1 INFO-VAX	Mon, 25 Aug 2003	Volume 2003 : Issue 469       Contents:! allocation classes and satellites % Re: allocation classes and satellites % Re: allocation classes and satellites % Re: allocation classes and satellites % Re: allocation classes and satellites % Re: allocation classes and satellites > Re: Difficulties with DPW 500au and StorageWorks components...> Re: Difficulties with DPW 500au and StorageWorks components...> Re: Difficulties with DPW 500au and StorageWorks components...> Re: Difficulties with DPW 500au and StorageWorks components... Re: Fork on OPEN VMS ftp access to itrc Re: Global Buffers and XFCJ Re: I need 4 DEC/BASIC OpenVMS programmers for perm positions immediately!" Re: Looking for TK50s in the UK...( Re: M7940/DHV-11, TU58, and a VAX 11/750. Re: Mount DQB0 gets 'allready allocated' error  Re: New kits of MySQL and Python% OT: 64 bit desktop computing is here. 
 Re: Praxis Print date/time and username+ Re: Reconstructing pdf file broken via mail + Re: Reconstructing pdf file broken via mail  Re: Strange programming problem  SYSn Re: SYSn Re: SYSn Re: SYSn/ Re: volume shadowing with 2 x IDE disks on Ds10   Re: XML Parser for VAX (OpenVMS) XML Parser for VAX (OpenVMS)  F ----------------------------------------------------------------------  + Date: Sun, 24 Aug 2003 19:48:17 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)* Subject: allocation classes and satellites$ Message-ID: <bib4q1$t29$1@online.de>  I In my hobbyist cluster, I use allocation classes, but only because I use  I volume shadowing.  I don't have any dual-ported disks.  No node has more   than one SCSI bus.  C To ensure unique names, each node has a different allocation class.    The documentation says:   A     1.  When serving satellites, the same nonzero node allocation =         class value must be assigned to the serving computers          and controllers.  I What does this mean?  Does it mean that if a satellite boots from a node  A with allocation class X, that the satellite has to have the same  < allocation class X?  If so, why?  If not, what does it mean?   ------------------------------  + Date: Sun, 24 Aug 2003 22:25:08 +0000 (UTC) 7 From: moroney@world.std.spaamtrap.com (Michael Moroney) . Subject: Re: allocation classes and satellites( Message-ID: <bibe04$q85$1@pcls4.std.com>  J Sounds like it may be related to an older restriction where a MSCP servingF node must have the same allocation class as the disks it serves.  WhatG version are you running?  I fixed this a while ago (in V7.1+patches and  in all later versions).    --   -Mike    ------------------------------    Date: 24 Aug 2003 18:42:06 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen) . Subject: Re: allocation classes and satellites3 Message-ID: <U7qyUqDbH$ZE@eisner.encompasserve.org>   w In article <bib4q1$t29$1@online.de>, helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) writes: K > In my hobbyist cluster, I use allocation classes, but only because I use  K > volume shadowing.  I don't have any dual-ported disks.  No node has more   > than one SCSI bus. > E > To ensure unique names, each node has a different allocation class.  >  > The documentation says:  > C >     1.  When serving satellites, the same nonzero node allocation ? >         class value must be assigned to the serving computers  >         and controllers. > K > What does this mean?  Does it mean that if a satellite boots from a node  C > with allocation class X, that the satellite has to have the same  > > allocation class X?  If so, why?  If not, what does it mean?  D I presume it means that if multiple boot nodes are able to serve theE system disk to a particular satellite, those boot nodes must have the $ same allocation class as each other.   ------------------------------  % Date: Sun, 24 Aug 2003 22:19:24 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> . Subject: Re: allocation classes and satellites' Message-ID: <3F49803C.8F3E5556@fsi.net>   / Phillip Helbig---remove CLOTHES to reply wrote:  > J > In my hobbyist cluster, I use allocation classes, but only because I useJ > volume shadowing.  I don't have any dual-ported disks.  No node has more > than one SCSI bus. > E > To ensure unique names, each node has a different allocation class.  >  > The documentation says:  > C >     1.  When serving satellites, the same nonzero node allocation ? >         class value must be assigned to the serving computers  >         and controllers. > J > What does this mean?  Does it mean that if a satellite boots from a nodeB > with allocation class X, that the satellite has to have the same> > allocation class X?  If so, why?  If not, what does it mean?  F I take it as meaning, for example, if you have NODEA with ALLOCLASS 1,D and it has CI paths to HSJ01/2 and HSJ03/4, then HSJ01/2 and HSJ03/4@ must use an allocation class of 1 for the disk units they serve.  H My take on it. I can't comprhend why this would be necessary, but that's how I read it...   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  + Date: Mon, 25 Aug 2003 05:23:37 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply). Subject: Re: allocation classes and satellites$ Message-ID: <bic6go$jte$1@online.de>  H In article <bibe04$q85$1@pcls4.std.com>, moroney@world.std.spaamtrap.com (Michael Moroney) writes:   L > Sounds like it may be related to an older restriction where a MSCP servingH > node must have the same allocation class as the disks it serves.  WhatI > version are you running?  I fixed this a while ago (in V7.1+patches and  > in all later versions).    It's still in the 7.2-1 HELP.    ------------------------------  + Date: Mon, 25 Aug 2003 05:24:07 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply). Subject: Re: allocation classes and satellites$ Message-ID: <bic6hm$jte$2@online.de>  3 In article <U7qyUqDbH$ZE@eisner.encompasserve.org>, 0 Kilgallen@SpamCop.net (Larry Kilgallen) writes:   E > >     1.  When serving satellites, the same nonzero node allocation A > >         class value must be assigned to the serving computers  > >         and controllers. > > M > > What does this mean?  Does it mean that if a satellite boots from a node  E > > with allocation class X, that the satellite has to have the same  @ > > allocation class X?  If so, why?  If not, what does it mean? > F > I presume it means that if multiple boot nodes are able to serve theG > system disk to a particular satellite, those boot nodes must have the & > same allocation class as each other.  $ In that case, no problem in my case.   ------------------------------  % Date: Sun, 24 Aug 2003 23:18:49 +0200  From: Dirk Munk <munk@home.nl>G Subject: Re: Difficulties with DPW 500au and StorageWorks components... 2 Message-ID: <biba5s$b5e$1@news2.tilbu1.nb.home.nl>   ohm62 wrote:K >>The differential adaptor is the KZPBA-CY. I have one (not in use), and it  >  > has a  > K >>sticker with that part number and anothor one with the Compaq replacement  >  > number >  >>(401922-001). L >>Furthermore if you look just above the external connector, there is a SCSIH >>symbol stamped in the metal. Just above the SCSI symbol the words SCSI > 
 > DIFF are > ' >>also stamped in the metal. Easy .....  >  >  > I > Geez!  I just glanced behind the PWS, and the differential SCSI logo is N > obvious enough indeed!  I can't believe I overlooked this; I guess I've beenK > a bit too enthusiastic unpacking it and plugging it in, as the seller was L > advertising it Single Ended, and it looked pretty much the same as the -CXL > ones I saw pictures of on eBay.  It doesn't pay to be overly confident ;-) > J > I guess that explains the incompatibility with the personality module...7 > The -CY is unable to function in S.E. mode, isn't it?   I No. it can't function in S.E. mode, you need another personality module.  G Personality modules are in fact SCSI bridges with their own electrical  : termination. I just hope your SCSI adaptor is still ok....   ------------------------------  # Date: Mon, 25 Aug 2003 01:47:41 GMT " From: "ohm62" <ohm62@coldmail.com>G Subject: Re: Difficulties with DPW 500au and StorageWorks components... 8 Message-ID: <1Vd2b.15997$kF1.15290@nwrdny02.gnilink.net>  I Well, as a matter of fact, the SCSI DIFF inscription and the differential H SCSI logo <<>- are stamped on a white sticker next to the connector, not really in the metal....   B The main Qlogic chip on board says ISP1040B and the following linkI http://www.qlogic.com/products/ultra_series/ultra_one_tech.asp  says that L processor is used in both QLA1040 (single ended) and QLA1041 (differential),J then mentions 3 connectors, 1 external 2 internals for QLA1040, and only 1I external connector for QLA1041... My board definitely has 3 connectors!!!   H Now I am really confused!!!  Could that sticker for differential type beL mistakenly applied there?  Am I faced with an alien replica?  How can I tellH these things apart for sure?  Does one of you have a QLA1041 or KZPBA-CB9 or -CY, who could tell me the number of connectors on it?   J The vendor proposes a QLA1080 (capable of both Ultra wide Single Ended andE Ultra2 LVD...) as a replacement or my money back... I am unsure it is I compatible with the PWS though, even with recent SRM versions.  There are I some warnings in older posts in this news group about its incompatibility J with OpenVMS, but maybe things have evolved...?   I don't know what is the  right thing to do at this point!  
 Thank You!  
   -- Olivier.    ------------------------------  # Date: Mon, 25 Aug 2003 03:52:26 GMT " From: "ohm62" <ohm62@coldmail.com>G Subject: Re: Difficulties with DPW 500au and StorageWorks components... 7 Message-ID: <_Jf2b.16077$kF1.7045@nwrdny02.gnilink.net>   4 Ok... I found the information at the following link:K http://lists.debian.org/debian-alpha/2003/debian-alpha-200302/msg00056.html   G The HVD version of the QLA board based on the ISP1040B controller has 8 J parallel yellow resistors, which the single ended version does not have...  L My board has these!  Bummer!!  This is the second board I have to return, asJ non-conformant to specifications...  And by the looks of it, it could have> damaged the BA personality module circuitry, and/or the disks!  H I hope this thread can help someone else out in the future, to avoid theC same mistake.  These things are easily mixed up one for another, as K descritions found here and there are confusing:  QLA1040B, as advertised by K some vendor here, has no reality.  QLA1040 (single ended) and QLA1041 (HVD) G although both based on the same Qlogic ISP1040B chip, are two different L beasts, and beware using one for the other may be harmful to your equipment!I You would think that Qlogic would have inscribed clearly HVD somewhere on I the board to help preventing such mistakes, as some others mention LVD... I Well, think again!  This is not a philantropic business.  And then do not 3 expect vendors themselves to know what they sell!!!   . Thanks to all who helped me unraveling this!!!  
   -- Olivier.    ------------------------------  # Date: Mon, 25 Aug 2003 04:55:31 GMT 6 From: "Craig A. Berry" <craigberry@mac.com.spamfooler>G Subject: Re: Difficulties with DPW 500au and StorageWorks components... @ Message-ID: <0d81b8e4ceb00a9f50eaee60042c2130@news.teranews.com>  7 In article <_Jf2b.16077$kF1.7045@nwrdny02.gnilink.net>, $  "ohm62" <ohm62@coldmail.com> wrote:  I > The HVD version of the QLA board based on the ISP1040B controller has 8 L > parallel yellow resistors, which the single ended version does not have... > N > My board has these!  Bummer!!  This is the second board I have to return, asL > non-conformant to specifications...  And by the looks of it, it could have@ > damaged the BA personality module circuitry, and/or the disks!  G Been there done that.  Luckily the only drive I ruined was something I  F paid $5.00 for on eBay.  This was one of my first hardware adventures G after the hobbyist program made it feasible for a software guy to bang  E about with old machines at home.  Now I'll never forget what HVD is,  ; though I think it's pretty much an obsolete technology now.   J > I hope this thread can help someone else out in the future, to avoid theE > same mistake.  These things are easily mixed up one for another, as M > descritions found here and there are confusing:  QLA1040B, as advertised by M > some vendor here, has no reality.  QLA1040 (single ended) and QLA1041 (HVD) I > although both based on the same Qlogic ISP1040B chip, are two different N > beasts, and beware using one for the other may be harmful to your equipment!K > You would think that Qlogic would have inscribed clearly HVD somewhere on K > the board to help preventing such mistakes, as some others mention LVD...   G Whoever decided to use the same 68-pin connector for differential SCSI  0 as the one used for single-ended should be shot.  C As a side note, if you can locate a DEC/Compaq-branded KZPCA, that  > should work in your 500au -- it works in mine and does either B singled-ended or LVD depending on what's connected.  Don't try an A off-brand but otherwise identical Symbios-based card; the driver  2 detects those and commits suicide during start-up.   ------------------------------  % Date: Sun, 24 Aug 2003 20:08:06 -0400   From: John Santos <JOHN@egh.com> Subject: Re: Fork on OPEN VMS 5 Message-ID: <1030824195841.6047A-100000@Ives.egh.com>   " On 19 Aug 2003, Jose Sancho wrote:   > Hi,=20 >=20 > Thank you for your answers.  >=20H > Well, I think that LIB$RUN_PROGRAM will do that I want. I have to testF > it, but I think it will work. After the tests, I will say something. >=20+ > >Please purge that "does not work" phrase H > I apologise. I'd like to say that I did not write "does not work", but; > "I cannot make it work", that assumes a error on my side.   > It's not a matter of assigning blame, either to you or to VMS!  G The problem with "does not work" is it isn't nearly specific enough for  anyone to be able to help you.   We need to know:  % 1) what you are trying to accomplish, I 2) how you are trying to do it (a small code fragment often helps a lot), ! 3) what you expect to happen, and F 4) what actually happens, including any error messages, status returns    from system calls, etc.   >=20H > Its hard to explain why I need to restart the same program. Basically,C > we are developing a Console that running in Windows, must control E > programs running on Open VMS, Linux, Unix, Windows. etc. One of the B > requirements is to restart these programs (they could be runningH > normally, at any point). We are doing it by using TIBCO RendezVous, soH > at a point, the program could have several threads running. Its not asE > simple as change the code or the system to assure the program keeps 
 > running. >=20, > Well, I'm going to try LIB$RUN_PROGRAM.=20 >=20 > Thanks again. 	 > Jos=E9.  >=20 >=20 >=20 >=20L > joseantonio_garcia@hotmail.com (Jose Sancho) wrote in message news:<f2e72=. 344.0308190112.5d40dce2@posting.google.com>... > > Hi,  > >=20G > > I'm porting a UNIX "C"; program to OPEN VMS. On UNIX, after a fork, F > > the child process continues existing when father dies. In OPEN VMSB > > (vfork) do not. I'd like to create a process (inside of a "C";/ > > program) that "survives"; when father dies.  > >=20D > > I've seen I can do that with SYS$CREPRC (...PRC$M_DETACH). But IJ > > cannot make it work. My problem is that I cannot achieve to assign theC > > terminal to the new process: father's sys$input, sys$output and E > > sys$error were the terminal. After creating the detached process, J > > father dies, and I'd like the new detached process to continue working > > whit the terminal. > >=20I > > The only think I want is to do a program restart: my program executes I > > a copy of itself, and then exits; the new process (that ALWAYS is the D > > same program, but a different process than the father) continues) > > running. Is there any way to do that?  > >=20 > > Thanks a lot.=20 > > Jos=E9 Sancho  > > Infineon - Portugal. >=20 >=20   --=20  John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  # Date: Mon, 25 Aug 2003 00:11:39 GMT 4 From: brad@.gateway.2wire.net (Bradford J. Hamilton) Subject: ftp access to itrc / Message-ID: <%uc2b.248008$o%2.112142@sccrnsc02>   N An earlier post in this NG (I think) complained of lack of access to the patchO list via FTP at ftp.itrc.hp.com.  When I saw the post, I checked, and found the L same problem that the poster complained about; no directories were available for viewing.  G I just checked the site again a few minutes ago, and all seems to be as J expected; directories are available for viewing, and the patch tree can be navigated as usual.   N I wanted to follow up to the original thread, but could not find it in Google,* so I am starting a new thread.  Apologies.  P ________________________________________________________________________________A Bradford J. Hamilton                    "All opinions are my own" K bMradAhamiPltSon-at-coMmcAast.nPeSt     "Lose the MAPS, and replace '-at-'  0                                          with @"   ------------------------------  % Date: Sun, 24 Aug 2003 21:55:12 -0400 $ From: Hein <hein_cov@eps.zk.dec.com># Subject: Re: Global Buffers and XFC . Message-ID: <3F496C80.DBE3A091@eps.zk.dec.com>   Wil Marsh wrote:  G > We are in the process of upgrading to VMS 7.3-1 from 7.2-2. Currently F > all of the production ISAM files have ample global buffers such thatF > the majority of reads are from data cached in the global buffers. WeF > are only using 30% of the memory, leaving lots of room for XFC to doG > its work. Our question is: Do we contiue to use the global buffers or 2 > let the production ISAM files use the XFC cache?  I With RMS Global Buffers the data will be truly shared: 1 buffer, mutliple  users.H With XFC you may be able to avoid re-reading data from the disk, but you
 will stillJ move it fromt the system buffer into the process RMS buffer, once for each users.  F So let's say you have an indexed file shared by 500 processes, and 200 global buffers. G Amongst those 200 buffers will for example be  the ROOT for the primary  index.I Without global buffers each process would likely have its own copy of the 
 root, plusC a copy in the XFX for 501 copies. So those 200 global buffers saved  activating 300J local buffer! A sngle gloabl buffer woudl have achieved that, but it would	 have been I flushed too often to make room for data and index. So you use a bunch and  you hopeH to catch many more frequently accessed VBNs and those 200 global buffers will likely I SAVE lots of MEMORY (500*4 buffers or so). Global buffers will  also SAVE  CPU J (no QIO, no buffer copy). Global buffers will NOT increase bucket / record locking,: in fact it may help lockign through concurrent read locks.  H Please do activate XFC (with all patches in place), but do NOT take that as a reason ; to disable global buffers. Those are two seperate decision.   
 Good luck, Hein.    ------------------------------  # Date: Sun, 24 Aug 2003 21:15:10 GMT 0 From: Randy Park <rjpark@mindspring.nospaam.com>S Subject: Re: I need 4 DEC/BASIC OpenVMS programmers for perm positions immediately! 8 Message-ID: <lcaikvgtv4dnptq4otkdpfe2j6fblfm389@4ax.com>  C On 22 Aug 2003 16:11:05 -0700, jcarl@paradigm-wa.com (Jason) wrote:   G >Currently interviewing and looking to fill ASAP.  If you fit the skill F >set listed above or know of someone that does please send your resume >to jcarl@paradigm-wa.com ASAP.  >  >thanks, >  >Jason Carl 	 >Paradigm 
 >253 887 8353   9 Hmmm.  Both myself and my wife have decades of expereince 5 with DEC/VAX Basic beginning with RSTS/E V6A (me) and / V6B (her).  We are both currently not employed.   9 We also live close enough that many area code 253 numbers  are a local call.   ? Any bets that this guy will respond to this post, or is he just  fishing for resumes.   ------------------------------    Date: 24 Aug 2003 14:02:55 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen) + Subject: Re: Looking for TK50s in the UK... 3 Message-ID: <Iyf4Spp3lLQq@eisner.encompasserve.org>   o In article <cf2c6063.0308232348.40d06f01@posting.google.com>, rod.chapman@praxis-cs.co.uk (Rod Chapman) writes: j > Kilgallen@SpamCop.net (Larry Kilgallen) wrote in message news:<dEY0phvoxDwl@eisner.encompasserve.org>... >> But for the size ofP >> demand you describe (i.e., not all customers), eBay would be a viable source. >  > yes - good idea. > H >> Rod, is Spark Inspector available for VAX VMS ?   What VMS versions ? > F > yes - it was the first platform supported.  The Examiner was largely  ) Whoops, sorry for getting the name wrong.   9 > developed on VMS.  VMS version 5.5-2 or better I think. ) >  - Rod Chapman, Praxis Critical Systems    ------------------------------  % Date: Mon, 25 Aug 2003 13:11:52 +1000 C From: "Christine Ricketts/Andrew Stewart" <u1276a@uxnxixtxe.com.au> 1 Subject: Re: M7940/DHV-11, TU58, and a VAX 11/750 , Message-ID: <3f497ee9_1@news.iprimus.com.au>  F "John Gemignani, Jr." <jon-nope@thiswontworkossc.net> wrote in message5 news:l7Y1b.6823$Nc.4290541@news1.news.adelphia.net...  > * > "msell" <msell@pdq.net> wrote in message) > news:bi9anp$qc1@library2.airnews.net...  > >  [cut] J > > I found a document describing how to configure an M7940 (which I have)K > > for operation with a TU58, and my QBus VAX (MicroVAX II) recognizes the F > > drives (the interface, actually) - so the card at least appears to > > partially operate. > > G > > However - when I try to "mount" a tape, there is no activity on the 3 > > interface (M7940) as tested with a logic probe.  [cut]  > = > If it's from an 11/03 then is it possible that it's limited  > to an 18-bit address space?   J The DLV11 doesn't do DMA so it will work in 16, 18 or 22 bit Qbus systems.  < My guess is that VMS expects that a terminal is on that port, and the Login type process is monitoring it.4 My VMS is rusty so I cannot suggest how to configure the DD (?) driver.   Regards, Andy.   ------------------------------  % Date: Mon, 25 Aug 2003 07:22:34 +0200 : From: Karl Rohwedder <extern.karl.rohwedder@volkswagen.de>7 Subject: Re: Mount DQB0 gets 'allready allocated' error , Message-ID: <bic6eb$oql3@doiweb4.wob.vw.vwg>   Larry Kilgallen wrote:  ] > In article <3F4692E6.BDA7C2BC@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:  >  >>Karl Rohwedder wrote:  >>j >>>I try to mount a CDROM (DQB0) on a DS10 and get the 'device already allocated to another user' - error, >  > F >>There is/was a known problem, especially with ISO-9660 CDs where theF >>volume label was longer than a number characters that I don't recall@ >>just now. Maybe someone else will chime in with more info. ... >  > E > Without looking it up, I expect that error would be "Another volume G > of this name is already mounted".  The volume label field for ISO9660 6 > is longer than supported by the VMS uniqueness code.  & The errors occurs during the ALLOCATE:  
 $ all dqb0< %SYSTEM-W-DEVALLOC, device already allocated to another user QAK01_Rohwedder. sh dev dqb0/fu   N Disk $1$DQB0: (QAK01), device type Compaq  CRD-8402B, is online, file-orientedG      device, shareable, available to cluster, error logging is enabled.   P      Error count                    0    Operations completed                907P      Owner process                 ""    Owner UIC                      [SYSTEM]P      Owner process ID        00000000    Dev Prot            S:RWPL,O:RWPL,G:R,WP      Reference count                1    Default buffer size                 512P      Total blocks             1290232    Sectors per track                     8P      Total cylinders            40320    Tracks per cylinder                   4%      Allocation class               1     S so media formt and/or labels could not be involved. The system is a standalone DS10 2 with VMS V7.3-1 and all level 1 patches installed. --    + mit freundlichen Gren | with best regards     Karl Rohwedder		iT-Ingenieurteam$ Ellernbruch 11		D-38112 Braunschweig  E mailto:it-ingteam(at)t-online.de | mailto:rohwedder(at)decus.decus.de - mailto:extern.karl.rohwedder(at)volkswagen.de    ------------------------------    Date: 24 Aug 2003 14:08:27 -0700& From: jordan@ccs4vms.com (Rich Jordan)) Subject: Re: New kits of MySQL and Python = Message-ID: <cc5619f2.0308241308.6a5c60ea@posting.google.com>   D I finally had a chance to install Python and MySQL yesterday (had to< wait on getting the PWS upgraded to VMS V7.3-1 and patched).  B The system is a PWS-600au (enhanced), 1GB RAM, OpenVMS V7.3-1 withF latest class 1 patches on an ODS-2 disk, SSL V1.1, Mozilla 1.4, PythonE and other products installed on an ODS-5 data disk.  No compilers are F currently installed.  The test user account has its UAF parameters set4 per the Mozilla and Java recommendations (sizeable).  B Python had a couple of unexpected hiccups when I ran the tests.  IA _did_ do the two optional compiles from the install instructions, D which apparently completed successfully.  None of the modules in theF 'install_dynamic_modules' procedure have been uncommented, so none are4 installed on the system, though the routine was run.  D TEST_SOCKET_SSL failed with the following error after a considerable delay:E    test test_socket_ssl crashed -- socket.sslerror: (8, 'EOF occurred  in violation of protocol')  B TEST_ZLIB failed after a _long_ 60+ minute delay at 100% CPU usage with: D    test test_zlib failed -- errors occurred; run in verbose mode for details   ? When I re-ran TEST_SOCKET_SSL in 'verbose' mode, I quickly got:       test_socket_ssl?    test_socket_ssl skipped -- Use of the `network' resource not  enabled     1 test skipped:       test_socket_ssl +    Those skips are all expected on OpenVMS.   B TEST_ZLIB in verbose mode showed all OK except the following threeE errors; as shown it completed in under 2 minutes, so no repeat of the  log delay from the test series: B    test_manydecompimax (test.test_zlib.CompressObjectTestCase) ... ERROR C    test_manydecompimaxflush (test.test_zlib.CompressObjectTestCase) 	 ... ERROR A    test_manydecompinc (test.test_zlib.CompressObjectTestCase) ...  ERROR   C The more detailed error is below; all were reported as MemoryError.   E I've not found a reference to these errors as yet, but I am looking.  
 Any thoughts?    Rich Jordan       F ======================================================================B ERROR: test_manydecompimax (test.test_zlib.CompressObjectTestCase)F ----------------------------------------------------------------------" Traceback (most recent call last):9   File "/python_root/Lib/test/test_zlib.py", line 383, in  test_manydecompimax      flush=False, cx=32, dcx=4)9   File "/python_root/Lib/test/test_zlib.py", line 225, in  test_decompimax *     bufs.append(co.compress(data[i:i+cx])) MemoryError   F ====================================================================== ERROR: test_manydecompimaxflush ' (test.test_zlib.CompressObjectTestCase) F ----------------------------------------------------------------------" Traceback (most recent call last):9   File "/python_root/Lib/test/test_zlib.py", line 389, in  test_manydecompimaxflu sh     flush=True, cx=32, dcx=4) 9   File "/python_root/Lib/test/test_zlib.py", line 225, in  test_decompimax *     bufs.append(co.compress(data[i:i+cx])) MemoryError   F ======================================================================A ERROR: test_manydecompinc (test.test_zlib.CompressObjectTestCase) F ----------------------------------------------------------------------" Traceback (most recent call last):9   File "/python_root/Lib/test/test_zlib.py", line 377, in  test_manydecompinc     flush=True, cx=32, dcx=4) 9   File "/python_root/Lib/test/test_zlib.py", line 188, in  test_decompinc*     bufs.append(co.compress(data[i:i+cx])) MemoryError   F ---------------------------------------------------------------------- Ran 27 tests in 100.783s   FAILED (errors=3) A test test_zlib failed -- errors occurred; run in verbose mode for  details    ------------------------------  % Date: Sun, 24 Aug 2003 20:31:21 -0400 * From: JF Mezei <jfmezei.spamnot@istop.com>. Subject: OT: 64 bit desktop computing is here.) Message-ID: <3F4958D6.D3D52D40@istop.com>   L While people were arguing whether AMD might survive long enough to introduce) its 64 bit 8086, Apple was hard at work.    I They have now introduced the G5, which is a 64 bit Power4 core with added C gadgets such as Apple's velocity engine. 13nm copper process. 2Ghz.    http://www.apple.com/ca/G5  N Oh, and it can run 32 bit MAC applications natively with no performance hit or need for an emulator.    Power Mac G5 Dual 2GHz PowerPC G5, SPECint_base2000: Integer calculations: 	8001 SPECfp_base2000: Floating-point calculations: 840 1 SPECint_rate_base2000: Integer calculations: 17.2 7 SPECfp_rate_base2000: Floating-point calculations: 15.7   K In the performance white paper, they only compare it to the 3ghz 8086. They K also compare real world applications perforance such as photoshop. It would D have been interesting to see comparisons with Alpha EV7 and whatever) version-du-jour IA64 has available today.   K Can anyone comment on whether Apple's G5 truly ranks as the world's fastest O desktop processor or will it be outranked by AMD's offering when it comes out ?   6 And how does it fare against IA64 based workstations ?  K Also, how does the 3mhz 8086 compare to whatever is currently available for * IA64 if you exclude non-CPU differences ?   F Interestingly, the G5 achieves its perforance with only 512k L2 cache.   ------------------------------  % Date: Sun, 24 Aug 2003 15:32:57 -0400 . From: Glenn Everhart <Everhart-nospam@gce.com> Subject: Re: Praxis + Message-ID: <bib33b$8dg$1@bob.news.rcn.net>    Bradford J. Hamilton wrote: r > In article <948f0720.0308230150.7ce94ec1@posting.google.com>, chris_doran@postmaster.co.uk (Chris Doran) writes:] > !"Barry Treahy, Jr." <Treahy@MMaz.com> wrote in message news:<3F4564E8.1080705@MMaz.com>...  > !> Tom Linden wrote: > !>  K > !> >He may have been referring to the programming language, which i think D > !> >came from somewhere in Germany.  Lasrt I saw of itwas ca. 1983 > !> >   > !> > > !> Yes, the language...  > !>  
 > !> Barry > ! B > !According to my venerable DECUS library catalogue, PRAXIS was aG > !block-structured language "similar to Ada" written for the Nova High I > !Energy Laser Project at Lawrence Livermore National Laboratory. It was G > !available as DECUS library item VS0040, and also on SIG tapes -- the H > !latest I have noted was "VAX85B", presumably Fall 1985, DECUS libraryE > !item VS0049. It worked on (or at least produced code for) VAX/VMS, - > !PDP11/RSX and LSI-11 standalone computers.  > ! F > !But none of this enables me to find it on any of the DECUS software > !web sites I know of.  > ! E > !I do have VAX85B here on 1/2" tape, but no way to read it (if it's H > !still readable after all this time), and I'm in the wrong country forE > !you. But maybe the above clues will enable someone closer to help.  > !  > !Chris > M > A quick trip (via FTP) to ftp.encompassus.org reveals the following, in the & > directory CD1/VAXSP46/VAX85B/PRAXIS: > 
 > FTP> dir# > 200 Port 24,62,69,12,49,255 Okay. 4 > 150 LIST of *.*; Started; Opening data connection. > . > ANONYMOUS:[000000.CD1.VAXSP46.VAX85B.PRAXIS] > E > AAAREADME.TXT;5          2 20-AUG-1985 15:57 [1,1] (RWED,RWED,RE,R) E > DIRTOT.LIS;1            15 20-AUG-1985 11:00 [1,1] (RWED,RWED,RE,R) F > LITTON.DIR;1             2 29-JUN-2000 16:57 [1,1] (RWED,RWED,RE,RE)E > LZDECOMP.EXE;1          68  1-MAY-1985 10:06 [1,1] (RWED,RWED,RE,R) F > TEXTIOSYS.DIR;1          1 29-JUN-2000 16:58 [1,1] (RWED,RWED,RE,RE) >  > O > Is this of use to the OP???  I have no clue whether this is useful, but if it N > is, thanks to folks at Encompass/Encompasserve for preserving this material! > K > I think there is also another (ex?)denizen of this NG who keeps a similar O > archive of old DECUS material, but his name and site escape me at the moment; 1 > perhaps if he reads this, he will come forward.  > C > _________________________________________________________________N2 > Bradford J. Hamilton			"All opinions are my own"A > bMradAhamiPltSon-at-coMmcAast.nPeSt	"Lose the MAPS, and replacea5 > (please note the new e-mail address)	'-at-' with @"i4 Have a look e.g. at mvb.saic.com/decus  and below...   ------------------------------    Date: 24 Aug 2003 20:21:31 -0700# From: CBarry34@hotmail.com (CBarry) % Subject: Print date/time and usernamel= Message-ID: <ed28678f.0308241921.18ca5e02@posting.google.com>l  E I am printing with DCPS and I am using the confidential.ps file setup @ for my queues.  I would like to add the username and date to theC confidential.ps file so that the top and bottom line looks like theO< below.  I am sure this is an easy task for someone who knows% postscript.  Any help is appreciated.-  2 Username:  XXX     Date/Time:  XXXX   Confidential   ------------------------------  % Date: Sun, 24 Aug 2003 21:38:30 -0400c$ From: Hein <hein_cov@eps.zk.dec.com>4 Subject: Re: Reconstructing pdf file broken via mail. Message-ID: <3F496896.C2F1DD3C@eps.zk.dec.com>   >aN > the MAIL application barfed on the message due to a record being longer thanN > some 5000 bytes, so I was not able to extract it. But I was able to find the# > actual VMS file and play with it.-  M > I used CONVERT to convert it to various formats, and then FTP to a mac that2S > has the acrobat reader, but alas, ACROBAT seems to complain about a corrupt file.i  M Convert will trust and use the file attributes as stored. Mostly likely those-
 attributesN do NOT match the contents. I tend to used DUMP [/RECORD] to verify the variousT formats (stream, stream_lf (unix), stream_cr (nt), variable (vms)). You then want toA allign the attributes with the fact trhough SET FILE/ATTR=RFM=xxx & After that FTP (as ASCII most likely).Q You may also want to check out EXCHANGE/NET as a tool to convert formats/deblock.o  O > Has anyone ever had success in reconstructing a PDF file ? Any tricks ? Or ist( > this a lost cause not worth pursuing ?  O It can be a fairly readable text file, where you can 'see' where records shoulde	 start/endo> But toss in a 'stream' or two and it becomes harder to decode.   %PDF-1.2 %- 2 0 obj- << /Length 3789 /Filter /FlateDecode >> stream HW[o! } ?A1SiڢH V d˲t CrRR0};Krk'y@ ^}NY??X7%/W7?.??̄ ?WPNuԮQv :0	 endstreams endobj 3 0 objn << /ProcSet [/PDF /Text ]    
 Good luck, Hein.h   ------------------------------  % Date: Mon, 25 Aug 2003 00:32:22 -0400t* From: JF Mezei <jfmezei.spamnot@istop.com>4 Subject: Re: Reconstructing pdf file broken via mail) Message-ID: <3F499144.42D85B17@istop.com>-   Hein wrote:-O > Convert will trust and use the file attributes as stored. Mostly likely thosel > attributes > do NOT match the contents.  N Well, they match the contents. The problem is that the contents were corrupted6 during transport because they were assumed to be text.  M And I couldn't just pay with the file, I had to edit the file first to remove L the RFC822 headers leaving only what seemed to be the PDF "binary" contents.I My mail program would have reconstitutedthe quoted printable stuff into a K single file, but it would have assumed it was a text file, hence created in M variable record format which MAIL likes. My guess is that some stray CR or LFrJ inside some stream would have been converted to CR-LF, thus corrupting the compressed contents.    5 > I tend to used DUMP [/RECORD] to verify the variousaE > formats (stream, stream_lf (unix), stream_cr (nt), variable (vms)).h  L One of the converts I did seemed to reconstitute the file with CR-LFs in theH right place, including when I used a hex editor on the mac after the FTP' transfer. But Acrobat still complained.u   ------------------------------  % Date: Sun, 24 Aug 2003 18:47:39 -0400c  From: John Santos <JOHN@egh.com>( Subject: Re: Strange programming problem5 Message-ID: <1030824183734.6047J-100000@Ives.egh.com>n  ( On Wed, 20 Aug 2003, Mike Bartman wrote:  E > On 20 Aug 2003 13:37:12 GMT, bill@gw5.cs.uofs.edu (Bill Gunshannon)m > wrote: > ; > >In article <1lu4kv0vfch61nnqn44ehtdtp5arqs440m@4ax.com>,n4 > >	Mike Bartman <omni@foolie.omniphile.com> writes:G > >> C is famous for stuff like this...uninitialized pointers, danglinguG > >> pointers, array out of bounds, bad pointer arithmetic, etc., etc.,a > > - > >It's a poor workman who blames his tools. h > D > And it's a foolish workman who removes all of the safety equipmentH > from his power tools, or chooses equipment that lacks them when better > is available.e  E Many times the workman doesn't make the choice.  IME, the old saying,aC "It's a poor workman who blames his tools" is usually said not by aiB more competent workman, but by the workman's boss who is trying to< get by on the cheap by providing him with second-rate tools.   > F > It's possible to cut many cords of firewood with a chainsaw that hasD > no anti-kickback devices, no guards and no auto-releasing throttleG > control.  It's also possible to cut your legs off with such a device.i > F > You can write perfectly good, easilly maintained code in C.  You canE > also write lousy, bug-ridden, almost impossible to debug code in C.oH > The language doesn't hinder the later much at all, though it does lackB > some things that would make the former a lot easier (such as theE > concept of namespaces so you can write sharable/re-usable librariestH > more easily).  It's just not a very good language, though if you thinkH > of it as a high level assembler, you are safer than if you think of it > as a high level language.e > @ > It's a poor workman who doesn't use the best tool for the job.? > Sometimes C is a good choice.  Just not when you have a large E > multiprogrammer system to develop, or when you have programmers wholE > don't have a decade or so of programming experience...that includes-C > maintaining large systems written by others in the distant past. e > E > All that said, most C code that I work with is a lot poorer than ituB > needed to be based on the language features available.  C *does*G > include comments for instance...and it doesn't *require* you to stuffiG > the entire function into the boolean expression of an if statement...g >  > -- Mike Bartmano   -- o John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  + Date: Sun, 24 Aug 2003 21:30:12 +0000 (UTC)oP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)
 Subject: SYSnt$ Message-ID: <bibap4$2s1$1@online.de>  D What is the maximum number of system roots one can have on a disk?    # When is it %XF?  When is it %XFFFF?a  I Which roots are used---officially or inofficially---for satellites?  For u' standalone backup?  For other purposes?e  H I'm finally biting the bullet and setting up multiple system roots on myG system disks, though since each system disk is only used by one system,@E most of them are dummies.  I want to make sure I don't go against anypC conventions, official or not.  The idea is to have a "master system6C disk" which is maintained, and then copies of this disk are used on G other systems.  Having multiple roots avoids the trouble of configuringn! the node-specific stuff by hand. n  G I'm still looking for something than a file-by-file comparison (in eacheC root!) of SYS$SPECIFIC before and after the upgrade as a means for aG making sure that the upgrade takes effect in each SYS$SPECIFIC without s# overwriting customised information.    ------------------------------  % Date: Mon, 25 Aug 2003 00:30:43 +0200o From: Dirk Munk <munk@home.nl> Subject: Re: SYSn 2 Message-ID: <bibecn$3k5$1@news2.tilbu1.nb.home.nl>  / Phillip Helbig---remove CLOTHES to reply wrote:lF > What is the maximum number of system roots one can have on a disk?   > % > When is it %XF?  When is it %XFFFF?  > K > Which roots are used---officially or inofficially---for satellites?  For  ) > standalone backup?  For other purposes?. > J > I'm finally biting the bullet and setting up multiple system roots on myI > system disks, though since each system disk is only used by one system,uG > most of them are dummies.  I want to make sure I don't go against anyjE > conventions, official or not.  The idea is to have a "master system-E > disk" which is maintained, and then copies of this disk are used onmI > other systems.  Having multiple roots avoids the trouble of configuringn# > the node-specific stuff by hand. 1 > I > I'm still looking for something than a file-by-file comparison (in eachiE > root!) of SYS$SPECIFIC before and after the upgrade as a means for  I > making sure that the upgrade takes effect in each SYS$SPECIFIC without p% > overwriting customised information.n  J The VMS installation CD goes up to SYSFF, so 255 cluster nodes. Should be  sufficient for a hobby cluster.L   ------------------------------  % Date: Sun, 24 Aug 2003 22:29:47 -0500s1 From: "David J. Dachtera" <djesys.nospam@fsi.net>x Subject: Re: SYSnr' Message-ID: <3F4982AB.C2BEDF34@fsi.net>n   Dirk Munk wrote: > 1 > Phillip Helbig---remove CLOTHES to reply wrote:iF > > What is the maximum number of system roots one can have on a disk? > > ' > > When is it %XF?  When is it %XFFFF?  > >oL > > Which roots are used---officially or inofficially---for satellites?  For+ > > standalone backup?  For other purposes?L > > L > > I'm finally biting the bullet and setting up multiple system roots on myK > > system disks, though since each system disk is only used by one system,wI > > most of them are dummies.  I want to make sure I don't go against anynG > > conventions, official or not.  The idea is to have a "master system G > > disk" which is maintained, and then copies of this disk are used on K > > other systems.  Having multiple roots avoids the trouble of configuring,$ > > the node-specific stuff by hand. > >oK > > I'm still looking for something than a file-by-file comparison (in each F > > root!) of SYS$SPECIFIC before and after the upgrade as a means forJ > > making sure that the upgrade takes effect in each SYS$SPECIFIC without' > > overwriting customised information.0 > K > The VMS installation CD goes up to SYSFF, so 255 cluster nodes. Should bel! > sufficient for a hobby cluster.r   Just wondering...n  5 Suppose one has a cluster with multiple system disks.e  G Normally, that would be a maintenance headache. However, if one were tofG automate the synchronization of roots among the system disks, could notaD one boot any node from any system disk and have it find the "latest" (most recently sync'd) files?e   Why do I bring that up?e  D The topic of rolling updates in our cluster as been raised more thanE once. The goal is to reduce/minimize cluster downtime on VMS/Multinetn
 upgrades.   , Oracle is quite another question, of course.   --   David J. Dachterae dba DJE Systemsg http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/u   ------------------------------    Date: 24 Aug 2003 21:00:21 -0700+ From: seanobanion@attbi.com (Sean O'Banion)i Subject: Re: SYSns= Message-ID: <f883d5a4.0308242000.36a5999b@posting.google.com>   A SYSF and SYSE are used for Standalone Backup on VAXen, and systemoB upgrades, respectively:  using these roots (on VAX or, in my case,D Alpha) for anything else causes problems.  I found this out the hard? way at HP ETS last year while setting up the Encompass Hands onmD Workshops.  As a workaround, I used SYSFF and SYSEE for the 15th andC 16th system: I had named the systems including 00 through 12 hex in@E the name, and intended to have that reflect the root that that system. booted from.  F Like you are doing, I configured one disk with all the roots I needed,E created an image saveset of that disk, did a MOP-based satellite boottF as "diskless" (no Votes) into the Cluster, and restored the saveset toF a local disk on the satellite.  When booting locally, I did not createE or join a Cluster.  This allowed each system to be nearly identicallyhB configured, and I only needed to install VMS once, but did need toF boot from each root to configure DECnet and TCP/IP before creating the saveset.  C Host-based Shadowing would have worked as well, but would have beenp@ slower, since each block of the destination disk would have been examined and possibly written.  F I think the recommended limit on the number of nodes in one cluster is$ 96, so that might be a useful limit.     Sean    | helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) wrote in message news:<bibap4$2s1$1@online.de>...F > What is the maximum number of system roots one can have on a disk?   > % > When is it %XF?  When is it %XFFFF?i > K > Which roots are used---officially or inofficially---for satellites?  For  ) > standalone backup?  For other purposes?a > J > I'm finally biting the bullet and setting up multiple system roots on myI > system disks, though since each system disk is only used by one system,hG > most of them are dummies.  I want to make sure I don't go against anyeE > conventions, official or not.  The idea is to have a "master systemoE > disk" which is maintained, and then copies of this disk are used onlI > other systems.  Having multiple roots avoids the trouble of configuringo# > the node-specific stuff by hand. - > I > I'm still looking for something than a file-by-file comparison (in eachwE > root!) of SYS$SPECIFIC before and after the upgrade as a means for tI > making sure that the upgrade takes effect in each SYS$SPECIFIC without 2% > overwriting customised information.S   ------------------------------  % Date: Sun, 24 Aug 2003 15:46:10 -04009. From: Glenn Everhart <Everhart-nospam@gce.com>8 Subject: Re: volume shadowing with 2 x IDE disks on Ds10+ Message-ID: <bib3so$b75$1@bob.news.rcn.net>B  A At least at one time vddriver (vdd64.zip vers.) did work with VMStA volume shadowing. Turned out the shdriver assumed that underlying B disks supported io$_dse so I added code to report success on such,E and returning the requested bytecount as done. This was used for somea< tests of a couple hundred shadowsets on one machine at once.  A Can't imagine what is wrong with use with IDE, unless nothing has"B issued io$_packack on the underlying drives to start them up. MostG asnvd programs don't do that (tho I did release one some time back thatwE does). $mount/foreign/system has the same effect, but I have no alpha)E that supports dqdriver AFAIK. (I have an ISA IDE controller card and nD could I suppose try it in my alphastation 200 4/233 but have no clueC if it would work. I am fairly sure I would need to turn off DMA forl such a beast at minimum.  @ Someone with listings might be needed to tell what happens today# with SHDRIVER and what it requires..   Glenn Everhart   Rob Young wrote:  e > In article <rwQQK6IDhZ4A@eisner.encompasserve.org>, Kilgallen@SpamCop.net (Larry Kilgallen) writes:d > [ >>In article <Rep1b.3093$DB6.966@news.cpqcorp.net>, "Safir" <axica_nopub@yahoo.com> writes:  >> >>>It is not supported >>>see5 >>>http://h18000.www1.hp.com/info/SP2729/SP2729PF.PDFo >>@ >>What line in that document makes you say it is not supported ? >  >  >  > 	Our old friend: > 3 > Volume Shadowing for OpenVMS provides support fors8 > SCSI disks and controllers on Alpha systems. Shadowing3 > can also be used with third-party SCSI disks that 9 > have READL (read long) and WRITEL (write long) commands + > implemented and that use the OpenVMS SCSIV; > disk driver. Features are restricted when SCSI disks thato4 > do not support READL and WRITEL are shadowed; disk8 > bad-block errors resulting from the use of these disks1 > can cause members to be removed from the shadowM > set. >  > B > 	Now maybe these IDE drives support SCSI READL/WRITEL?  The only@ > 	way I would do shadowing without READL/WRITEL is if there wasH > 	a worthy (or better) substitute for READL/WRITEL (i.e. you never get 1 > 	bad blocks, for instance - many months uptime:t >  > 	$ show error0 >  > 	Looks like this:n >  >  > <NODE1>$ show err . > Device                           Error Count, > $20$DQA1: (NODE1)                        1, > $20$DQB0: (NODE1)                        1, > $20$DQB1: (NODE1)                        1 >  > 	 > 				Robh >    ------------------------------  # Date: Sun, 24 Aug 2003 21:26:10 GMTa6 From: "Craig A. Berry" <craigberry@mac.com.spamfooler>) Subject: Re: XML Parser for VAX (OpenVMS)-@ Message-ID: <5b3cc9530549e70081587031c419b652@news.teranews.com>  8 In article <mm9ikvknijs0d3pm2m7bv9fd0keggn1ggl@4ax.com>,2  Randy Park <rjpark@mindspring.nospaam.com> wrote:  4 > Does anyone know of any XML parsers, callable from4 > a VMS application, that are available for both Vax2 > and Alpha systems.  I have C and Basic compilers8 > currently available.  I have found numerous references1 > to parsers for OpenVMS Alpha using C++ or Java,w > but nothing for Vax.  5 There's a non-validating parser in C called expat at  ? <http://sourceforge.net/projects/expat>.  The Perl XML::Parser eD extension is based on expat.  There's also a validating parser in C ? called libxml2 that's part of the Gnome project.  More info at w <http://www.xmlsoft.org/>.    D Both of these parsers have been known to work on VMS.  There may be F additional porting required for VAX support (or maybe not -- I simply  haven't tried to build in VAX).S   ------------------------------  # Date: Sun, 24 Aug 2003 21:04:49 GMT 0 From: Randy Park <rjpark@mindspring.nospaam.com>% Subject: XML Parser for VAX (OpenVMS)08 Message-ID: <mm9ikvknijs0d3pm2m7bv9fd0keggn1ggl@4ax.com>  7 I have been asked to enhance my application software soz7 that it can process (read/write) data from and to files 2 formatted with XML tags.  After reading through an5 introductory book on XML, writing data files with XMLe tags seems fairly easy.i  2 Does anyone know of any XML parsers, callable from2 a VMS application, that are available for both Vax0 and Alpha systems.  I have C and Basic compilers6 currently available.  I have found numerous references/ to parsers for OpenVMS Alpha using C++ or Java,i but nothing for Vax.   Thanks in advance.   ------------------------------   End of INFO-VAX 2003.469 ************************