1 INFO-VAX	Tue, 05 Oct 2004	Volume 2004 : Issue 553       Contents: Advanced Server problem  Re: Advanced Server problem  Re: Advanced Server problem  Re: Advanced Server problem  Re: As seen in WSJ Re: As seen in WSJ Re: As seen in WSJ Re: As seen in WSJ Re: As seen in WSJ% RE: C Decompiler for HP Alpha OpenVMS 4 CDRECORD alters process priority and leaves it high?8 Re: CDRECORD alters process priority and leaves it high?8 Re: CDRECORD alters process priority and leaves it high?8 Re: CDRECORD alters process priority and leaves it high? CMS V3.3 Re: CMS V3.3 DCL-like syntax standard DEC C: printf %.*s versus %*.s" Re: DEC C: printf %.*s versus %*.s" Re: DEC C: printf %.*s versus %*.s" Re: DEC C: printf %.*s versus %*.s, DECC: sizeof assigned to single byte integer0 Re: DECC: sizeof assigned to single byte integer* Re: DecWindows performance on old hardware% DECwindows: UIL multilanguage support ) Re: DECwindows: UIL multilanguage support ) Re: DECwindows: UIL multilanguage support  Disabling Video Card for X Boot # Re: Disabling Video Card for X Boot % Re: Hobbyist site licenses DOWN AGAIN  HP trying to defend against Sun # Re: HP trying to defend against Sun # Re: HP trying to defend against Sun ; Re: Kyocera printer on two (2) IP OVMS 7.2-1 doesn't print.  Re: light browser? More Intel/HP Developer Forums  Re: MS to lose its patent on FAT Re: Newbie DEBUG question  Re: Newbie DEBUG question  Re: Newbie DEBUG question B Re: Obsucure Inquier article about Intel mentions VMS (and not anyP Re: Obsucure Inquier article about Intel mentions VMS (and not any other OS') otP Re: Obsucure Inquier article about Intel mentions VMS (and not any other OS') ot Re: OpenVMS newbie Re: OT: Sun's fighting chance  Re: OT: Sun's fighting chance + Rules for commenting out continuation lines / Re: Rules for commenting out continuation lines / Re: Rules for commenting out continuation lines ' Sun can't invent - resorts to stealing! + Re: Sun can't invent - resorts to stealing!  Tel. Co. Sys. Op.?  Update Multiple decnet databases$ Re: VAX/VMS in industrial automation$ Re: VAX/VMS in industrial automation$ Re: VAX/VMS in industrial automation$ Re: VAX/VMS in industrial automation$ Re: VAX/VMS in industrial automationP Re: www.hp.com/go... URL links (was: Re: Maximum file size, FAQ v. "Guide to OpeP Re: www.hp.com/go... URL links (was: Re: Maximum file size, FAQ v. "Guide to Ope@ [OpenVMS Alpha V7.3-2] What Priv for SET SECURITY/CLASS=DEVICE ?. [OT]: Software Disasters Often People Problems2 Re: [OT]: Software Disasters Often People Problems  F ----------------------------------------------------------------------  $ Date: Tue, 5 Oct 2004 12:00:11 +0300- From: "Kari Keronen" <kari.keronen@digita.fi>   Subject: Advanced Server problem5 Message-ID: <tot8d.151$Gz.38@reader1.news.jippii.net>   G I have two member OpenVMS 7.3-2 cluster with Advanced Server 7.3A ECO 3 J running in both of them. Both servers are members in Windows domain. Other is fine:   TKJR\\AHMA> sho comp ahma    Computers in domain "TKJR":    Computer Type Description   L -------------------- --------------------------- --------------------------- --  C [SV] AHMA OpenVMS (NT 4.0) Server Advanced Server V7.3A for OpenVMS   
 Other is not:    TKJR\\AHMA> sho comp susi   > %PWRK-E-ERRGETSRV, error getting server information for "SUSI"  9 -LM-E-ERROR_BAD_NETPA, the network path cannot be located    And looking from SUSI:   TKJR\\SUSI> sho comp susi    Password for server SUSI:   > %PWRK-E-ERRGETSRV, error getting server information for "SUSI"  1 -LM-E-ERROR_INVALID_P, invalid password specified   L It's asking password, so I suppose there is something wrong in communicationC between SUSI and PDC. Does anyone know how to reset the situation ?    -Kari-   ------------------------------  $ Date: Tue, 5 Oct 2004 09:38:48 -0400$ From: "PEN" <paul.nuneznosp@mhp.com>$ Subject: Re: Advanced Server problem, Message-ID: <cju85c$q5n$1@hplms2.hpl.hp.com>   Hi Kari,  5 First check that the pwrk$lmsrv process exists using:   # $ @sys$startup:pwrk$define_commands  $ pwver   B If it's not running, check the tail end of these files for errors:  1 $ type/tail pwrk$lmlogs:pwrk$lmsrv_<nodename>.log 1 $ type/tail pwrk$lmlogs:pwrk$lmmcp_<nodename>.log   < as well as the TCP/IP helper process - pwrk$knbdaemon - log:  . $ type pwrk$logs:pwrk$knbdaemon_<nodename>.log  L If all processes are running, check whether the NETLOGON service started on  that cluster member:   $ admin show services   L If it's not in the list displayed, it didn't start and there should be some I accompanying error in the System event log.   But even if it is Started,  M check the system event log (it's common to the cluster so you can check from   the working node):  . $ admin show event/full/since=yesterday  ! Or 0 /since=<date-AdvSrv-was-started on problem node>  I Also check the common event log (it too is shared among cluster members):   J $ admin/analyze/since=yesterday   ! Or /since=<date-AdvSrv-was-started on 
 problem node>      HTH,   Paul    9 "Kari Keronen" <kari.keronen@digita.fi> wrote in message  / news:tot8d.151$Gz.38@reader1.news.jippii.net... H >I have two member OpenVMS 7.3-2 cluster with Advanced Server 7.3A ECO 3L > running in both of them. Both servers are members in Windows domain. Other
 > is fine: >  > TKJR\\AHMA> sho comp ahma  >  > Computers in domain "TKJR":  >  > Computer Type Description  > N > -------------------- --------------------------- --------------------------- > -- > E > [SV] AHMA OpenVMS (NT 4.0) Server Advanced Server V7.3A for OpenVMS  >  > Other is not:  >  > TKJR\\AHMA> sho comp susi  > @ > %PWRK-E-ERRGETSRV, error getting server information for "SUSI" > ; > -LM-E-ERROR_BAD_NETPA, the network path cannot be located  >  > And looking from SUSI: >  > TKJR\\SUSI> sho comp susi  >  > Password for server SUSI:  > @ > %PWRK-E-ERRGETSRV, error getting server information for "SUSI" > 3 > -LM-E-ERROR_INVALID_P, invalid password specified  > A > It's asking password, so I suppose there is something wrong in   > communication E > between SUSI and PDC. Does anyone know how to reset the situation ?  >  > -Kari- >  >  >  >  >  >    ------------------------------  % Date: Tue, 05 Oct 2004 09:08:11 -0500 & From: Thomas Wirt <twnews@kittles.com>$ Subject: Re: Advanced Server problem* Message-ID: <4162AACB.7020308@kittles.com>   Kari Keronen wrote:   I > I have two member OpenVMS 7.3-2 cluster with Advanced Server 7.3A ECO 3 L > running in both of them. Both servers are members in Windows domain. Other
 > is fine: >  > TKJR\\AHMA> sho comp ahma  >  > Computers in domain "TKJR":  >  > Computer Type Description  > N > -------------------- --------------------------- --------------------------- > -- > E > [SV] AHMA OpenVMS (NT 4.0) Server Advanced Server V7.3A for OpenVMS  >  > Other is not:  >  > TKJR\\AHMA> sho comp susi  > @ > %PWRK-E-ERRGETSRV, error getting server information for "SUSI" > ; > -LM-E-ERROR_BAD_NETPA, the network path cannot be located  >  > And looking from SUSI: >  > TKJR\\SUSI> sho comp susi  >  > Password for server SUSI:  > @ > %PWRK-E-ERRGETSRV, error getting server information for "SUSI" > 3 > -LM-E-ERROR_INVALID_P, invalid password specified  > N > It's asking password, so I suppose there is something wrong in communicationE > between SUSI and PDC. Does anyone know how to reset the situation ?  >  > -Kari-  F As always, Paul's advice in the other thread is good and sound, but I D have a couple of quick and simple suggestions.  First, I would just H restart AS.  If you are anywhere as busy as me, you would rather fix it I than know why it was broke.  Until I see a problem like this twice, I do  ! not worry about understanding it.   I Second thing that I would try if an AS restart did not work is to remove  G SUSI from the Domain and then add it back in.  This can be done easily  H with the Server manager.  I have not done this with AS on the VMS side. I   I can not address the risks of doing this, but I have had very similar  G problems with regular old Windows servers that were solved by removing  1 the member from the Domain and adding it back in.   G You might also delete any SUSI entries on your WINS server.  And check  9 your DNS to make sure that the entries there are correct.   
 Good luck, --     Thomas Wirt  Systems Manager  Kittle's Home Furnishings  Indianapolis, IN   ------------------------------  $ Date: Tue, 5 Oct 2004 11:13:05 -0400$ From: "PEN" <paul.nuneznosp@mhp.com>$ Subject: Re: Advanced Server problem, Message-ID: <cjudm5$3rp$1@hplms2.hpl.hp.com>   Hello again,  4 "Thomas Wirt" <twnews@kittles.com> wrote in message $ news:4162AACB.7020308@kittles.com... > M > As always, Paul's advice in the other thread is good and sound, but I have  M > a couple of quick and simple suggestions.  First, I would just restart AS.  K > If you are anywhere as busy as me, you would rather fix it than know why  M > it was broke.  Until I see a problem like this twice, I do not worry about   > understanding it.   M Appreciate that, Thomas :).   And I can certainly appreciate your suggestion   above...   > K > Second thing that I would try if an AS restart did not work is to remove  I > SUSI from the Domain and then add it back in.  This can be done easily  L > with the Server manager.  I have not done this with AS on the VMS side. I G > can not address the risks of doing this, but I have had very similar  M > problems with regular old Windows servers that were solved by removing the  / > member from the Domain and adding it back in.  >   > Stop.  Don't do it.  Actually, you'll likely find there is no 1 computer/machine account for SUSI (nor for AHMA).   L The one (and only) computer account for an Advanced Server _cluster_ (event K a one-node cluster) will always be the name of the Advanced Server cluster  K alias (the cluster is regarding as a single domain member in this regard).  L If computer accounts for SUSI and AHMA exist, you might as well just delete ( 'em, cuz Advanced Server won't use them.  J If you remove the computer account for the Advanced Server cluster alias, M then none of the cluster members running Advanced Server will function again  G (well, NETLOGON won't start which means no user can get validated when  J mapping drives) until you run pwrk$config (which retrieves and stores the J computer account password/secret in the local SAM so when Advanced Server L starts its password is in sync with that of the account created on the PDC).  M Since the account is used by all cluster members running Advanced Server and  L since AHMA is functioning, I don't think the machine account is the problem = (but ya never know, because AHMA may have started before the  L account/password went bad).   The machine account password is changed every H 7 days (Advanced Server and Windows NT.  W2K/XP/W2K3 change it every 30  days).     Paul     ------------------------------  % Date: Tue, 05 Oct 2004 11:10:25 +0100 O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com>  Subject: Re: As seen in WSJ 0 Message-ID: <cjtrui$3uu$2@new-usenet.uk.sun.com>  
 dooley wrote: J > fabiopenvms@yahoo.com.br (Fabio Cardoso) wrote in message news:<f30679f=, b.0409290342.2dd606f5@posting.google.com>... >=20J >>Keith Parris <keithparris_NOSPAM@yahoo.com> wrote in message news:<Oql6=% d.11915$m86.8560@news.cpqcorp.net>...  >> >>>John Smith wrote: >>> 2 >>>>Hewlett-Packard Co. said it will stop offeringA >>>>desktop workstations based on Intel's Itanium microprocessor.  >>> 9 >>>Business as usual, and zero impact on OpenVMS users...  >>> J >>>We haven't had OpenVMS workstations for many years on Alpha, either, b= ut=20 J >>>that hasn't stopped people from putting graphics cards into servers (o= r=20G >>>simply leaving in the graphics card which comes with every server=20 G >>>anyway) and making their own workstations. Folks can and will use=20 J >>>rx1600, rx2600, even rx4640 systems as workstations under OpenVMS, and= =20 G >>>they'll continue to be supported by OpenVMS Engineering in doing so.  >>J >>Workstations would be an interesting choice for Industrial Automation, =  D >>LIMS, etc ...but for me it=B4s important to have Itanium blades=20E >>to run OpenVMS clusters. These blades could be used as workstations 5 >>in a BladePC-like aproach. Any date of release ?=20  >> >=20. > HPs current "blade pc" uses transmeta chips,* > a Transmeta Efficeon TM8000 to be exact,8 > is there going to be an opteron-like version of these? > Phil  9 HP are apparently developing Opteron based blades, Sun is % definitely developing Opteron blades.    Regards  Andrew Harrison    ------------------------------  % Date: Tue, 05 Oct 2004 12:12:17 +0100 O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com>  Subject: Re: As seen in WSJ 0 Message-ID: <cjtvii$5h1$1@new-usenet.uk.sun.com>   Bob Ceculski wrote:  > Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<cjrkht$7nr$1@new-usenet.uk.sun.com>...  > @ >>>>>you are in left field as usual Andrew ... what will make or6 >>>>>break itanium as a viable chip are two things ... >>>>>  >>>>>1. OpenVMS 3 >>>>>2. the ev8 alpha team to make itanium ev8 like  >>>>> D >>>>>OpenVMS will make or break itanium ... that is the bottom lime! >>>>< >>Really so imagine how your theory will work with Intel the# >>owners and developers of Itanium.  >>= >>Guess what an argument that follows the lines of it doesn't / >>matter about volumes will play with Intel ???  >>9 >>Remember Intel doesn't have a platform (or much of one) > >>or services or OS an revenue stream that gets pulled through7 >>by each Itanium chip that is sold. They do make money 8 >>on support chipsets but HP, by far in away the largest= >>OEM doesn't buy from them, they also make some money out of * >>compilers but again HP supply their own. >>9 >>So all on all what Intel is interested in is volume and ; >>if they don't get it they won't be interested any longer.  >>? >>But thanks for lettimg me know that I failed to get it again.  >>	 >>Regards  >>Andrew Harrison  >  > : > I said vms would be the spark that would lead to volume!9 > Something has to spark interest before there are volume  > sales ... hello!    5 So how do you think this spark is going to ignite the 6 Itanium when the same skark cannot even ignite its own market ?   Regards  Andrew Harrison    ------------------------------   Date: 5 Oct 2004 08:17:28 -0700 ' From: icerq4a@spray.se (David Svensson)  Subject: Re: As seen in WSJ = Message-ID: <734da31c.0410050717.314c604c@posting.google.com>    Andrew Harrison SUNUK Consultancy <Andrew No.Harrison No@nospamn.sun.com> wrote in message news:<cjtrui$3uu$2@new-usenet.uk.sun.com>...  > dooley wrote: K > > fabiopenvms@yahoo.com.br (Fabio Cardoso) wrote in message news:<f30679f / >  b.0409290342.2dd606f5@posting.google.com>...  > > K > >>Keith Parris <keithparris NOSPAM@yahoo.com> wrote in message news:<Oql6 ( >  d.11915$m86.8560@news.cpqcorp.net>... > >> > >>>John Smith wrote: > >>> 4 > >>>>Hewlett-Packard Co. said it will stop offeringC > >>>>desktop workstations based on Intel's Itanium microprocessor.  > >>> ; > >>>Business as usual, and zero impact on OpenVMS users...  > >>> K > >>>We haven't had OpenVMS workstations for many years on Alpha, either, b  >  ut K > >>>that hasn't stopped people from putting graphics cards into servers (o  >  r  G > >>>simply leaving in the graphics card which comes with every server  G > >>>anyway) and making their own workstations. Folks can and will use  K > >>>rx1600, rx2600, even rx4640 systems as workstations under OpenVMS, and  >   I > >>>they'll continue to be supported by OpenVMS Engineering in doing so.  > >>K > >>Workstations would be an interesting choice for Industrial Automation,   >   B > >>LIMS, etc ...but for me it s important to have Itanium blades G > >>to run OpenVMS clusters. These blades could be used as workstations 5 > >>in a BladePC-like aproach. Any date of release ?   > >> > > 0 > > HPs current "blade pc" uses transmeta chips,, > > a Transmeta Efficeon TM8000 to be exact,: > > is there going to be an opteron-like version of these? > > Phil > ; > HP are apparently developing Opteron based blades, Sun is ' > definitely developing Opteron blades.   9 Sun has dropped it's development of Opteron based blades.    ------------------------------   Date: 5 Oct 2004 08:13:54 -0700 ' From: icerq4a@spray.se (David Svensson)  Subject: Re: As seen in WSJ = Message-ID: <734da31c.0410050713.3468e0e8@posting.google.com>    Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<cjjnra$k0u$1@new-usenet.uk.sun.com>...  > 2.	The EV8 team. > 7 > Some of the EV8 team have made it to Intel, most have 5 > however have left to joind AMD etc and to be honest 8 > whatever they do will be too late. The fate of Itanium; > will be decided in the next 18 months and any impact from 3 > the EV8 team will be well outside that timeframe.   L Not true. Most (I heard from a good source, _all_) EV8 people work at Intel.  P Alpha people in general before EV7 have gone to many companies during the years.   ------------------------------   Date: 5 Oct 2004 15:34:16 GMT ( From: bill@cs.uofs.edu (Bill Gunshannon) Subject: Re: As seen in WSJ + Message-ID: <2sft7oF1kg52nU1@uni-berlin.de>   = In article <734da31c.0410050717.314c604c@posting.google.com>, * 	icerq4a@spray.se (David Svensson) writes: > Andrew Harrison SUNUK Consultancy <Andrew No.Harrison No@nospamn.sun.com> wrote in message news:<cjtrui$3uu$2@new-usenet.uk.sun.com>...  >> dooley wrote:L >> > fabiopenvms@yahoo.com.br (Fabio Cardoso) wrote in message news:<f30679f0 >>  b.0409290342.2dd606f5@posting.google.com>... >> >  L >> >>Keith Parris <keithparris NOSPAM@yahoo.com> wrote in message news:<Oql6) >>  d.11915$m86.8560@news.cpqcorp.net>...  >> >>  >> >>>John Smith wrote:  >> >>>5 >> >>>>Hewlett-Packard Co. said it will stop offering D >> >>>>desktop workstations based on Intel's Itanium microprocessor. >> >>>< >> >>>Business as usual, and zero impact on OpenVMS users... >> >>>L >> >>>We haven't had OpenVMS workstations for many years on Alpha, either, b >>  ut  L >> >>>that hasn't stopped people from putting graphics cards into servers (o >>  r H >> >>>simply leaving in the graphics card which comes with every server H >> >>>anyway) and making their own workstations. Folks can and will use L >> >>>rx1600, rx2600, even rx4640 systems as workstations under OpenVMS, and >>    J >> >>>they'll continue to be supported by OpenVMS Engineering in doing so. >> >> L >> >>Workstations would be an interesting choice for Industrial Automation,  >>  C >> >>LIMS, etc ...but for me it s important to have Itanium blades  H >> >>to run OpenVMS clusters. These blades could be used as workstations6 >> >>in a BladePC-like aproach. Any date of release ?  >> >>  >> >  1 >> > HPs current "blade pc" uses transmeta chips, - >> > a Transmeta Efficeon TM8000 to be exact, ; >> > is there going to be an opteron-like version of these? 	 >> > Phil  >>  < >> HP are apparently developing Opteron based blades, Sun is( >> definitely developing Opteron blades. > ; > Sun has dropped it's development of Opteron based blades.   D Care to justify this?  A quick search using Google turned up severalD pages of articles announcing Sun's plans for Opteron Blades and none( announcing they had abandoned this plan.   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------   Date: 5 Oct 2004 08:02:09 -0500 ; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) . Subject: RE: C Decompiler for HP Alpha OpenVMS3 Message-ID: <o3yHeHJN+P1M@eisner.encompasserve.org>    In article <6A998621840A594EAFD672A543B14A0408F35E9F@usmbcusmkexmb01.mbc.com>, "Hipenbecker, Doug" <hipenbecker.doug@mbco.com> writes: > C > OK...curiosity killed the cat.  I ran into a bug where some old C N > executables compiled/linked as debug from 1995 won't run when switching from: > an Alpha EV56 to Alpha EV67.  I'm going down the path toL > recompile/relink...but...I'm just curious on the assembler option.  By the7 > way, it is Macro-64...we have the compiler installed.   F    Is it C or is it Macro-64?  This really looks like you're trying toE    dissasemble some code and then get it to build.  Are you sure your 0    license doesn't prohibit reverse engineering?   > K > Am I missing the boat here?...maybe you could advise.  I put the 1st four G > statements at the top and the one at the bottom...not sure if they're  > correct.  Here's a sample:  E    This code is full of errors and I've never seen any version of the A    macro assembler which would assemble it.  The notation for hex C    constants is ^X, not 0X.  There are only two operands for a JSR, E    the second should be a label, we've no way to know what label from     what you've given us.  %       jsr            R26,^x2B0C0(R26)   F    and the BSR is trying to use a hard coded offset.  It, too, should G    branch to a label, but I can't tell which from what you've given us.    ------------------------------  * Date: Tue, 5 Oct 2004 01:12:41 -0500 (CDT) From: sms@antinode.org= Subject: CDRECORD alters process priority and leaves it high? ) Message-ID: <04100501124167@antinode.org>   H    I've been using CDRECORD 1.8.1 for a fairly long time, and I recentlyF noticed that it uses nice( -20) to raise its process base priority (toE help prevent buffer exhaustion).  I discovered this because I noticed H that my process base priority was unusually high (24, up from 4) _after_C I had run it.  (Other activity on the system was strangely slow...)   A    On a UNIX system, this would not be a problem, as each program H execution gets its own process, so there the altered priority disappears> when CDRECORD exits.  On VMS it persists after CDRECORD exits.  G    I have not looked at any later versions of CDRECORD, so I don't know E if this behavior continues in current versions.  I've written a VMS.C H and VMS.H which, with minor modifications to CDRECORD.C (and the compileF and link procedure), establish an exit handler, which uses nice( 0) to= restore the original process base priority as CDRECORD exits.   A   I normally link CDRECORD /NOTRACEBACK, and then install it with H ALTPRI, DIAGNOSE, and PHY_IO, so this might be a more dangerous thing if normal users can do it, too.  F    If this code might be useful to anyone, let me know and I'll put itD somewhere public.  (If this wheel's already been invented, I'll just leave it where it lies.)  H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-org     Saint Paul  MN  55105-2547    ------------------------------   Date: 5 Oct 2004 10:42:47 +0200 C From: vaxinf@chclu.chemie.uni-konstanz.de (Eberhard Heuser-Hofmann) A Subject: Re: CDRECORD alters process priority and leaves it high? - Message-ID: <41625e87$1@news.uni-konstanz.de>   B In article <04100501124167@antinode.org>, sms@antinode.org writes:I >   I've been using CDRECORD 1.8.1 for a fairly long time, and I recently G >noticed that it uses nice( -20) to raise its process base priority (to F >help prevent buffer exhaustion).  I discovered this because I noticedI >that my process base priority was unusually high (24, up from 4) _after_ D >I had run it.  (Other activity on the system was strangely slow...) > B >   On a UNIX system, this would not be a problem, as each programI >execution gets its own process, so there the altered priority disappears ? >when CDRECORD exits.  On VMS it persists after CDRECORD exits.  > H >   I have not looked at any later versions of CDRECORD, so I don't knowF >if this behavior continues in current versions.  I've written a VMS.CI >and VMS.H which, with minor modifications to CDRECORD.C (and the compile G >and link procedure), establish an exit handler, which uses nice( 0) to > >restore the original process base priority as CDRECORD exits. > B >  I normally link CDRECORD /NOTRACEBACK, and then install it withI >ALTPRI, DIAGNOSE, and PHY_IO, so this might be a more dangerous thing if  >normal users can do it, too.  > G >   If this code might be useful to anyone, let me know and I'll put it E >somewhere public.  (If this wheel's already been invented, I'll just  >leave it where it lies.)   E I never saw any improvement when I altered the priority. Nowadays all E recorders have a buffer underrun option which starts working when the I input stream is empty. This helps you to burn DVDs/CDs even on a MicroVAX  II/III.   G If your recorder is an very old one (<1999) it might help in some very, C very rare circumstances so there is need to implement new VMS code.   B One task I'm still planning to do: an official tested and acceptedC cdrecord/mkisofs/readcd contribution. Who ever had contact with the = official author of these tools know what challenge this is...   	 eberhard     ------------------------------   Date: 5 Oct 2004 03:32:57 -0700 % From: Bart.Zorn@xs4all.nl (Bart Zorn) A Subject: Re: CDRECORD alters process priority and leaves it high? = Message-ID: <a98cd882.0410050232.2835f95b@posting.google.com>    Yes, I have seen that too.  ( Somewhere in SYS$MANAGER:CDRECORD.COM, a  # $ cur_priority = f$getjpi("","PRI")    is done, which should be  % $ cur_priority = f$getjpi("","PRIB").   ) Just adding this 'B' solves this problem.    Regards,  	 Bart Zorn   G sms@antinode.org wrote in message news:<04100501124167@antinode.org>... G > I've been using CDRECORD 1.8.1 for a fairly long time, and I recently H > noticed that it uses nice( -20) to raise its process base priority (toG > help prevent buffer exhaustion).  I discovered this because I noticed J > that my process base priority was unusually high (24, up from 4) _after_E > I had run it.  (Other activity on the system was strangely slow...)  > C >    On a UNIX system, this would not be a problem, as each program J > execution gets its own process, so there the altered priority disappears@ > when CDRECORD exits.  On VMS it persists after CDRECORD exits. > I >    I have not looked at any later versions of CDRECORD, so I don't know G > if this behavior continues in current versions.  I've written a VMS.C J > and VMS.H which, with minor modifications to CDRECORD.C (and the compileH > and link procedure), establish an exit handler, which uses nice( 0) to? > restore the original process base priority as CDRECORD exits.  > C >   I normally link CDRECORD /NOTRACEBACK, and then install it with J > ALTPRI, DIAGNOSE, and PHY_IO, so this might be a more dangerous thing if > normal users can do it, too. > H >    If this code might be useful to anyone, let me know and I'll put itF > somewhere public.  (If this wheel's already been invented, I'll just > leave it where it lies.) > J > ------------------------------------------------------------------------ > 6 >    Steven M. Schweda               (+1) 651-699-98185 >    382 South Warwick Street        sms@antinode-org  >    Saint Paul  MN  55105-2547    ------------------------------  * Date: Tue, 5 Oct 2004 08:33:30 -0500 (CDT) From: sms@antinode.orgA Subject: Re: CDRECORD alters process priority and leaves it high? ) Message-ID: <04100508333081@antinode.org>   C From: vaxinf@chclu.chemie.uni-konstanz.de (Eberhard Heuser-Hofmann)   G > I never saw any improvement when I altered the priority. Nowadays all G > recorders have a buffer underrun option which starts working when the K > input stream is empty. This helps you to burn DVDs/CDs even on a MicroVAX 	 > II/III.   H    I have an older Yamaha drive which still can make a bad disc when itsH buffer is exhausted.  Before I installed the CDRECORD image with ALTPRI,F I had seen problems as an unprivileged user which went away when I wasE SYSTEM.  (This is on an AlpSta 200 4/233, but a busy system is a busy  system, I reckon.)  I > If your recorder is an very old one (<1999) it might help in some very, E > very rare circumstances so there is need to implement new VMS code.   F    Not so rare, in my experience.  Also, there's no new code needed toA get the priority _up_.  Building with CC /DEFINE = "HAVE_NICE" is H enough for that.  (I didn't really look at the code, I just figured thatC I had nice(), so it was true.)  The problem (on VMS) is getting the  priority back _down_.   F    Did you say "very old"?  Around here, 1999 is brand-spanking _new_.  E    If it's possible to establish the exit handler in a LIB$INITIALIZE F gizmo, that would obviate VMS.H and changes to CDRECORD.C.  You'd needE only to compile VMS.C and link it in.   Hmmm.  CDRECORD seems to have G one upper-case option ("-V"), so perhaps there's a good excuse for this - (to get "DECC$ARGV_PARSE_STYLE") in any case.   % From: Bart.Zorn@xs4all.nl (Bart Zorn)   * > Somewhere in SYS$MANAGER:CDRECORD.COM, a% > $ cur_priority = f$getjpi("","PRI")  > is done, which should be' > $ cur_priority = f$getjpi("","PRIB"). + > Just adding this 'B' solves this problem.   D    It might if you use SYS$MANAGER:CDRECORD.COM, which I don't.  HowH does SYS$MANAGER:CDRECORD.COM deal with CRTL/Y?  Is it as reliable as an exit handler in CDRECORD?   = >    Privileges required to write cd's: diag, phy_io, sysprv. < >    Privilege required to alter execution priority: altpri.  H    Not much help for a peon.  And if the image is installed with ALTPRI,7 an exit handler is pretty much a necessity, I'd expect.   G    Do I really need SYSPRV, by the way?  Or is that for something else?   H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-org     Saint Paul  MN  55105-2547    ------------------------------  % Date: Tue, 05 Oct 2004 12:33:04 -0400 2 From: "Stanley F. Quayle" <squayle@insight.rr.com> Subject: CMS V3.3 / Message-ID: <41629480.13393.1CFC4FD6@localhost>   B I need the installation savesets for CMS V3.3 (VAX).  I have only  V3.4 and later...   
 --Stan Quayle  Quayle Consulting Inc.  
 ----------- Stanley F. Quayle, P.E. N8SQ  +1 614-868-1363 3 8572 North Spring Ct., Pickerington, OH  43147  USA 0 stan-at-stanq-dot-com       http://www.stanq.com   ------------------------------  * Date: Tue, 5 Oct 2004 16:54:27 +0000 (UTC)6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) Subject: Re: CMS V3.3 1 Message-ID: <newscache$jie45i$ru41$1@news.sil.at>   d In article <41629480.13393.1CFC4FD6@localhost>, "Stanley F. Quayle" <squayle@insight.rr.com> writes:B >I need the installation savesets for CMS V3.3 (VAX).  I have only >V3.4 and later...   Which DECset version is this ?   --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------   Date: 5 Oct 2004 05:27:49 -0700 . From: fabiopenvms@yahoo.com.br (Fabio Cardoso)! Subject: DCL-like syntax standard = Message-ID: <f30679fb.0410050427.141089f4@posting.google.com>    I have a doubt:   + Is the DCL-like syntax standard patented ?    = Can I develop a utility for another plaform (Linux) using it?    Example:  + > COMMAND /PARAMETER  i.e.  SHOW /PARAMETER   < The Lantronix software is similar to the DECserver software.7 But I dont know if DEC and Lantronix had an agreement.      Regards    FC   ------------------------------  % Date: Tue, 05 Oct 2004 02:43:33 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> ' Subject: DEC C: printf %.*s versus %*.s , Message-ID: <41624266.22D70AA9@teksavvy.com>  ( Ok, just spent many hours wondering why:  . printf("String=>%*.s<=\n" , length, address)    J didn't work in a program when that construct seemed to work in a differentI program. (unfortunatly, Google doesn't allow one to search on an esoteric  string like %*.s).  6 %*.s would print a blank string of the correct length.  K Turns out I had to use %.*s to get it working (and not understanding at all 0 why the %*.s seemed to work in another program).  N Does anyone have the exact explanation of what "%*.s" does versus "%.*s"  ????  H I am not comfortable using it if I don't understand why it works in someJ places and not other places. The DEC C booreader files I have don't really0 mention the * operator with strings for output.)   ------------------------------  % Date: Tue, 05 Oct 2004 10:29:12 +0100e- From: John Laird <nospam@laird-towers.org.uk> + Subject: Re: DEC C: printf %.*s versus %*.se8 Message-ID: <mhp4m0dp0v58c4p8pch627904ermrdusgn@4ax.com>  K On Tue, 05 Oct 2004 02:43:33 -0400, JF Mezei <jfmezei.spamnot@teksavvy.com>i wrote:  O >Does anyone have the exact explanation of what "%*.s" does versus "%.*s"  ????O  I The first is the same as %*.0s, the second %0.*s, the number before the .iF being of course the minimum field width, and the number afterwards theI precision, which for s fields is maximum number of chars transferred fromvF the string argument.  It should be clear then why %*.0s yields a blank? string in the output, of the length given by the next argument.    -- d+ Zen order: Make me none with everything...     Mail john rather than nospam...u   ------------------------------  % Date: Tue, 05 Oct 2004 05:52:57 -0400r- From: JF Mezei <jfmezei.spamnot@teksavvy.com> + Subject: Re: DEC C: printf %.*s versus %*.sy, Message-ID: <41626EF5.ECBDFA11@teksavvy.com>   John Laird wrote:eK > The first is the same as %*.0s, the second %0.*s, the number before the .2H > being of course the minimum field width, and the number afterwards theK > precision, which for s fields is maximum number of chars transferred fromqH > the string argument.  It should be clear then why %*.0s yields a blankA > string in the output, of the length given by the next argument.I  V many thanks. The above should go to the DEC C documentation if it isn't there already.  K Now I have to figure out how %*.s worked in another program :-) :-) :-) :-)g   ------------------------------  % Date: Tue, 05 Oct 2004 14:46:04 +0100t- From: John Laird <nospam@laird-towers.org.uk>u+ Subject: Re: DEC C: printf %.*s versus %*.sn8 Message-ID: <s495m0pmooobcn1g0hn86apobvthmlv99j@4ax.com>  K On Tue, 05 Oct 2004 05:52:57 -0400, JF Mezei <jfmezei.spamnot@teksavvy.com>i wrote:   >John Laird wrote:L >> The first is the same as %*.0s, the second %0.*s, the number before the .I >> being of course the minimum field width, and the number afterwards theML >> precision, which for s fields is maximum number of chars transferred fromI >> the string argument.  It should be clear then why %*.0s yields a blankxB >> string in the output, of the length given by the next argument. >fW >many thanks. The above should go to the DEC C documentation if it isn't there already.   F I found it in HELP CC with no trouble.  (But please don't ask me which version compiler, etc !)  L >Now I have to figure out how %*.s worked in another program :-) :-) :-) :-)  J Seems unlikely it did, to be honest.  I can't quite see the point of widthG and precision for strings unless the two are equal - which guarantees as/ fixed-width output regardless of string length.b   -- t) I can't remember if I used to know that. m   Mail john rather than nospam...    ------------------------------  % Date: Tue, 05 Oct 2004 04:14:38 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com>-5 Subject: DECC: sizeof assigned to single byte integerM, Message-ID: <416257BA.251763E9@teksavvy.com>   If I have this:n   char buffer[255];h unsigned char mylength ;  S mylength = sizeof(buffer); 	-> this generates warning that value would be truncatedoF mylength = 255 ;			-> compiler is a happy camper and doesn't complain.  J Shouldn't "sizeof" result in the compiler generating the "mylength = 255"  code ?  K What is it that makes sizeof complain ?  (I can understand if it were a 256tM bytes array, at which point sizeof woudl generate a number that could not fitw in a byte).P  N Sicne this is all executed at compile time with  known static values, is there@ a reason for the compiler to warn me about possible truncation ?   ------------------------------   Date: 5 Oct 2004 08:25:03 -0500g; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)h9 Subject: Re: DECC: sizeof assigned to single byte integera3 Message-ID: <Zq9lkG1dhmrP@eisner.encompasserve.org>u  \ In article <416257BA.251763E9@teksavvy.com>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes: > If I have this:t >  > char buffer[255];p > unsigned char mylength ; > U > mylength = sizeof(buffer); 	-> this generates warning that value would be truncateddH > mylength = 255 ;			-> compiler is a happy camper and doesn't complain. >   G    sizeof is an operator which returns an int.  We can see that in this E    case the int will have the value 255, but the compiler does't knowpH    that.  It should:  in this case sizeof should be evaluated at compile>    time and the constant value inserted in the resulting code.  E    I can't reporduce this with what you've got.  Maybe somehow you're .    hiding the constant size from the compiler?  E    OBTW, sizeof is an operator, not a function.  You can code this ase      mylength = sizeof buffer;  E    This is a common confusion because "sizeof (int)" is how you cast  F    "nothing" to int and then determine it's size.  In this case (int) 0    is a cast operator, not a function parameter.   ------------------------------   Date: 5 Oct 2004 01:37:06 -0700r. From: martinkirby12@yahoo.co.uk (Martin Kirby)3 Subject: Re: DecWindows performance on old hardwaren< Message-ID: <224291b.0410050037.2e77d07d@posting.google.com>  s Ulrich Bellgardt <Ulrich.bellgardt@debitel.net> wrote in message news:<416180f1$0$9618$9b622d9e@news.freenet.de>...- > Martin Kirby wrote:S > > Ulrich,M > > G > > During the V1.3 development there was an issue with the performance I > > creating a DECterm remotely. That was fixed. After the fix I timed it H > > as 2.4 seconds to create a DECterm, which was the same as on V1.2-6. > > I > How was that fixed? Is there a newer version around than the one I have  > at home (V1.3-1)?e >   5 The fix was included in all V1.3 and later versions. m   > C > You are right.Now that I think of it, the PF1-PF3 sequence is not G > a good example of poor DecWindows performance. I further studied that E > phenomenon and realized that in DWMOTIF V1.2-5 the sequence PF1-PF3h= > did exactly what I expected even if the PF1 key was not yet-< > released when pressing PF3, while V1.3 seems to expect theG > distinct sequence "PF1 - Release_PF1 - PF3 - Release_PF3 - [type whatpA > I'm looking for]". Maybe I should exercise precision in typing.w >   E I don't see this on the version ( a bit later than V1.3-1 but I can't C think of any relevant change since then ) that I have. I display tosE eXcursion and use EDIT/EDT, change to screen mode and press PF1 down,pF PF3 down and at that point I get the search prompt. My terminal is set; as a VT300_Series and the DECterm as VT300, 7-bit controls.-   > D > > Regarding running the 1.2-5 CDE login screen on V7.3-2. Try "runJ > > CDE$PATH:DECW$LOGINOUT". It should load and ACCVIO (because that isn't  > > how it is supposed to run).  > G > I tried this, and DECW$LOGINOUT crashes with ACCVIO as you predicted. C > Afterwards, DTGREET is active but no login screen. Apart from theh@ > stack dump, which does not tell me very much, I do not see any > further messages.s >   E Interesting. However, the accounting log shows that LOGINOUT couldn'tp> start. That suggests that the mismatch is between LOGINOUT andA CDE$PATH:DECW$LOGINOUT in which case I think you are out of luck.n   Martin Kirby   ------------------------------  % Date: Tue, 05 Oct 2004 03:01:35 -0400e- From: JF Mezei <jfmezei.spamnot@teksavvy.com>q. Subject: DECwindows: UIL multilanguage support, Message-ID: <4162469F.6BEB03C6@teksavvy.com>  J OK, An MRM based application starts.  As it initialises it specifies a UID< file to be loaded which contains the user interface details.  M What if I had a french and english set of UID files, and I wanted the correct : one to be loaded depending on user preference/environment.  I Is there a standard way for the application to obtain the user's preferedMK language ? (which would allow the application to start an automatically usem the right language UID).  K Or must some application specific logical/symbol be set to tell my code thea prefered language of the user ?    ------------------------------   Date: 5 Oct 2004 08:03:28 -0700 . From: martinkirby12@yahoo.co.uk (Martin Kirby)2 Subject: Re: DECwindows: UIL multilanguage support< Message-ID: <224291b.0410050703.7328d8c5@posting.google.com>  a JF Mezei <jfmezei.spamnot@teksavvy.com> wrote in message news:<4162469F.6BEB03C6@teksavvy.com>...hL > OK, An MRM based application starts.  As it initialises it specifies a UID> > file to be loaded which contains the user interface details. > O > What if I had a french and english set of UID files, and I wanted the correctd< > one to be loaded depending on user preference/environment. > K > Is there a standard way for the application to obtain the user's preferedyM > language ? (which would allow the application to start an automatically uses > the right language UID). > M > Or must some application specific logical/symbol be set to tell my code the,! > prefered language of the user ?p   I believe it will look in:    []     DECW$USER_DEFAULTS:    DECW$SYSTEM_DEFAULTS:  > in order and you handle this by changing the definition of theC logicals. Typically using a search list for DECW$SYSTEM_DEFAULTS so.# when French is selected it becomes:   4           "SYS$SYSROOT:[DECW$DEFAULTS.FRENCH.USER]" 5         = "SYS$SYSROOT:[DECW$DEFAULTS.FRENCH.SYSTEM]" -         = "SYS$SYSROOT:[DECW$DEFAULTS.USER]" g.         = "SYS$SYSROOT:[DECW$DEFAULTS.SYSTEM]"         = "SYS$LIBRARY:"  8 Then you install the UID file for your French version inB "SYS$SYSROOT:[DECW$DEFAULTS.FRENCH.SYSTEM]" and the English one in$ "SYS$SYSROOT:[DECW$DEFAULTS.SYSTEM]"  F Not having a localized DECwindows I may be wrong in just where the UID3 files are placed but I think that is the principle.n   Martin Kirby   ------------------------------  $ Date: Tue, 5 Oct 2004 11:30:43 -0400< From: "Peter Weaver" <WeaverConsultingServices@sympatico.ca>2 Subject: Re: DECwindows: UIL multilanguage support+ Message-ID: <2sft15F1ks0rrU1@uni-berlin.de>n   JF Mezei wrote:rF > OK, An MRM based application starts.  As it initialises it specifiesD > a UID file to be loaded which contains the user interface details. > G > What if I had a french and english set of UID files, and I wanted theuD > correct one to be loaded depending on user preference/environment. >pB > Is there a standard way for the application to obtain the user'sD > prefered language ? (which would allow the application to start an, > automatically use the right language UID). >kD > Or must some application specific logical/symbol be set to tell my* > code the prefered language of the user ?  C OK, I do not know what a MRM is, and I assume you are talking aboutsE DECWindows which I know nothing about so my answer may be 100% wrong,s but this is Usenet after all.-  A Chapter 27 of the Programming Concepts Manual talks about settingrD SYS$LANGUAGE and SYS$LANGUAGES for dates to display correctly. We do this on our boxes in Quebec    (LNM$SYSTEM_TABLE)  ' "LIB$DT_FORMAT" = "LIB$DATE_FORMAT_012"e        = "LIB$TIME_FORMAT_011"8 "LIB$DT_INPUT_FORMAT" = "!DB-!MAAU-!Y4 !H04:!M0:!S0.!C2" "SYS$LANGUAGE" = "CANADIAN"m "SYS$LANGUAGES" = "ENGLISH"e         = "FRENCH"         = "CANADIAN"   $ dir/date login.com   Directory USAGER_DIR:[me]   % LOGIN.COM;9 juillet 29, 2002 10:47 AMl   Total of 1 file.  F So I would hope that a package would look at SYS$LANGUAGE to determine what language I want to use.  G (But do not try to search for SYS$LANGUAGE on the broken ITRC, you willaC get 22,247 hits. The majority of these hits have nothing to do with> SYS$LANGUAGE.)   -- o Peter Weaver Weaver Consulting Services Inc.= Canadian VAR for CHARON-VAXt www.weaverconsulting.cam   ------------------------------  $ Date: Tue, 5 Oct 2004 10:34:32 -0400* From: "Bobby Coleman" <colemanrl@ornl.gov>( Subject: Disabling Video Card for X Boot( Message-ID: <cjubdo$hrk$1@sws1.ornl.gov>  M I have recently installed a Radeon 7500 video card into a DS20E workstation.  J The system originally used the Permedia graphics on the 3X-DEPVZ-AA combo J SCSI/Ether/Video card.  I moved the jumper on the combo card to "disable" J the video and installed the service packs for support of the Radeon card. I The system will boot with the Radeon card without issue (X boots fine as hL well) only if I remove the combo card first.  However, if I install both at M the same time (with the video disabled on the combo card) the X-Windows boot sJ procedure still finds the Permedia card and tries to initialize it.  This F returns an unrecoverable error (according to DECW$SERVER_0_ERROR.LOG).  L Is there any way to use this card (for ethernet/SCSI) alongside the Radeon? K Or... must I buy a separate ethernet NIC.  Thank you for any help that can eL be provided.  In case it is useful, I've pasted the last few lines from the K DECW$Server_0_error.log below.  Note that everything looks normal up until $
 this section..   Bobby.    ) -------------------- Exerpts from end of r3  DECW$Server_0_error.log --------------------------b  ! Permedia2: Init the backing store>. Permedia2: Exit from p2ScreenInit with success& Permedia2: Exit from init with success! Unsupported Depth 4 in Create GC!hK %SYSTEM-F-OPCDEC, opcode reserved to Digital fault at PC=000000000000FFFF, i PS=0l  K Unrecoverable server internal error (error code = 1084) found, terminating o all h Mapped Images...  -   START       END        LENGTH    IMAGE NAME -   -----       ---        ------    ----------o2        0      301ff      301ff    DECW$SERVER_MAIN1    32000     1e3bff     1b1bff    DECW$SERVER_DIX 3 7bbbe000   7bbeffff      31fff    DECW$SECURITY_VMSc+ 7b88c000   7b8fdfff      71fff    SECURESHR>, 7b374000   7b3f5fff      81fff    SECURESHRP7 7c846000   7c851fff       bfff    DECW$TRANSPORT_COMMON.0 81020120   81032bf0      12ad0    SYS$BASE_IMAGE/ 7bfa8000   7c03bfff      93fff    DECC$SHR_EV56 * 7bc7a000   7bcbffff      45fff    DPML$SHR- 7b92e000   7b93dfff       ffff    CMA$TIS_SHRa# ... plus a bunch more image info...o  " Exception Call stack dump follows:         PC     IMAGE+offset of call       --     --------------------!        4     DECW$SERVER_MAIN + 4 $    e817c     DECW$SERVER_DIX + b617c$    e82ec     DECW$SERVER_DIX + b62ec$   1044c4     DECW$SERVER_DIX + d24c4 PC     IMAGE+offset of call        --     --------------------!        4     DECW$SERVER_MAIN + 40$    e817c     DECW$SERVER_DIX + b617c$    e82ec     DECW$SERVER_DIX + b62ec$   1044c4     DECW$SERVER_DIX + d24c4  8 ********** marking the end of call stack dump **********   ------------------------------  % Date: Tue, 05 Oct 2004 17:51:43 +0100n- From: John Laird <nospam@laird-towers.org.uk> , Subject: Re: Disabling Video Card for X Boot8 Message-ID: <thj5m01lh7cb3cmrj6fr6ude0j9sj15agb@4ax.com>  G On Tue, 5 Oct 2004 10:34:32 -0400, "Bobby Coleman" <colemanrl@ornl.gov>  wrote:  M >Is there any way to use this card (for ethernet/SCSI) alongside the Radeon?    L Do both devices configure during startup ?  You're probably looking for GxA0F for some values of x.  If so, then there should be a documented way ofE asking configure (tis done with SYSMAN IO AUTO xxx) to ignore certaineI devices.  Alternatively, and this is probably easier, I am almost certainAJ you have more control over what Decwindows does when it starts much later.L (He says, quickly scanning sys$startup:decw*.com and not seeing anything tooI obvious...)  You could hack the appropriate decw$device*.com file to exito7 immediately as if the device didn't exist, for example.x  D Or wait until Fred tells you the proper way to achieve your goal ;-)   -- o9 Recent studies show that recent studies are meaningless.     Mail john rather than nospam...i   ------------------------------   Date: 5 Oct 2004 12:06:24 -0500r4 From: kuhrt@nospammy.encompasserve.org (Marty Kuhrt). Subject: Re: Hobbyist site licenses DOWN AGAIN3 Message-ID: <wjJv7wPRXUiD@eisner.encompasserve.org>t  O In article <_zr7d.650$1g5.150@trnddc07>, "operagost" <operagost@og.com> writes:  > E > "Bradford J. Hamilton" <brad@rabbit.dnsalias.org> wrote in message o* > news:puw6d.137314$D%.115608@attbi_s51...? >> In article <Tyq6d.113$Za.6@fe40.usenetserver.com>, Alphaman i* >> <alphaman-nix-spam@alphant.com> writes: >> !Alan Frisbie wrote:oA >> !> I had the same problem about a month ago when I was getting B >> !> a friend's system set up.   After waiting a week, I e-mailedD >> !> David Cathay and he replied that the procedure had been broken? >> !> but was now fixed.   I'm guessing that it's broken again.d >> !>H
 >> !> Alan >> !L >> !I left a message for David Cathey this afternoon regarding this issue; IJ >> !hope we can get to the bottom of it soon!  I'd hate to lose my 50 days, >> !of uptime due to expired licenses... 8^) >>2 >> David's usually been responsive - not to worry. >>E >> My license expired recently, but my system stayed up until I got a * >> "replacement" from openvmshobbyist.org. >>B >> http://uptimes.hostingwired.com/account.php?op=details&hid=8347 >> >> !F >> !  http://uptimes.hostingwired.com/account.php?op=details&hid=10205 >> >> UPS - no fair!!! :-)- > M > Neat - they didn't have a VAX client last time I checked (the Alpha client i > wouldn't compile).* > You'll just have to take my word for it: >  > Orff> show sys /nopr /fuF > OpenVMS V7.3 on node ORFF 2-OCT-2004 02:03:13.93 Uptime 314 09:36:19 >  > VAXstation 4000-60  G The client should work for VAXen.  If you look at the top thirty activenH hosts via http://uptimes.hostingwired.com/stats.php?op=active you'll seeC that (at this writing) number six is a VAX with 524 days of uptime.m  > One way to increase the uptime value is to use the client withH clustering turned on.  By default when you have clustering turned on theE client will report the cluster foundation time instead of the current>G system uptime.  It will also report itself as OpenVMSCluster instead oftE just plain OpenVMS for the OS.  This way you can reboot a node in theoG cluster, and still keep the uptime number.  Think of it as "applications availability uptime".e  A My home system is a single node cluster.  If I want to work on it H without clobbering my uptime number I bring another non-voting node intoF the cluster (which I do everyday anyway for the machine in my office),E crank the RECNXINTERVAL up to however long I think the fix will take,.D and fix the problem.  The non-voting node will stall while I fix theH main node, but the cluster foundation time will remain intact.  When the0 fixed main node rejoins the cluster all is well.   ------------------------------  % Date: Tue, 05 Oct 2004 02:27:37 -0400n- From: JF Mezei <jfmezei.spamnot@teksavvy.com>o( Subject: HP trying to defend against Sun, Message-ID: <41623EAA.90C04998@teksavvy.com>  b > http://news.com.com/HP+tells+Sun+to+get+its+facts+straight/2100-1016_3-5396064.html?tag=nefd.top  N HP has sent a letter asking Sun to stop denigrating HP's strategy with regards	 to HP-UX.i  M This means that Sun attacks against HP-UX/IA64 have had a noticeable negativer
 impact on HP.v  2 Sun said it woudln't widthdraw its ads/starements.  L Of course, the article does not mention VMS (no surprise there). But it doesH show that HP has realised that it is now on the defensive and that if itH doesn't defend itself, Sun's attacks will hurt HP's enterprise business.  K Sun's new focus on the 8086, combined with HP's statement that it would not J port its enterprise systems to the 8086 will give a lot of fuel to Sun and% hurt HP's momentum (or lack thereof).s   ------------------------------  % Date: Tue, 05 Oct 2004 10:09:16 -0400 ( From: David Froble <davef@tsoft-inc.com>, Subject: Re: HP trying to defend against Sun, Message-ID: <4162AB0C.4010904@tsoft-inc.com>   JF Mezei wrote:b  b >>http://news.com.com/HP+tells+Sun+to+get+its+facts+straight/2100-1016_3-5396064.html?tag=nefd.top >> > P > HP has sent a letter asking Sun to stop denigrating HP's strategy with regards > to HP-UX.e > O > This means that Sun attacks against HP-UX/IA64 have had a noticeable negative  > impact on HP.e > 4 > Sun said it woudln't widthdraw its ads/starements. > N > Of course, the article does not mention VMS (no surprise there). But it doesJ > show that HP has realised that it is now on the defensive and that if itJ > doesn't defend itself, Sun's attacks will hurt HP's enterprise business. > M > Sun's new focus on the 8086, combined with HP's statement that it would notbL > port its enterprise systems to the 8086 will give a lot of fuel to Sun and' > hurt HP's momentum (or lack thereof).m >   O When AMD began to 'HAMMER' on Intel's image they were at least smart enough to aO recognize the danger and executed a 'U' turn, regardless of the extremely sour sP taste it might have.  If the best HP can do ask Sun to quit trying to take HP's P customers then stubbornness and stupidity have hit a new low.  I guess when you 3 have a nice printer/ink business you can be stupid.    Dave   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Roads Vanderbilt, PA  15486l   ------------------------------  $ Date: Tue, 5 Oct 2004 10:35:58 -0400# From: "John Smith" <a@nonymous.com>2, Subject: Re: HP trying to defend against Sun, Message-ID: <ZcGdndUJr77OLP_cRVn-vQ@igs.net>   David Froble wrote:e > JF Mezei wrote:1 >6 >>>iL http://news.com.com/HP+tells+Sun+to+get+its+facts+straight/2100-1016_3-53960 64.html?tag=nefd.top >>>s >>D >> HP has sent a letter asking Sun to stop denigrating HP's strategy >> with regards to HP-UX.t >>G >> This means that Sun attacks against HP-UX/IA64 have had a noticeable  >> negative impact on HP.u >>5 >> Sun said it woudln't widthdraw its ads/starements.  >>G >> Of course, the article does not mention VMS (no surprise there). ButtD >> it does show that HP has realised that it is now on the defensiveE >> and that if it doesn't defend itself, Sun's attacks will hurt HP'sv >> enterprise business.  >>D >> Sun's new focus on the 8086, combined with HP's statement that itG >> would not port its enterprise systems to the 8086 will give a lot ofa8 >> fuel to Sun and hurt HP's momentum (or lack thereof). >> >iF > When AMD began to 'HAMMER' on Intel's image they were at least smartG > enough to recognize the danger and executed a 'U' turn, regardless ofBD > the extremely sour taste it might have.  If the best HP can do askA > Sun to quit trying to take HP's customers then stubbornness and = > stupidity have hit a new low.  I guess when you have a nice ) > printer/ink business you can be stupid.     K HP could sell off its computing division and then it wouldn't have to worry-= its pretty little head about what McNeely is saying about it.e  J Actually a better strategy would be to spin-off the entire computing groupK to existing shareholders in a 1:1 ratio. The new computing company could be-$ named Digital Equipment Corporation.   ------------------------------  % Date: Tue, 05 Oct 2004 08:28:06 +0200 . From: "Hans M. Aus" <aus@vim.uni-wuerzburg.de>D Subject: Re: Kyocera printer on two (2) IP OVMS 7.2-1 doesn't print.7 Message-ID: <aus-24D743.08280605102004@news.cis.dfn.de>o   > > H > > $init/que/start/on=V99::"10.60...:9102"/proc=tcpip$telnetsym KYOCERA  E A port scan of the new IP=10... revealed that the active TCP port is   9100.h  E $init/que/start/on=V99::"10.60...:9100"/proc=tcpip$telnetsym KYOCERA a starts the queue correctly.s   ------------------------------  # Date: Mon, 04 Oct 2004 20:54:43 GMT ! From: Nigel Barker <nigel@hp.com>) Subject: Re: light browser?h8 Message-ID: <qrjvl09b4kcsvi8ci8sgtpt9c0i0vpat30@4ax.com>  H On Fri, 01 Oct 2004 07:33:05 -0700, "Tom Linden" <tom@kednos.com> wrote:  ; >I gave up trying to run a browser under VMS, way too slow.a  P Mozilla performs just fine on a reasonably modern Alpha with adequate memory. So a 500MHz EV56 with 256MB.n   -- Nigel Barker Live from the sunny Cote d'Azurr   ------------------------------  $ Date: Tue, 5 Oct 2004 10:05:02 -0400# From: "John Smith" <a@nonymous.com>e' Subject: More Intel/HP Developer Forumsa, Message-ID: <osKdnfg4w9iON__cRVn-qg@igs.net>  E Interested in taking home an HP Integrity Server featuring the IntelS Itanium 2 processor?m  E If you are an enterprise-focused independent software vendor (ISV) or & developer, then this forum is for you.  J To help you keep your enterprise-class software solutions in sync with theK newest technology, the HP and Intel Developer Forum provides everything youvK need to design, deploy, and deliver next-generation products and solutions.eG For the special price of $2,000 USD ($2,900 CAD for the Toronto Forum),eJ you'll receive the HP Integrity rx2600 server and the software developmentE tools as well as the foundation, tips, and interaction with technicalmK contacts to help you rapidly migrate your applications to the Intel ItaniumaG 2 microarchitecture. Due to an overwhelming response to the first threetL forums, we're adding two more forum dates: one in Toronto, Canada and one in Mahwah, New Jersey.s  I Act now and register for the HP and Intel Developer Forum to receive this2! unique set of benefits including:nL   a.. Your very own HP Integrity rx2600 server featuring the Intel Itanium 2& processor (yes, this is yours to keep)I   b.. 3 days of interactive lectures and proctored hands-on labs in HP-UXS- 11i, Linux*, OpenVMS or Windows* environmentsfA   c.. Membership into the HP Developer & Solution Partner Programf5   d.. Membership into the Intel Early Access Programs  4 You've got two more opportunities to register today.  9 Registration for the Toronto, Canada forum, please visit:'* www.hp.com/go/itaniumdeveloperforumscanada  < Registration for the Mahwah, New Jersey forum, please visit:$ www.hp.com/go/itaniumdeveloperforums Register Today!eD You've got two more opportunities to participate in the HP and Intel Developer Forum 2004:s       OCTOBER 19 - 21, 2004f Toronto, Canada  Sheraton Gateway Hotel   DECEMBER 1 - 3, 2004 Mahwah, New Jersey Sheraton Crossroads Hotell  H **Note: The Canadian dollar price is about $400 more in Canadian dollarsL than the value of $2000 USD converted to CAD at prevailing exchange rates. IK guess that's how Intel/HP plan to make money to continue working on Itanic.    ------------------------------  # Date: Mon, 04 Oct 2004 22:08:24 GMTo4 From: "Fred Kleinsorge" <fred.nospam@nospam.dec.com>) Subject: Re: MS to lose its patent on FATm- Message-ID: <sRj8d.13$FQ2.0@news.cpqcorp.net>h  9 "JF Mezei" <jfmezei@spamnotteksavvy.com> wrote in messages- news:415CE084.7693F08E@spamnotteksavvy.com...o > ##B > As part of a re-examination, the U.S. Patent Office has issued a preliminaryHK > rejection for a patent previously granted to Microsoft for a Windows fileo	 > format.  > L > The agency ruled that, based on existing inventions at the time, the ideasK > behind the FAT (File Allocation Table) patent would have been obvious andi aresJ > therefore not subject to patent. Microsoft now has 90 days to respond to the-& > filing and make its case for why the > patent should be upheld. > ## >WK > Does this mean that HP can now lower the price of VMS because part of VMS-K > revenus won't have to go to Microsoft becayuse VMS now has FAT partitionsr form > the IA64 version of VMS ?C  - I guess that you are just being... well, you.h  K If you had been paying attention, you would have noted that MS had provided-L a royalty free license for the purpose of booting Itanium/EFI based systems.A of course, you might just not have the :-) keys on your keyboard.e   ------------------------------   Date: 4 Oct 2004 23:29:58 -0700T# From: dooleys@snowy.net.au (dooley)n" Subject: Re: Newbie DEBUG question= Message-ID: <1ca82fc6.0410042229.1a785ad8@posting.google.com>s  ` JF Mezei <jfmezei.spamnot@teksavvy.com> wrote in message news:<416206E1.C0DBBE3@teksavvy.com>...4 > Sorry for such a basic question, but I am stomped. >  > C program: >  > struct NAM mynam;u > 
 > HOWEVER: >  > Inside debug,n= > I can "examine mynam.nam$l_dir" and I get an address value.l >  > But why can't I:& > examine/ascii:20 mynam.nam$l_dir  or% > examine/ascii:20 *mynam.nam$l_dir  aH > both of which yield an "%DEBUG-E-INVCHAR, invalid character" error ???% exa/hex always tells you the contentsA then you have to work out why  Phil   ------------------------------   Date: 5 Oct 2004 08:09:22 -0500 ; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)c" Subject: Re: Newbie DEBUG question3 Message-ID: <yi9KbCtsjAvw@eisner.encompasserve.org>   [ In article <416206E1.C0DBBE3@teksavvy.com>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes:   D > printf("Whole shabang=>%s", mynam.nam$l_dev) prints the whole file  E    Since printf is from the C RTL %s assumes null terminated strings.)I    So that statement says:  start printing at address in mynam.nam$l_dev EJ    and stop when you get to a null byte.  Since NAM blocks don't use null J    terminated strings you are continuing until you run into a random 0 in 
    memory.   > BUT:K > printf("Just the device:%*.s", mynam.nam$b_dev, mynam.nam$l_dev) prints 8p( > characters (the b_dev has value of 8).  ;    Now you are telling printf to stop at a maximum width inR8    mynam.nam$b_dev, unless it found a null byte earlier.   ------------------------------   Date: 5 Oct 2004 08:16:25 -0500h; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)i" Subject: Re: Newbie DEBUG question3 Message-ID: <F$gXyjqkU3HV@eisner.encompasserve.org>,  [ In article <416206E1.C0DBBE3@teksavvy.com>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes:e4 > Sorry for such a basic question, but I am stomped. >  > C program: >  > struct NAM mynam;- > 
 > HOWEVER: >  > Inside debug,e= > I can "examine mynam.nam$l_dir" and I get an address value.  >  > But why can't I:& > examine/ascii:20 mynam.nam$l_dir  or% > examine/ascii:20 *mynam.nam$l_dir  ,H > both of which yield an "%DEBUG-E-INVCHAR, invalid character" error ???  F    I can't reproduce that.  What version of VMS, C compiler, and DEBUGC    are you using?  Is it possible that you're using a telnet and/orlH    terminal emulator that isn't passing the characters to DEBUG that you    think you are typing?   ------------------------------   Date: 5 Oct 2004 12:00:31 -05004; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) K Subject: Re: Obsucure Inquier article about Intel mentions VMS (and not any)3 Message-ID: <+zDeU2eYFQiB@eisner.encompasserve.org>E  R  Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> writes:  c > Bob Koehler wrote:B >>    With what for an Xserver?  My VXT 2000+ is a bit long in the? >>    tooth already and I don't intent to spend my time mapping  >>    keyboards.          hA > You need something like a SunRay, secure mobile sessions, sounde6 > local USB device support, low bandwidth clients etc.  <    Read carefully:  "I don't intent to spend my time mapping    keyboards."          G    SunRay's keyboards are no better than any others.  And anything fromi.    Sun with the word "secure" is still a joke.   ------------------------------  % Date: Tue, 05 Oct 2004 11:04:48 +0100.O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> Y Subject: Re: Obsucure Inquier article about Intel mentions VMS (and not any other OS') otP0 Message-ID: <cjtrk1$3uu$1@new-usenet.uk.sun.com>   Bob Koehler wrote:o > In article <f30679fb.0410040823.88cf934@posting.google.com>, fabiopenvms@yahoo.com.br (Fabio Cardoso) writes:- > D >>As I said in another post: wait for the Itanium blades. May be you8 >>can use them as workstations in a Blade-PC-like model. >  > G >    With what for an Xserver?  My VXT 2000+ is a bit long in the toothsC >    already and I don't intent to spend my time mapping keyboards.     ? You need something like a SunRay, secure mobile sessions, soundw4 local USB device support, low bandwidth clients etc.   Regardso Andrew Harrisonn   ------------------------------  % Date: Tue, 05 Oct 2004 23:16:47 +0800r From: prep@prep.synonet.comaY Subject: Re: Obsucure Inquier article about Intel mentions VMS (and not any other OS') ot - Message-ID: <87hdp9b3fk.fsf@prep.synonet.com>a  Q Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> writes:i   > Bob Koehler wrote:? >> In article <f30679fb.0410040823.88cf934@posting.google.com>,e3 >> fabiopenvms@yahoo.com.br (Fabio Cardoso) writes:r  E >>>As I said in another post: wait for the Itanium blades. May be you,9 >>>can use them as workstations in a Blade-PC-like model.d  B >>    With what for an Xserver?  My VXT 2000+ is a bit long in the? >>    tooth already and I don't intent to spend my time mappingO >>    keyboards.  A > You need something like a SunRay, secure mobile sessions, sounde6 > local USB device support, low bandwidth clients etc.  B Sunrays and itanic blades... Colour me breathless, a match made in heaven!f   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.m@                                              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: 5 Oct 2004 07:42:47 -0500e; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)p Subject: Re: OpenVMS newbie:3 Message-ID: <3jcc1Hfm0xsy@eisner.encompasserve.org>6  c In article <53cc0efd.0410041053.52e7fb89@posting.google.com>, Stitzman@gmail.com (stitzman) writes:   E    Please start with the VMS User's Guide.  You should not be playingt9    consultant on a system you don't even know how to use.s  C > My first (of many) questions is about naming conventions and files+ > structure.  I look at a VMS filename like F > "PUB$:[ANONYMOUS.SDSCPUB.NEXT]README.TXT;1" and it makes no sense toE > me.  I've read about file structure in the system manager's manual,hD > but without a system to play on I can't seem to grasp the concept.  A    This is the most common way to specify a file name.  PUB$ is a,G    logical name that points to a disk device.  [ANONYMOUS.SDSCPUB.NEXT]kI    is the directory path on that disk.  README is the name.  .TXT is the cC    extension.  ;1 is the version.  Generally each disk has its own  C    directory tree, like DOS.  You cannot (yet) mount a device in a :    directory like UNIX.e  B    There are variations on this; system routines like SYS$FILESCAND    can parse all native VMS naming formats.  C routines like fopen()G    can also parse UNIX style filenames (and therefor C++ and Java sincee%    their routines sit on top of C's).h  G > Also, what's with all the special characters ($ and @ specifically)? bD > I see all sorts of documentation referencing sys$system, sys$root,E > sys$startup etc., but I can't find anything that tells me what thati > weird "$" is for.p  I    $ is reserved to DEC/Compaq/HP.  That is:   when VMS Engineering adds uH    new names in the OS (new routines, new logical names, ...) they will G    include $.  The end user should not include $ in names thay make up oF    so that they will never have a conflict.  sys$system, sys$sysroot, H    sys$startup are all defined by VMS so they have $ in them.  PUB$ was G    defined by the end user and should not have a $ in it.  Someday VMS -E    Engineering may add a PUB$ to VMS at which time this system could e    unecessarily break.  C    It is all too common to find system managers (admins in the UNIX C    world) who break the $ is reserved rule, like the one who set up5    PUB$ on your system.g  F    @ is the command for executing a command script (usually .COM file)    in the forground.   ------------------------------  # Date: Tue, 05 Oct 2004 16:48:06 GMTh1 From: Keith Parris <keithparris_NOSPAM@yahoo.com>t& Subject: Re: OT: Sun's fighting chance. Message-ID: <afA8d.64$LM3.52@news.cpqcorp.net>   Here's what I meant to say:a   Keith Parris wrote:fJ > And even Sun could take down Linux, what about OpenVMS, NetBSD, FreeBSD?9           /if\                                    OpenBSDt  B Funny how the fingers often automatically complete what the brain + starts. I meant OpenBSD, not OpenVMS there.    ------------------------------  # Date: Tue, 05 Oct 2004 16:57:21 GMTB1 From: Keith Parris <keithparris_NOSPAM@yahoo.com> & Subject: Re: OT: Sun's fighting chance. Message-ID: <RnA8d.66$IN3.13@news.cpqcorp.net>   JF Mezei wrote:@0 > Since the POSIX product is no longer available  D Check out GNV. That, plus many improvements in the CRTL, seem to be C meeting the needs today of most of the folks who used POSIX before.a  D (Another factor making POSIX less important these days is the trend 7 toward writing Java code, which runs just fine on VMS.)s   ------------------------------   Date: 5 Oct 2004 01:06:16 -0700 . From: spamsink2001@yahoo.com (Alan E. Feldman)4 Subject: Rules for commenting out continuation lines= Message-ID: <b096a4ee.0410050006.7dcba7e1@posting.google.com>l   Hello,  7 Why does commenting out continuation lines work here...w  	 DCL> @DIRt+ $    DIRECTORY/NOHEAD/NOTRAI F:LOGIN.COM; -a /SIZE -r /DATE -m /FILE   DISK$DATA1:[FELDMAN]LOGIN.COM;82>                      (26655,16,0)               1  25-APR-2003 01:34:25.72T $ + $    DIRECTORY/NOHEAD/NOTRAI F:LOGIN.COM; -i /SIZE -  !/DATE /FILEh  DISK$DATA1:[FELDMAN]LOGIN.COM;821                      (26655,16,0)               1, $t+ $    DIRECTORY/NOHEAD/NOTRAI F:LOGIN.COM; -- /SIZE -- $!/DATE  /FILE0  DISK$DATA1:[FELDMAN]LOGIN.COM;821                      (26655,16,0)               1i $w   but not here...d   DCL> @A  $  A = "A" +  -9        "B" +  -7
        "C" $ SHOW SYMBOL Al   A = "ABC"t- $ IF (F$TYPE(A).NES."") THEN DELETEZ/SYMBOL Ae $w $  A = "A" +  -8 !      "B" +  -e> %DCL-W-EXPSYN, invalid expression syntax - check operators and operands
        "C"F %DCL-W-IVVERB, unrecognized command verb - check validity and spelling  \"C"\ $ SHOW SYMBOL At= %DCL-W-UNDSYM, undefined symbol - check validity and spelling - $ IF (F$TYPE(A).NES."") THEN DELETEZ/SYMBOL Aa $h $  A = "A" +  -e $!     "B" +  -a= %DCL-W-UNDSYM, undefined symbol - check validity and spelling   \$\
        "C"F %DCL-W-IVVERB, unrecognized command verb - check validity and spelling  \"C"\ $ SHOW SYMBOL Ai= %DCL-W-UNDSYM, undefined symbol - check validity and spellingh- $ IF (F$TYPE(A).NES."") THEN DELETEZ/SYMBOL At $  $ EXIT DCL>   ?s  E I have done the same commenting out in both procedures, yet one worksoF but the other doesn't. What are the rules for this? Is this documentedD somewhere? I looked in the User's Manual but could not find anything about this there.]  D And while $! and ! both work in the DIR case, I would think that the latter is preferable.    Thanks.V   ------------------------------  % Date: Tue, 05 Oct 2004 14:48:45 +0200I9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>m8 Subject: Re: Rules for commenting out continuation lines' Message-ID: <4162982D.7D911EA3@aaa.com>s   "Alan E. Feldman" wrote: >  > $  A = "A" +  -t > !      "B" +  -eI > %DCL-W-EXPSYN, invalid expression syntax - check operators and operandsH  9 You can not do that (mixing continuation and commenting).t  	 Jan-Erik.i   ------------------------------  % Date: Tue, 05 Oct 2004 10:18:02 -0400C( From: David Froble <davef@tsoft-inc.com>8 Subject: Re: Rules for commenting out continuation lines, Message-ID: <4162AD1A.6090208@tsoft-inc.com>   Alan E. Feldman wrote:    N  > And while $! and ! both work in the DIR case, I would think that the latter  > is preferable.g  G The way I understand it the '$' should be used.  What works and what's h) documented doesn't always match with DCL.a   Dave   -- -4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin RoadB Vanderbilt, PA  15486    ------------------------------   Date: 5 Oct 2004 09:32:47 -0700a( From: bob@instantwhip.com (Bob Ceculski)0 Subject: Sun can't invent - resorts to stealing!= Message-ID: <d7791aa1.0410050832.71b49e41@posting.google.com>o   shame, shame, shame Andrew! :)  ) http://www.theinquirer.net/?article=18864    ------------------------------  $ Date: Tue, 5 Oct 2004 13:29:39 -0400# From: "John Smith" <a@nonymous.com>u4 Subject: Re: Sun can't invent - resorts to stealing!, Message-ID: <yb2dnaHSLuqUR__cRVn-hw@igs.net>   Bob Ceculski wrote:   > shame, shame, shame Andrew! :) >-+ > http://www.theinquirer.net/?article=188640     Bob,    1) Software patents are a crock.' 2) All of this is covered by prior art.dJ 3) The case was heard in Rochester, NY by a jury composed of retired Kodak	 janitors..  K The oft-cited index case which seems to permit software patents to exist is.F not about software patents at all, but is about embedded software in aC device, ie. a controller, and not about general purpose programmingtG languages and concepts. Shit, I was doing stuff that Kodak says Sun wasoJ infringing upon at least 15 years ago, and others years before me. Maybe IK should sue Kodak for $2 billion and ask for it all in small unmarked bills.o0 At the very least, it invalidates Kodaks patent.   ------------------------------   Date: 5 Oct 2004 07:59:58 -0700h' From: mark_tarka@yahoo.com (Mark Tarka)i Subject: Tel. Co. Sys. Op.?l= Message-ID: <6b70c71c.0410050659.64c27d64@posting.google.com>p  - Is there a telephone company systems operatorS, in the house (or could you forward this post/ to one, or anyone who might choose to respond)?o  , Telephone companies store massive amounts of/ date (terebytes?), sure to increase as cellularl+ expands.  I've found a small amount of info[1 on the Net.  I'm interested in what is stored andT what it costs to retrieve it.   1 First, what's the depth of the data in years from ) today (for how long is information kept)?   . The classifications of data and storage length2 depends on location and state/federal regulations.1 How far back could someone go to find information 1 using _all_ the available sources, not just thosea2 available to lawyers or P.I.s (say regarding white/ collar criminal activity not discovered until ap$ "mole" retired, got fired, or quit)?  0 What information is stored (what a consumer sees6 on the monthly statement -- date, time, number dialed,) city, state, duration of call, and cost)?U  4 Or is is other than complete billing data, something abbreviated/condensed?  4 Then, what's your typical search and retrieval costs for out-going calls?  3 And, what's your typical search and retrieval costs 1 for _in-coming_ calls (calls made _to_ a specificd phone number)?  2 Now, to keep in touch with the theme of the group,0 could anyone tell me what o.s. Cray used, and if0 the company/project is still alive (like for DOD- Star Wars or CIC/AWACS/3-D 360 ground/air/seaw$ battle-field info systems/displays)?    1   Mark (No, I am not an Al Kaduh operative, but Ir1         used to play on on c.o.v. many years ago.e5         David P. Murphy seems to be in an interesting?;         position now-a-days; ouch, _that's_ gotta hurt  :-)k  :   mark_tarka@yahoo.comNOTTOBEUSEDFORSPAMORSTALKINGBEHAVIOR   ------------------------------  * Date: Tue, 5 Oct 2004 09:43:37 -0700 (PDT) From: da2341@cedar.ca.sbc.comu) Subject: Update Multiple decnet databasesf1 Message-ID: <04100509433718_1A1@cedar.ca.sbc.com>o   Hi,Q  : Does anyone have a command procedure to update new decnet , addresses accross multiple decnet databases?   Thanks!_   ------------------------------   Date: 5 Oct 2004 07:28:38 -0500N; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) - Subject: Re: VAX/VMS in industrial automationr3 Message-ID: <3apt6Fj3rv4z@eisner.encompasserve.org>:  U In article <1tmdnbtFh-QcG_zcRVn-vg@bresnan.com>, GreyCloud <mist@cumulus.com> writes:o > H > In DEC schools, they told us to set the process priority to 16.  That K > forces it into real time.  All other processes are pre-empted until that eH > process finishes.  But that was in the days of the VAX.  I don't know & > how the Alpha hardware handles that. >   F    That is a VMS software design, independent of what hardware it runs    on.   ------------------------------  # Date: Tue, 05 Oct 2004 12:39:30 GMTo# From: "Phil" <dooleys@snowy.net.au> - Subject: Re: VAX/VMS in industrial automation>= Message-ID: <6Cw8d.15211$5O5.8387@news-server.bigpond.net.au>S  & "asm" <arut@post.com> wrote in message6 news:c8b782e3.0410040507.6ab70c5@posting.google.com...	 > Hi All,d >cB > I have come across Vax/VMS being used in steel plant automation,8 > but couldn't quite understand if it is an RTOS or not. >d > VMS - virtual memory systems. > Not normally encountered in realtime systems >l? > What kind of scheduling does VAX/VMS use and if it is an RTOSo  > what features make it an RTOS. >  > Please clarify.b4 If you do a $ show system on vms it lists processes,6 each process has a "base" and "current" priority level; (by convention batch has base priority 3 and interactive 4)L9 and the vms scheduler can increase the "current" priorityI6 in response to interrupts (amongst other things), this4 provides good "timesharing" response and throughput./ However the swapper process runs at priority 15t- so any other process can have its working sete0 reduced or be swapped out, so cannot be regarded> as "realtime". Processes started with a priority of 16 or more5 cannot be pre-empted by the swapper, so can provide a-. realtime response (with a few qualifications).* There was some marketing-speak around that. suggested that vms was "soft" realtime, rather than "hard" realtime Phil   ------------------------------  % Date: Tue, 05 Oct 2004 14:52:46 +0200F9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>c- Subject: Re: VAX/VMS in industrial automationd' Message-ID: <4162991E.A7FD1E36@aaa.com>N   Phil wrote:M > ; > and the vms scheduler can increase the "current" priority 8 > in response to interrupts (amongst other things), this6 > provides good "timesharing" response and throughput.  9 Rather while *waiting* for an interrupt (such as terminal 8 I/O from a user doing editing). As soon as the interrupt> actuly has happend, the priority is lowered again to "normal".  	 Jan-Erik.f   ------------------------------  % Date: Tue, 05 Oct 2004 09:16:14 -0400E* From: Joel Loveless <joell@mindspring.com>- Subject: Re: VAX/VMS in industrial automation_8 Message-ID: <or55m0pa1hkm405e5rim39fffvle6h3mgo@4ax.com>  F I have been involved in many  industrial automation systems using VMS.A I worked in the textile industry with materials handling systems, C automated storage and retreival systems (ASRS), guided vehicles and4@ the like. VMS has many system functions built into the operatingF system to help facilitate real time processing environments. To name a= few, global pages, process and global event flags and mailboxbC communications. My last employer was doing stock trading on VMS fore> over 20 years. So, VMS is well suited to handle many real time
 applications.,  8 On 4 Oct 2004 06:07:43 -0700, arut@post.com (asm) wrote:   >Hi All, > A >I have come across Vax/VMS being used in steel plant automation,17 >but couldn't quite understand if it is an RTOS or not.4 >: >VMS - virtual memory systems - >Not normally encountered in realtime systems  >u> >What kind of scheduling does VAX/VMS use and if it is an RTOS >what features make it an RTOS.t >? >Please clarify. >k >Thanks, >ASM   ------------------------------  % Date: Tue, 05 Oct 2004 12:33:04 -0400g2 From: "Stanley F. Quayle" <squayle@insight.rr.com>- Subject: Re: VAX/VMS in industrial automation . Message-ID: <41629480.1498.1CFC4F86@localhost>  # On 5 Oct 2004 at 12:39, Phil wrote:eF > However the swapper process runs at priority 15 so any other processB > can have its working set reduced or be swapped out, so cannot beD > regarded as "realtime". Processes started with a priority of 16 orE > more cannot be pre-empted by the swapper, so can provide a realtimee( > response (with a few qualifications).   B This assumes, of course, that you don't change the priority of the. swapper, like we do with the CRISP/32 product.  F There's more going on with processes at and above priority 16 -- they = don't get time-slices -- they run until they block for I/O.  OE Actually, the whole scheduler setup is too complicated to cover in a t
 few emails.  $  
 --Stan Quayle, Quayle Consulting Inc.  
 ----------- Stanley F. Quayle, P.E. N8SQ  +1 614-868-1363n3 8572 North Spring Ct., Pickerington, OH  43147  USA%0 stan-at-stanq-dot-com       http://www.stanq.com   ------------------------------  $ Date: Mon, 4 Oct 2004 15:13:41 -0400, From: "warren sander" <warren.sander@hp.com>Y Subject: Re: www.hp.com/go... URL links (was: Re: Maximum file size, FAQ v. "Guide to Opef, Message-ID: <4161a1df$1@usenet01.boi.hp.com>  L one more thing.. currently there are over 21,000 marketing urls on hp.com...# (those are www.hp.com/go etc url's)n -warrenH  0 "Hoff Hoffman" <hoff@hp.nospam> wrote in message- news:8RG6d.12050$4p7.4415@news.cpqcorp.net...>> > In article <415A1D13.D3475B84@comcast.net>, David J Dachtera# <djesys.nospam@comcast.net> writes:d > :n) > :Any chance for an index of "go" links?p >iL >   I am aware of no such list, and would expect there can be both generallyI >   announced, and unannounced or audience-targeted /go links configured./ >iD >   I would also not expect to see a list of these jumpstation links
 maintained7 >   anywhere other than on the jumpstations themselves.  >UB >   The core relevent go links that I am immediately aware of are: >i >     www.hp.com/go/openvms/ >     www.hp.com/go/openvms/docl >     www.hp.com/go/openvms/faq:" >     www.hp.com/go/openvms/wizard >nH >   I fully expect there are other links around, but I do not know that. >aJ >   But you are correct, this is not my baliwick -- otherwise, structures,H >   navigation and organization would differ somewhat, rather obviously. >o >l( >  ---------------------------- #include' <rtfaq.h> -----------------------------a4 >     For additional, please see the OpenVMS FAQ --  www.hp.com/go/openvms/faqB, >  --------------------------- pure personal# opinion ---------------------------uG >         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com  >e   ------------------------------  $ Date: Mon, 4 Oct 2004 15:11:47 -0400, From: "warren sander" <warren.sander@hp.com>Y Subject: Re: www.hp.com/go... URL links (was: Re: Maximum file size, FAQ v. "Guide to Opem, Message-ID: <4161a16e$1@usenet01.boi.hp.com>  G These are the current OpenVMS marketing url's (plus Doc, rtr and spd's)iE I'm the person who requests them for the OpenVMS site.. If there is an6 shortcut url you'd like to have created let me know...   www.hp.com/go/bridgeworks  www.hp.com/go/openvmsh! www.hp.com/go/openvms/alphaserverI www.hp.com/go/openvms/argusc www.hp.com/go/openvms/bootcamp www.hp.com/go/openvms/ebusinessn www.hp.com/go/openvms/faq  www.hp.com/go/openvms/freeware www.hp.com/go/openvms/journal  www.hp.com/go/openvms/oracle! www.hp.com/go/openvms/solution-cdm www.hp.com/go/openvms/wizard www.hp.com/go/openvmsedu www.hp.com/products/bridgeworks> www.hp.com/products/openvmst$ www.hp.com/products/openvms/attunity' www.hp.com/products/openvms/bridgeworksh www.hp.com/products/openvms/como/ www.hp.com/products/openvms/enterprisedirectoryX$ www.hp.com/products/openvms/netbeans  www.hp.com/products/openvms/perl www.hp.com/products/openvms/phpu, www.hp.com/products/openvms/securewebbrowser+ www.hp.com/products/openvms/securewebserver   www.hp.com/products/openvms/soap+ www.hp.com/products/openvms/swbreleasenoteso" www.hp.com/products/openvms/tomcat www.hp.com/products/openvms/xmlw www.hp.com/software/bridgeworks. www.hp.com/go/openvms/doc8 www.hp.com/go/openvms/doc/order  www.hp.com/go/rtr  www.hp.com/go/spd  www.hp.com/info/spdi  ) These are the AlphaServer Marketing URL'so www.hp.com/go/alphaserver-ds15 www.hp.com/go/alphastation-ds15a www.hp.com/go/ev7alphaserver www.hp.com/go/fortune_ad www.hp.com/go/new-alpha-systemsR  www.hp.com/products/alphastation% www.hp.com/products/alphastation/ds10n% www.hp.com/products/alphastation/ds25n% www.hp.com/products/alphastation/es47t www.hp.com/servers/alphaserver# www.hp.com/servers/alphaserver/ds10n$ www.hp.com/servers/alphaserver/ds10l$ www.hp.com/servers/alphaserver/ds20e$ www.hp.com/servers/alphaserver/ds20l# www.hp.com/servers/alphaserver/ds25i# www.hp.com/servers/alphaserver/es40 # www.hp.com/servers/alphaserver/es45s# www.hp.com/servers/alphaserver/es47e# www.hp.com/servers/alphaserver/es80p% www.hp.com/servers/alphaserver/gs1280P$ www.hp.com/servers/alphaserver/gs160$ www.hp.com/servers/alphaserver/gs320# www.hp.com/servers/alphaserver/gs80'$ www.hp.com/servers/alphaserver/linux www.hp.com/storage/alphaserver      0 "Hoff Hoffman" <hoff@hp.nospam> wrote in message- news:8RG6d.12050$4p7.4415@news.cpqcorp.net... > > In article <415A1D13.D3475B84@comcast.net>, David J Dachtera# <djesys.nospam@comcast.net> writes:g > :y) > :Any chance for an index of "go" links?  > L >   I am aware of no such list, and would expect there can be both generallyI >   announced, and unannounced or audience-targeted /go links configured.t >rD >   I would also not expect to see a list of these jumpstation links
 maintained7 >   anywhere other than on the jumpstations themselves.- > B >   The core relevent go links that I am immediately aware of are: >p >     www.hp.com/go/openvms/ >     www.hp.com/go/openvms/doc  >     www.hp.com/go/openvms/faqm" >     www.hp.com/go/openvms/wizard >5H >   I fully expect there are other links around, but I do not know that. >0J >   But you are correct, this is not my baliwick -- otherwise, structures,H >   navigation and organization would differ somewhat, rather obviously. >e > ( >  ---------------------------- #include' <rtfaq.h> -----------------------------g4 >     For additional, please see the OpenVMS FAQ --  www.hp.com/go/openvms/faqi, >  --------------------------- pure personal# opinion ---------------------------eG >         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comv >m   ------------------------------  * Date: Tue, 5 Oct 2004 14:15:39 +0000 (UTC)6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)I Subject: [OpenVMS Alpha V7.3-2] What Priv for SET SECURITY/CLASS=DEVICE ?d1 Message-ID: <newscache$w5745i$zy31$1@news.sil.at>S   I've a code snippet which does  )   $ priv=F$SETPRV("CMKRNL,SYSLCK,SYSPRV")o8   $ MCR SYSMAN IO CONNECT xxy0: /DRIVER=SYS$xxDRIVER.EXE0   $ SET SECURITY/CLASS=DEVICE xxy0: /OWNER=[uic]  2 where the SET SECURITY fails with %SYSTEM-F-NOPRIV  N Between the SYSMAN IO CONNECT and the SET SECURITY the device looks like this:   $ SHOW DEVICE xxy0:/FULLJ Device xxy0:, device type local memory mailbox, is online, record-oriented@     device, sharable, mailbox device, device is a template only.  L ....                                Owner UIC                          [0,0]L ....                                Dev Prot S:RWPL, O:RWPL, G:RWPL, W: RWPL  F Protection seems ok and Help tells SET SEC/OWN requires GRPPRV/SYSPRV.F But it is obviously not enough for this SET SECURITY to be successful.  O Maybe a privilege is missing (besides the "OPER" for the SYSMAN ;-) but which ?oF I tried all of them and didn't succeed (except with BYPASS or course).   But:E After the first modification of this (template) device, every further-9 SET SECURITY/OWNER succeeded (with GRPPRV/SYSPRV only !!)A   Is this a bug or a feature ?   TIA:   -- o Peter "EPLAN" LANGSTOEGERO% Network and OpenVMS system specialista E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  $ Date: Tue, 5 Oct 2004 09:44:55 -0400# From: "John Smith" <a@nonymous.com>i7 Subject: [OT]: Software Disasters Often People ProblemsK, Message-ID: <Jd6dneEIr4PAOP_cRVn-uw@igs.net>  L http://story.news.yahoo.com/news?tmpl=story&cid=562&ncid=738&e=3&u=/ap/20041" 005/ap_on_hi_te/software_disasters    J Not a bad article, but somebody ought to tell this guy that the managementH 'people problems' also involve a flawed choice of operating system -- he2 only alludes to it near the bottom of the article.    ( Software Disasters Often People Problems   By MATTHEW FORDAHL,i AP Technology Writer    J SAN JOSE, Calif. - New software at Hewlett-Packard Co. was supposed to getK orders in and out the door faster at the computer giant. Instead, a botched F deployment cut into earnings in a big way in August and executives got fired.  I Last month, a system that controls communications between commercial jetsHH and air traffic controllers in southern California shut off because someD maintenance had not been performed. A backup also failed, triggering potential peril.  I Computer code foul-ups also recently held Tacoma, Wash.'s budget hostage, K delayed financial aid to university students in Indiana and caused retailers8 Ross Stores Inc.'s profits to plummet 40 percent after a# merchandise-tracking system failed.   L Such disasters are often blamed on bad software, but the cause is rarely badH programming. As systems grow more complicated, failures instead have farG less technical explanations: bad management, communication or training.c  H "In 90 percent of the cases, it's because the implementer did a bad job,L training was bad, the whole project was poorly done," said Joshua Greenbaum,H principal analyst at Enterprise Applications Consulting in Berkeley. "At> which point, you have a real garbage in, garbage out problem."  I As governments, businesses and other organizations become more reliant onrL technology, the consequences of software failures are rarely trivial. Entire+ businesses - and even lives - are at stake.   I Many experts believe the situation will only worsen as software automatesVI new tasks and more systems interconnect with and rely on other computers.tF Technical challenges may be surmounted, but managing people never gets easier.o  H "The limit we're hitting is the human limit, not the limit of software,"F Greenbaum said. "Technology has gotten ahead of our organizational andK command capabilities in many cases. It's amazing when you go into companies.) and see the kinds of battles that go on."T  H Big software projects - whether to manage supply chains, handle payroll,L track inventory, prepare finances - tend to begin with high expectations and@ the best intentions.  They're all about efficiency, reliability, cost-savings, competitiveness.  K Companies might develop their own programs internally, outsource the job orvB buy from a company such as SAP AG, Oracle Corp. or PeopleSoft Inc.G Regardless of the route, it's usually a major undertaking to get thingsc right.  H Often, however, the first step toward total disaster is taken before theK first line of code is drawn up. Organizations must map out exactly how theyaH do business, refining procedures along the way. All this must be clearly( explained to a project's technical team.  F "The risk associated with these projects is not around software but isH around the actual business process redesign that takes place," said BillE Wohl, an SAP spokesman. "These projects require very strong executiveoK leadership, very talented consulting resources and a very focused effort ife4 the project is to be successful and not disruptive."  D A 2002 study commissioned by the National Institute of Standards andH Technology found software bugs cost the U.S. economy about $59.5 billionJ annually. The same study found that more than a third of that cost - about9 $22.2 billion - could be eliminated by improving testing.M  J A lack of strong leadership appears to have been a factor in HP's problem,L which led to the dismissal of three top executives in its server and storageI business hours after the company announced disappointing earnings on Aug.S 12.p  J HP did not return a telephone call seeking comment but has said previouslyL that its problems have been resolved. Wohl said the software, made by SAP, w as not at fault.  K Big projects also can sour during development, particularly when not enough(K resources are allocated, the people who will have a stake in the new system9H don't participate in planning and executives don't care. All can lead to% miscommunication with the developers.@  G "Mistakes hurt, but misunderstandings kill," said John Michelsen, chief L executive of iTKO Inc., which makes software that helps companies manage bigI software projects and test them automatically as they're being developed.h  L Too often, he said, programmers are handed a lengthy document explaining theF business requirements for a software project and left to interpret it.  K "Developers are least qualified to validate a business requirement. They're C either nerds and don't get it, or they're people in another culture9G altogether," said Michelsen, referring to cases where development takes  place offshore.i  J The Dallas-based company's LISA software attempts to reduce the complexityK of testing, so nontechnical executives in charge of major software projectsm3 can ensure the actual code adheres to their vision.$  A The lack of robust testing during and after such a project likelyaJ contributed to the Sept. 14 radio system outage over the skies of parts of California, Nevada and Arizona.   L Though there were a handful of close calls, all 403 planes in the air duringF the incident managed to land safely, said FAA spokesman Donn Walker. AH handful violated rules that dictate how close they are allowed to fly to? each other - but the FAA maintains there were no "near misses.".  L The genesis of the problem was the transition in 2001 by Harris Corp. of theE Federal Aviation Administration's Voice Switching Control System from-F Unix-based servers to Microsoft Corp.'s off-the-shelf Windows Advanced Server 2000.  K By most accounts, the move went well except the new system required regular@F maintenance to prevent data overload. When that wasn't done, it turnedL itself off as it was designed to do. But the backup also failed. In all, theE southern California system was down for three hours, though other FAA < centers restored communications within seconds, Walker said.  H The FAA's investigation is continuing, and Harris Corp. did not return a call seeking comment.   5 Michelsen said the failure was in inadequate testing.c  I "On a regular basis, the FAA should have been downing that primary systemoK and watching that backup system come up," he said. "If it doesn't go up and I stay up, they would have known they had a problem to fix long before theys needed to rely on it."  L Another common theme in failures lies in the ranks of employees who actually must use the systems.:  H Often they're not given proper training. There's also a chance that theyK don't want the project to succeed, especially if they see it as a threat to  employment.+  H "It becomes a major role of (management) to kind of herd the cats in andL make them all line up in a reasonable way," said Barry Wilderman, an analyst6 at the Meta Group. "That's why this stuff is so hard."   ------------------------------  % Date: Tue, 05 Oct 2004 08:32:04 -0700 # From: "Tom Linden" <tom@kednos.com>e; Subject: Re: [OT]: Software Disasters Often People Problems ( Message-ID: <opsfelfqlkzgicya@hyrrokkin>  E On Tue, 5 Oct 2004 09:44:55 -0400, John Smith <a@nonymous.com> wrote:g  N > http://story.news.yahoo.com/news?tmpl=story&cid=562&ncid=738&e=3&u=/ap/20041$ > 005/ap_on_hi_te/software_disasters >" >uC > Not a bad article, but somebody ought to tell this guy that the  0 > managementJ > 'people problems' also involve a flawed choice of operating system -- he4 > only alludes to it near the bottom of the article.   And programming language.(   >o >n* > Software Disasters Often People Problems >- > By MATTHEW FORDAHL,o > AP Technology Writer >c ><J > SAN JOSE, Calif. - New software at Hewlett-Packard Co. was supposed to   > getrG > orders in and out the door faster at the computer giant. Instead, a  d	 > botchedyH > deployment cut into earnings in a big way in August and executives got > fired. >rK > Last month, a system that controls communications between commercial jetsnJ > and air traffic controllers in southern California shut off because someF > maintenance had not been performed. A backup also failed, triggering > potential peril. > K > Computer code foul-ups also recently held Tacoma, Wash.'s budget hostage,0F > delayed financial aid to university students in Indiana and caused  
 > retailer: > Ross Stores Inc.'s profits to plummet 40 percent after a% > merchandise-tracking system failed.s >nL > Such disasters are often blamed on bad software, but the cause is rarely   > bad J > programming. As systems grow more complicated, failures instead have farI > less technical explanations: bad management, communication or training.r > J > "In 90 percent of the cases, it's because the implementer did a bad job,E > training was bad, the whole project was poorly done," said Joshua  a > Greenbaum,J > principal analyst at Enterprise Applications Consulting in Berkeley. "At@ > which point, you have a real garbage in, garbage out problem." >gK > As governments, businesses and other organizations become more reliant onoI > technology, the consequences of software failures are rarely trivial.  s > Entire- > businesses - and even lives - are at stake.  > K > Many experts believe the situation will only worsen as software automates K > new tasks and more systems interconnect with and rely on other computers. H > Technical challenges may be surmounted, but managing people never gets	 > easier.h > J > "The limit we're hitting is the human limit, not the limit of software,"H > Greenbaum said. "Technology has gotten ahead of our organizational andE > command capabilities in many cases. It's amazing when you go into  R > companiesw+ > and see the kinds of battles that go on."a >rJ > Big software projects - whether to manage supply chains, handle payroll,L > track inventory, prepare finances - tend to begin with high expectations   > andoB > the best intentions.  They're all about efficiency, reliability,  > cost-savings, competitiveness. >nL > Companies might develop their own programs internally, outsource the job   > orD > buy from a company such as SAP AG, Oracle Corp. or PeopleSoft Inc.I > Regardless of the route, it's usually a major undertaking to get things' > right. > J > Often, however, the first step toward total disaster is taken before theJ > first line of code is drawn up. Organizations must map out exactly how   > theyJ > do business, refining procedures along the way. All this must be clearly* > explained to a project's technical team. >eH > "The risk associated with these projects is not around software but isJ > around the actual business process redesign that takes place," said BillG > Wohl, an SAP spokesman. "These projects require very strong executivekL > leadership, very talented consulting resources and a very focused effort   > if6 > the project is to be successful and not disruptive." >aF > A 2002 study commissioned by the National Institute of Standards andJ > Technology found software bugs cost the U.S. economy about $59.5 billionH > annually. The same study found that more than a third of that cost -   > about'; > $22.2 billion - could be eliminated by improving testing.p > E > A lack of strong leadership appears to have been a factor in HP's  s
 > problem,H > which led to the dismissal of three top executives in its server and  	 > storageeK > business hours after the company announced disappointing earnings on Aug.v > 12.  >oC > HP did not return a telephone call seeking comment but has said    > previouslyI > that its problems have been resolved. Wohl said the software, made by  $ > SAP, w > as not at fault. >eH > Big projects also can sour during development, particularly when not   > enoughH > resources are allocated, the people who will have a stake in the new   > systemJ > don't participate in planning and executives don't care. All can lead to' > miscommunication with the developers.t > I > "Mistakes hurt, but misunderstandings kill," said John Michelsen, chief-L > executive of iTKO Inc., which makes software that helps companies manage   > bigrK > software projects and test them automatically as they're being developed.a >dL > Too often, he said, programmers are handed a lengthy document explaining   > the H > business requirements for a software project and left to interpret it. > G > "Developers are least qualified to validate a business requirement.   	 > They'reeE > either nerds and don't get it, or they're people in another cultureOI > altogether," said Michelsen, referring to cases where development takes1 > place offshore.  >tC > The Dallas-based company's LISA software attempts to reduce the  t > complexityF > of testing, so nontechnical executives in charge of major software  
 > projects5 > can ensure the actual code adheres to their vision.p > C > The lack of robust testing during and after such a project likelyfK > contributed to the Sept. 14 radio system outage over the skies of parts  s > of! > California, Nevada and Arizona.a >wI > Though there were a handful of close calls, all 403 planes in the air  - > duringH > the incident managed to land safely, said FAA spokesman Donn Walker. AJ > handful violated rules that dictate how close they are allowed to fly toA > each other - but the FAA maintains there were no "near misses."o >cL > The genesis of the problem was the transition in 2001 by Harris Corp. of   > theOG > Federal Aviation Administration's Voice Switching Control System fromIH > Unix-based servers to Microsoft Corp.'s off-the-shelf Windows Advanced > Server 2000. > G > By most accounts, the move went well except the new system required  K	 > regularDH > maintenance to prevent data overload. When that wasn't done, it turnedL > itself off as it was designed to do. But the backup also failed. In all,   > the[G > southern California system was down for three hours, though other FAA > > centers restored communications within seconds, Walker said. > J > The FAA's investigation is continuing, and Harris Corp. did not return a > call seeking comment.  >=7 > Michelsen said the failure was in inadequate testing.e >oK > "On a regular basis, the FAA should have been downing that primary systemiK > and watching that backup system come up," he said. "If it doesn't go up  L > andYK > stay up, they would have known they had a problem to fix long before theyH > needed to rely on it." >AG > Another common theme in failures lies in the ranks of employees who  a
 > actually > must use the systems." >DJ > Often they're not given proper training. There's also a chance that theyL > don't want the project to succeed, especially if they see it as a threat   > to
 > employment.. >.J > "It becomes a major role of (management) to kind of herd the cats in andH > make them all line up in a reasonable way," said Barry Wilderman, an  	 > analyst 8 > at the Meta Group. "That's why this stuff is so hard." >e >e >  >u       -- dC Using Opera's revolutionary e-mail client: http://www.opera.com/m2/k   ------------------------------   End of INFO-VAX 2004.553 ************************