1 INFO-VAX	Tue, 01 Apr 2003	Volume 2003 : Issue 180       Contents: Re: Alpha emulator for x86?  Re: Alpha emulator for x86?  Re: Alpha emulator for x86?  Re: Alpha emulator for x86?  Re: Alpha emulator for x86?  Re: Alpha emulator for x86?  Re: Alpha emulator for x86?  Re: Alpha emulator for x86? ; Re: Apache %SYSTEM-F-NOPRIV on RIGHTSLIST.DAT after upgrade  Re: Battery for 500au  Re: Battery for 500au  Re: Default Gateway  Re: Demagnetize Disk Drives  Re: Demagnetize Disk Drives  Re: Demagnetize Disk Drives  Re: Disk cache with shadow sets ' Re: Earth to Andrew, do you get it now? ' Re: Earth to Andrew, do you get it now? ' Re: Earth to Andrew, do you get it now? ' Re: Earth to Andrew, do you get it now? ' Re: Earth to Andrew, do you get it now? ' Re: Earth to Andrew, do you get it now?  Re: Exporting users? Re: Fortran Guru requested RE: Fortran Guru requested Re: Fortran Guru requested Re: Fortran Guru requested Re: Fortran Guru requested Re: Fortran Guru requested Re: Fortran Guru requested Re: Fortran Guru requested Re: Fortran Guru requested Re: Fortran Guru requested Re: Fortran Guru requested Re: Fortran Guru requested Re: Fortran Guru requested Re: High RMS file activity RE: High RMS file activity Re: HP doesn't know OpenVMS  Re: Logical names in decwindows  Re: Macs vs. Pathworks et. al.$ Re: Netscape gags on VMS FTP Servers) Re: New additions to decus.acornsw.com...  New SPAWN behavior with V7.3-1" Re: New SPAWN behavior with V7.3-1" Re: New SPAWN behavior with V7.3-1P Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha RetaiP Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retai pathworks administer issue2 PicoVax and OpenVMS hobbyist CD beginners question6 Re: PicoVax and OpenVMS hobbyist CD beginners question6 Re: PicoVax and OpenVMS hobbyist CD beginners question6 Re: PicoVax and OpenVMS hobbyist CD beginners question6 RE: PicoVax and OpenVMS hobbyist CD beginners question6 Re: PicoVax and OpenVMS hobbyist CD beginners question0 Re: Poka Yoke on loppy drive flat cable is wrong/ Re: Remotely opening a window on my workstation / Re: Remotely opening a window on my workstation  Re: Rich Marcello / Re: RMS Tune Check (was High RMS file activity) / Re: RMS Tune Check (was High RMS file activity)  Re: Second IP address  SYSDUMP.DMP corruption Re: SYSDUMP.DMP corruption Re: SYSDUMP.DMP corruption Re: SYSDUMP.DMP corruptionD Re: SYSTARTUP_VMS (was: Three questions re: the VMS startup process)D Re: SYSTARTUP_VMS (was: Three questions re: the VMS startup process)& Re: TCPIP Services and IP masquerading The 15th greatest Hoax Re: VMS Upgrade Questions 1 Re: [MOZILLA V1.3] Compatibility with JAVA V1.4 ? 1 Re: [MOZILLA V1.3] Compatibility with JAVA V1.4 ?   F ----------------------------------------------------------------------   Date: 1 Apr 2003 00:33:38 -0800 # From: dooleys@snowy.net.au (dooley) $ Subject: Re: Alpha emulator for x86?= Message-ID: <1ca82fc6.0304010033.7b4c5d95@posting.google.com>   d "Tom Linden" <tom@kednos.com> wrote in message news:<CIEJLCMNHNNDLLOOGNJIIEJKGOAA.tom@kednos.com>...E > Digital made a product called FX!32 which ran under NT on Alpha and 7 > was available at one time on ftp.digital... something - FX32 was (sort of) an X86 emulator for alpha, 8 it allowed you to run win32 binaries on Alpha using NT4. Phil  
 <evilthought> 9 If there was an Alpha emulator for x86, you would be able - to use NT & FX32 to run notepad really slowly  </evilthought>   ------------------------------  # Date: Tue, 01 Apr 2003 10:27:48 GMT ) From: P.Lj <plj@NOSPAMbyron.ext.telia.se> $ Subject: Re: Alpha emulator for x86?8 Message-ID: <e7qi8vkcvrj9onodvsgfg03nahn44vieuj@4ax.com>  3 On Mon, 31 Mar 2003 19:25:32 -0500, "Timothy Stark" # <sword7nospam@speakeasy.org> wrote:   
 >Hello folks,  > G >Ok, I am curious about that Alpha emulator for x86 platforms.  Is that M >possible to emulate the Alpha system on x86?  I was looking for that through I >google and web searches and found not much news but only few articles on  >this newsgroup. >  >Thank you. 
 >Tim Stark >   F If I remember right, there was a Alpha emulator called 'mannequin' (?)A or something similar, used by Digital to port VMS while the Alpha  hardware was still developed.   1 It is descsribed in one of the Techinal Journals.   , Someone around here must know more about it.  	 >>> ^P.Lj    ------------------------------  # Date: Tue, 01 Apr 2003 10:29:07 GMT ) From: P.Lj <plj@NOSPAMbyron.ext.telia.se> $ Subject: Re: Alpha emulator for x86?8 Message-ID: <teqi8vkf35k721mkf2h11koefuvt3qdr1u@4ax.com>  E On Tue, 01 Apr 2003 10:27:48 GMT, P.Lj <plj@NOSPAMbyron.ext.telia.se>  wrote:  4 >On Mon, 31 Mar 2003 19:25:32 -0500, "Timothy Stark"$ ><sword7nospam@speakeasy.org> wrote: >  >>Hello folks, >>H >>Ok, I am curious about that Alpha emulator for x86 platforms.  Is thatN >>possible to emulate the Alpha system on x86?  I was looking for that throughJ >>google and web searches and found not much news but only few articles on >>this newsgroup.  >> >>Thank you. >>Tim Stark  >> > G >If I remember right, there was a Alpha emulator called 'mannequin' (?) B >or something similar, used by Digital to port VMS while the Alpha >hardware was still developed. > 2 >It is descsribed in one of the Techinal Journals. > - >Someone around here must know more about it.  > 
 >>>> ^P.Lj  0 Must add, not sure which plattform it ran under.  	 >>> ^P.Lj    ------------------------------  * Date: Tue, 1 Apr 2003 11:25:44 +0000 (UTC)+ From: david20@alpha1.mdx.ac.uk (David Webb) $ Subject: Re: Alpha emulator for x86?+ Message-ID: <b6bsvo$lpc$1@aquila.mdx.ac.uk>   Z In article <3E88B840.23663.D21AA3@localhost>, "Stanley F. Quayle" <stan@stanq.com> writes:) >On 1 Apr 2003 at 0:59, Spud Demon wrote: M >> Have you heard of "Charon-VAX"?  It simulates a VAX instead, which can run ! >> VMS and probably Digital Unix.  > G >VMS, Digital UNIX, and even -- VAXELN.  I'm planning on testing it on  F >Linux, once they get a little farther with the Linux for VAX project. >   7 Are you sure you mean Digital Unix rather than Ultrix ? G As far as I am aware Digital Unix/TRU64 has only run on the Alpha chip.   
 David Webb VMS and Unix team leader CCSS Middlesex University    6 >> "Timothy Stark" <sword7nospam@speakeasy.org> writesJ >> >Ok, I am curious about that Alpha emulator for x86 platforms.  Is that0 >> >possible to emulate the Alpha system on x86? > G >Technically, it's possible.  It's even been done.  But emulating a 64- B >bit machine on a 32-bit machine leads to a REALLY SLOW emulation. > G >If you have to have VMS on a x86, have it emulate a VAX.  Version 7.3  A >is available for VAX, including CHARON-VAX.  Works quite nicely.  > C >(And please pardon the shameless plug from a CHARON-VAX reseller.)  >  >--Stan Quayle >Quayle Consulting Inc.  >  >---------- D >Stanley F. Quayle, P.E. N8SQ  +1 614-868-1363  Fax: +1 614 868-16712 >8572 North Spring Ct. NW, Pickerington, OH  43147> >Preferred address:  stan@stanq.com       http://www.stanq.com >  >    ------------------------------  # Date: Tue, 01 Apr 2003 13:50:24 GMT 2 From: "Robert Boers" <Robert.boers@softresint.com>$ Subject: Re: Alpha emulator for x86?( Message-ID: <3e899a8e@news.deckpoint.ch>  I The SimpleScalar emulator has an Alpha mode for the 21064 (and 21164A and L 21264A in the latest version). It can run unmodified statically-linked AlphaE binaries. There is a specially compiled version of Quake to run on it  (slowly). Here is the URL:, http://www.cs.wisc.edu/~plakal/simplescalar/   Regards, Robert   = "Timothy Stark" <sword7nospam@speakeasy.org> wrote in message , news:n72cnRaGopXgQRWjXTWc2Q@speakeasy.net... > Hello folks, > H > Ok, I am curious about that Alpha emulator for x86 platforms.  Is thatF > possible to emulate the Alpha system on x86?  I was looking for that through J > google and web searches and found not much news but only few articles on > this newsgroup.  >  > Thank you. > Tim Stark  >  >    ------------------------------   Date: 1 Apr 2003 14:21:56 GMT ( From: bill@cs.uofs.edu (Bill Gunshannon)$ Subject: Re: Alpha emulator for x86?5 Message-ID: <b6c7a3$3qmb5$1@ID-135708.news.dfncis.de>   - In article <3E88C2B5.29880.FAF2B0@localhost>, - 	"Stanley F. Quayle" <stan@stanq.com> writes:  > A >                                                  Heck, I still  F > encounter people who think that spending thousands for 15+ year old + > used VAXes is "better" than a new system.   C Wow!!!  Wish i could encounter them.  I would likely sell one of my @ VAXen for thousands and then I could finish restoring my classicC Triumph Spitfire.  But the few times I have had offers they usually  ran closer to beer money.    bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  $ Date: Tue, 1 Apr 2003 10:21:50 -0500. From: "Douglas H. Quebbeman" <dougq@iglou.com>$ Subject: Re: Alpha emulator for x86?) Message-ID: <3e89ae8f$1_1@news.iglou.com>   5 "Stanley F. Quayle" <stan@stanq.com> wrote in message ' news:3E88B840.23663.D21AA3@localhost... * > On 1 Apr 2003 at 0:59, Spud Demon wrote:J > > Have you heard of "Charon-VAX"?  It simulates a VAX instead, which can run " > > VMS and probably Digital Unix. > G > VMS, Digital UNIX, and even -- VAXELN.  I'm planning on testing it on G > Linux, once they get a little farther with the Linux for VAX project.  > 7 > > "Timothy Stark" <sword7nospam@speakeasy.org> writes K > > >Ok, I am curious about that Alpha emulator for x86 platforms.  Is that 1 > > >possible to emulate the Alpha system on x86?  > H > Technically, it's possible.  It's even been done.  But emulating a 64-C > bit machine on a 32-bit machine leads to a REALLY SLOW emulation.  > G > If you have to have VMS on a x86, have it emulate a VAX.  Version 7.3 B > is available for VAX, including CHARON-VAX.  Works quite nicely. > D > (And please pardon the shameless plug from a CHARON-VAX reseller.)  > As was written about, Tim, the OP, has written a VAX emulator,> part of a package that includes PDP-11 and PDP-10 emulators asA well. I use his PDP-10 emulator to run TOPS-10 from time to time.   9 And his emulator package is Open-Source, and free. But of 7 course, no warrantee accompanies such software, and not ; the kind of support paying customers expect, so don't sweat  it...    -dq    ------------------------------  % Date: Tue, 01 Apr 2003 09:10:11 -0700 + From: "Barry Treahy, Jr." <Treahy@MMaz.com> $ Subject: Re: Alpha emulator for x86?' Message-ID: <3E89B9E3.4070004@MMaz.com>   
 dooley wrote:   e >"Tom Linden" <tom@kednos.com> wrote in message news:<CIEJLCMNHNNDLLOOGNJIIEJKGOAA.tom@kednos.com>...  >    > E >>Digital made a product called FX!32 which ran under NT on Alpha and 7 >>was available at one time on ftp.digital... something  >>     >>. >FX32 was (sort of) an X86 emulator for alpha,9 >it allowed you to run win32 binaries on Alpha using NT4.  >Phil  >  ><evilthought>: >If there was an Alpha emulator for x86, you would be able. >to use NT & FX32 to run notepad really slowly ></evilthought>  >    >   H Though it would be scary if it did work, how about the old SoftPC which H ran on VAX/VMS, running Linux, running VMware for Linux, to run Windows E 2000 running Charon-VAX to run VAX/VMS...  Talking about long coffee  G breaks for boot up times, this would have to exceed any TU58 boot by a   mile...    Barry      >  >    >    --    @ Barry Treahy, Jr  *  Midwest Microwave  *  Vice President & CIO   A E-mail: Treahy@mmaz.com * Phone: 480/314-1320 * FAX: 480/661-7028    ------------------------------   Date: 1 Apr 03 20:46:44 +0200 ) From: p_sture@elias.decus.ch (Paul Sture) D Subject: Re: Apache %SYSTEM-F-NOPRIV on RIGHTSLIST.DAT after upgrade) Message-ID: <bMBUcaJ2R5E0@elias.decus.ch>   R In article <b68k5c$cb$1@news4.tilbu1.nb.home.nl>, Dirk Munk <munk@home.nl> writes:Q > Did you run the config procedure after the upgrade? After each upgrade of CSWS  V > you MUST run the config procedure, otherwise you get all kind of Privilege problems. >   G That appears to fix it. FWIW I did run APACHE$CONFIG after the upgrade, " but not after the security update.  O I see today there's a new update. I'll see if I can reproduce the problem after L applying that (the update instructions only contain instructions to shutdown
 and startup).    --  
 Paul Sture   ------------------------------  $ Date: Tue, 1 Apr 2003 09:40:13 -0500, From: "Island" <dbturner@nospamislandco.com> Subject: Re: Battery for 500au/ Message-ID: <v8j96prt7f3d59@news.supernews.com>    We have them for $5  NEW    David    ------------------------------  # Date: Tue, 01 Apr 2003 15:00:35 GMT ' From: Colin Blake <colin@theblakes.com>  Subject: Re: Battery for 500au, Message-ID: <3E89A993.8030703@theblakes.com>   Keith A. Lewis wrote:   M >Symptom:  Every time I take a power hit or turn off the power, my 500au goes J >into ARC mode.  I have to set it into SRM and then use the "reset" button& >(not the power button) to restart it. > D Mine did exactly this for months. I thought I needed a new battery. I Finally I broke down and ripped the system open to find out what sort of  D battery I needed to get, and its been fine ever since. I think just F taking the battery out and reinstalling it (the same one!) cured some 9 kind of bad connection problem. That was over a year ago.    Colin.   ------------------------------  % Date: Tue, 01 Apr 2003 09:16:18 +0100 ( From: Nic Clews <sendspamhere@127.0.0.1> Subject: Re: Default Gateway) Message-ID: <3E894AD2.A306699F@127.0.0.1>    Hamlyn Mootoo wrote: > [ > Nic Clews <sendspamhere@127.0.0.1> wrote in message news:<3E8818EE.F3F17DDA@127.0.0.1>...  > > routerwoman wrote: > > > R > > > Is it possible to set up a default gateway on a VMS system.  I have a systemR > > > that's running RIPv1 just to talk to the one and only router in the network.K > > > Seems like a static route to the Gateway would stop all that chatter.  > >  > > All "what" chatter?  > > > > > Connected an analyser to determine what the "chatter" is ? > H > The "chatter" that she speaks of is RIP (Routing Information Protocol)* > updates being exchanged with the router.  H Just because that is the protocol talked about, that may not be the onlyF thing causing chatter. Is she running DECnet? Does she have a cluster?E Unless I've seen a SH LAN/FULL from SDA I reserve judgement. I've had B people before reckon they can tell me what protocol is running and1 between where just by watching activity lights...    --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot com    ------------------------------  $ Date: Tue, 1 Apr 2003 09:20:01 -0800* From: "Jack Peacock" <peacock@simconv.com>$ Subject: Re: Demagnetize Disk Drives2 Message-ID: <f0udnUl-Kc3fVxSjXTWcpg@mpowercom.net>  . "John Smith" <a@nonymous.com> wrote in message= news:la7ia.235$Mt1.15@news01.bloor.is.net.cable.rogers.com...   >  Be sure to have your welder's! > goggles handy when you go to it  > K A college professor once explained to me his method of securing old drives. J One of his early jobs working for an intelligence agency in Vietnam was toG dismantle retired Univac FASTRAND drums and literally grind off all the L oxide from the surface.  Though it surely induces heart attacks for computerF museum operators these days I suspect you could not find a more secureK method of ensuring data is not recovered by a third party.  However it does 1 affect resale value of the equipment on ebay.  :)     Jack Peacock    ------------------------------  % Date: Tue, 01 Apr 2003 11:22:39 -0600  From: brandon@dalsemi.com $ Subject: Re: Demagnetize Disk Drives1 Message-ID: <03040111223962@dscis6-0.dalsemi.com>   M > A college professor once explained to me his method of securing old drives. L > One of his early jobs working for an intelligence agency in Vietnam was toI > dismantle retired Univac FASTRAND drums and literally grind off all the N > oxide from the surface.  Though it surely induces heart attacks for computerH > museum operators these days I suspect you could not find a more secureM > method of ensuring data is not recovered by a third party.  However it does 3 > affect resale value of the equipment on ebay.  :)  >    Jack Peacock   O I think you could take the oxide dust, put it in small glass vials, label it as ? "classified data" and get a great price on e-bay...  marketing!        John Brandon VMS Systems Administrator  Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fx    ------------------------------  % Date: Tue, 01 Apr 2003 10:32:28 -0700 + From: "Barry Treahy, Jr." <Treahy@MMaz.com> $ Subject: Re: Demagnetize Disk Drives% Message-ID: <3E89CD2C.50205@MMaz.com>    brandon@dalsemi.com wrote:  M >>A college professor once explained to me his method of securing old drives. L >>One of his early jobs working for an intelligence agency in Vietnam was toI >>dismantle retired Univac FASTRAND drums and literally grind off all the N >>oxide from the surface.  Though it surely induces heart attacks for computerH >>museum operators these days I suspect you could not find a more secureM >>method of ensuring data is not recovered by a third party.  However it does 3 >>affect resale value of the equipment on ebay.  :)  >>   Jack Peacock  >>     >> > P >I think you could take the oxide dust, put it in small glass vials, label it as@ >"classified data" and get a great price on e-bay...  marketing! >    > F Remember the pet rock?  Just like you said, packaging and marketing...   Barry    --    @ Barry Treahy, Jr  *  Midwest Microwave  *  Vice President & CIO   A E-mail: Treahy@mmaz.com * Phone: 480/314-1320 * FAX: 480/661-7028    ------------------------------  % Date: Tue, 01 Apr 2003 09:44:42 +0100 ( From: Nic Clews <sendspamhere@127.0.0.1>( Subject: Re: Disk cache with shadow sets) Message-ID: <3E89517A.283E1617@127.0.0.1>    Jesse Chambless wrote: > G > We are running VMS V7.2-2 and volume shadowing ( 2 member sets ).  We G > write inhibit the application, and break the secondary member out for 	 > backup. G > My question is, when in this process is the disk cache flushed to the  > member we are breaking out?  > F > Since the secondary member will not dismount until all IO operations: > are complete, does the cache represent an incomplete IO?  . Rob has provided some very useful information.  = VMS as an operating system does not have write caching (yet).   H The cache you speak of, is the controller cache, which is to some degreeF independent of the operating system. Without the $FLUSH, the operatingG system signals that an operation has completed when the data arrives in A the cache, ready to be written (hence the importance of the cache @ battery). When you then read the data, if it hasn't been flushedE (written) to disk, it doesn't matter because your read of the disk is E the content of the disk and the unflushed cache, so you read back the F data you wrote, regardless if its on physical disk. The data itself isC written either when the cache becomes "full", or the flush interval  expires.  F That IO of which you speak is not "incomplete", however, that IO could= be a partial transaction of a (now) inconsistent data record.   H Take note of Rob's discussion of quiescing applications, also important.   --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot com    ------------------------------  $ Date: Tue, 1 Apr 2003 14:22:34 +01003 From: "Tom Wade" <t.wade@vms.eurokom.removespam.ie> 0 Subject: Re: Earth to Andrew, do you get it now?* Message-ID: <b6c3n6$s9u$1@kermit.esat.net>  5 "Bob Ceculski" <bob@instantwhip.com> wrote in message 7 news:d7791aa1.0303301807.7412168f@posting.google.com... J > One of the good things in OpenVMS are the Descriptors. With this OpenVMS' >is protected against "buffer overflow"   K Another useful feature is the use of PSECTs.  A good compiler will put code I in a non-writeable PSECT by default, thus making it possible to overwrite J data with buffer overruns, but not code.  Does the C compiler do this sort+ of thing ?  Does Unix have such a concept ?   D --------------------------------------------------------------------< Tom Wade             | EMail: tee dot wade at eurokom dot ie? EuroKom              | X400:  g=tom;s=wade;o=eurokom;p=eurokom; / Unit A2              |        a=eirmail400;c=ie / Nutgrove Office Park | Tel:   +353 (1) 296-9696 / Rathfarnham          | Fax:   +353 (1) 296-9697 < Dublin 14            | Disclaimer:  This is not a disclaimerC Ireland              | Tip:   "Friends don't let friends do Unix !"    ------------------------------   Date: 1 Apr 2003 07:39:56 -0600 ; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) 0 Subject: Re: Earth to Andrew, do you get it now?3 Message-ID: <zuHiDGY1jYNj@eisner.encompasserve.org>   h In article <d7791aa1.0303310709.5b4f3636@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes: > @ > but even if that is true, after reading others posts here, you@ > only get accvio errors or at worst maybe denial of service ...< > you can't just get on a vms system and do the nasty thingsA > that you can on unix/linux ... I think that has been made clear A > here and many times in the past, which means vms is more secure A > than unix/linux, and even the number of certs the past 10 years 9 > prove it ... why don't you just admit the truth Andrew?   F    Because it's simply not true.  C programs with buffer overflow bugsB    have been sucessfully ported to VMS with the bugs.  A notoriousD    example was provided by an early vendor in TCP/IP stacks for VMS.(    IIRC they aren't in business anymore.   ------------------------------   Date: 1 Apr 2003 07:43:05 -0600 ; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) 0 Subject: Re: Earth to Andrew, do you get it now?3 Message-ID: <dYIjJX0U+nOb@eisner.encompasserve.org>   ` In article <b6c3n6$s9u$1@kermit.esat.net>, "Tom Wade" <t.wade@vms.eurokom.removespam.ie> writes: >  > M > Another useful feature is the use of PSECTs.  A good compiler will put code K > in a non-writeable PSECT by default, thus making it possible to overwrite L > data with buffer overruns, but not code.  Does the C compiler do this sort- > of thing ?  Does Unix have such a concept ?  >   E    Doesn't matter.  That's now how the overrun exploits work.  And if A    you think I'm going to give you technical background on buffer B    overruns on USENET then I've got a bridge I'd like to sell you.   ------------------------------  % Date: Tue, 01 Apr 2003 16:04:02 +0200 E From: Jan C. =?iso-8859-1?Q?Vorbr=FCggen?= <jvorbrueggen@mediasec.de> 0 Subject: Re: Earth to Andrew, do you get it now?+ Message-ID: <3E899C52.CF974D9C@mediasec.de>   @ >    Doesn't matter.  That's now how the overrun exploits work.   F But PSECTs help in that the linker and IMGSTA could cooperate and markF the stack section NOEXE. Although I seem to remember that no processor: on which VMS runs honours the (NO)EXE attribute...too bad.   	Jan   ------------------------------  % Date: Tue, 01 Apr 2003 18:03:15 +0100 ' From: Andrew Harrison SUNUK Consultancy 0 Subject: Re: Earth to Andrew, do you get it now?. Message-ID: <3E89C653.1050206@nospamn.sun.com>   Bob   
 Two questions   8 1. Why did you re-post an article that has just appeared     in this thread.   6 2. Why did you re-post an article that doesn't support%     your claim but does support mine.   5 Note I said OpenVMS does not insulate you from buffer 4 overruns, it provides more insulation. Read the post1 you seemed to think we all needed to get a second 4 bite at comprehend the post that you seemed to think0 we need a second bite at and then admit that you
 are mistaken.    regards  Andrew Harrison      Bob Ceculski wrote:  > Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<3E88202E.4020009@nospamn.sun.com>...  >  >>Bob Ceculski wrote:  >> >>>Andrew writes:  >>>  >>>  >>> ; >>>>There is no OpenVMS magic bullet get used to it, I have 8 >>>>provided you with enough examples over the years and9 >>>>to be frank the argument is getting pretty pointless.  >>>  >>><<< >>> N >>>One of the good things in OpenVMS are the Descriptors. With this OpenVMS isM >>>protected against "buffer overflow" Also there was a better security model K >>>(e.g. xhost + versus security tripple transport, user, host). But in the N >>>newer times there are a lot of tools and programs without the philosophy ofG >>>OpenVMS. The most one came from UNIX. So a lot of buggy tools become  >>>standard in OpenVMS.  >>E >>No this is untrue you are not insulated against buffer overflows in @ >>OpenVMS, this is just another myth. You are more insulated but >>not insulated. >>4 >>If you need confirmation of this ask Hoff Hoffman. >>	 >>Regards  >>Andrew Harrison  >  > E > and here's his answer ... it may be a bit over your head though ...  > K > The descriptor construct actually tends to improve the general resistence L >   to buffer overruns -- both accidents and attacks -- while programs basedK >   on the standard C library interfaces tend to be far more vulnerable (on L >   OpenVMS, UNIX, Linux, Solaris, and other platforms) to these errors, dueJ >   in no small part to the lack of a descriptor-like construct within the >   standard C specifications. > H >   With the core of OpenVMS continuing to use descriptors and itemlistsL >   for argument passing, the numbers of APIs on OpenVMS that are vulnerableN >   to overruns is arguably rather smaller than that of UNIX.  A specific partK >   of the layering of OpenVMS that can be overlooked here: the standard C  N >   library on OpenVMS is not privileged, and -- unlike most UNIX boxes -- theO >   OpenVMS C library is not implemented as an integrated part of the operating Q >   sysyem kernel.  Further, the OpenVMS C library tends to use descriptor-based  N >   system  services and descriptor-based LIB run-time library (LIBRTL) calls. > N >   The relative difficulty of the buffer overrun on an OpenVMS system is thatM >   you not only have to blow out the buffer, you have to blow out the buffer J >   in a security-relevent environment -- privileged-mode or running with M >   privileges -- where something useful for a nefarious-minded programmer is L >   actually possible.  Within the Trusted Computing Base (TCB), to use some >   security jargon.   ------------------------------  # Date: Tue, 01 Apr 2003 17:36:26 GMT " From:   VAXman-  @SendSpamHere.ORG0 Subject: Re: Earth to Andrew, do you get it now?0 Message-ID: <00A1DBD3.C997A8A2@SendSpamHere.ORG>   In article <3E89C653.1050206@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> writes: >  >  >Bob >  >Two questions > 9 >1. Why did you re-post an article that has just appeared  >    in this thread. > 7 >2. Why did you re-post an article that doesn't support & >    your claim but does support mine.  	 A: Andrew  B: Bob  
 B: (Knock)
 A:   Come in. 0 B:   Ah, Is this the right room for an argument? A:   I told you once.  B:   No you haven't. A:   Yes I have.
 B:   When? A:   Just now. B:   No you didn't.  A:   Yes I did.  B:   You didn't  A:   I did!  B:   You didn't! A:   I'm telling you I did!  B:   You did not!!D A:   Oh, I'm sorry, just one moment. Is this a five minute argument       or the full half hour?  B:   Oh, just the five minutes. " A:   Ah, thank you. Anyway, I did.  B:   You most certainly did not.C A:   Look, let's get this thing clear; I quite definitely told you.  B:   No you did not. A:   Yes I did.  B:   No you didn't.  A:   Yes I did.  B:   No you didn't.  A:   Yes I did.  B:   No you didn't.  A:   Yes I did.  B:   You didn't.	 A:   Did. % B:   Oh look, this isn't an argument.  A:   Yes it is. * B:   No it isn't. It's just contradiction. A:   No it isn't.  B:   It is!  A:   It is not. $ B:   Look, you just contradicted me. A:   I did not.  B:   Oh you did!!  A:   No, no, no. B:   You did just then.  A:   Nonsense! B:   Oh, this is futile! A:   No it isn't. % B:   I came here for a good argument. 6 A:   No you didn't; no, you came here for an argument.* B:   An argument isn't just contradiction. A:   It can be. B B:   No it can't. An argument is a connected series of statements )      intended to establish a proposition.  A:   No it isn't. , B:   Yes it is! It's not just contradiction.D A:   Look, if I argue with you, I must take up a contrary position. 3 B:   Yes, but that's not just saying 'No it isn't.'  A:   Yes it is!  B:   No it isn't!    A:   Yes it is! D B:   Argument is an intellectual process. Contradiction is just the B      automatic gainsaying of any statement the other person makes.   (short pause)  A:   No it isn't.  B:   It is.  A:   Not at all. B:   Now look.   (Rings bell) A:   Good Morning.
 B:   What? A:   That's it. Good morning. # B:   I was just getting interested. # A:   Sorry, the five minutes is up. ! B:   That was never five minutes!  A:   I'm afraid it was.  B:   It wasn't.     Pause5 A:   I'm sorry, but I'm not allowed to argue anymore.  B:   What?! E A:   If you want me to go on arguing, you'll have to pay for another        five minutes.@ B:   Yes, but that was never five minutes, just now. Oh come on!
 A:  (Hums) B:   Look, this is ridiculous.@ A:   I'm sorry, but I'm not allowed to argue unless you've paid! B:   Oh, all right.    (pays money) A:   Thank you. 
   short pause 
 B:   Well? A:   Well what? / B:   That wasn't really five minutes, just now. = A:   I told you, I'm not allowed to argue unless you've paid.  B:   I just paid!  A:   No you didn't.  B:   I DID!  A:   No you didn't. , B:   Look, I don't want to argue about that. A:   Well, you didn't pay.: B:   Aha. If I didn't pay, why are you arguing? I Got you! A:   No you haven't.5 B:   Yes I have. If you're arguing, I must have paid. : A:   Not necessarily. I could be arguing in my spare time.  B:   Oh I've had enough of this. A:   No you haven't. B:   Oh Shut up.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM              5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------   Date: 1 Apr 2003 07:47:18 -0600 ; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)  Subject: Re: Exporting users? 3 Message-ID: <2YSYLyZk9$If@eisner.encompasserve.org>   d In article <pan.2003.03.31.14.12.39.659297@hotmail.com>, "DigiDemon" <digidemon@hotmail.com> writes: > Hello all! > K > I'll be totally redoing this OpenVMS machine from 7.2-1 to 7.3.  Is there L > any way I can save the users that are registered on this machine?  Thanks!  6    Send them to the nearest Starbucks for a few hours?  D    Actually, why are you "totally redoing", why not just an upgrade?  ?    If you really must reinstall, there's no reason to touch the G    non-system disks, and you can certainly backup any user files on the     systems disk.  B    As far as account settings, etc., just backup everything that'sG    listed as a cluster-common file in the system manager's manual, even     if you don't have a cluster.    ------------------------------  % Date: Tue, 01 Apr 2003 09:16:04 +0200 4 From: Didier Morandi <Didier.Morandi.nospam@Free.fr># Subject: Re: Fortran Guru requested & Message-ID: <3E893CB4.1080700@Free.fr>   Carl Perkins wrote:     D > My suggestion: go to INTEGER*4 TIME(2) instead of double precisionA > (or INTEGER*8 if you don't care about VAX compatability). There E > is some slight chance that the floating point is messing things up, G > especially if it now defaults to an IEEE data type (which can do some  > weird things).  	 I changed  	call sleep('05:00') to 	character duration*5  	duration='05:00'  	call sleep(duration)  and it is ok, now. Thanks,    D.   ------------------------------  $ Date: Tue, 1 Apr 2003 10:22:55 +0200( From: "Bruin, J.M. de" <Bruin@WT.TNO.NL># Subject: RE: Fortran Guru requested C Message-ID: <6B80E71673E6D611AC1D0008C7F37BC2737C85@wt15.wt.tno.nl>   & And that's what I've stated all along:  E passing a character STRING differs from passing a character VARIABLE. 5 A string isn't passed as a descriptor, a variable is.   G Probably (!?) doing CALL SLEEP(%DESCR('05:00')) does the trick as well.T   Mark   -----Original Message-----; From: Didier Morandi [mailto:Didier.Morandi.nospam@Free.fr]m$ Sent: Tuesday, April 01, 2003 09:16  To: Info-VAX@Mvb.Saic.Comr# Subject: Re: Fortran Guru requested6     Carl Perkins wrote:r    D > My suggestion: go to INTEGER*4 TIME(2) instead of double precisionA > (or INTEGER*8 if you don't care about VAX compatability). There E > is some slight chance that the floating point is messing things up,zG > especially if it now defaults to an IEEE data type (which can do some  > weird things).  	 I changedt 	call sleep('05:00') to 	character duration*5  	duration='05:00'  	call sleep(duration)a and it is ok, now. Thanks,n   D.   ------------------------------  % Date: Tue, 01 Apr 2003 17:23:20 +1000t1 From: Paddy O'Brien <paddy.o'brien@tg.nsw.gov.au>e# Subject: Re: Fortran Guru requestedr, Message-ID: <3E893E68.3060908@tg.nsw.gov.au>   Ken Fairfield wrote:   [snips]u  D > 	Paddy (and John Santos) correctly question using DOUBLE PRECISIONJ > for the TIME variable.  Use INTEGER*4 TIME(2) instead.  The issue is theJ > _potential_ for an HPARITH error for some bit patterns in those 64 bits. > : Ken, (and Hi it's a long time since we had a conversation)  B Your attributions are a little bit wrong.  It was John Santos who E originally postulated the problem that you describe.  In my reply to gA him, I suggested that there should be no problem with TIME being  E declared REAL*8 (though I entirely agree that the correct definition   should have been used).7  H My postulation was that TIME was not "touched" in the SLEEP routine and H should therefore maintain the same bit pattern between the calls to the  system routines..   D > 	However, the address of %x30 looks to me like the TAG variable isJ > being passed by address, that is %x30 is 48 decimal, which is "0" ASCII.C > I'd suspect the calling routine, SPECIAL_GRAPHIC in this case, iso	 > callingeC > sleep with a Hollerith argument or a numeric variable that's beeni
 > initialized G > with character data (non-standard, common place old usage).  Find allf > thelA > callers of SLEEP and make sure that the TAG argument is of typee > CHARACTER,F > or is specified as a character literal (e.g., CALL SLEEP('10:00') ). > ? Yes, this seems to be the real (pun) problem.  But the call in iH SPECIAL_GRAPHIC uses a descriptor, as it should.  At least my knowledge = of VMS Fortran says that CALL SLEEP('10.00') is a descriptor.n  D Along with (IIRC) Carl Perkins, I cannot see a problem with Fortran  V7.5. on VMS 7.2  H The release notes of the FORRTL do not address this, nor do the Fortran  release notes.  I My only suggestion now is that Didier bundles up a replay of what caused iF the problem and passes it to fortran@compaq.com (might be hp.com, but D they still receive my compiler internal errors) and giving them his  version of Fortran.p   Regards, Paddy      G ***********************************************************************?  C "This electronic message and any attachments may contain privilegedm> and confidential information intended only for the use of the B addressees named above.  If you are not the intended recipient of C this email, please delete the message and any attachment and advisefB the sender.  You are hereby notified that any use, dissemination, 7 distribution, reproduction of this email is prohibited.   A If you have received the email in error, please notify TransGrid  A immediately.  Any views expressed in this email are those of the  = individual sender except where the sender expressly and with 0C authority states them to be the views of TransGrid.  TransGrid uses > virus scanning software but excludes any liability for viruses contained in any attachment.  < Please note the email address for TransGrid personnel is now$ firstname.lastname@transgrid.com.au"  G ***********************************************************************m   ------------------------------   Date: 1 Apr 2003 10:29:09 +0200o' From: huber@mppmu.mpg.de (Joseph Huber)e# Subject: Re: Fortran Guru requestedl+ Message-ID: <FCTDj3YgUfsq@vms.mppmu.mpg.de>a  ] In article <3E893CB4.1080700@Free.fr>, Didier Morandi <Didier.Morandi.nospam@Free.fr> writes:) >  > I changedm > 	call sleep('05:00') > to > 	character duration*5d > 	duration='05:00'p > 	call sleep(duration)i > and it is ok, now.	 > Thanks,   , Glad it works - but be warned, it just hidesM some error, most probably overriding local variable space (calling sequences,0J descriptors). You had a similar shift of the error, when You inserted some other variables yesterday.  K I compiled Your code just with a later compiler- and forrtl version without L catching an error (I did a /check=bounds, so it is not simply array boundaryH violation). And changing character constants to variables should make no3 difference, it just moves locations of descriptors.t   Good luck , it's just a game !   -- eN Joseph "Sepp" Huber   mailto:joseph.huber@web.de   http://www.huber-joseph.de/   ------------------------------  % Date: Tue, 01 Apr 2003 17:31:47 +1000w1 From: Paddy O'Brien <paddy.o'brien@tg.nsw.gov.au>I# Subject: Re: Fortran Guru requested , Message-ID: <3E894063.3000308@tg.nsw.gov.au>   Didier,u  G There should be no difference.  Take up my suggestion in a mail sent a  H minute ago (in reply to Ken Fairfield) that you should send this to the B compiler team.  If this is a compiler bug that none of us has yet G encountered, you will be helping the VMS Fortran community.  I'm part, -! and I shall thank you for it. :-)2   Regards, Paddy   Didier Morandi wrote:a > Carl Perkins wrote:j >  > E >> My suggestion: go to INTEGER*4 TIME(2) instead of double precision B >> (or INTEGER*8 if you don't care about VAX compatability). ThereF >> is some slight chance that the floating point is messing things up,H >> especially if it now defaults to an IEEE data type (which can do some >> weird things).a >  >  > I changed  >     call sleep('05:00')v > to >     character duration*5 >     duration='05:00' >     call sleep(duration) > and it is ok, now.	 > Thanks,  >  > D.      G ***********************************************************************t  C "This electronic message and any attachments may contain privilegedn> and confidential information intended only for the use of the B addressees named above.  If you are not the intended recipient of C this email, please delete the message and any attachment and advisetB the sender.  You are hereby notified that any use, dissemination, 7 distribution, reproduction of this email is prohibited.2  A If you have received the email in error, please notify TransGrid aA immediately.  Any views expressed in this email are those of the l= individual sender except where the sender expressly and with nC authority states them to be the views of TransGrid.  TransGrid uses > virus scanning software but excludes any liability for viruses contained in any attachment.  < Please note the email address for TransGrid personnel is now$ firstname.lastname@transgrid.com.au"  G ***********************************************************************-   ------------------------------  % Date: Tue, 01 Apr 2003 17:43:32 +1000"1 From: Paddy O'Brien <paddy.o'brien@tg.nsw.gov.au>e# Subject: Re: Fortran Guru requested8, Message-ID: <3E894324.5040502@tg.nsw.gov.au>   Mark,n  F Again I must beg to differ.  On VMS, CALL SLEEP('05:00') is passed by F descriptor.  If it were not, it would kill all us Fortranners who use E the CXML (based on LAPACK) routines, at least.  Also all the calls I RG have to NAG library routines.  Also all the calls I have to home grown T	 routines.k  G A literal constant passed in quotes is passed by descriptor.  Not %VAL 2 or %LOC.  F Note also that if you read Didier's program that he posted, there are - many calls to SLEEP with CHARACTER constants.l  H There is something strange about the SPECIAL_GRAPHICS call that none of   us have yet been able to fathom.   Regards, Paddy     Bruin, J.M. de wrote:p( > And that's what I've stated all along: > G > passing a character STRING differs from passing a character VARIABLE.i7 > A string isn't passed as a descriptor, a variable is.a > I > Probably (!?) doing CALL SLEEP(%DESCR('05:00')) does the trick as well.c >  > Mark >  > -----Original Message-----= > From: Didier Morandi [mailto:Didier.Morandi.nospam@Free.fr] & > Sent: Tuesday, April 01, 2003 09:16  > To: Info-VAX@Mvb.Saic.Comw% > Subject: Re: Fortran Guru requested+ >  >  > Carl Perkins wrote:t >  >  > D >>My suggestion: go to INTEGER*4 TIME(2) instead of double precisionA >>(or INTEGER*8 if you don't care about VAX compatability). TherenE >>is some slight chance that the floating point is messing things up,eG >>especially if it now defaults to an IEEE data type (which can do some  >>weird things). >  >  > I changed  > 	call sleep('05:00') > to > 	character duration*5e > 	duration='05:00'r > 	call sleep(duration)c > and it is ok, now.	 > Thanks,a >  > D.      G ***********************************************************************u  C "This electronic message and any attachments may contain privileged.> and confidential information intended only for the use of the B addressees named above.  If you are not the intended recipient of C this email, please delete the message and any attachment and advise0B the sender.  You are hereby notified that any use, dissemination, 7 distribution, reproduction of this email is prohibited.r  A If you have received the email in error, please notify TransGrid 'A immediately.  Any views expressed in this email are those of the e= individual sender except where the sender expressly and with oC authority states them to be the views of TransGrid.  TransGrid usesp> virus scanning software but excludes any liability for viruses contained in any attachment.  < Please note the email address for TransGrid personnel is now$ firstname.lastname@transgrid.com.au"  G ***********************************************************************    ------------------------------  % Date: Tue, 01 Apr 2003 18:00:59 +1000l1 From: Paddy O'Brien <paddy.o'brien@tg.nsw.gov.au>p# Subject: Re: Fortran Guru requesteda, Message-ID: <3E89473B.7040506@tg.nsw.gov.au>  D Picked up from another mail 'cos I lost this one, Carl -- so you're  double quoted :-)1   >> Carl Perkins wrote: >>F >>> My suggestion: go to INTEGER*4 TIME(2) instead of double precisionC >>> (or INTEGER*8 if you don't care about VAX compatability). TheremG >>> is some slight chance that the floating point is messing things up,cI >>> especially if it now defaults to an IEEE data type (which can do somee >>> weird things).  D IEEE is done by a qualifier.  The default on Alpha is F-floating or B G-floating.  The default on VAX is F-floating or D-floating.  The D default on IA64 could well be IEEE, but I don't know.  It does make I sense though since the F2K standard is geared to IEEE mathematics, as is  ! I believe the Intel architecture.a   Regards, Paddy      G ***********************************************************************F  C "This electronic message and any attachments may contain privilegedn> and confidential information intended only for the use of the B addressees named above.  If you are not the intended recipient of C this email, please delete the message and any attachment and advisehB the sender.  You are hereby notified that any use, dissemination, 7 distribution, reproduction of this email is prohibited.u  A If you have received the email in error, please notify TransGrid gA immediately.  Any views expressed in this email are those of the e= individual sender except where the sender expressly and with  C authority states them to be the views of TransGrid.  TransGrid uses > virus scanning software but excludes any liability for viruses contained in any attachment.  < Please note the email address for TransGrid personnel is now$ firstname.lastname@transgrid.com.au"  G ***********************************************************************    ------------------------------  % Date: Tue, 01 Apr 2003 12:44:16 +0100o+ From: John Laird <john@laird-towers.org.uk>i# Subject: Re: Fortran Guru requestedt8 Message-ID: <3oui8v4jtd6q0jk8mdbtm4ragtfd7rts0p@4ax.com>  2 On Mon, 31 Mar 2003 09:07:43 +0200, Didier Morandi& <Didier.Morandi.nospam@Free.fr> wrote:  P >Yeah man, but how do you explain that some of you here did compile/link/run it  >without pb???  H The source you posted doesn't correspond to the stack dump :-))  Perhaps: this explains why no-one else could reproduce your error !  E Hint:  in the source, MOVE does *not* call SPECIAL_GRAPHIC.  It calls- SCR$SET_CURSOR only.  E [Don't worry, I've had more than my fair share of compiling the wrong.J source, quitting from the editor instead of exiting, forgetting to compile= and wondering why the relinked image still crashed, etc etc.]-     	John-   ------------------------------   Date: 1 Apr 2003 07:32:04 -0600c From: briggs@encompasserve.org# Subject: Re: Fortran Guru requested83 Message-ID: <+1aagHNZFOPq@eisner.encompasserve.org>   U In article <FCTDj3YgUfsq@vms.mppmu.mpg.de>, huber@mppmu.mpg.de (Joseph Huber) writes: _ > In article <3E893CB4.1080700@Free.fr>, Didier Morandi <Didier.Morandi.nospam@Free.fr> writes:r >> t >> I changed >> 	call sleep('05:00')t >> toh >> 	character duration*5 >> 	duration='05:00' >> 	call sleep(duration) >> and it is ok, now.s
 >> Thanks, > . > Glad it works - but be warned, it just hidesO > some error, most probably overriding local variable space (calling sequences,sL > descriptors). You had a similar shift of the error, when You inserted some > other variables yesterday.  C Changing character constants to variables does plausibly change oner? thing -- the black magic that deals with Hollerith literals and.> procedure calls.  The linker has the ability to change Fortran! calling conventions at link time.s   	John Briggs   ------------------------------   Date: 1 Apr 2003 16:08:19 +0200i' From: huber@mppmu.mpg.de (Joseph Huber) # Subject: Re: Fortran Guru requested + Message-ID: <HNJ4k4lLIAVl@vms.mppmu.mpg.de>a  T In article <+1aagHNZFOPq@eisner.encompasserve.org>, briggs@encompasserve.org writes:W > In article <FCTDj3YgUfsq@vms.mppmu.mpg.de>, huber@mppmu.mpg.de (Joseph Huber) writes:d` >> In article <3E893CB4.1080700@Free.fr>, Didier Morandi <Didier.Morandi.nospam@Free.fr> writes: >>> 
 >>> I changede >>> 	call sleep('05:00') >>> to >>> 	character duration*5e >>> 	duration='05:00'  >>> 	call sleep(duration)h >>> and it is ok, now. >>> Thanks,E >>  / >> Glad it works - but be warned, it just hidesAP >> some error, most probably overriding local variable space (calling sequences,M >> descriptors). You had a similar shift of the error, when You inserted somes >> other variables yesterday.s > E > Changing character constants to variables does plausibly change one-A > thing -- the black magic that deals with Hollerith literals and0@ > procedure calls.  The linker has the ability to change Fortran# > calling conventions at link time.l  F '05:00' is a character string, not a Hollerith literal. We are dealingJ with Fortran77 and later. It's always a descriptor, and the difference is J only the location of the string itself (.e.g. in a read-only or rw psect).    -- sN Joseph "Sepp" Huber   mailto:joseph.huber@web.de   http://www.huber-joseph.de/   ------------------------------   Date: 1 Apr 2003 11:10:25 -0600e From: briggs@encompasserve.org# Subject: Re: Fortran Guru requestedr3 Message-ID: <h1qi4C8L19eL@eisner.encompasserve.org>k  E [references header broken -- my reader doesn't like it when it wraps]aU In article <HNJ4k4lLIAVl@vms.mppmu.mpg.de>, huber@mppmu.mpg.de (Joseph Huber) writes:eV > In article <+1aagHNZFOPq@eisner.encompasserve.org>, briggs@encompasserve.org writes:X >> In article <FCTDj3YgUfsq@vms.mppmu.mpg.de>, huber@mppmu.mpg.de (Joseph Huber) writes:a >>> In article <3E893CB4.1080700@Free.fr>, Didier Morandi <Didier.Morandi.nospam@Free.fr> writes:l >>>> f >>>> I changed >>>> 	call sleep('05:00')c >>>> to  >>>> 	character duration*5 >>>> 	duration='05:00' >>>> 	call sleep(duration) >>>> and it is ok, now.e >>>> Thanks, >>> 0 >>> Glad it works - but be warned, it just hidesQ >>> some error, most probably overriding local variable space (calling sequences,oN >>> descriptors). You had a similar shift of the error, when You inserted some >>> other variables yesterday. >> .F >> Changing character constants to variables does plausibly change oneB >> thing -- the black magic that deals with Hollerith literals andA >> procedure calls.  The linker has the ability to change Fortranr$ >> calling conventions at link time. > H > '05:00' is a character string, not a Hollerith literal. We are dealingL > with Fortran77 and later. It's always a descriptor, and the difference is L > only the location of the string itself (.e.g. in a read-only or rw psect).  F We have strong evidence that the '05:00' is being passed by reference.C Its location in a read only or read write psect is irrelevant if wefB try to do indirect reference based on a descriptor whose value is:  ? '05:0' : 303A0500	A 1280 byte string with bogus descriptor typea+ '0   ' : 00000030	The address of the stringm  A Yes, I agree that in my experience, character string literals are E always passed by descriptor.  But in this case, that's not happening.d@ Whether that is a compiler bug, a linker bug or a combination of the two is not clear.d  > There is a competing explanation -- that the '05:00' parameterB value is somehow overwritten across Fortran's statically allocatedA string descriptor due to a program bug.  Why would you think thatl< this behavior would be influenced by the psect attributes of: the address space within which the '05:00' parameter value
 is allocated?    	John Briggs   ------------------------------   Date: 1 Apr 2003 11:40:01 -0600  From: briggs@encompasserve.org# Subject: Re: Fortran Guru requestedA3 Message-ID: <n9+GjZC7Y1nS@eisner.encompasserve.org>   T In article <h1qi4C8L19eL@eisner.encompasserve.org>, briggs@encompasserve.org writes:A > '05:0' : 303A0500	A 1280 byte string with bogus descriptor type1- > '0   ' : 00000030	The address of the stringa  A Correcting my translation of character string to descriptor, this 
 should be:  @ '05:0' : 303A3530	A 13616 byte string with bogus descriptor type, '0'    : 00000030	The address of the string.  B Since we actually saw reference to address 00000030, we may safely@ assume that the three bytes allocated following the '05:00' were zeroes at run time.o   	John Briggs   ------------------------------  % Date: Tue, 01 Apr 2003 20:29:20 +0200 4 From: Didier Morandi <Didier.Morandi.nospam@Free.fr># Subject: Re: Fortran Guru requested & Message-ID: <3E89DA80.4050305@Free.fr>   Paddy O'Brien wrote: ../..e, >  Didier bundles up a replay of what causedH > the problem and passes it to fortran@compaq.com (might be hp.com, but F > they still receive my compiler internal errors) and giving them his  > version of Fortran.    I will.t   D.   ------------------------------  % Date: Tue, 01 Apr 2003 13:10:30 -0500e9 From: Hein van den Heuvel <hein_netscape@eps.zko.dec.com>e# Subject: Re: High RMS file activityy/ Message-ID: <3E89D616.FA4CA0B8@eps.zko.dec.com>    Chris Sharman wrote:  
 > Hein wrote:aX > > Still, for grins, run my rms_tune_check (script in the .txt file) on a critical disk   > Some feedback for you: > Some 'BUG!!' messages below.  " > $1$DKA500:[OTHER]DESPATCH.DAT;31  2 > - BUG!! Bad record flag 3 at byte 49 in bucket 4  < Oops... I forgot a condition. If duplicates are allowed on a= prologue 3 file then for deleted records the keys stay behind-: and the data gets squished out. This only is a problem for? this tool if no compression is active with fixed length recordsu= because otherwise RMS will have a record length field to use.c8 This file must have had a 26 byte fixed record: 49 minus0 record_overhead (= 9) minus bucket_header (= 14)    Thanks for the report! Will fix.   Hein.t   ------------------------------  $ Date: Tue, 1 Apr 2003 10:38:02 -08001 From: "Farrell, Michael" <MFarrell@voltdelta.com>f# Subject: RE: High RMS file activitycP Message-ID: <025766C9BBC5D511A4ED00B0D0F08C23163B24@zny_exchange1.maintech1.com>   Hein,r  K Can you let the group know when these fixes are available?  Can you tell uso9 where to find them before the next freeware CD is issued?-   Thanks,-   Mike Farrell    -----Original Message------B From: 	Hein van den Heuvel [mailto:hein_netscape@eps.zko.dec.com] % Sent:	Tuesday, April 01, 2003 1:11 PMo To:	Info-VAX@Mvb.Saic.ComA# Subject:	Re: High RMS file activityc       Chris Sharman wrote:  
 > Hein wrote:2J > > Still, for grins, run my rms_tune_check (script in the .txt file) on a
 critical diske   > Some feedback for you: > Some 'BUG!!' messages below.  " > $1$DKA500:[OTHER]DESPATCH.DAT;31  2 > - BUG!! Bad record flag 3 at byte 49 in bucket 4  < Oops... I forgot a condition. If duplicates are allowed on a= prologue 3 file then for deleted records the keys stay behind?: and the data gets squished out. This only is a problem for? this tool if no compression is active with fixed length recordsn= because otherwise RMS will have a record length field to use. 8 This file must have had a 26 byte fixed record: 49 minus0 record_overhead (= 9) minus bucket_header (= 14)    Thanks for the report! Will fix.   Hein.    ------------------------------  $ Date: Tue, 1 Apr 2003 12:37:53 -0500; From: "Brian Tillman" <Brian.Tillman@notnoone.notnohow.com>s$ Subject: Re: HP doesn't know OpenVMS$ Message-ID: <3e89ce6e$1@news.si.com>  L >first, if you look at the support page you will notice that the first thing) >out of your mouth should be OpenVMS whenu" >calling the support hotlines etc.   Did that.  Didn't help.  -- uI Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot comk5 Smiths Aerospace  Addresses modified to prevent SPAM. @ 3290 Patterson Ave. SE, MS Replace "at" with "@", "dot" with "." Grand Rapids, MI 49512-1991.8        This opinion doesn't represent that of my company   ------------------------------  $ Date: Tue, 1 Apr 2003 13:55:07 -0500; From: "Brian Tillman" <Brian.Tillman@notnoone.notnohow.com> ( Subject: Re: Logical names in decwindows$ Message-ID: <3e89e087$1@news.si.com>  K >However, when I start an application such as MAIL from the session managerb asF >a detached process, is there any way to give it some logicals names ?  G Define your Mail menu item run a command procedure that will define the ) logical names you want and then run mail.t --  I Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot comt5 Smiths Aerospace  Addresses modified to prevent SPAM. @ 3290 Patterson Ave. SE, MS Replace "at" with "@", "dot" with "." Grand Rapids, MI 49512-1991e8        This opinion doesn't represent that of my company   ------------------------------  % Date: Tue, 01 Apr 2003 16:23:39 +0100r9 From: Alan Adams <alan.adams@orchard-way.freeserve.co.uk>a' Subject: Re: Macs vs. Pathworks et. al.n? Message-ID: <25c461dc4b.Alan.Adams@orchard-way.freeserve.co.uk>   + In message <b6akqk02g87@enews2.newsguy.com> =           "Zane H. Healy" <healyzh@shell1.aracnet.com> wrote:s  7 > Carl Karcher <karcher@thuria.waisman.wisc.edu> wrote:aE > > Last time I tried this from OS 10.2 to Pathworks V6.1 I received:1 > ) > >        "An error has occurred: -5000"s > J > > Support will say "MacIntosh clients are not supported by PATHWORKS forI > > OpenVMS V6.1". Have them escalate it anyway. If enough of us to that o@ > > it might get fixed. I've not tried it with AS 7.3A-ECO1 yet. > K > Unless Pathworks supports disks drastically different than Samba then therM > problem is probably that DAVE expects to be able to create directories withtK > names that aren't valid for OpenVMS.  At least that was the problem I hado? > when attempting to use DAVE to access an OpenVMS Samba share.i > 	 > 			Zanee  K Pathworks has a renaming / name-hiding scheme to allow PC-legal names to beoF stored on ODS-2 disks. The problem here may be due to the Mac allowing characters which a PC does not.   . Not sure what Pathworks does on an ODS-5 disk.  G Pathworks/MAC also had a renaming scheme, completely different from the D Pathworks/PC one. This made life fun serving the same files to both.   Alan   -- w
 Alan Adams& alan.adams@orchard-way.freeserve.co.uk http://www.nckc.org.uk/n   ------------------------------  $ Date: Tue, 1 Apr 2003 13:20:02 -0500; From: "Brian Tillman" <Brian.Tillman@notnoone.notnohow.com>a- Subject: Re: Netscape gags on VMS FTP Serversr$ Message-ID: <3e89daf2$1@news.si.com>  G >Anyone know if Opera works better against VMS FTP sites than these twoe dogs?a  G Many of the true FTP clients work well with VMS.  In particular, I likeoL WS_FTP LE.  (www.ftpplanet.com)  I use it for my home FTP needs.  At work, IK use the FTP that is part of Exceed (an X Window server package).  Why use a>" browser to do an FTP client's job? -- rI Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot comh5 Smiths Aerospace  Addresses modified to prevent SPAM. @ 3290 Patterson Ave. SE, MS Replace "at" with "@", "dot" with "." Grand Rapids, MI 49512-1991c8        This opinion doesn't represent that of my company   ------------------------------  $ Date: Tue, 1 Apr 2003 13:38:47 -0500; From: "Brian Tillman" <Brian.Tillman@notnoone.notnohow.com>o2 Subject: Re: New additions to decus.acornsw.com...$ Message-ID: <3e89dcb4$1@news.si.com>   >Check it out at:  >n >    http://decus.acornsw.com/   I tried.  % Unable to connect to the destination.e -- aI Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot comi5 Smiths Aerospace  Addresses modified to prevent SPAM. @ 3290 Patterson Ave. SE, MS Replace "at" with "@", "dot" with "." Grand Rapids, MI 49512-1991A8        This opinion doesn't represent that of my company   ------------------------------  # Date: Tue, 01 Apr 2003 16:07:16 GMTg" From:   VAXman-  @SendSpamHere.ORG' Subject: New SPAWN behavior with V7.3-1e0 Message-ID: <00A1DBC7.54E20983@SendSpamHere.ORG>  I This is a question for the VMS engineers reading this newsgroup.  I'd ask J it directly of VMS engineering via WIS but the question and its answer may be of use/interest to others.i  I VMS V7.3-1 changes aspects of SPAWNing subprocesses.  My question is:  IsvI there any mechanism (SYSGEN parameter or other device) to revert back thet old behavior?  n  I I don't have time at the moment to scour the V7.3-1 listings.  I have onesJ customer with an old homegrown menuing package that no longer functions inH the same fashion as it used to on machine.  Because of this, a piece of I software that I support and that they use in their subprocesses cannot be K used.  The problem has nothing to do with my software directly but the fact J that their menuing software relied upon the old process naming scheme.  ItK is my guess that there is nobody willing to make the changes to the menuingN/ system so how can the old behavior be returned?o   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMs             5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------  # Date: Tue, 01 Apr 2003 16:34:45 GMTc" From:   VAXman-  @SendSpamHere.ORG+ Subject: Re: New SPAWN behavior with V7.3-1 0 Message-ID: <00A1DBCB.2BBC0961@SendSpamHere.ORG>  S In article <00A1DBC7.54E20983@SendSpamHere.ORG>, VAXman-  @SendSpamHere.ORG writes: J >This is a question for the VMS engineers reading this newsgroup.  I'd askK >it directly of VMS engineering via WIS but the question and its answer may  >be of use/interest to others. >,J >VMS V7.3-1 changes aspects of SPAWNing subprocesses.  My question is:  IsJ >there any mechanism (SYSGEN parameter or other device) to revert back the >old behavior?   > J >I don't have time at the moment to scour the V7.3-1 listings.  I have oneK >customer with an old homegrown menuing package that no longer functions in:I >the same fashion as it used to on machine.  Because of this, a piece of sJ >software that I support and that they use in their subprocesses cannot beL >used.  The problem has nothing to do with my software directly but the factK >that their menuing software relied upon the old process naming scheme.  ItNL >is my guess that there is nobody willing to make the changes to the menuing0 >system so how can the old behavior be returned?   I must have read too quickly...u  I I received a private email from Richard Brodie (thanks Richard) about then SYSGEN parameter DCL_CTLFLAGS.   That saved my time.o --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMi            L5   "Well my son, life is like a beanstalk, isn't it?" i   ------------------------------  # Date: Tue, 01 Apr 2003 16:37:24 GMT + From: "Mike Kier" <michael.kier@compaq.com> + Subject: Re: New SPAWN behavior with V7.3-1t/ Message-ID: <8fjia.124$mk1.75@news.cpqcorp.net>   , <VAXman- @SendSpamHere.ORG> wrote in message* news:00A1DBC7.54E20983@SendSpamHere.ORG...K > This is a question for the VMS engineers reading this newsgroup.  I'd askWL > it directly of VMS engineering via WIS but the question and its answer may > be of use/interest to others.  >:K > VMS V7.3-1 changes aspects of SPAWNing subprocesses.  My question is:  Is K > there any mechanism (SYSGEN parameter or other device) to revert back the  > old behavior?t > K > I don't have time at the moment to scour the V7.3-1 listings.  I have oneYL > customer with an old homegrown menuing package that no longer functions inI > the same fashion as it used to on machine.  Because of this, a piece ofsK > software that I support and that they use in their subprocesses cannot berH > used.  The problem has nothing to do with my software directly but the factL > that their menuing software relied upon the old process naming scheme.  ItE > is my guess that there is nobody willing to make the changes to the  menuingo1 > system so how can the old behavior be returned?u >d > --4 > VAXman- OpenVMS APE certification number: AAA-0001 VAXman(at)TMESIS(dot)COM >.6 >   "Well my son, life is like a beanstalk, isn't it?" >:  
 Hey Brian!  I Set bit 0 in DCL_CTLFLAGS - I dunno offhand if it is a dynamic parameter,  though I suspect it is not.    --	 Mike KierA0 Consultant, HP Consulting & Integration Services Cincinnati, OH, USAo mike.kier@hp.com  ' "Practice random acts of VMS marketing"k   ------------------------------  % Date: Tue, 01 Apr 2003 12:02:39 +01001' From: Andrew Harrison SUNUK ConsultancyoY Subject: Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retain. Message-ID: <3E8971CF.6030409@nospamn.sun.com>   jlsue wrote:G > On Mon, 31 Mar 2003 11:59:00 +0100, Andrew Harrison SUNUK Consultancyl0 > <Andrew_No.Harrison_No@nospamn.sun.com> wrote: >  >  >> >>Main, Kerry wrote: >  > K >>>So, are you saying that these GS Series Customers with Oracle FinancialsS0 >>>never considered performance as a criteria??? >>> B >>>Come on, you can do better fud than that .. Lets not get silly. >>>I/ >>>While a tad dated now, here are some quotes: > >>>http://h18002.www1.hp.com/alphaserver/gs/quotes/oracle.html >>>r >>9 >>So you have some nice sounding testimonials from Oracleg= >>execs, in return for your porting and joint marketing fees.c >>= >>Incedentally they arn't very current, one of the two Oraclea? >>products headlined is Oracle 8i which is in maintenance mode,  >>9i is the current DBMS.r >  > D > But the point is still that there are testimonials to counter your@ > argument that the GS160/320 is not scalable enough for runningG > transaction/database systems.  Apparently in the real world there arep$ > ample examples of happy customers. >   5 Really so have you actually read the testimonials ???c  A Acxiom. Avista Advantage, BlueSky Studios, ClearData, Employease,n@ First Health, International Truck and Engine Corporation, Lycos,< Memorial Herman Healthcare, Parexcel, Southern Freight Lines> are all quotes from companies saying how much they are lookingA forward to getting their GS boxes not how well the boxes perform.u  < SAIC are a Compaq Integrator, they currently have three jobs= requiring OpenVMS experience only one of which is a full timeE@ OpenVMS placement, no Tru64 jobs, they do on the other hand have. they have 131 jobs that require Solaris skills  < That leaves Bank Austria which isn't a good reference unless= you are hell bent in illustrating how poorly GS boxes performo" when compared with the TurboLaser.  ? Celera Genomics who are so committed to GS boxes that they havei bought IBM P690's instead.  G Northern Light were aquired by Divine and the public search engine thatXC the reference refers to no longer exists because it was a financials drain.  C E*TRADE who are apparently moving because of Sybase support issues.i  ? And thats your list of GS customer references, sorry this is sor; bad that its embarassing and not one of them actually referg; to the "performance" advantage that the GS boxes gave them.    Regards  Andrew Harrisone         > : >>Does this mean that Oracle are fully endorsing customers; >>who want to run 8i on GS boxes but are less interested ink >>9i.  >  > G > Now, why would it have to mean that?  It means that you are incorrect2F > if you assert that Oracle apps do not perform well on the GS160/320. >  >    ------------------------------  % Date: Tue, 01 Apr 2003 18:20:10 +0100r' From: Andrew Harrison SUNUK Consultancy Y Subject: Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha RetaiE, Message-ID: <3E89CA4A.10808@nospamn.sun.com>   Main, Kerry wrote: > Andrew, Andrew ..sigh .. > G > <<< The last benchmark of any significance that was published for thetF > GS320 was your TPC-C result which was submitted on 18/06/2001 over a) > year before the GS1280 announcement.<<<y > ; > And when was the last Sun TPC benchmark - for any system?d >   ! What has that got to do with it ?d  2 We don't do TPC-C mainly because of the lead taken6 by Compaq in shared nothing TPC-C benchmarkology which3 has destroyed the remaining value that it had after 6 Compaq nad Sequent had started the rot with the OPS in
 a box scheme.h  7 I said the last major benchmark you published was TPC-Co8 and it was a pretty terrible result unless being 4th was a good position.  9 There are other benchmarks, Oracle Apps, Peoplesoft, SAP, 6 TPC-H etc but all your GS benchmark runs for those are older than the TPC-C number.  : And the last major E10K benchmark result Sun published was7 SPECjbb on the 20th Nov 2001 somewhat after the last GSF/ 320 result and the E10K is a much older system.e  7 So I guess that wraps that up, silly point on your partp best forgotten.a    H > [insert never ending argument about how the world should adapt to whatJ > you feel is the right way to do benchmarks - even though the TPC council > disagrees with you]n >   8 We obviously have a cultural comprehension gap here. You2 keep banging on about TPC-C because you think that+ Sun not doing TPC-C is somehow significant.2  1 While I cannot understand why you keep mentioningi- a benchmark which you ran it on the GS320 butl0 which at the time you ended up with a 4th placed2 result for. 4th isn't leading it isn't even on the
 medal podium.   1 Would you like to captain Englands Cricket team ?r! you seem like an ideal candidate.i  @ > <<< Keep it run your existing apps on it but for the new stuff > ?????????<<< > G > Actually, given that Oracle 9I runs better than 8i, if a Customer wastI > happy with their GS series system before with Oracle 8i, why would thenyI > not consider going to Oracle 9i (other than normal test time issues) ??r >   8 Does it how interesting, but that isn't what your Oracle* reference page refers to, it refers to 8i.  F > It also tends to confirm the pessimistic views of your customers who; > seem less than enamoured by HP's strategy for OpenVMS.<<<i > E > Of course, what would a direct competitor state?  Spread the fud ..i >   9 No I am simply reporting what your web site shows and then' inevitable conclusion you have to draw.s  ? If you cannot even get your paid for Oracle VP to say somethingg> nice about 9i running on AlphaServers and get it onto your web  site then what hope do you have.   Regardss Andrew Harrisone   ------------------------------  % Date: Tue, 01 Apr 2003 17:37:00 +0200*8 From: "serge.zangheri" <szangheri@grenoble.sema.slb.com># Subject: pathworks administer issue 5 Message-ID: <3E89B21B.B6355B87@grenoble.sema.slb.com>o   Hi,b* I installed Pathworks for the second time.3 But when I start admin I got the following answer :t+ Cannot continue - the server is not runningd  F I can connect pwrk$root:netlogon etc, on my pc, I can do admin/config. No problem about licence.t swapfile put to 150000 pagefile put to 300000 80Mb RAM# other system paramters semmes good.    Someone could help ? Thanks Sergeu   ------------------------------  $ Date: Tue, 1 Apr 2003 15:17:27 +0200 From: atilla <na@pandinus.com>; Subject: PicoVax and OpenVMS hobbyist CD beginners questiont7 Message-ID: <MPG.18f3afd2f9e7ed16989680@news.xs4all.nl>h   Hi all,m  K i'm interested in picking up my VMS knowladge again(it's been 10 years) So oG i downloaded picovax and want to order the hobbyist cd. Is this a good t: 'home' solution? does anyone have some experience with it?  F host machine is a intele celeron 1.7 Ghz with 384 MB internal running 
 Windows XP   all help/info appreciated.  
 best regards,u   A.   ------------------------------   Date: 1 Apr 2003 08:04:09 -06001; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler).? Subject: Re: PicoVax and OpenVMS hobbyist CD beginners question.3 Message-ID: <i+85$S9zdZ7y@eisner.encompasserve.org>   X In article <MPG.18f3afd2f9e7ed16989680@news.xs4all.nl>, atilla <na@pandinus.com> writes:	 > Hi all,o > M > i'm interested in picking up my VMS knowladge again(it's been 10 years) So uI > i downloaded picovax and want to order the hobbyist cd. Is this a good m< > 'home' solution? does anyone have some experience with it? >   ?    Yes.  If you get tired of renewing Pico-VAX (it has built ins&    timeouts), you can do simh instead.  7    Or you can get a really cheap VAX or Alpha off eBay.t   ------------------------------  $ Date: Tue, 1 Apr 2003 16:19:16 +0200 From: atilla <na@pandinus.com>? Subject: Re: PicoVax and OpenVMS hobbyist CD beginners questione7 Message-ID: <MPG.18f3be51492433f6989681@news.xs4all.nl>e  4 In article <i+85$S9zdZ7y@eisner.encompasserve.org>, / koehler@eisner.nospam.encompasserve.org says... Z > In article <MPG.18f3afd2f9e7ed16989680@news.xs4all.nl>, atilla <na@pandinus.com> writes: > > Hi all,* > > O > > i'm interested in picking up my VMS knowladge again(it's been 10 years) So  K > > i downloaded picovax and want to order the hobbyist cd. Is this a good a> > > 'home' solution? does anyone have some experience with it? > >  > A >    Yes.  If you get tired of renewing Pico-VAX (it has built inh( >    timeouts), you can do simh instead. > 9 >    Or you can get a really cheap VAX or Alpha off eBay.  >  >  Ok thnx, btw what is simh? url?p   regards,   A.   ------------------------------  % Date: Tue, 01 Apr 2003 08:25:34 -0600 ( From: Michael Rice <marice@whiteice.com>? Subject: Re: PicoVax and OpenVMS hobbyist CD beginners questiona/ Message-ID: <v8j8b2lncar740@corp.supernews.com>   " On 4/1/2003 8:19 AM, atilla wrote:6 > In article <i+85$S9zdZ7y@eisner.encompasserve.org>, 1 > koehler@eisner.nospam.encompasserve.org says...a > Z >>In article <MPG.18f3afd2f9e7ed16989680@news.xs4all.nl>, atilla <na@pandinus.com> writes: >>
 >>>Hi all, >>>-N >>>i'm interested in picking up my VMS knowladge again(it's been 10 years) So J >>>i downloaded picovax and want to order the hobbyist cd. Is this a good = >>>'home' solution? does anyone have some experience with it?t >>>t >>A >>   Yes.  If you get tired of renewing Pico-VAX (it has built int( >>   timeouts), you can do simh instead. >>9 >>   Or you can get a really cheap VAX or Alpha off eBay.o >> >> > ! > Ok thnx, btw what is simh? url?s > 
 > regards, >  > A.  . Simulators for various systems, including VAX.   http://simh.trailing-edge.comd   ------------------------------  $ Date: Tue, 1 Apr 2003 06:25:54 -0800# From: "Tom Linden" <tom@kednos.com>y? Subject: RE: PicoVax and OpenVMS hobbyist CD beginners questions9 Message-ID: <CIEJLCMNHNNDLLOOGNJIGEKJGOAA.tom@kednos.com>v    Try to google (new verb) on simh   >-----Original Message-----o& >From: atilla [mailto:na@pandinus.com]& >Sent: Tuesday, April 01, 2003 6:19 AM >To: Info-VAX@Mvb.Saic.Com@ >Subject: Re: PicoVax and OpenVMS hobbyist CD beginners question >  >a4 >In article <i+85$S9zdZ7y@eisner.encompasserve.org>,0 >koehler@eisner.nospam.encompasserve.org says...A >> In article <MPG.18f3afd2f9e7ed16989680@news.xs4all.nl>, atillad ><na@pandinus.com> writes: >> > Hi all, >> >B >> > i'm interested in picking up my VMS knowladge again(it's been
 >10 years) So K >> > i downloaded picovax and want to order the hobbyist cd. Is this a goodM? >> > 'home' solution? does anyone have some experience with it?r >> > >>B >>    Yes.  If you get tired of renewing Pico-VAX (it has built in) >>    timeouts), you can do simh instead.i >>: >>    Or you can get a really cheap VAX or Alpha off eBay. >> >>  >Ok thnx, btw what is simh? url? >l	 >regards,) >  >A.  >a >---' >Incoming mail is certified Virus Free.l; >Checked by AVG anti-virus system (http://www.grisoft.com)..A >Version: 6.0.459 / Virus Database: 258 - Release Date: 2/25/2003* >* ---*& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.459 / Virus Database: 258 - Release Date: 2/25/2003   ------------------------------  $ Date: Tue, 1 Apr 2003 17:34:03 +0200 From: atilla <na@pandinus.com>? Subject: Re: PicoVax and OpenVMS hobbyist CD beginners questiona7 Message-ID: <MPG.18f3cfd6b326e4c1989682@news.xs4all.nl>b   Cool, thank you alle   got something to do ...p     l8ri   A.   ------------------------------  % Date: Tue, 01 Apr 2003 09:12:55 +0200e4 From: Didier Morandi <Didier.Morandi.nospam@Free.fr>9 Subject: Re: Poka Yoke on loppy drive flat cable is wrongr$ Message-ID: <3E893BF7.60504@Free.fr>   Steve wrote:d > Didier Morandi <Didier.Morandi.nospam@Free.fr> wrote in message news:<3E87F38C.1020206@Free.fr>... > O >>in an AlphaStation 250, the pokayoke of the floppy drive is upside down. The *N >>piece of plastic is upwards and the hole where it should fit is downwards... >> >>:-(e >> >>D. >>M >>PS to French speaking people : comment on dit "dtrompeur" en anglais ? :-)c >  >  > KeyG  	 merci :-)a   D.   ------------------------------  $ Date: Tue, 1 Apr 2003 12:24:27 -0500; From: "Brian Tillman" <Brian.Tillman@notnoone.notnohow.com> 8 Subject: Re: Remotely opening a window on my workstation$ Message-ID: <3e89cb47$1@news.si.com>  K >I would assume in the next quarterly release.  You can get it [Motif V1.3]s
 as a seperateaK >product right now, and it will be the standard release in the next OpenVMSd >Alpha release.E   How about on a VAX?o -- eI Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot comi5 Smiths Aerospace  Addresses modified to prevent SPAM.f@ 3290 Patterson Ave. SE, MS Replace "at" with "@", "dot" with "." Grand Rapids, MI 49512-1991k8        This opinion doesn't represent that of my company   ------------------------------  $ Date: Tue, 1 Apr 2003 12:37:12 -0500A From: "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com>*8 Subject: Re: Remotely opening a window on my workstation. Message-ID: <3e89ce49$1_2@hpb10302.boi.hp.com>  K There are no plans to make any significant changes to the window components  on VAX.n  F "Brian Tillman" <Brian.Tillman@notnoone.notnohow.com> wrote in message news:3e89cb47$1@news.si.com...G > >I would assume in the next quarterly release.  You can get it [Motifd V1.3]o > as a seperaterE > >product right now, and it will be the standard release in the nexta OpenVMS  > >Alpha release.i >  > How about on a VAX?  > --K > Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot com 7 > Smiths Aerospace  Addresses modified to prevent SPAM.sB > 3290 Patterson Ave. SE, MS Replace "at" with "@", "dot" with "." > Grand Rapids, MI 49512-1991m: >        This opinion doesn't represent that of my company >s   ------------------------------  # Date: Tue, 01 Apr 2003 15:13:45 GMT*# From: "John Smith" <a@nonymous.com>* Subject: Re: Rich MarcelloF Message-ID: <J0iia.2346$xD7.1285@news01.bloor.is.net.cable.rogers.com>  0 "Marty Kuhrt" <marty@barra.com> wrote in message7 news:71c2e778.0303311426.29d35bd7@posting.google.com...7/ > "Hans Vlems" <hvlems@iae.nl> wrote in message22 news:<b5hgck$28q5it$1@ID-143435.news.dfncis.de>...D > > "Phillip Helbig" <HELBPHI@sysdev.deutsche-boerse.com> schreef in berichti9 > > news:01KTSHDD82U49H1MPS@sysdev.deutsche-boerse.com... E > > > > As an ISV interested in growing Customer interest in OpenVMS,s as aF > > > > minimum I would expect that when you get these questions, thatD > > > > regardless of your personal opinions, you would at least say ".. Here isn" > > > > HP's official position .." > > >rC > > > Yes, but what is HP's "official position" with regard to VMS?e We willdE > > > help you port to HPUX?  That might not be wise to state.  Or do- we have-' > > > a choice of "official positions"?: > > >0E > > > What if the customer says "ah yes, DEC's official position usede to besF > > > "NT on ALPHA is the future", and I invested millions in that and gotv > > > burned"? > > >@F > > > (I suddenly started hearing Nazareth's "Love Hurts" in my head.) > >,@ > > Exactly. The company I work for is looking at an alternative platform for aF > > large Sybase application. The current platform options are Windows	 (on whichaD > > it runs) or port it to DB/2 on our mainframe (big $$). We're not	 afraid to C > > look at other platforms and of course AS/400 and unix have beenl
 discussed.F > > At one time the name VMS popped up and all but one had at one time workedD > > with that platform. They were all still favorably impressed with itD > > (stability, reliability, ease of use etc.). The majority thought that VMS/ > > was gone (together with the VAX platform!).aA > > If I had had any positive signal from HP over the past monthsi (and, earlier,A > > from Compaq for that matter) then I'd have had no problems in 
 pursuing that0C > > road. My fear is exactly the one Philip mentions in his post: IG	 propose aeA > > completely new platform and within a year I get burned by HP.a > >O > > Hans Vlems > >v > > PSB > > This same company was once a DEC customer: they went for NT on	 Alpha and 7 > > Wolfpack in '97 or '98. That memory is still there.  >8 >B@ > If the current version of Sybase products were running on VMS,@ > I'd probably be able to eventually convince the powers that beC > here to put it (back) on VMS.  When Sybase dropped VMS support weu? > pretty much dropped VMS for everything _but_ our stock marketw> > stuff.  Before then most of our in house production was done< > using Sybase, SAS and a gob of home grown DCL and FORTRAN. >p> > As a side story, I ran into a Sybase guy at a local pub.  HeC > actually made contact with me because I was wearing a DECUS polo.nA > He asked about DECUS and VMS and said he missed working on VMS.g >t= > I get that a lot at some of the parties I go to.  I'll wear A > something that mentions VMS on it, and the conversation usuallyl > goes like this...  >a< > OtherGuy: Hiya.  Noticed your VMS shirt.  Cool.  I used to- >           work with that.  It still around?r= > Me: Yep.  We use it in production everyday for stock marketf4 >     stuff at work where downtime is not an option.> > OG: Really?  I thought it was dead.  They stopped developing >     it, right?> > Me: Nope.  As a matter of fact they're finishing the port to  >     the Intel 64 bit platform.> > OG: Geez, the place I used to work for was convinced that it< >     was a goner and ported to (insert lame solution here).! > Me: So how is that working out? ? > OG: After (insert large multiyear time reference) were almosto( >     to where we were when we left VMS.@ > Me: So you can do multi-site, multi-city, single coherent data3 >     and system image with (insert lame solution)?r@ > OG: Kind of.  After spending (insert dollar amount that in the@ >     long run is probably about four to six times the VMS cost)0 >     we have replication and a hotsite working.- > Me: How long does it take you to fail over?> > OG: We don't know, really. > Me: Sounds nerve racking.iB > OG: Now that you mention it, it is.  You guys wouldn't be hiring >     would you? > Me: 'fraid not.t  F Write a letter to Marcello/Gorham/Blackmore/Stallard (cc them all) andD ask them what HP is doing with Sybase to ensure that VMS is returnedC to/kept as a GA release with all Sybase products (not just Adaptive 2 Server). Let us know what the HP brain trust says.   ------------------------------  % Date: Tue, 01 Apr 2003 10:07:11 +0100 0 From: Chris Sharman <chris.sharman@sorry.nospam>8 Subject: Re: RMS Tune Check (was High RMS file activity)4 Message-ID: <b6bks0$94o$1$830fa7b3@news.demon.co.uk>   Hein van den Heuvel wrote: >  > Chris Sharman wrote:G >>Nice alert to massively duplicated key values - will be using that toa" >>fix a few (with null keys, etc). > % > Show us an example or two? How bad?a> > (Btw... it has to be pretty darn bad to quilify for a prize.C > We have 4104 buckets of dups for one value as the winner so far).c   Nope, I can't compete. My worst is:   Key 20  ) Found 36789 sidr records in 36726 buckets . There were 10520354 user data record pointers.; There were 0 deleted, 0 RU_delete and 0 RU_update pointers.   4    Top Ten Table of Sidrs with more than 1 duplicate  &    Duplicate count, Buckets, Key value8 --------------------------------------------------------(      456290         1579              38(      331298         1147            1074'      227924         789              29c'      210009         727              30a'      189442         655            1107 '      178323         617            1115p'      172815         598              39h'      162345         562            2301a'      139736         484            1065-'      119596         414              31:F It's a poor choice of key, and I don't believe RMS supports segmented H integer keys ? (Although I've seen structure declarations which suggest H it could ?) I'll be trying to find out what (if anything) uses the key,  with a view to removing it.2   >>Also, we do: >>$ anal/rms/fdl, >>$ edi/fdl/nointer/gran=double/emph=smaller >>$ conv/fdl >>on a weekly basis.  = > Hmm, I've never heard of anyone using that edit/fdl option.:B > Most folks have their buckets too small, but small may be betterW > for some high contention sitiations. As long as you index trees do not get too deep..,  >* > Imho a full anal/RMS is a waste of time.@ > Just feed last weeks convert record count into this weeks fdl.< > You may want to perform anal/rms on the old files / backup= > just to make sure you do not have a silent data corruption,@O > but that would be a background job not in the prime backup/reorganize window.e  I This all goes on unattended on a Sunday, so time isn't precious (for the aG peak season when we're working Sunday, it doesn't go on at all - can't  ? unlock the files to rebuild). The actual code is a little more gG complicated - bucket size changes are referred to me, and approved for -? implementation the following Sunday. What qualifiers would you nG recommend, & I'll give them a try. [We use those above up to 5 keys, & iI /gran=4 thereafter (we saw some problem some very long time ago with 6/7 r! keys, gran=double, and convert).]t   ------------------------------  % Date: Tue, 01 Apr 2003 13:04:29 -0500a9 From: Hein van den Heuvel <hein_netscape@eps.zko.dec.com> 8 Subject: Re: RMS Tune Check (was High RMS file activity)/ Message-ID: <3E89D4AD.9D4A4D61@eps.zko.dec.com>.   Chris Sharman wrote:  I > >>Nice alert to massively duplicated key values - will be using that tok   > My worst is:	 >   Key 2a >t( >    Duplicate count, Buckets, Key value: > --------------------------------------------------------* >      456290         1579              38* >      331298         1147            1074) >      227924         789              29e   :o  G > It's a poor choice of key, and I don't believe RMS supports segmented.I > integer keys ? (Although I've seen structure declarations which suggest:
 > it could ?)   8 You de-duplicate using segments in a clumsy/tricky  way.I You can define 4 single byte segements over the 32-bit integer 'the wrong@H way around' to honor the little endian ordering of the original integer.I Then you can then add a 5th segment deduplicate. For the file on hand you C need a factor of 1000x, so a single random binary byte will not do.gI You'll need an unrelated 2 byte binary field or 3 byte ascii number/text.oC (Maybe a 2 byte ascii text, if it is mixed case and nicely spread).S  I Now as long as the application just points to the integer KBF=xxx, KSZ=32 L and does not verify exact key types then you are all set. Cobol will verify.N Basic will verify IF you tell it about the key on the open. If you just removeM the keys from the 'open for input access modiy' then it wil tolerate any key.t  = > I'll be trying to find out what (if anything) uses the key,e > with a view to removing it.f  H That's the better approach. You really have to ask the applicaiton folksT the question 'what is the business function of finding the first of the half-millionN records with alternate key 2 value = 38. The quite possibly is no value today.F They may have been a use many moons ago when the business was smaller.  T If a certain report uses that key, then either teach it select and/or sort, or point7 it to a report-only convert  of the file with that key.:  J > This all goes on unattended on a Sunday, so time isn't precious (for theH > peak season when we're working Sunday, it doesn't go on at all - can't@ > unlock the files to rebuild). The actual code is a little moreH > complicated - bucket size changes are referred to me, and approved for& > implementation the following Sunday.   Excellent approach!.   > What qualifiers would you H > recommend, & I'll give them a try. [We use those above up to 5 keys, &J > /gran=4 thereafter (we saw some problem some very long time ago with 6/7# > keys, gran=double, and convert).]   G Hmm, I don't think areas will do much good and rarely worry about them.-E (Yes I know when/where to use them, notably when using bound volumes, / and when multipel bucket sizes are appropriate)R .L@ The important qualifiers for me are: /NOSORT/FAST/STAT/EXCEP=x.x   ------------------------------  + Date: Tue, 01 Apr 2003 11:50:56 +0100 (MET)s9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>i Subject: Re: Second IP address; Message-ID: <01KU7MREYZMAA9LL1B@sysdev.deutsche-boerse.com>i  K > Is it possible to add a second IP address on a single ethernet interface?r   Yes.  # > I run the following combinations:s > / > AXP/VMS 7.3 with  TCPIP services for VMS V5.10. > VAX/VMS 7.2 with TCPIP services for VMS V5.1  G > I want to add connectivity to network 10.0.0.0 mask 255.255.255.0 forrF > the new ADSL router. The other options are either adding an ethernetA > interface on the AXP or modify all IP addresses on my systems. w   Been there, done that.  H Note that this might not do what you want.  If you have VMS talk to the H router via the 10 network, the other network (the one you already have) D will probably become useless.  The reason is that you have to set a G default gateway for TCPIP.  Unless there is an explicit route set, ALL rD outgoing traffic will go through the default gateway, even if it is 3 return traffic which came in via another interface.s  D There are basically three workarounds.  The first is to add explicit@ routes for connections which shouldn't use the default gateway. D However, this will prevent them from working via the normal gateway.  F In my hobbyist setup, I am moving from ISDN with several static publicB addresses to DSL with one dynamic public IP address and private IPF addresses on the LAN.  This in itself is pretty straightforward.  The D difficult bit is to have both interfaces working for a transitional  period.e  H In one flat (A), I have the old ISDN connection and a DSL connection.  IH still have the ISDN router as the default gateway.  In my other flat (B)F I have just DSL.  Most of the traffic is my own between the two flats.G Thus, I set a route in flat A for the IP address in flat B.  That way, eC I can connect via the DSL router to flat A.  This is fine for this f! purpose but is rather inflexible.   G This weekend, I will implement at least the first of the following two w workarounds.  G Build a cluster.  Have one node use one network for its default gateway0E and the other the other network.  That way, from the point of view of:E TCPIP, the two have nothing to do with each other.  However, from the.H point of view of VMS, they are practically the same system (they should 0 share as many resources as possible, of course).  ? If you have a NAT/PAT DSL router, you can forward the incoming tG connections to the TCPIP cluster alias.  Thus, you don't have to worry aE about a machine going down, if you have more than one using the same n cluster alias.  F SIDEBAR: Can I have a four-node cluster with two nodes on one network G using one TCPIP cluster alias and two nodes on another network using a P different TCPIP cluster alias?  F The other solution involves introducing an additional router and doingA reverse NAT.  This is a BRILLIANT suggestion from JF.  Search the-G archives for posts discussing this.  Basically, you set up a second NATiA router behind the first one, but backwards.  You forward incomingeH connections to the first DSL NAT router to the LAN address of the secondG NAT router.  The WAN address of this is on the same network (10.0.0.0)  H as your VMS machine.  Then, set an explicit route to the WAN address of I the second router.  With the disadvantage that incoming DSL traffic will nD all appear to come from the same address, you then have a system on F which you can have connections through both interfaces and everything 
 will work.   ------------------------------  $ Date: Tue, 1 Apr 2003 11:07:11 +0100- From: "Steve Spires" <Steve.Spires@torex.com>u Subject: SYSDUMP.DMP corruption.E Message-ID: <91947A84607D9D48B8E674A5FAB54DA63CAE78@tahiti.tinuk.com>s  B We had a crash on one of our systems which produced the following;   CENTRAL $ anal/crash _Dump File: sysdump.dmpn      ' OpenVMS (TM) Alpha system dump analyzeri1 ..analyzing a compressed selective memory dump...s  # %SDA-W-LINKTIMEMISM, link time of =e. SYS$COMMON:[SYS$LDR]SYS$BASE_IMAGE.EXE;2 (28-MA AR-2002 14:19) does not match link time of image in system dump =  (23-JAN-2001 08: 35)tF %SDA-W-SDALINKMISM, link time of SYS$BASE_IMAGE built into SDA$SHARE = (28-MAR-200cH 2 14:19) does not match link time of image in system dump (23-JAN-2001 = 08:35). %SDA-W-VERSMISM, version mismatch with image =# SYS$COMMON:[SYS$LDR]SYS$BASE_IMAGE.T EXE;2l% Dump taken on 15-MAY-**** 05:40:33.27i ** Invalid bugcheck code **e   SDA> clue crash                  Crashdump Summary Information:  -----------------------------* Crash Time:        15-MAY-**** 05:40:33.27; Bugcheck Type:     INVEXCEPTN, Exception while above ASTDELmD Node:              00  `%=A5=D7)=F1=A0=E9V7.2-1  =FDComp(Standalone)* CPU Type:          Compaq AlphaServer ES40 VMS Version:       -1  ...." Current Process:   == MS_DSP_48....................................................  ..................: Current Image:     DSA100:[SYS0.SYSCOMMON.][SYSEXE]DSM.EXE@ Failing PC:        FFFFFFFF.800EA728    PROCESS_MANAGEMENT+1C728$ Failing PS:        20000000.00000800G Module:            PROCESS_MANAGEMENT    (Link Date/Time: 23-JAN-2001 =d 08:37:04.46)       Anyone seen this before?  J Alphaserver ES40 running VMS 7.3 with the following patches [I know some =G of them are replaced now - I have yet to fit in updating the patches, = - but will do so if required for this problem];G   VMS73_BACKUP-V0100 VMS73_CLUSTER-V0200   =20  VMS73_DCL-V0200=20 VMS73_DDTM-V0100       =20 VMS73_DRIVER-V0200 VMS73_F11X-V0100       =20 VMS73_FIBRE_SCSI-V0300 VMS73_INIT-V0100     =20 VMS73_LIBRTL-V0200 VMS73_LMF-V0100      =20 VMS73_RMS-V0300n VMS73_SHADOWING-V0200  =20 VMS73_SYS-V0400f VMS73_SYSINI-V0100     =20 VMS73_SYSLOA-V0200 VMS73_UPDATE-V0100   Cheers   Steve Spires Technical Consultant Torex Health [T] +44 (0) 1295 274200  [F] +44 (0) 1295 275131l
 www.torex.comd   ------------------------------  % Date: Tue, 01 Apr 2003 08:11:56 -0500m2 From: rdeininger@mindspring.com (Robert Deininger)# Subject: Re: SYSDUMP.DMP corruptiontL Message-ID: <rdeininger-0104030811560001@user-105n8n8.dialup.mindspring.com>  E In article <91947A84607D9D48B8E674A5FAB54DA63CAE78@tahiti.tinuk.com>,m. "Steve Spires" <Steve.Spires@torex.com> wrote:  C >We had a crash on one of our systems which produced the following;y >n >CENTRAL $ anal/crasha >_Dump File: sysdump.dmp >- >- >-( >OpenVMS (TM) Alpha system dump analyzer2 >..analyzing a compressed selective memory dump... > $ >%SDA-W-LINKTIMEMISM, link time of =/ >SYS$COMMON:[SYS$LDR]SYS$BASE_IMAGE.EXE;2 (28-MoB >AR-2002 14:19) does not match link time of image in system dump = >(23-JAN-2001 08:a >35)G >%SDA-W-SDALINKMISM, link time of SYS$BASE_IMAGE built into SDA$SHARE =. >(28-MAR-200I >2 14:19) does not match link time of image in system dump (23-JAN-2001 =  >08:35)s/ >%SDA-W-VERSMISM, version mismatch with image =a$ >SYS$COMMON:[SYS$LDR]SYS$BASE_IMAGE. >EXE;2  J The warnings above are side-effects of applying ECOs (and not rebooting?),> I think.  These are annoying, but not fatal to crash analysis.    & >Dump taken on 15-MAY-**** 05:40:33.27 >** Invalid bugcheck code **  O This could be an invalid dump file, or a may be due to a mis-matched SDA image.o     >SDA> clue crash >i >a >e >o >  >o >a >t >Crashdump Summary Information:n >------------------------------ + >Crash Time:        15-MAY-**** 05:40:33.27s< >Bugcheck Type:     INVEXCEPTN, Exception while above ASTDELE >Node:              00  `%=A5=D7)=F1=A0=E9V7.2-1  =FDComp(Standalone)e+ >CPU Type:          Compaq AlphaServer ES40  >VMS Version:       -1  .... >Current Process:   =e> >MS_DSP_48.................................................... >..................M    F More indications that SDA doesn't understand this dump file very well.    ; >Current Image:     DSA100:[SYS0.SYSCOMMON.][SYSEXE]DSM.EXEiA >Failing PC:        FFFFFFFF.800EA728    PROCESS_MANAGEMENT+1C728e% >Failing PS:        20000000.00000800eH >Module:            PROCESS_MANAGEMENT    (Link Date/Time: 23-JAN-2001 =
 >08:37:04.46)m >e >g >  >Anyone seen this before?-  B Don't know offhand.  If you can log a service call, HP can check aF database to see if this crash signature has been seen before.  If theyF find a match, they can often supply a fix right away.  If the crash isI new, of course it takes a while to find the cause and make a fix.  If thewI problem isn't reproducible, it can take a very long time.  Self-diagnosistE of these crashes is not recommended.  If you have a support contract,t don't be shy, log the call.t  M But since you're asking in C.O.V, you probably don't have a support contract.C  D INVEXCEPTN crashes are often caused by corruption of privileged dataG structures.  The crash PC inside PROCESS_MANAGEMENT, along with map andcJ listing files for that image, would likely indicate what data structure isI busted.  But finding the cause of the corruption is the fun part.  If youn< don't have the listings kit, you don't really have a chance.  B SHOW CRASH usually has useful summary information for this type ofF bugcheck.  The invalid exception is often an ACCVIO, and SHOW CRASH is helpful there.  F In these crashes, the "current process" and "current image" are likely@ just victims, and might not be directly responsible for the bug.  D Some images have "_MON" variants that have additional diagnostics orC tracing built in.  PROCESS_MANAGEMENT has a _MON variant on my V7.3 H system.  The SYSTEM_CHECK parameter can be used to activate all the _MONG images (and more goodies besides), but often slows performance too muchaI for production systems, and may change timing so much the bug goes away. eI Another alternative is to rename selected SYS$COMMON:[SYS$LDR]..._MON.EXEoJ images to SYS$SPECIFIC:[SYS$LDR]....EXE (without the _MON) so they will beI active after the next reboot.  All of this is usually done at the requestm3 of, and with assistance from, the HP support folks.o  H And you'd want to make sure your dump file is big enough, working, etc. G I'd verify this, reboot, and force a crash to verify that the dump filei
 is  readable.     J With all of your diagnostics in place, you sit back and wait for the crashE to recur.  When it happens, you hope there's enough clues to find thet2 cause.  Lather, Rinse, Repeat, as the saying goes.  E If you go through all this on your own, you'll probably decide the HPc/ support contract is not so expensive after all.r  K >Alphaserver ES40 running VMS 7.3 with the following patches [I know some =fH >of them are replaced now - I have yet to fit in updating the patches, =. >but will do so if required for this problem]; >" >VMS73_BACKUP-V0100S >VMS73_CLUSTER-V0200   =20 >VMS73_DCL-V0200=20  >VMS73_DDTM-V0100       =20t >VMS73_DRIVER-V0200e >VMS73_F11X-V0100       =20t >VMS73_FIBRE_SCSI-V0300t >VMS73_INIT-V0100     =20> >VMS73_LIBRTL-V0200w >VMS73_LMF-V0100      =20a >VMS73_RMS-V0300 >VMS73_SHADOWING-V0200  =20a >VMS73_SYS-V0400 >VMS73_SYSINI-V0100     =20h >VMS73_SYSLOA-V0200  >VMS73_UPDATE-V0100-  I Offhand, I'd guess at least half of the "rating 1" patches might fix bugsUJ that could cause an INVEXCEPTN bugcheck.  Best advice for a do-it-yourselfF shop would be to install each rating 1 patch, unless the release notes  show a reason not to install it.  D (Actually, best advice would be to upgrade to V7.3-1, plus patches.)     -- Roberts   ------------------------------   Date: 1 Apr 2003 08:02:42 -0600/; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) # Subject: Re: SYSDUMP.DMP corruptionc3 Message-ID: <wfZ2xenSOe1E@eisner.encompasserve.org>d  u In article <91947A84607D9D48B8E674A5FAB54DA63CAE78@tahiti.tinuk.com>, "Steve Spires" <Steve.Spires@torex.com> writes:h  F > Node:              00  `%=A5=D7)=F1=A0=E9V7.2-1  =FDComp(Standalone)  H    The last time I had really screwy data like this in a dump it was dueF    to a broken backplane on an 11/780.  Microdiagnostics wouldn't find.    it, but UETP agreed something wasn't right.   ------------------------------  % Date: Tue, 01 Apr 2003 11:52:33 -0500o% From: "Mark D. Jilson" <jilly@hp.com>e# Subject: Re: SYSDUMP.DMP corruptiona& Message-ID: <3E89C3D1.5A6A4A9D@hp.com>  F If this system dumps to the system disk and it appears that the systemF disk is a shadow set, review the console log to determine which memberG of the system disk shadow set was the dump target and then dismount alliG other members out of the shadow set and try the dump again.  If you caniE analyze it successfully now then copy it to another disk and then addh! those dismounted members back in.u  G This situation can happen since the dump drivers are working outside of C shadowing.  I'm not sure of the status of the work in this area buta6 contact the CSC for some tools to help this situation.   Steve Spires wrote:S > D > We had a crash on one of our systems which produced the following; >  > CENTRAL $ anal/crash > _Dump File: sysdump.dmp  > ) > OpenVMS (TM) Alpha system dump analyzer 3 > ..analyzing a compressed selective memory dump...i > R > %SDA-W-LINKTIMEMISM, link time of SYS$COMMON:[SYS$LDR]SYS$BASE_IMAGE.EXE;2 (28-MR > AR-2002 14:19) does not match link time of image in system dump (23-JAN-2001 08: > 35)sR > %SDA-W-SDALINKMISM, link time of SYS$BASE_IMAGE built into SDA$SHARE (28-MAR-200O > 2 14:19) does not match link time of image in system dump (23-JAN-2001 08:35) R > %SDA-W-VERSMISM, version mismatch with image SYS$COMMON:[SYS$LDR]SYS$BASE_IMAGE. > EXE;2e' > Dump taken on 15-MAY-**** 05:40:33.27  > ** Invalid bugcheck code **P >  > SDA> clue crash. >   > Crashdump Summary Information:  > ------------------------------, > Crash Time:        15-MAY-**** 05:40:33.27= > Bugcheck Type:     INVEXCEPTN, Exception while above ASTDEL : > Node:              00  `%) V7.2-1  Comp(Standalone), > CPU Type:          Compaq AlphaServer ES40 > VMS Version:       -1  ....bR > Current Process:   MS_DSP_48.................................................... > ..................< > Current Image:     DSA100:[SYS0.SYSCOMMON.][SYSEXE]DSM.EXEB > Failing PC:        FFFFFFFF.800EA728    PROCESS_MANAGEMENT+1C728& > Failing PS:        20000000.00000800T > Module:            PROCESS_MANAGEMENT    (Link Date/Time: 23-JAN-2001 08:37:04.46) >  > Anyone seen this before? >  > Alphaserver ES40 running VMS 7.3 with the following patches [I know some of them are replaced now - I have yet to fit in updating the patches, but will do so if required for this problem]; >  > VMS73_BACKUP-V0100 > VMS73_CLUSTER-V0200d > VMS73_DCL-V0200  > VMS73_DDTM-V0100 > VMS73_DRIVER-V0200 > VMS73_F11X-V0100 > VMS73_FIBRE_SCSI-V0300 > VMS73_INIT-V0100 > VMS73_LIBRTL-V0200 > VMS73_LMF-V0100d > VMS73_RMS-V0300D > VMS73_SHADOWING-V0200s > VMS73_SYS-V0400d > VMS73_SYSINI-V0100 > VMS73_SYSLOA-V0200 > VMS73_UPDATE-V0100 >  > Cheers >  > Steve Spires > Technical Consultant > Torex Health > [T] +44 (0) 1295 274200e > [F] +44 (0) 1295 275131d > www.torex.como   --  C Jilly	- Working from Home in the Chemung River Valley - Waverly, NYr0 	- jilly@clarityconnect.com			- Brett Bodine fan+ 	- Mark.Jilson@hp.com				- since 1975 or sor 	- http://www.jilly.baka.com   ------------------------------  % Date: Tue, 01 Apr 2003 09:45:28 -0600s From: brandon@dalsemi.comKM Subject: Re: SYSTARTUP_VMS (was: Three questions re: the VMS startup process)s1 Message-ID: <03040109452828@dscis6-0.dalsemi.com>   I > We must keep the environment in the state that the vendor will support.o  O No disagreement here.  However if a vendor requires priv's for a user to accesstO data I would create captive accounts for those users.  I would also address the K issue with the vendor and work with them on making the changes necessary to? resolve the matter.o  ? > Given the dictates of support contracts, there is very littleo* > alternative without the vendor's buy-in.  * I believe there is.  See captive accounts.  K > I am not aware of any LPs that do, either. Only third-party applications.a  M LP typo - I meant to include TP-LP.  We have a number of LP and TP-LP that dooO not require users to have privileges.  The images are installed with privilegesSG and where there is critical data involved, password protection enabled.r    J As with any startup, measures should be put in place that allow for sanityL checks.  Be it policy or procedure.  If a startup failure is detected, usersK can be disabled from access using combinations of SET INT/LOG=0 or checks &wO disables in SYLOGIN.COM.  There can also be startup checks in applicaitons - ashG suggested by checking the BG drivers.  This can be elaborate or simple.n     John Brandon VMS Systems Administratort Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wke 972.371.4003 fxf   ------------------------------  # Date: Tue, 01 Apr 2003 19:22:53 GMT-& From: jlsue <jefflsxxxz@sbcglobal.net>M Subject: Re: SYSTARTUP_VMS (was: Three questions re: the VMS startup process)c8 Message-ID: <hamj8v8jmhlcostp30rlq6jv96rcp7tmql@4ax.com>  > On Mon, 31 Mar 2003 16:47:22 -0600, brandon@dalsemi.com wrote:  I >> I would certainly not wait for the startup of DECevent and DESTA, but  K >> just SUBMIT them. Both products have a very high Unix flavor and I have 8B >> seen at least once that the startup of one of them just hanged. > L >All the jobs are started using SUBMIT.  Some of the jobs are submitted with	 >/prio=10   A Just out of curiosity, why "/prior=10"?  This does not affect the.* process priority, only the queue priority.   ------------------------------   Date: 1 Apr 03 20:37:53 +0200e) From: p_sture@elias.decus.ch (Paul Sture)i/ Subject: Re: TCPIP Services and IP masquerading ) Message-ID: <Ht4jVZDuglbf@elias.decus.ch>l  b In article <3E872341.5E69BEF3@vl.videotron.ca>, JF Mezei <jfmezei.spamnot@vl.videotron.ca> writes: > Paul Sture wrote: @ >> Message read in bedroom, answer posted on balcony, VMS system& >> in my den doing some downloads. :-) > * > Bragging about how big your antenna is ? > 
 > :-) :-) :-):  N Nope. How long my ethernet cable is. The problem with trailing that around theL appartment is that doors will slam shut when windy, potentially severing the cable.   -- -
 Paul Sture   ------------------------------  # Date: Tue, 01 Apr 2003 18:36:54 GMT ( From: "Joel Berman" <berman@zk3.dec.com> Subject: The 15th greatest Hoaxe0 Message-ID: <a%kia.130$QD1.107@news.cpqcorp.net>  / From  http://www.museumofhoaxes.com/top100.htmlm  
  #16: KremvaxgL In 1984, back in the Stone Age of the internet, a message was distributed toI the members of Usenet (the online messaging community that was one of theiK first forms the internet took) announcing that the Soviet Union was joiningoH Usenet. This was quite a shock to many, since most assumed that cold warL security concerns would have prevented such a link-up. The message purportedK to come from Konstantin Chernenko (from the address chernenko@kremvax.UUCP)iJ who explained that the Soviet Union wanted to join the network in order toF "have a means of having an open discussion forum with the American andK European people." The message created a flood of responses. Two weeks latersI its true author, a European man named Piet Beerma, revealed that it was a F hoax. This is believed to be the first hoax on the internet. Six yearsL later, when Moscow really did link up to the internet, it adopted the domain$ name 'kremvax' in honor of the hoax.   ------------------------------  $ Date: Tue, 1 Apr 2003 11:00:49 +0100* From: "Richard Brodie" <R.Brodie@rl.ac.uk>" Subject: Re: VMS Upgrade Questions, Message-ID: <b6bo0i$11ha@newton.cc.rl.ac.uk>  D "Bradford J. Hamilton" <brad@.homeportal.2wire.net> wrote in message' news:WZ2ia.281788$F1.48877@sccrnsc04...   Q > Unless I am badly mistaken, you can go from 7.2 directly to 7.3-1.  The UpgraderD > and Installation Manual is your friend here; check it for details.  N I thought you could as well; then I checked and came up with the wrong answer!8 You can upgrade directly. Sorry for confusing the issue.   ------------------------------  # Date: Tue, 01 Apr 2003 14:50:24 GMT2' From: Colin Blake <colin@theblakes.com>p: Subject: Re: [MOZILLA V1.3] Compatibility with JAVA V1.4 ?* Message-ID: <3E89A72F.20301@theblakes.com>   Rich Jordan wrote:  G >Also now have TCPIP V5.3 media (and patches) available so I could movec0 >to that if there's a possibility of it helping. >-/ I think that would be worthwhile. Just in case.7  G >I'm not fluent in bugzilla.  Should I be entering any of this info, orc >other data there? > E You need a bugzilla account before you can post or add yourself as a wG "cc" to an existing bug. But getting an account is real simple, all it  ! needs is your email address. See  - http://bugzilla.mozilla.org/createaccount.cgi   F I'll post the info over in bug 199932. If you can get yourself set up - with bugzilla we'll continue this over there.c   Colin.   ------------------------------  % Date: Tue, 01 Apr 2003 19:55:06 +0200e4 From: Didier Morandi <Didier.Morandi.nospam@Free.fr>: Subject: Re: [MOZILLA V1.3] Compatibility with JAVA V1.4 ?& Message-ID: <3E89D27A.9080802@Free.fr>   Colin Blake wrote:  H > I know this sounds stupid, but make sure you ran the Java 1.4-0 setup G > procedure before trying to run Mozilla. I heard of people having the rC > 1.3-1 setup in their login.com and forgetting to change it after   > upgrading to Java 1.4-0. >  > Colin.  - You are on all my terminal screens, Colin :-)m   D.   ------------------------------   End of INFO-VAX 2003.180 ************************