vax4000J fopen(): which shr= option do I need to open a 2nd time for append/extend?4 Re: Frustration! EB64+ too old to run OpenVMS 7.3-1?4 Re: Frustration! EB64+ too old to run OpenVMS 7.3-1?4 Re: High File Watermarking question for Hoff Hoffman4 Re: High File Watermarking question for Hoff Hoffman1 Re: Hobbyists - you have a friend in the business 1 Re: Hobbyists - you have a friend in the business  loging in and out user?  Re: loging in and out user?   Re: Multiprocessor Microvax II ?  Re: Multiprocessor Microvax II ?  Re: Multiprocessor Microvax II ?  Re: Multiprocessor Microvax II ?$ Re: poor disk I/O performace on ds25$ Re: poor disk I/O performace on ds25 Re: Site down? Re: Site down? Re: Site down? Sort Directory by Date Re: Sort Directory by Date Re: Sort Directory by Date Re: Suggestion for FAQ? Re: TCO Study Rates HP OpenVMS on HP Integrity over IBM and Sun ? Re: TCO Study Rates HP OpenVMS on HP Integrity over IBM and Sun ? Re: TCO Study Rates HP OpenVMS on HP Integrity over IBM and Sun   RE: VMS 7.3-2 with 48 MB memory?F Re: Which release notes say sts$manager:utc$configure_tdf  is obsoleteP Re: Which release notes say sts$manager:utc$configure_tdf  is obsolete obsoleteoP Re: Which release notes say sts$manager:utc$configure_tdf  is obsolete obsoleteoE Re: Which release notes say sts$manager:utc$configure_tdf is obsolete E Re: Which release notes say sts$manager:utc$configure_tdf is obsolete   Re: [TCPIP V5.4 ECO4] SFTP rants  F ----------------------------------------------------------------------  % Date: Thu, 07 Apr 2005 20:53:39 +0100 & From: Elliott Roper <nospam@yrl.co.uk>G Subject: Re: A blast from the past - PDP history, and a bit of TECO too 1 Message-ID: <070420052053399545%nospam@yrl.co.uk>   ? In article <z9NFE8GohqF7@eisner.encompasserve.org>, Bob Koehler 0 <koehler@eisner.nospam.encompasserve.org> wrote:  G >     In what repository can I find VTEDIT?  Does it run on VT100 (last " >    one I had was VT52 specific).  , Yep, the VT52 one was VT52.tec I'll swap ya!   I have a copy, and a manual too ' http://www.yrl.co.uk/elliott/vtedit.tec ' http://www.yrl.co.uk/elliott/vtedit.doc @ http://www.yrl.co.uk/~elliott/vtedit.tes (that's the VTEDIT with' comments - for the *truly* masochistic)   % http://www.yrl.co.uk/elliott/teco.zip  and B http://www.yrl.co.uk/elliott/vtedit.zip for collections thereof (I think)  % For variations on the Dec teco manual + http://www.yrl.co.uk/elliott/teco/teco.html - inspect the top of the page for other formats ? (We used the teco manual as a test document for JAF's AsctoHTM) E The webby manual needed hardly any hand work once we had the AsctoHTM F policy file right. If you have any runoff-ish output like the originalE teco manual, grab AsctoHTM. It's PFM [1] The VMS version is freeware. D The Windows version is shareware, and I'm still trying to con me JAF into doing a Mac version.       1. Pure F***** Magic   --  1 To de-mung my e-mail address:- fsnospam$elliott$$ C PGP Fingerprint: 1A96 3CF7 637F 896B C810  E199 7E5C A9E4 8E59 E248    ------------------------------   Date: 7 Apr 2005 15:54:26 -0500 ; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) G Subject: Re: A blast from the past - PDP history, and a bit of TECO too 3 Message-ID: <VV2HCLfCMGCN@eisner.encompasserve.org>   Z In article <070420052053399545%nospam@yrl.co.uk>, Elliott Roper <nospam@yrl.co.uk> writes:   > . > Yep, the VT52 one was VT52.tec I'll swap ya!  B    I no longer have it.  I think it was on the TOPS-20 tools tape.   ------------------------------  # Date: Thu, 07 Apr 2005 18:53:12 GMT ( From: Mark Schafer <mark.schafer@hp.com> Subject: Re: Broken vax4000 2 Message-ID: <skf5e.3231$Iv3.2512@news.cpqcorp.net>  H I didn't read anything about that in the guide.  There is an audio self B test, so if you hear something, that means the audio unit is good.  : > During the self tests, what am I supposed to hear if the > unit is good??   ------------------------------   Date: 7 Apr 2005 12:30:20 -0700 " From: chris_doran@postmaster.co.uk Subject: Re: Broken vax4000 C Message-ID: <1112902220.540191.112270@f14g2000cwb.googlegroups.com>    Mark Schafer wrote: D > I didn't read anything about that in the guide.  There is an audio selfD > test, so if you hear something, that means the audio unit is good. > < > > During the self tests, what am I supposed to hear if the > > unit is good??  C OK, I've tried booting a couple of old friends. It seems I have one F good one and one sick. There's no audio output on either until the endA of tests when you get a rapid series of half a dozen tones. Since ? they're the same on both goodie and baddie, it doesn't tell you  anything useful.  F Neither machine has a VMS disk, but I'll try putting one in and report back later.   G As to the cable, I haven't tried it as I have a houseful of MMJs, but I C believe you can make an RJ11 go in by just filing off the clip. You > don't need to get a special tool -- just use the RJ11 end of a? 'phone/modem cable and put a 9 pin D-type on the other. The pin  connections are in the FAQ.   A Does anyone have any ideas on my "baddie"? It says nothing on the D console until I hit Break, when it says ?? CRPT reenter bit clr. The9 diags end with F3, which the book is not very helpful on.    Chris    ------------------------------  % Date: Thu, 07 Apr 2005 16:08:16 -0400 ' From: Dave Froble <davef@tsoft-inc.com>  Subject: Re: Broken vax4000 0 Message-ID: <115b46sjc3ug472@corp.supernews.com>  # chris_doran@postmaster.co.uk wrote:  > Mark Schafer wrote:  > D >>I didn't read anything about that in the guide.  There is an audio >  > self > D >>test, so if you hear something, that means the audio unit is good. >> >>; >>>During the self tests, what am I supposed to hear if the  >>>unit is good??  >  > E > OK, I've tried booting a couple of old friends. It seems I have one H > good one and one sick. There's no audio output on either until the endC > of tests when you get a rapid series of half a dozen tones. Since A > they're the same on both goodie and baddie, it doesn't tell you  > anything useful.  D I don't think the audio tones are useful.  What if the audio output 2 itself is inoperatable, but everything else is Ok?  I You need to connect the console output to some type of terminal, real or  % emulated, and see the console output.   H My VAXstation 4000 model 90A will not auto boot.  After a power failure E I must manually boot it.  (It has a TurboChannel SCSI adapter, which  I needs special drivers that I don't have, and the adapter causes an error  F in the console diagnostics.  A bad video card could do the same, even ) though the computer will work just fine.)   4 Don't count on auto-booting straight through to VMS.  H > Neither machine has a VMS disk, but I'll try putting one in and report
 > back later.  > I > As to the cable, I haven't tried it as I have a houseful of MMJs, but I A > believe you can make an RJ11 go in by just filing off the clip.   H The cable end that plugs into the VAXstation 4000 is an MMJ.  3 & 4 are H a common ground.  2 & 5 are xmit and recv, I forget which is which, but C the FAQ I think has the MMJ pin-outs.  It needs to be a null modem  G cable, ie; xmit on one end goes to recv on the other end, and the recv    goes to the xmit on the far end.   > You @ > don't need to get a special tool -- just use the RJ11 end of aA > 'phone/modem cable and put a 9 pin D-type on the other. The pin  > connections are in the FAQ.  > C > Does anyone have any ideas on my "baddie"? It says nothing on the F > console until I hit Break, when it says ?? CRPT reenter bit clr. The; > diags end with F3, which the book is not very helpful on.  >  > Chris  >    ------------------------------  % Date: Thu, 07 Apr 2005 16:14:19 -0400 ' From: Dave Froble <davef@tsoft-inc.com>  Subject: Re: Broken vax4000 0 Message-ID: <115b4i7pi39un42@corp.supernews.com>   Tom Linden wrote: 9 > On Wed, 06 Apr 2005 21:45:51 -0400, Stanley F. Quayle   ! > <squayle@insight.rr.com> wrote:  > - >> On 6 Apr 2005 at 21:21, Dave Froble wrote:  >>F >>> Come on Stan, shameless plugs for Charon are one thing.  $25 for aJ >>> cable with $3 of parts is another.  The newsgroup isn't for commercial
 >>> stuff. >> >>E >> It's sort of like the engineer joke about knowing where to put the D >> chalk mark.  Actually, I've written up the details on my web site >> (for free): > E > Allright Dave you have the specs, courtesy of Stan, would you do it 
 > for $25? >  >>$ >>   http://www.stanq.com/cable.html >> >> --Stan Quayle >> Quayle Consulting Inc.  >>
 >> ---------- 0 >> Stanley F. Quayle, P.E. N8SQ  +1 614-868-13636 >> 8572 North Spring Ct., Pickerington, OH  43147  USA3 >> stan-at-stanq-dot-com       http://www.stanq.com  >> >> >> > I I've had the specs for years.  I also have the tool for crimping modular  F adapters, RJ45, MMJ, etc, and the tool for crimping pins for 'D' type  connectors.   E If someone local needs a cable, I just make it.  Not worth charging.  I It's 4 miles to the nearest post office, and further to UPS.  If a cable  D isn't available closer to home, sure, I'll make one for the cost of G shipping and parts, as I mentioned, no more then $3.  Shipping is more  C then the cable.  Only trouble is, the last cables I gave away, the  0 recipient never re-imbursed me for the shipping.  E Hobbyists get turned off by $25 cables.  Commercial is another story.    ------------------------------  % Date: Thu, 07 Apr 2005 16:21:01 -0400 ' From: Dave Froble <davef@tsoft-inc.com>  Subject: Re: Broken vax4000 0 Message-ID: <115b4upljmu9i3e@corp.supernews.com>   Barry Treahy, Jr. wrote: > Dave Froble wrote: >  >> Stanley F. Quayle wrote:  >>@ >>>> Getting the appropriate connector is going to be a problem. >>>  >>>  >>> G >>> I have a whole bag of MMJ connectors, and the appropriate crimping  : >>> tool.  I could make you a cable with a DB-9 for $25... >>> F >> Come on Stan, shameless plugs for Charon are one thing.  $25 for a J >> cable with $3 of parts is another.  The newsgroup isn't for commercial 	 >> stuff.  >> > Dave,  > I > you miss two very fundamental aspects regarding Stan's offer that that  J > should be seriously considered before take a shot charging $25 to fab a ! > cable and help this person out. H > First, answering a question on COV for 'free' is one thing.  Actually J > doing work for someone is a totally different story and though I'm sure J > you love what you do, I doubt you would do it for free and his offer is F > no different that any merchant offering products or services on COV.K > Second, the person that needed the cable stated that obtaining one would  E > be a problem; presumably because they did not have the connectors,  J > cable, crimper, and soldering gear.  I would challenge you to purchase, J > not pull out of your drawer, all of that for $25 but even if you could, K > then what is your time worth?  I suspect that most IT folks on this list  J > have a wage that is between $25 to $80/hour, so when it is all said and K > done, Stan's offer to help this person out for $25 could be considered a   > bargain... >  > Barry  >   H We're all entitled to opinions, and I do respect yours, and Stan's.  As I I posted elsewhere, for commercial work, expect to pay for the expertise   and the tools.  H I PERSONALLY, and I don't hold anyone else to my opinions, think that a I hobbyist is different.  I think it's encourging the use/learning of VMS.  I   I think it's about the same as loaning a media CD rather than charging   for the use of it.  H Also, like Stan, I run my own business, and I don't have an employer to I keep me in mortgage money while I do favors.  And right now VMS business   sucks!   That's just me.    Dave   ------------------------------   Date: 7 Apr 2005 13:50:34 -0700 " From: chris_doran@postmaster.co.uk Subject: Re: Broken vax4000 C Message-ID: <1112907034.341020.294340@f14g2000cwb.googlegroups.com>   # chris_do...@postmaster.co.uk wrote:  > Mark Schafer wrote: F > > I didn't read anything about that in the guide.  There is an audio > selfF > > test, so if you hear something, that means the audio unit is good. > > > > > > During the self tests, what am I supposed to hear if the > > > unit is good?? > E > OK, I've tried booting a couple of old friends. It seems I have one D > good one and one sick. There's no audio output on either until the end C > of tests when you get a rapid series of half a dozen tones. Since A > they're the same on both goodie and baddie, it doesn't tell you  > anything useful. > A > Neither machine has a VMS disk, but I'll try putting one in and  report
 > back later.  >   C I've now booted VMS on "goodie", and as I predicted, it hangs at an C "enter date and time" prompt having been off for a year or more. At C this point all the lights are off. Nothing more from the sound port  even while VMS is running.  B I've just checked the postage rate, and I can air mail a cable for@ about $3. Of course, it'll take a while to arrive, especially ifE customs think it's a WMD, but if you're having to order one anyway...    Chris    ------------------------------  % Date: Thu, 07 Apr 2005 14:11:16 -0700 # From: "Tom Linden" <tom@kednos.com>  Subject: Re: Broken vax4000 ( Message-ID: <opsovrs2zgzgicya@hyrrokkin>  G On Thu, 07 Apr 2005 16:14:19 -0400, Dave Froble <davef@tsoft-inc.com>    wrote:  L > I've had the specs for years.  I also have the tool for crimping modular  I > adapters, RJ45, MMJ, etc, and the tool for crimping pins for 'D' type   
 > connectors. I >  If someone local needs a cable, I just make it.  Not worth charging.   L > It's 4 miles to the nearest post office, and further to UPS.  If a cable  G > isn't available closer to home, sure, I'll make one for the cost of   J > shipping and parts, as I mentioned, no more then $3.  Shipping is more  F > then the cable.  Only trouble is, the last cables I gave away, the  2 > recipient never re-imbursed me for the shipping.H >  Hobbyists get turned off by $25 cables.  Commercial is another story.  I I, for one think that is a very generous offer.  I regard it as virtually K free and a matter of community rather than commerce.  I think Stan is to be 
 commended.   ------------------------------  # Date: Thu, 07 Apr 2005 22:55:16 GMT   From: John Santos <john@egh.com> Subject: Re: Broken vax4000 ) Message-ID: <oTi5e.1048$hB6.873@trnddc06>   # chris_doran@postmaster.co.uk wrote: % > chris_do...@postmaster.co.uk wrote:  >  >>Mark Schafer wrote:  >>E >>>I didn't read anything about that in the guide.  There is an audio  >> >>self >>E >>>test, so if you hear something, that means the audio unit is good.  >>>  >>> < >>>>During the self tests, what am I supposed to hear if the >>>>unit is good?? >>E >>OK, I've tried booting a couple of old friends. It seems I have one D >>good one and one sick. There's no audio output on either until the >  > end  > C >>of tests when you get a rapid series of half a dozen tones. Since A >>they're the same on both goodie and baddie, it doesn't tell you  >>anything useful. >>A >>Neither machine has a VMS disk, but I'll try putting one in and  >  > report > 
 >>back later.  >> >  > E > I've now booted VMS on "goodie", and as I predicted, it hangs at an E > "enter date and time" prompt having been off for a year or more. At E > this point all the lights are off. Nothing more from the sound port  > even while VMS is running.  E Most (all?) Vaxes have a hardware time-of-year clock that is good for G about 15 months.  (After that it overflows.)  The system knows the time C at boot by adding the time-of-year clock to the base date stored in D SYS.EXE, which is the VMS time (64-bit quadword time) of midnight on> January 1st of the year in which $ SET TIME was last executed.  H If you last booted the system last year, and haven't done a "$ set time"A since Jan 1 this year, about the end of march or early April, the H time-of-year clock overflows, which has no immediate effect on anything.F Normally during shutdown, SYS$SYSTEM:SHUTDOWN.COM does a "$ SET TIME",A which notices the base time was last year, updates it to Jan 1 of E this year, and resets the time-of-year clock to the correct offset of D "now" from midnight January 1st of this year, so you are now good toH go for another year.  However, if the system crashes after the TOY clockE has overflowed, it gets really confused.  (I think the exact behavior D is VMS version-dependent.)  I've seen the time wrong by a year, or aD year and a day (I think this depends on if last year was a leap yearA and this year isn't or vice versa.)  Sometimes SYSBOOT recognizes C there's a problem and prompts for the time.  Eventually, in about a D million microFortnights :-) it times out and continues with its best guess of the time.  D The cure for this is to always do a "$ set time" sometime in JanuaryB or February each year.  This will fix the base time in SYS.EXE and? reset the TOY clock, and you don't have to worry about crashes.   E (I think if it crashes in January or February, before you've done the F set time, the TOY clock is still valid and VMS fixes things during the reboot.)  + I don't think any Alphas have this problem.   > Note: if you have DTSS enabled, you can't use set time, so you? have to disable and delete DTSS first.  ($ mcr ncl disable dtss > followed by $ mcr ncl delete dtss)  I don't know if DTSS takes7 care of this situation for you, making this restriction  irrelevant.     D > I've just checked the postage rate, and I can air mail a cable forB > about $3. Of course, it'll take a while to arrive, especially ifG > customs think it's a WMD, but if you're having to order one anyway...  >  > Chris  >      --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Thu, 07 Apr 2005 19:53:13 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com>  Subject: Re: Broken vax4000 B Message-ID: <1112917956.81f79a08b3099cdcc76d49f9f0b10857@teranews>   Dave Froble wrote:E > isn't available closer to home, sure, I'll make one for the cost of 7 > shipping and parts, as I mentioned, no more then $3.    D If I have a big inventory of cable and connectors, then parting withG some hardware for someone which has provided me help (on-line, replying + to questions etc) over the years may be OK.   H But if I have to go and buy the hardware and assemble it, I doubt that I could make a cable for $3.005 One connector alone is more than that at Radio Shack.    ------------------------------  % Date: Thu, 07 Apr 2005 19:33:09 -0700  From: Z <Z@no.spam> S Subject: fopen(): which shr= option do I need to open a 2nd time for append/extend? * Message-ID: <H1m5e.7744$LV2.4481@fe03.lga>  G A C function, f1(), opens a file with fopen(fnam, "r+b", "shr=get") it  E then calls another function, f2(), which determines that it needs to   extend fnam.   To extend fnam, f2() tries to:    1. fopen(fnam, "a+b", "shr=get")H 2. fwrite() a number of blank, fixed-size records to the end of the fnam 3. fclose()   H The fopen() in f2()/step1 fails, presumable because the file was opened  in f1() with shr=get.    Does that sound correct?  E If so, which of the various shr= options do I need to use in f1() to  E allow f2()'s fopen() to succeed?  There's shr=put, shr=upd, shr=upi,   shr=mse ...    ------------------------------  $ Date: Thu, 7 Apr 2005 23:37:50 +0200" From: "Bill" <balings@zfree.co.nz>= Subject: Re: Frustration! EB64+ too old to run OpenVMS 7.3-1? < Message-ID: <b7858$4255a90b$513b9e33$7624@news.versatel.net>  - <johnhreinhardt@yahoo.com> schreef in bericht = news:1112838199.059904.272630@f14g2000cwb.googlegroups.com...  >  > Larry Kilgallen wrote:H > > In article <d301m2$gc7$1@online.de>, helbig@astro.multiCLOTHESvax.de4 > (Phillip Helbig---remove CLOTHES to reply) writes:F > > > In article <3Yy4e.2985$vH.1818@news.cpqcorp.net>, hoff@hp.nospam > (Hoff  > > > Hoffman) writes: > > > H > > >>   I would personally aim for an AlphaStation series, including anG > > >>   XP900, XP1000 or equivalent box.  I would not acquire an Alpha H > > >>   system with less than an EV56 processor, and I would prefer and: > > >>   would look for an EV6-class processor, or better. > > > C > > > Why?  Obviously, these are more powerful machines.  Assuming, 
 > however,H > > > that one doesn't need the power, is there a reason to say "EV56 or& > > > later, preferably EV6 or later"? > > C > > For one thing, for current used equipment prices on eBay, exact C > > cpu contents seems to matter a lot less than phase of the moon.  >  > I > As one who has been hoping to catch an XP1000 I can tell you that while E > a 256MB PWS600AU usually goes for less than $300, the XP1000/500MHz G > rarely go for less than $700 and the 667's for less than $1K.  Also a B > decent AS1200 runs between $300-$600 but I've seen DS20's go forI > between $1250 and $2500.  Forget a DS10 (not the "L") for less than $2k G > usually.  Since the PWS and AS1200's are EV56 and the XP's and DS20's E > are EV6 (or EV67) in basically the same boxes it seems the CPU type + > does make a big difference at the moment.  >   K The Digital Server 5305 is the white box equivalent of the AS1200. Since it 
 is well known J that these systems have no offical VMS support, just for NT4, they tend to be fairly cheap.9 Must have something to do with support for the latter os. L However, with minor (read: two SRM command additions) changes, a DS5305 does run J VMS as well as the next AS1200. For a hobbyist it probably does not matter that the system is not supported.   Hans   ------------------------------   Date: 7 Apr 2005 15:01:23 -0700 ; From: "johnhreinhardt@yahoo.com" <johnhreinhardt@yahoo.com> = Subject: Re: Frustration! EB64+ too old to run OpenVMS 7.3-1? C Message-ID: <1112911283.863352.228790@l41g2000cwc.googlegroups.com>    Bill wrote:  > > E > > As one who has been hoping to catch an XP1000 I can tell you that  while G > > a 256MB PWS600AU usually goes for less than $300, the XP1000/500MHz G > > rarely go for less than $700 and the 667's for less than $1K.  Also  a D > > decent AS1200 runs between $300-$600 but I've seen DS20's go forG > > between $1250 and $2500.  Forget a DS10 (not the "L") for less than  $2k B > > usually.  Since the PWS and AS1200's are EV56 and the XP's and DS20'sG > > are EV6 (or EV67) in basically the same boxes it seems the CPU type - > > does make a big difference at the moment.  > >  > D > The Digital Server 5305 is the white box equivalent of the AS1200. Since it > is well known D > that these systems have no offical VMS support, just for NT4, they tend to  > be fairly cheap.; > Must have something to do with support for the latter os. B > However, with minor (read: two SRM command additions) changes, a DS5305 does  > run E > VMS as well as the next AS1200. For a hobbyist it probably does not  matter
 > that the > system is not supported. >  > Hans    ? Yep. I have one. Sits right beside my UWS 533au (also an AS1200 A "work-a-like") and chugs along on VMS just fine. ;-)  I also have E another UWS 533au shell (no cpu/memory/disk) that supplies parts when , necessary (so far only a P/S on the DS5305).   ------------------------------  % Date: Thu, 07 Apr 2005 16:36:30 -0400 ' From: Dave Froble <davef@tsoft-inc.com> = Subject: Re: High File Watermarking question for Hoff Hoffman 0 Message-ID: <115b5roiqnu58b6@corp.supernews.com>   Larry Kilgallen wrote:\ > In article <11592lu8evgkd38@corp.supernews.com>, Dave Froble <davef@tsoft-inc.com> writes: >  > G >>The bottom line, if the system is in trusted hands, and super secret  J >>data is not at issue, turn off the highwater marking.  What do you care < >>what's on the disk if you're about to overwrite it anyway? >  > B > The bottom line is, for most normal use, there is no performanceC > penalty for highwater marking, and there _is_ a security benefit.   @ Ok, I'll confess, I do have an application that can suffer from 7 highwater marking, and I still leave it turned on.  :-)    ------------------------------  % Date: Thu, 07 Apr 2005 21:01:14 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>= Subject: Re: High File Watermarking question for Hoff Hoffman + Message-ID: <4255E5E9.F517A6E0@comcast.net>    Rob Brooks wrote:  > 8 > "Peter Weaver" <newsgroup@weaverconsulting.ca> writes: > > Jilly wrote: > / >         [stuff about the DKLOG SDA extension]  > >>A > >> Now DKLOG.LOG will have all the IO activity to the testdisk.  > >>...  > > ! > > Is DKLOG documented anywhere?  > P > No, but there is not much to it; the command syntax is pretty straightforward.5 > Type DKLOG at the SDA> prompt, and you'll get . . .  >  > DKLOG command format: O >   DKLOG SHOW  {<devnam>|/ADDRESS=<ucbaddr>} [/DIRECTION=[FORWARD|BACKWARD(D)] G >   DKLOG START {<devnam>|/ADDRESS=<ucbaddr>}[/ENTRIES=[<count>|128(D)] - >   DKLOG STOP  {<devnam>|/ADDRESS=<ucbaddr>} - >   DKLOG CLEAR {<devnam>|/ADDRESS=<ucbaddr>} N >   <devnam> acts on all matching device UCBs, <ucbaddr> acts on a single UCB. > G > SYS$DKDRIVER has several points during the life of an I/O at which it P > stores info.  An understanding of the SCSI spec is helpful when looking at the > data.  >  > It's a great tool!  E When did it first appear? U get his on my V7.2-2 system here at home:    DJAS01::DDACHTERA$ ana/sys      " OpenVMS (TM) Alpha system analyzer  
 SDA> dklog$ %CLI-W-SYNTAX, error parsing 'DKLOG' SDA>     --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Thu, 07 Apr 2005 16:38:52 -0400 ' From: Dave Froble <davef@tsoft-inc.com> : Subject: Re: Hobbyists - you have a friend in the business0 Message-ID: <115b606ngk6ca5f@corp.supernews.com>   JF Mezei wrote:  > DavidT wrote:  >   >>www.islandco.com/hobbyist.html >  > J > As a suggestion, you should open your program to any DECUS member or itsG > descendant organisation around the world, not just the US orgaisation H > formerly known as DECUS (I think this week they are called Encompass).  & I'd bet that he would if anyone asked.   ------------------------------  % Date: Thu, 07 Apr 2005 21:03:31 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>: Subject: Re: Hobbyists - you have a friend in the business+ Message-ID: <4255E673.B74D3711@comcast.net>    Dave Froble wrote: >  > JF Mezei wrote:  > > DavidT wrote:  > > " > >>www.islandco.com/hobbyist.html > >  > > L > > As a suggestion, you should open your program to any DECUS member or itsI > > descendant organisation around the world, not just the US orgaisation J > > formerly known as DECUS (I think this week they are called Encompass). > ( > I'd bet that he would if anyone asked.   (I think he just did...)   --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  # Date: Thu, 07 Apr 2005 23:52:37 GMT / From: "Craig Marburger" <craigm2@earthlink.net>   Subject: loging in and out user?@ Message-ID: <9Jj5e.2416$An2.81@newsread2.news.pas.earthlink.net>  L     Is there documentation of the auditing utlitity to tell when an account 9 was logged into and out off?  How do you turn it on?  ect   	 Thank you    Craig Marburger    ------------------------------  % Date: Thu, 07 Apr 2005 21:11:22 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>$ Subject: Re: loging in and out user?+ Message-ID: <4255E84A.6CB4FB5C@comcast.net>    Craig Marburger wrote: > M >     Is there documentation of the auditing utlitity to tell when an account ; > was logged into and out off?  How do you turn it on?  ect    HELP SET AUDIT?   C The ACCOUNTNG utility will tell you (with /FULL) when a process was ? logged out (terminated) and how long it ran. From there you can  extrapolate the login time.   	 See also:   http://www.hp.com/go/openvms/doc   --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Thu, 07 Apr 2005 21:21:57 +0200 + From: Robert Boers <r_boers@softresint.com> ) Subject: Re: Multiprocessor Microvax II ? 6 Message-ID: <42558858$0$1160$5402220f@news.sunrise.ch>  F There was a 2 and a 4 CPU MicroVAX (3520 and 2540), but the CPUs were C interconnected via a separate bus, and the Qbus was for peripheral  G connects as usual. A MicroVAX with multiple CPUs on a Qbus might work,  K but was never implemented. I don't think VMS would know what to do with it.    Robert   JF Mezei wrote: 5 > I found a copy of the KA630 CPU motherboard manual.  > J > Low and behold, it mentions that the all mighty Microvax II was designedH > to have multiple CPU boards on the single Q bus, with one board acting7 > as the arbitrator and other acting as auxiliary CPUs.  > H > I was flaberghasted ! Does this mean that by just setting a bit in theE > console memory that I could then insert a second MVII board into my I > all-mighty Microvax II and magically find myself with an SMP system ??? K > (could I insert it on a running system without shutting it down ? :-) :-)  > F > Seriously, this is a most interesting engineering capability. Has itG > ever been put to use ? If so, what sort of uses ? The text mentions a 0 > type of interCPU interconnect similar to a CI. >  > : > Oh, and those big purple chips are called "Gate Arrays". > 3 > I put up a page showing the MV-II motherboard at:  > 2 > http://www.vaxination.ca/vms/microvax/mv_ii.html > I > (At the bottom is also a link to the microvax resources home page which 8 > contains many links to microvax resources on the net). > J > BTW, the manual talks about the 78032 CPU chip, but on my board (versionI > 02 of the board) the chip is labeled as 21-22797-01. Are these the same H > or a different generation ? (or does the manual uses KA630-AA to refer > to a Microvax 1 ?)   ------------------------------  # Date: Thu, 07 Apr 2005 20:58:05 GMT / From: "JOHN GEMIGNANI" <jgemignani@verizon.net> ) Subject: Re: Multiprocessor Microvax II ? & Message-ID: <x9h5e.86$ff4.67@trndny08>  ; "JF Mezei" <jfmezei.spamnot@teksavvy.com> wrote in message  < news:1112867637.a98ecd3b253054750b3bb119c7c6218a@teranews...4 >I found a copy of the KA630 CPU motherboard manual. > J > Low and behold, it mentions that the all mighty Microvax II was designedH > to have multiple CPU boards on the single Q bus, with one board acting7 > as the arbitrator and other acting as auxiliary CPUs.  > H > I was flaberghasted ! Does this mean that by just setting a bit in theE > console memory that I could then insert a second MVII board into my I > all-mighty Microvax II and magically find myself with an SMP system ??? K > (could I insert it on a running system without shutting it down ? :-) :-)  > F > Seriously, this is a most interesting engineering capability. Has itG > ever been put to use ? If so, what sort of uses ? The text mentions a 0 > type of interCPU interconnect similar to a CI. >  > : > Oh, and those big purple chips are called "Gate Arrays". > 3 > I put up a page showing the MV-II motherboard at:  > 2 > http://www.vaxination.ca/vms/microvax/mv_ii.html > I > (At the bottom is also a link to the microvax resources home page which 8 > contains many links to microvax resources on the net). > J > BTW, the manual talks about the 78032 CPU chip, but on my board (versionI > 02 of the board) the chip is labeled as 21-22797-01. Are these the same2H > or a different generation ? (or does the manual uses KA630-AA to refer > to a Microvax 1 ?)  K Well, there is a little bit of an issue in placing a second CPU board onto e the bus.M From what I recall, there was a jumper on the board to make it a slave (i.e. v 2nd0J CPU), but I believe that the Q-bus only supported three slots for CPU and  memory.eH The CPU had a special connector over the top to a maximum of two memory  boards.tK So, with the parts that were available, you could only connect one CPU and u
 two memoryJ boards.  This is based on what I recall as I was looking into trying this  out.   -Johnt   ------------------------------  # Date: Fri, 08 Apr 2005 01:11:35 GMT % From: Roger Ivie <rivie@ridgenet.net>e) Subject: Re: Multiprocessor Microvax II ?o3 Message-ID: <slrnd5bmi7.nq2.rivie@Stench.no.domain>I  ; On 2005-04-07, Robert Boers <r_boers@softresint.com> wrote:aH > There was a 2 and a 4 CPU MicroVAX (3520 and 2540), but the CPUs were E > interconnected via a separate bus, and the Qbus was for peripheral wI > connects as usual. A MicroVAX with multiple CPUs on a Qbus might work,  M > but was never implemented. I don't think VMS would know what to do with it.o  E The Firefox (3520 and 3540) connected the CPUs, memory, and I/O via atB write-back cache bus called MBus. The QBus adapter (there were twoE varieties: the general-purpose FQAM and the FTAM, which was supportedeG only for TK70 tapes) plugged into the MBus in much the same manner thats0 a UNIBUS adapter on a /780 plugged into the SBI. -- o
 Roger Ivie rivie@ridgenet.net http://anachronda.webhop.org/_ -----BEGIN GEEK CODE BLOCK-----h
 Version: 3.12eH GCS/P d- s:+++ a+ C++ UB--(++++) !P L- !E W++ N++ o-- K w O- M+ V+++ PS+? PE++ Y+ PGP t+ 5+ X-- R tv++ b++ DI+++ D+ G e++ h--- r+++ z+++ C ------END GEEK CODE BLOCK-------   ------------------------------  # Date: Fri, 08 Apr 2005 01:09:07 GMTp% From: Roger Ivie <rivie@ridgenet.net>o) Subject: Re: Multiprocessor Microvax II ?t3 Message-ID: <slrnd5bmdj.nq2.rivie@Stench.no.domain>   = On 2005-04-07, JF Mezei <jfmezei.spamnot@teksavvy.com> wrote:<H > I was flaberghasted ! Does this mean that by just setting a bit in theE > console memory that I could then insert a second MVII board into myDI > all-mighty Microvax II and magically find myself with an SMP system ???,  A SMP? No. An attached processor for specific tasks (like the KA800a2 running the graphics engine on a VAXstation 8000).  G IIRC, the KA630s would need separate memory busses. This makes them nottG symmetric. My KA630 manual is in a box somewhere, though, so I could be 
 mistaken.  -- .
 Roger Ivie rivie@ridgenet.net http://anachronda.webhop.org/. -----BEGIN GEEK CODE BLOCK-----h
 Version: 3.12iH GCS/P d- s:+++ a+ C++ UB--(++++) !P L- !E W++ N++ o-- K w O- M+ V+++ PS+? PE++ Y+ PGP t+ 5+ X-- R tv++ b++ DI+++ D+ G e++ h--- r+++ z+++ t ------END GEEK CODE BLOCK------    ------------------------------  % Date: Thu, 07 Apr 2005 17:09:47 -0400t' From: Dave Froble <davef@tsoft-inc.com>a- Subject: Re: poor disk I/O performace on ds25o0 Message-ID: <115b7q8opj1kt56@corp.supernews.com>   John Santos wrote:  I > Here's a snippet of my code (DEC BASIC V1.5A, Alpha VMS V7.3-2 with alla8 > patches except the recent C RTL and GRAPHICS patches): > I >    open filnam$ for output as file 1%, organization sequential fixed, &9@ >       filesize sz, access scratch, allow none, map test_buffer   John,e  F As old RSTS programmers, we both have the same bad habits on VMS.  :-)  F You use the FILESIZE to pre-allocate the file space, usually to get a D contiguous file.  However, it's not filesize we need for contiguous F storage, just allocation of the storage.  Wasn't any such concepts on # RSTS, and we're creatures of habit.t  H BASIC, and I'd bet every other language, but don't know, actually makes I the file SIZE whatever you specify, and therefore the EOF is at the last  8 block of the file, and HWM must zero out all the blocks.  H If you could create a file with the allocation whatever contiguous size I you wanted, but leave the actual filesize zero, then HWM would only come e3 into play when you accessed a block beyond the EOF.   D Same problem when we open a file, and then PUT the highest block in H order to get a single, and hopefully contiguous, allocation of storage. 1   In such a case, HWM must then zero every block._  H I've got a very bad example of this.  In the reorganization utility for G my DAS database system, (similar to RMS's CONVERT), I do everything to  F cause problems with HWM.  I create the file with the final filesize I G want, I think it then does a PUT to the highest block, I then populate  D the file, and finally write a pattern of 255's in all unused blocks.  ; Such practices pre-date HWM, so we can be slightly excused.   G A rewrite whth current knowledge and implementing async I/O would be a h? great help, but it ain't broken, and nobody's complaining.  :-)d   Dave   ------------------------------  # Date: Thu, 07 Apr 2005 23:33:01 GMTo  From: John Santos <john@egh.com>- Subject: Re: poor disk I/O performace on ds25i( Message-ID: <Nqj5e.825$0c2.594@trnddc08>   Dave Froble wrote: > John Santos wrote: > J >> Here's a snippet of my code (DEC BASIC V1.5A, Alpha VMS V7.3-2 with all9 >> patches except the recent C RTL and GRAPHICS patches):1 >>J >>    open filnam$ for output as file 1%, organization sequential fixed, &A >>       filesize sz, access scratch, allow none, map test_bufferd >  >  > John,i > H > As old RSTS programmers, we both have the same bad habits on VMS.  :-) > H > You use the FILESIZE to pre-allocate the file space, usually to get a F > contiguous file.  However, it's not filesize we need for contiguous H > storage, just allocation of the storage.  Wasn't any such concepts on % > RSTS, and we're creatures of habit.w > J > BASIC, and I'd bet every other language, but don't know, actually makes K > the file SIZE whatever you specify, and therefore the EOF is at the last  : > block of the file, and HWM must zero out all the blocks. > J > If you could create a file with the allocation whatever contiguous size K > you wanted, but leave the actual filesize zero, then HWM would only come n5 > into play when you accessed a block beyond the EOF.e > F > Same problem when we open a file, and then PUT the highest block in J > order to get a single, and hopefully contiguous, allocation of storage. 2 >  In such a case, HWM must then zero every block. > J > I've got a very bad example of this.  In the reorganization utility for I > my DAS database system, (similar to RMS's CONVERT), I do everything to  H > cause problems with HWM.  I create the file with the final filesize I I > want, I think it then does a PUT to the highest block, I then populate  F > the file, and finally write a pattern of 255's in all unused blocks. > = > Such practices pre-date HWM, so we can be slightly excused.c > I > A rewrite whth current knowledge and implementing async I/O would be a mA > great help, but it ain't broken, and nobody's complaining.  :-)  >  > Dave  	 Hi Dave -t  F I've been assuming all along that there is a separate "Highwater Mark"G in the file header, so why should it care about logical EOF, just checkiF the reads and writes for any that extend past the HWM and erase blocksD as necessary, but maybe there isn't a separate field and it uses theE logical EOF (as stored in the file attributes area of the header in amD format inherited from RSX/ODS-1!)  If so, then I know understand whyG pre-extending the file using the BASIC open (which sets the logical EOFeH to the end of the block specified in the "size" clause) forces an erase.  F I think a call to FDL$CREATE with a FDL specifying an empty sequentialE file with the desired allocation, followed by an OPEN FOR INPUT would F do the trick, or maybe use size 0 and a USEROPEN that jams the desiredH size into FAB$L_ALQ, but leaves EBK and FFB at zero, would do the trick.  E I don't want to initially create the file with 0 blocks allocated andnD then force an extend with each put, or create with 0 blocks and thenH immediately PUT the last one.  Those would be no better than the currentA program and possibly much worse, depending on file fragmentation.9   -- r John Santoso Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Thu, 07 Apr 2005 18:55:11 -0400>' From: Dave Froble <davef@tsoft-inc.com>e Subject: Re: Site down? / Message-ID: <115bdvrncdffc5@corp.supernews.com>r   Jean-Franois Pironne wrote:  > Dave Weatherall a crit :  > ; >> On Wed, 6 Apr 2005 18:04:18 UTC, Jean-Franois Pironne n >> <jf.pieronne@laposte.net>
 >>  wrote: >> >> >>> issinoho a crit : >>>n$ >>>> Anyone know what's happened to 3 >>>> http://www.pi-net.dyndns.org/anonymous/kits/ ?r >>>>J >>>> I've been trying to get the latest version of MySQL for several days 	 >>>> now.  >>>S >>>uF >>> My new ADSL modem has crashed and I was on another site for a few @ >>> days, I have rebooted (power off/on) it and this work again. >>> = >>> I have never thought that a ADSL modem can crash/hang :-(. >> >> >>! >> It's got software in it... :-)a >>F > Yes, but the previous one had never hang during more than 7 years...  B You were lucky.  It's good to be lucky.  It sucks when you're not.  I > One of my friend which do plane model-making has told me a story about tE > new remote control which use windows CE with many new features but iI > randomly crash. This not a problem, you can reboot the remote control, rI > but generally in meantime your plane has also crash and when this is a OG > jet flighting a few hundred km/h this can also be very dangerous for bK > present people which have sometime to throw to the ground... And it is a cG > little more difficult to repair a crash plane than to reboot a crash   > remote control :-))   1 There is a lesson there, but nobody is listening.R   ------------------------------  % Date: Thu, 07 Apr 2005 20:58:56 -0500r2 From: David J Dachtera <djesys.nospam@comcast.net> Subject: Re: Site down? + Message-ID: <4255E560.4D5D4691@comcast.net>    Jean-Franois Pironne wrote:  >  > Paul Sture a crit :! > > Jean-Franois Pironne wrote:p > >  > >> issinoho a crit :c > >>$ > >>> Anyone know what's happened to4 > >>> http://www.pi-net.dyndns.org/anonymous/kits/ ? > >>>hJ > >>> I've been trying to get the latest version of MySQL for several days
 > >>> now. > >> > >> > >>F > >> My new ADSL modem has crashed and I was on another site for a fewA > >> days, I have rebooted (power off/on) it and this work again.a > >>> > >> I have never thought that a ADSL modem can crash/hang :-( > >a > >iC > > My experience is that they crash (in a visble way) but they cankH > > certainly hang - I've no idea whether that was the modem itself or aJ > > problem at the ISAP side.  Sometimes rebooting the modem (router in my< > > case) is enough; at other times a power off is required. > ? > So now the problem is to automatically handle such situation.wA > Is is fearly easy to detect the problem, but to cure it is morenH > difficult... I have to find a way to power off/power on the modem from > VMS...  F It may have a web interface that is not publicised. I found out fairlyA recently that I can get to my cable modem at 192.168.100.1 from a E browser. It may be possible to use WGET to send whatever is needed toi9 cause it to restart, if it is not completely hung (hard).n   -- r David J Dachtera dba DJE Systemsg http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:r" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/h  " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/t   Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Thu, 07 Apr 2005 22:15:10 -0400a- From: JF Mezei <jfmezei.spamnot@teksavvy.com>f Subject: Re: Site down?IB Message-ID: <1112926463.c0a6a64d55e49e9dc180ae99865e27e3@teranews>   David J Dachtera wrote:vH > It may have a web interface that is not publicised. I found out fairlyC > recently that I can get to my cable modem at 192.168.100.1 from arG > browser. It may be possible to use WGET to send whatever is needed toD; > cause it to restart, if it is not completely hung (hard).     D A DSL modem behind a NAT router that has PPPoE thing mames it nearlyH impossible to access from a real machine on the lan side of the  router.  F My modem has a serial port, and so does my router, and I find those to( be the only reliable way to access them.   ------------------------------  $ Date: Thu, 7 Apr 2005 16:32:49 -04004 From: "Peter Weaver" <newsgroup@weaverconsulting.ca> Subject: Sort Directory by Dateu, Message-ID: <3bljnlF6gktqgU1@individual.net>  B Last year the subject of sorting a directory by date came up and IH posted a little routine that used LIB$DT_FORMAT and a sort specificationD file.  I think a few people thought it was useful and someone postedA that it was a good example of how to use /include statements in adB specification file. Someone found it recently and asked if I couldE modify it to work without the LIB$DT_FORMAT so here is a version thatNA does not use the LIB$DT_FORMAT if anyone else wants it. Be reallye? careful with the lines that specify a certain number of spaces.b   $!" $!      Sorts a directory by date. $!      PRW 10-MAY-2004t $!D $!      Modified by Peter Weaver 7-APR-2005 to not use LIB$DT_FORMAT $! $! ASSIGN LIB$DATE_FORMAT_037,-  $!   LIB$TIME_FORMAT_001 - $!   LIB$DT_FORMAT/USER_MODE $!% $ directx -  ! Ignore any DIR symbolsa    /date=modified -o     /width=(file:80,display:132)-6    /out=sys$scratch:sort_date_'f$getjpi("","PID")'.tmp $! $!: $ sort sys$scratch:sort_date_'f$getjpi("","PID")'.tmp tt:-     /specification=sys$input:  /collating_sequence=(sequence=( = "AN","EB","AR","PR","AY","UN","UL","UG","EP","CT","OV","EC","I ","0"-"9"))a* /field=(name=filename1,position:1,size:19)+ /field=(name=filename2,position:20,size:60)o) /field=(name=dayfield,position:83,size:4) + /field=(name=monthfield,position:87,size:2)l* /field=(name=yearfield,position:90,size:4)* /field=(name=timefield,position=94,siz:11)/ /condition=(name=shortname,test=(filename2 eq " *                            ")) ! 60 spaces. /condition=(name=longname,test=(filename2 ne ")                           ")) ! 60 spaces  /include=(condition=shortname, key=yearfield, key=monthfield,e
 key=dayfield,I key=timefield, data=filename1,d data="  ", ! 2 spacesr data=dayfield, data=monthfield,	 data="-",  data=yearfield,- data=timefield)- /include=(condition=longname,0 key=yearfield, key=monthfield,o
 key=dayfield,> key=timefield, data=filename1,e data=filename2,1* data="                      ", ! 22 spaces data=dayfield, data=monthfield,	 data="-",c data=yearfield,  data=timefield)o $ delete/nolog/noconfirm0 sys$scratch:sort_date_'f$getjpi("","PID")'.tmp.*     --   Peter Weaver Weaver Consulting Services Inc.t Canadian VAR for CHARON-VAXo www.weaverconsulting.cai   ------------------------------  $ Date: Thu, 7 Apr 2005 17:49:46 -0400 From: norm.raphael@metso.com# Subject: Re: Sort Directory by DatevQ Message-ID: <OF57CE3FE3.B8E3596B-ON85256FDC.00765F14-85256FDC.0077CCB0@metso.com>   F I find that I have to add 1 position to the timefield field to get the full time_spec.h* /field=(name=timefield,position=94,siz:12). and also that if I modify the key lines thusly   key=(yearfield,descending),t key=(monthfield,descending), key=(dayfield,descending), key=(timefield,descending),   0 I can produce the directory in descending order.  5 Note:  <> is null or meta-notation and never appears.a   I am presuming that in the linee\ /collating_sequence=(sequence=("AN","EB","AR","PR","AY","UN","UL","UG","EP","CT","OV","EC","
  ~","0"-"9"))hB which my email client split, (and which may be mangled here again)D that there is a single space between the Quote-marks just before the'  <,"0"-"9"))> near the end of the line.t  3 The trailer and a few extraneous lines seem to fallr> where they may.  Is this expected.  That is, in the .tmp file, the first three lines: <>$ Directory <device>:[directory_spec]> <> and the last two lines:m <> Total of n...n files.a  * do not seem to be handled by this routine.   ------------------------------   Date: 7 Apr 2005 15:57:03 -0700.$ From: "AEF" <spamsink2001@yahoo.com># Subject: Re: Sort Directory by DatehC Message-ID: <1112914623.022576.112300@l41g2000cwc.googlegroups.com>s   Peter Weaver wrote:iD > Last year the subject of sorting a directory by date came up and I< > posted a little routine that used LIB$DT_FORMAT and a sort
 specification F > file.  I think a few people thought it was useful and someone postedC > that it was a good example of how to use /include statements in a'D > specification file. Someone found it recently and asked if I couldG > modify it to work without the LIB$DT_FORMAT so here is a version that.C > does not use the LIB$DT_FORMAT if anyone else wants it. Be reallyhA > careful with the lines that specify a certain number of spaces.m >o > $!$ > $!      Sorts a directory by date. > $!      PRW 10-MAY-2004  > $!F > $!      Modified by Peter Weaver 7-APR-2005 to not use LIB$DT_FORMAT > $!! > $! ASSIGN LIB$DATE_FORMAT_037,-  > $!   LIB$TIME_FORMAT_001 - > $!   LIB$DT_FORMAT/USER_MODE > $!' > $ directx -  ! Ignore any DIR symbolsu >    /date=modified - " >    /width=(file:80,display:132)-8 >    /out=sys$scratch:sort_date_'f$getjpi("","PID")'.tmp > $! > $!< > $ sort sys$scratch:sort_date_'f$getjpi("","PID")'.tmp tt:- >     /specification=sys$input: ! > /collating_sequence=(sequence=(s? > "AN","EB","AR","PR","AY","UN","UL","UG","EP","CT","OV","EC"," 
 > ","0"-"9"))   A What is this line supposed to be? Can you repost using VMS_SHARE?    [...]t  D It seems to work fine except that I get the date/time too far to the left:   + NETSERVER.LOG;117    24-MAR-2005 23:25:24.8v+ NETSERVER.LOG;118    24-MAR-2005 23:31:05.4 + NETSERVER.LOG;119    28-MAR-2005 00:22:46.3 + NETSERVER.LOG;120    28-MAR-2005 00:34:41.2n FTPSERVER_DTP.LOG;171.3                               9-MAR-2005 18:09:33.2-+ NOTIFY-637.LOG;1      1-APR-2005 02:39:51.7 + NOTIFY-642.LOG;1      1-APR-2005 02:57:52.7  FTPSERVER_DTP.LOG;172a3                               6-APR-2005 15:24:40.49+ IDS03000078AC.COM;1   7-APR-2005 19:31:40.9 + DIRBYDATE.COM;1       7-APR-2005 22:18:31.0k+ DIRBYDATE2.COM;1      7-APR-2005 22:19:14.9 + DIRBYDATE3.COM;1      7-APR-2005 22:21:37.5v  * And the hundreths are missing a character.     Bonus question!:  @ What is the point of LIB$DT_FORMAT? All it seems to do is affectG directory command date-time output. Now in LIB$DT_STARTUP.COM there areuF 41 date formats and 20 time formats. I like the ones that let you sortF with a simple SORT command. But there is even more stuff in this file.F Why all this work just to produce non-standard output of the DIRECTORY? command? I'm not complaining, but why so much work just for oneo command? Just curious.   ------------------------------   Date: 7 Apr 2005 20:40:22 -0700n$ From: "AEF" <spamsink2001@yahoo.com> Subject: Re: Suggestion for FAQsC Message-ID: <1112931622.260370.188840@l41g2000cwc.googlegroups.com>f  
 AEF wrote: > Dave Froble wrote: > > AEF wrote: > >"G > > > So symbols are variables (in DCL), and logical names are pointerse > (in G > > > the system) that point to generalized file-specs (including disksa > and F > > > directories), queues, mail addresses, or other logical names (or > > > anything else I forgot).    A The "generalized file-sepcs" include node names, of course. Also,?B logical names are automatically translated in the context of queueG forms, queue characteristics, and queue manager names! And logical name<B table names, of course, including the process and system directory tables.7  G Oddly, it is usually best to include colons when logical names are usedhD as sets of file-spec components, but without the colons when used in queue-related commands.u    ? > > If you call them pointers you miss some of their uses.  For1 example,> > > programs checking the value, or existance, of a logical to	 determinee > theiD > > state of an application, even for a type of cooperative locking. > >m > > Dave >  > 7 > Man, you just can't please some people!!! :-) :-) :-)  > G > I must say that when I listed generalized file-specs, mail addresses,tE > and queues, I meant that logical names are translated automatically D > whenever they are used in any of those contexts. Sure, they can be usedE > to store data such as configuration data, status data, flags, misc.oG > data, what have you. In this case, logical names are pointers to such > > data! And such data can be other logical names in which case	 iterativet > translation is used.    - How about generalized environment variables?!x  @ Anh (pronounced like en in en masse, except with a short a soundD instead), there are just too many useful aspects to logical names to( sum them up in a short descriptive term.   ------------------------------  % Date: Thu, 07 Apr 2005 21:23:05 +0100a& From: Elliott Roper <nospam@yrl.co.uk>H Subject: Re: TCO Study Rates HP OpenVMS on HP Integrity over IBM and Sun1 Message-ID: <070420052123055446%nospam@yrl.co.uk>t  H In article <4254df6c$0$308$4d4ef98e@read.news.ch.uu.net>, Soterro wrote:   > Hello, > ' > I just read this in the DSPP Journal.y? > http://h71000.www7.hp.com/openvms/whitepapers/alinean_tco.pdfnJ > There are a few things where I'm a bit surprised. Could anybody comment 
 > on them?  . It's great to see such gung-ho puffery for VMS  C I'm particularly impressed with the TARDIS [1] time travel module:-a  D > With a similar case study analyses comparing the TCO of HP OpenVMSE > running on Integrity servers  versus Sun Solaris, savings were moreuE > than $1.3 million over the five year analysis period, over 17%, in  D > direct savings, and an additional $8.6 million in availability andA > security benefits, yielding a significant  $10.0 million of TCOn > benefits.o  G Now, hands up! Who's been running VMS for 5 years on Integrity servers?r  E 1. TCO Analysis Retroactively Determined from Imaginary Statistics ==g
 TARDIS ;-)   -- i1 To de-mung my e-mail address:- fsnospam$elliott$$ C PGP Fingerprint: 1A96 3CF7 637F 896B C810  E199 7E5C A9E4 8E59 E248u   ------------------------------  % Date: Thu, 07 Apr 2005 19:56:53 -0400a- From: JF Mezei <jfmezei.spamnot@teksavvy.com> H Subject: Re: TCO Study Rates HP OpenVMS on HP Integrity over IBM and SunB Message-ID: <1112918183.84bdde382cd31248b9cfb9ee7f7e2e9a@teranews>   Elliott Roper wrote:I > Now, hands up! Who's been running VMS for 5 years on Integrity servers?  > G > 1. TCO Analysis Retroactively Determined from Imaginary Statistics ==  > TARDIS ;-)  H I heard that HP is interested in making animated plastic mannequins. :-) ;-) :-) :-) :-)0   ------------------------------  % Date: Thu, 07 Apr 2005 21:05:04 -0500p2 From: David J Dachtera <djesys.nospam@comcast.net>H Subject: Re: TCO Study Rates HP OpenVMS on HP Integrity over IBM and Sun+ Message-ID: <4255E6D0.9C21F00E@comcast.net>s   JF Mezei wrote:5 >  > Elliott Roper wrote:K > > Now, hands up! Who's been running VMS for 5 years on Integrity servers?e > >yI > > 1. TCO Analysis Retroactively Determined from Imaginary Statistics ==m > > TARDIS ;-) > J > I heard that HP is interested in making animated plastic mannequins. :-) > ;-) :-) :-) :-)   # BLAST! I just sold my motorcycle...s   -- w David J Dachtera dba DJE Systemso http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/H  " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  $ Date: Thu, 7 Apr 2005 19:33:07 -0400' From: "Main, Kerry" <kerry.main@hp.com>s) Subject: RE: VMS 7.3-2 with 48 MB memory?eR Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB595890@tayexc19.americas.cpqcorp.net>   > -----Original Message-----3 > From: Steve Thompson [mailto:smt@vgersoft.com]=20e > Sent: April 7, 2005 11:43 AM > To: Info-VAX@Mvb.Saic.Come+ > Subject: Re: VMS 7.3-2 with 48 MB memory?e >=20( > On Thu, 7 Apr 2005, Bob Koehler wrote: >=20+ > > In article <d2h4j8$hes$3@online.de>,=20o= > helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove=20  > CLOTHES to reply) writes:a > > >a@ > > > However, the release notes say that 7.3-2 has a minimum=20 > memory of 64 MB. > >o@ > >    64 MB was the minimum for Alpha at least back to 6.0. =20 > Probably for: > >    1.5 .  I don't recall DEC ever offering a system=20 > building block with  > >    less than 64 MB.G >=20A > When the 3000/400 came out with AXP/VMS 1.5 back in 1992, it=20t > was sold with A > a base of 32MB. We bought several of them, and upgraded them=20h
 > all to 64MB 7 > almost immediately since it was soooooo slow with 32.y >=20 > Steve  >=20   Well,d  ? I have a 32MB AXP 150 running VMS V7.3-2 with no issues. Course<E depending on the App, you will have prob's. As an example I could notOH load Adv Server with less than 64MB memory. However, for misc DCL /small! programming type stuff, it is ok.>   RegardsO  
 Kerry Main Senior Consultantd HP Services Canada Voice: 613-592-4660e Fax: 613-591-4477> kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  $ "OpenVMS has always had integrity .. Now, Integrity has OpenVMS .."   ------------------------------  % Date: Thu, 07 Apr 2005 20:55:20 -0500t2 From: David J Dachtera <djesys.nospam@comcast.net>O Subject: Re: Which release notes say sts$manager:utc$configure_tdf  is obsoleteM* Message-ID: <4255E487.6AB5C32@comcast.net>   JF Mezei wrote:r > [snip]A > Is there really a point in having NTP run continuously ? [snip]w  H If you've never seen what goes on in healthcare, here's a brief insight:H the days when everyone bought "best of breed" before they considered howH well it would "play nice with others" (other software) created a monsterF out of what we casually call "transaction processing". These days, theD number of hard/software systems generating transactions at any givenF installation can be truly staggering. You wouldn't expect many changesH to be occuring in patients' records and such in the wee hours, but you'dB be surprised (alarmed, really!) at how many transactions cross theB "clearing house" when everyone except the medical staff and the ERH should be sound asleep. Even a few seconds of drift can wreak havoc when7 so many systems are passing information back and forth.o  H So, yes - making small clock adjustments a few times a day is not at all	 uncommon.a   -- s David J Dachtera dba DJE Systems- http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:<" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/D  " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/a   Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Fri, 08 Apr 2005 01:46:06 +0800m From: prep@prep.synonet.comoY Subject: Re: Which release notes say sts$manager:utc$configure_tdf  is obsolete obsoleteot- Message-ID: <873bu2zdi9.fsf@prep.synonet.com>   5 hammond@not@peek.ssr.hp.com (Charlie Hammond) writes:u  C > Yes, if things were different they wouldn't be the same.  <smile>y  B > My vote would be to keep ALL time in UTC and display it in local+ > time when and as necessary. {sigh} If....p  / No, use TAI and avoid the leap seconds as well.    -- i< 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: Fri, 08 Apr 2005 01:53:41 +0800a From: prep@prep.synonet.comoY Subject: Re: Which release notes say sts$manager:utc$configure_tdf  is obsolete obsoleteo-- Message-ID: <87y8buxyl6.fsf@prep.synonet.com>r  / JF Mezei <jfmezei.spamnot@teksavvy.com> writes:u  ; > Is there really a point in having NTP run continuously ? B  > Yes, the very core of NTP is a phase lock loop, plus a pile of? selection and voting algorithms to select the best time source.a> NTP tunes the clock speed to get the clock accurate, then keep= it running at the correct rate. Once stable, it backs off the)< interval between update of its servers. Mine is currently at< 128 sec to the upstream servers (dialup delay/variation) and 1024 to the alpha.   Note well, yet again:-  B NTP uses UTC, UTC and only UTC. I knows zip about daylight saving, and cares even less ;)   --  < 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, 07 Apr 2005 20:46:06 -0500D2 From: David J Dachtera <djesys.nospam@comcast.net>N Subject: Re: Which release notes say sts$manager:utc$configure_tdf is obsolete+ Message-ID: <4255E25D.248E5B2D@comcast.net>e   tadamsmar wrote: > [snip]@ > If the time does not change immediately, then what does it do? > F > And, if it does not hange immediately, then how does turning off NTP > make a difference?  E Again, the issue is synchronization. *YOU* are changing the time. YouoC cannot guarantee when NTP will attempt its next time adjustment. TovG prevent "surprises", best to have only one "hand in the pot" at a time.e  D ...that means stopping NTP until after you've completed/verified the changes.  A You may "get lucky" and never have a problem letting NTP stay on,  until...  D ...and at that hour of the morning, I'm more concerned about drivingD home safely than I am with gambling whether or not NTP will cause an issue.   ...IMHO, YMMV...   -- w David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/u   Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Thu, 07 Apr 2005 22:12:31 -0400o- From: JF Mezei <jfmezei.spamnot@teksavvy.com>cN Subject: Re: Which release notes say sts$manager:utc$configure_tdf is obsoleteB Message-ID: <1112926311.5d1c5d7b340d0047b8ea60f6b117468c@teranews>   David J Dachtera wrote:bG ~> ...that means stopping NTP until after you've completed/verified theo	 > changesm    F No. it means haveing NTP software that is truly integrated into VMS soF that changing the UTC offset would be done though the SET TIME commandH which would also change the time offset logicals and data structures, as, well as tell NTP to adapt to the new offset.  G SET TIME should also have a /NTP=SUSPEND so that you could temporartilye? set the time to some arbirtary value (for instance, to test forK! something in the past or future).    ------------------------------  # Date: Thu, 07 Apr 2005 19:10:02 GMT " From:   VAXman-  @SendSpamHere.ORG) Subject: Re: [TCPIP V5.4 ECO4] SFTP rantst0 Message-ID: <00A41F0C.562B3CCA@SendSpamHere.ORG>  e In article <425572e0$1@NEWS.LANGSTOEGER.AT>, peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) writes:o ><RANT>iO >Maybe I repeat myself, but I'd like to ask, can we have a better SSH/SFTP/SCP,y	 >please ?n >P7 >1) SFTP doesn't exit with ^Z/F10 (in TCPware it does).tI >With F10 it brings spurious chars (1~), with ^Z it cries ** Interrupt **nD >It also has problems with command abbreviation and command recalls.F >It also has problems with prompt reappearance after a longer filelist1 >(you need to type a key to get the prompt back).c >sH >2) SFTP doesn't support ascii/auto/binary commands (in TCPware it does); >Is there no requirement to differentiate on VMS nowadays ?.3 >It also has no support for VMS file transfer mode.o > J >3) SFTP doesn't support OpenVMS VAX (TCPIP V5.3 without SSH is the latestM >or last for the VAX and, you guessed it, TCPware does support VAX with all).  >rA >4) SFTP has a lot less supported commands (than in TCPware) likeeG >"buffersize", "debug", "delete", "directory", "status", "verbose", ...0 > K >5) SFTP "ls -l" doesn't list extended informations (like size, owner, ...)p@ >though it is documented that it should !! (works OK in TCPware) > I >6) SFTP "ls -R" tries to list all files as directories - giving messagesiL >"No such file or directory" (TCPware does list directories as [.directory]) > H >7) Win32 App FILEZILLA doesn't list files from VMS/TCPIP via SFTP/SSH2.5 >Could be the "ls -l" problem or yet another problem.n >tB >And the list goes on (eg. haven't talked about performance, yet).K >(And I also haven't seen improvement in V5.5 yet, but I need to test more)- ></RANT> >eG >Please consider paying some bucks to PSC and get a better toolset than G >continue porting yourself for yet another couple of years. Or at leastiN >coordinate with them so that we can run TCPWARE$SFTP2.EXE on top of TCPIP ;-)    , ... and ssh -X support would be so nice too.     -- )K VAXman- A Bored Certified VMS Kernel Mode Hacker   VAXman(at)TMESIS(dot)COMo            w5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------   End of INFO-VAX 2005.195 ************************