1 INFO-VAX	Thu, 15 Jun 2000	Volume 2000 : Issue 333       Contents:* 2nd request, why can't I get to COMPAQ.COM. Re: 2nd request, why can't I get to COMPAQ.COM. Re: 2nd request, why can't I get to COMPAQ.COM AXP VOLATILE WARNINGS  Re: AXP VOLATILE WARNINGS  Re: AXP VOLATILE WARNINGS  Re: Backup performance Re: Backup performance# Re: C 6.2-007 compilation annoyance / Re: C bashing (was Re: VMS File Caching Futures > Re: Canadian Association of Compaq Users Symposium rescheduled Cisco Configs? Re: Cisco Configs?" Re: CMQ/Dec with a sense of Humor?" Re: CMQ/Dec with a sense of Humor?" Re: CMQ/Dec with a sense of Humor?" Re: CMQ/Dec with a sense of Humor?" Re: CMQ/Dec with a sense of Humor?" Re: CMQ/Dec with a sense of Humor?" Re: CMQ/Dec with a sense of Humor?" Re: CMQ/Dec with a sense of Humor?" Re: CMQ/Dec with a sense of Humor? Computer Prosperity > DEC Pascal 5.7 Alignment warnings (was: AXP Volatile Warnings)B Re: DEC Pascal 5.7 Alignment warnings (was: AXP Volatile Warnings)B Re: DEC Pascal 5.7 Alignment warnings (was: AXP Volatile Warnings)B Re: DEC Pascal 5.7 Alignment warnings (was: AXP Volatile Warnings)B Re: DEC Pascal 5.7 Alignment warnings (was: AXP Volatile Warnings) DPW433 and graphic's card * Re: Files-11 ODS-2 Readability for FreeBSD Re: Fun VMS Facts? Re: Fun VMS Facts? Re: Fun VMS Facts? re: Fun VMS Facts? re: Fun VMS Facts? Re: Fun VMS Facts? Re: Fun VMS Facts? Re: Fun VMS Facts? Re: Fun VMS Facts? Re: Fun VMS Facts? Re: Fun VMS Facts? Re: GS320/GS160 , Help! VAX emulator on Intel with Hobbyist CDG Re: Large amount of files in one directory causes very slow performance  Lock Statistics  Looking for Jnet software  Mail batch log file  RE: Mail batch log file  RE: Mail batch log file  patches, graphic cards Re: patches, graphic cards Re: patches, graphic cards5 Re: Proxy problem: Why does node:: work but 0:: fail? . Re: Removing/Stripping RFC headers from emails RE: Secondary Passwords  Re: Secondary Passwords  Re: Secondary Passwords  Re: Some HW questions ' Stealing the best- (was Fun VMS Facts?) # System call to get CPU utilization? ' Re: System call to get CPU utilization? ' Re: System call to get CPU utilization? ' Re: System call to get CPU utilization?  Techwise report on availability  Re: VAX on Intel?  Re: VAX on Intel?  Re: VAX on Intel? ; Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha? & WTB: VAX-bootable external SCSI CD-ROM* Re: WTB: VAX-bootable external SCSI CD-ROM* Re: WTB: VAX-bootable external SCSI CD-ROM WTD: VAX parts in OTTAWA  F ----------------------------------------------------------------------  # Date: Thu, 15 Jun 2000 12:36:49 GMT / From: "John Nixon" <jorlnixon@worldnet.att.net> 3 Subject: 2nd request, why can't I get to COMPAQ.COM F Message-ID: <B5425.2699$Uw3.175151@bgtnsc04-news.ops.worldnet.att.net>  ' Nobody replied to my previous question.   H Starting one day about a week ago, I was unable to connect to COMPAQ.COMK from my LAN connection at work.  Since I was able to connect from my dialup I connection from my PC at home, I assumed we were having a problem where I L work. However, on this newsgroup several other people claimed to be having a? similar problem, and there was some reference to a DNS problem.   J Now, since it was a known problem outside of our company, I cannot get ourL busy network people to work on this since I am the only one that cares (I amK also practically the only one pushing to keep COMPAQ products in our shop!! L In fact, I may be on the verge of a major breakthrough for OPENVMS there!!!.C THIS IS NOT HELPING ME PROMOTE THE IMAGE OF COMPAQ OR OPENVMS!!!!!!     2 ISN'T ANYONE ELSE HAVING A PROBLEM WITH COMPAQ.COM. Or if you fixed the problem HOW DID YOU FIX IT? Or if you know what the problem was or is, DOES IT STILL EXIST?    ------------------------------  % Date: Thu, 15 Jun 2000 09:03:26 -0400 , From: Howard S Shubs <hshubs@mindspring.com>7 Subject: Re: 2nd request, why can't I get to COMPAQ.COM > Message-ID: <hshubs-7E9558.09032615062000@news.mindspring.com>  G In article <B5425.2699$Uw3.175151@bgtnsc04-news.ops.worldnet.att.net>,  0 "John Nixon" <jorlnixon@worldnet.att.net> wrote:  3 >ISN'T ANYONE ELSE HAVING A PROBLEM WITH COMPAQ.COM / >Or if you fixed the problem HOW DID YOU FIX IT @ >Or if you know what the problem was or is, DOES IT STILL EXIST?  1 Haven't you tried it?  That cleared up last week.    --   Howard S Shubs, the Denim Adept    ------------------------------  # Date: Thu, 15 Jun 2000 13:22:25 GMT / From: "John Nixon" <jorlnixon@worldnet.att.net> 7 Subject: Re: 2nd request, why can't I get to COMPAQ.COM F Message-ID: <lM425.3207$Xx5.196545@bgtnsc06-news.ops.worldnet.att.net>  J I have tried it several times a day.   It has NOT cleared up for me.  That is why I keep asking. K If the problem has been correctly fixed, then I have a problem at work, but  since my problemK at work started at the same time as the overall problem, our network people 
 are reluctant  to spend valuable time on it.   K But thanks for responding.  That does give me a little leverage to use with  our network  folks.  9 "Howard S Shubs" <hshubs@mindspring.com> wrote in message 8 news:hshubs-7E9558.09032615062000@news.mindspring.com...H > In article <B5425.2699$Uw3.175151@bgtnsc04-news.ops.worldnet.att.net>,2 > "John Nixon" <jorlnixon@worldnet.att.net> wrote: > 5 > >ISN'T ANYONE ELSE HAVING A PROBLEM WITH COMPAQ.COM 1 > >Or if you fixed the problem HOW DID YOU FIX IT B > >Or if you know what the problem was or is, DOES IT STILL EXIST? > 3 > Haven't you tried it?  That cleared up last week.  >  > --! > Howard S Shubs, the Denim Adept    ------------------------------  # Date: Thu, 15 Jun 2000 13:08:16 GMT  From: paul_hallam@hotmail.com  Subject: AXP VOLATILE WARNINGS) Message-ID: <8iakfk$obe$1@nnrp1.deja.com>   E When compiling code on a V7.1 ALPHA, Pascal 5.7 I am getting warnings  about unaligned variables G BUT as far as I can see all references are VOLATILE and I don't believe # I should be getting these warnings.   E I know I must be doing something wrong and I would like to know what, + thats if anyone out there knows the answer.   E I know these are only warnings but if we don't understand the cause I - can't completely guarantee atomicity when the  processes finally run.  F HERES THE VARIABLE DEFINITIONS (and I can't change those starting with  XMI) followed by a sample error.   ***** 1 xmi_uword           = [VOLATILE, WORD] 0..65535 ;     xmi_rtl_type        = xmi_UWORD;  / xmi_ubyte           = [VOLATILE, BYTE] 0..255 ;   xmi_buf_elem        = xmi_ubyte; xmi_max_buf_len     = 5400; > xmi_buf_type        = [VOLATILE] ARRAY [0..xmi_max_buf_len] OF>                                                  xmi_buf_elem;  ' xmi_rec_type        = [VOLATILE] RECORD %                       CASE BOOLEAN OF ;                            TRUE  : (data   : xmi_buf_type); G                            FALSE : (tpl_no : [POS (0)] xmi_tpl_no_type; E                                     rtl    : [POS (8)] xmi_rtl_type);                        END;  / xmi_rec_ptr         = [VOLATILE] ^xmi_rec_type;   " reqst_record        : XMI_rec_ptr;" rtl                 : xmi_rtl_type   ******  F WHY, WHEN I COMPILE ON THE V7.1 ALPHA WITH THE PASCAL-V57-073 COMPILER DO I GET THE FOLLOWING WARNING.         reqst_record^.rtl := 1; .....^C %PASCAL-W-UNAVOLACC, volatile access appears unaligned, but must be < aligned at run-time to ensure atomicity and byte granularity at line number 6421   !     reqst_record^.rtl    := rtl ;  ....^ C %PASCAL-W-UNAVOLACC, volatile access appears unaligned, but must be < aligned at run-time to ensure atomicity and byte granularity at line number 5356    Many thanks               Paul     & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------  % Date: Thu, 15 Jun 2000 11:41:55 -0400 = From: "Stephane Paquin" <stephane.paquin@nospam.ispatnet.com> " Subject: Re: AXP VOLATILE WARNINGS1 Message-ID: <8P625.4091$227.94285@nnrp1.uunet.ca>   K     I am not a PASCAL specialist but on the Alpha the compilers checks that L each variable memory allocation starts on a byte multiple of the size of theI variable. For 8 bytes variable, it would be on multiple of 8. For 4 bytes F variable, it would be on multiple of 4 and so forth. For instance, theK following declaration in a common or structure delcaration would produce an = alignment error since i is not on a 4 byte multiple boundary.    character*1 c1 integer*4 i   > The follwing would be ok since i starts at a correct multiple.   character*4 c4 integer*4 i   K This is a well known problem/feature of the Alpha architecture. See the VMS J FAQ for more info(http://www.openvms.digital.com/wizard/openvms_faq.html).H Also the compiler reference should contain a section on alignment on the Alpha.  	 Good luck    Stphane      * <paul_hallam@hotmail.com> wrote in message# news:8iakfk$obe$1@nnrp1.deja.com... G > When compiling code on a V7.1 ALPHA, Pascal 5.7 I am getting warnings  > about unaligned variables I > BUT as far as I can see all references are VOLATILE and I don't believe % > I should be getting these warnings.  > G > I know I must be doing something wrong and I would like to know what, - > thats if anyone out there knows the answer.  > G > I know these are only warnings but if we don't understand the cause I / > can't completely guarantee atomicity when the  > processes finally run. > H > HERES THE VARIABLE DEFINITIONS (and I can't change those starting with" > XMI) followed by a sample error. >  > ***** 3 > xmi_uword           = [VOLATILE, WORD] 0..65535 ;  > " > xmi_rtl_type        = xmi_UWORD; > 1 > xmi_ubyte           = [VOLATILE, BYTE] 0..255 ; " > xmi_buf_elem        = xmi_ubyte; > xmi_max_buf_len     = 5400; @ > xmi_buf_type        = [VOLATILE] ARRAY [0..xmi_max_buf_len] OF@ >                                                  xmi_buf_elem; > ) > xmi_rec_type        = [VOLATILE] RECORD ' >                       CASE BOOLEAN OF = >                            TRUE  : (data   : xmi_buf_type); I >                            FALSE : (tpl_no : [POS (0)] xmi_tpl_no_type; G >                                     rtl    : [POS (8)] xmi_rtl_type);  >                       END; > 1 > xmi_rec_ptr         = [VOLATILE] ^xmi_rec_type;  > $ > reqst_record        : XMI_rec_ptr;$ > rtl                 : xmi_rtl_type >  > ****** > H > WHY, WHEN I COMPILE ON THE V7.1 ALPHA WITH THE PASCAL-V57-073 COMPILER! > DO I GET THE FOLLOWING WARNING.  >  >      reqst_record^.rtl := 1; > .....^E > %PASCAL-W-UNAVOLACC, volatile access appears unaligned, but must be > > aligned at run-time to ensure atomicity and byte granularity > at line number 6421  > # >     reqst_record^.rtl    := rtl ;  > ....^ E > %PASCAL-W-UNAVOLACC, volatile access appears unaligned, but must be > > aligned at run-time to ensure atomicity and byte granularity > at line number 5356  > 
 > Many thanks  >              Paul  >  > ( > Sent via Deja.com http://www.deja.com/ > Before you buy.    ------------------------------  % Date: Thu, 15 Jun 2000 12:36:36 -1300 - From: John Reagan <reagan@hiyall.zko.dec.com> " Subject: Re: AXP VOLATILE WARNINGS2 Message-ID: <3948CDD4.7B748B2B@hiyall.zko.dec.com>   paul_hallam@hotmail.com wrote: > G > When compiling code on a V7.1 ALPHA, Pascal 5.7 I am getting warnings  > about unaligned variables I > BUT as far as I can see all references are VOLATILE and I don't believe % > I should be getting these warnings.  > G > I know I must be doing something wrong and I would like to know what, - > thats if anyone out there knows the answer.   H As others have pointed out, VOLATILE asks the compiler to make sure thatG all reads/writes to the variable are atomic in nature.  However, if the H variable/field/etc. is not aligned on a proper boundary, there may be noF Alpha instructions to perform the read/write operation.  In this case,= the compiler is telling you that it cannot do what you ask.     @ You have 3 choices, 1) remove the VOLATILE attribute and let theG compiler generate the multi-instruction sequence, 2) align the field in H the structure so the compiler can generate an atomic sequence, or 3) useC the /USAGE=NOVOLATILE command qualifier to suppress these messages.   B Since xmi_rtl_type is a word, it should be on a word boundary.  ItE currently isn't.  Unaligned atomic word updates are not possible with 
 the Alpha.  C Many times people use VOLATILE when they don't really mean it.  Are < these data structures really volatile?  Can they be modified$ asychronously from the main program?   --   John Reagan  Compaq Pascal Project Leader   ------------------------------  % Date: Thu, 15 Jun 2000 09:26:35 +0100 2 From: Chris Sharman <Chris.Sharman@CCAgroup.co.uk> Subject: Re: Backup performance . Message-ID: <3948A14B.7E5C9F86@CCAgroup.co.uk>  E > > If the tape and source drive both have a higher priority than the F > > quorum disk, then you'll see timeouts during backups. This is alsoD > > likely the reason you see bad performance in other things duringG > > backup, but that may just be because a half-zillion IOs are already  > > queued up.  F I wondered if performance was impacted by the loss of the quorum disk,H and its 1 vote (of 3): I'd previously assumed it was ignoreable chatter.  F > One should 'de-tune' BACKUP. It really doesn't make sense to try andD > fire hundredths or thousands of DIOs (IIRC the default DIOLM for aE > BACKUP account 'should' be something like 4096) to a single disk at  > the same time.  B Having not so long ago tuned it, I can report that we now get moreA backed up & faster: a full disk backup does go faster if backup's E allowed to saturate the disk. Perhaps I'll play with fastback off the H freeware disk: maybe back/phy will enable me to get full speed backup as  well as reasonable user access ?  7 We have a TZ87 (mka500) on our Alphaserver 2100A 5/300. E We have a RAID 230, also on the 2100A, containing 6 2Gb RAID 1 pairs, E and 2 9Gb JBOD disks. We also have 7 2Gb DSSI compatible disks, slung ! between the 2100A and an old VAX. G We have a development Alphastation with a 1Gb, a 2Gb, & a 9Gb disk, all 7 on one bus (the 9Gb is scratch - not backed up at all).   C The quorum disk that we do the disk-disk image backup on is a DSSI.   H > > >The problem is our time 'window' isn't really there for much of theC > > >year any more - no quiet times, not at night, not at weekends.  > > D > > Consider switching over to storage hanging off a HSx controller.E > > The HSx firmware has a CLONE function that lets the controller do E > > a drive cloning (maintaining consistency until the clone is done, G > > which is nice) which you can use--you CLONE then back the clone up. F > > You'll also get more SCSI channels (up to six, if you spring for aG > > full complement of cables and shelves) to the HSx controller, which & > > should cut down on bus contention. > D > Please remember that (last time I checked) CLONING didn't work for@ > RAID-5 storage sets, SNAPSHOT does. I don't know if SNAPHOT isD > available for HSZ controllers; I have played with it on the HSG80,D > however it requires a special (more expensive controller firmware)A > and you need 2 (two) controllers with 512 MByte mirrored cache. ? > I have tried this with a single controller and my command was  > rejected.   H HSx has always been very nice, but way out of our league. I guess that's, probably no longer true. We don't use RAID5.   ------------------------------  % Date: Thu, 15 Jun 2000 11:18:17 -0400 " From: Dan Sugalski <dan@sidhe.org> Subject: Re: Backup performance : Message-ID: <4.3.2.7.0.20000615111637.00e4e5f0@24.8.96.48>  / At 09:26 AM 6/15/00 +0100, Chris Sharman wrote: F > > Please remember that (last time I checked) CLONING didn't work forB > > RAID-5 storage sets, SNAPSHOT does. I don't know if SNAPHOT isF > > available for HSZ controllers; I have played with it on the HSG80,F > > however it requires a special (more expensive controller firmware)C > > and you need 2 (two) controllers with 512 MByte mirrored cache. A > > I have tried this with a single controller and my command was 
 > > rejected.   L SNAPSHOT isn't in the HSx50s. Might be in the x70s. I expect it would be in 3 all the x80s, but I've not checked the HSZ80s here.   I >HSx has always been very nice, but way out of our league. I guess that'su- >probably no longer true. We don't use RAID5.u  5 Go check. Used HSZ controllers aren't that expensive.c   					Dan  I --------------------------------------"it's like this"-------------------A2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and evenw;                                       teddy bears get drunk    ------------------------------   Date: 15 Jun 2000 13:35:19 GMT= From: jlw@psulias.psu.edu (j.lance wilkinson, (814) 865-1818)e, Subject: Re: C 6.2-007 compilation annoyance, Message-ID: <8iam2n$1bfe@r02n01.cac.psu.edu>  U In article <3947D7FD.1F58455C@compaq.com>, Ed Vogel <edward.vogel@compaq.com> writes:o >u@ >    Just the same, the compiler should not be emitting messages@ >    that detect unused include files from header files supplied> >    by the compiler.  We will try to correct this is a future: >    release of Compaq C.  Thanks for bringing this to our >    attention.s  M         If you're going to be opening this up, enhance the warning message toeM         show the "INCLUDE HIERARCHY" that resulted in the unused include file M         -- an include file might be unused, but the include file that invoked M         it might be, and while in some cases a programmer NEEDS the enclosing M         include file and has no control over its content, in other cases they M         do have such control and might edit/modify the enclosing include file L         to suppress the problem.  But if they have no idea where that unusedK         include file is invoked (maybe 7 levels down!) they can't tell what          they might want to do.  K         Needless to say, if in picking up on this thread without thoroughlygN         reading it, and not having the current complier anyway to test it out,F         if the message already says that include file X (included by Y.         (included by Z...) ), then never mind.  M +-"Never Underestimate the bandwidth of a station wagon full of mag tapes"--+aN | J.Lance Wilkinson ("Lance")            InterNet:  Lance.Wilkinson@psu.edu | M | Systems Design Specialist - Lead       AT&T:      (814) 865-1818          |iM | Library Computing Services             FAX:       (814) 863-3560          |tM | E3 Paterno Library                     "I'd rather be dancing..."         |AM | Penn State University         A host is a host from coast to coast,       |SM | University Park, PA 16802     And no one will talk to a host that's close |AM | <postmaster@psulias.psu.edu>  Unless the host that isn't close            |aM | VMS GopherMeister             Is busy, hung or dead.                      |nM +------"He's dead, Jim. I'll get his tricorder. You take his wallet."-------+29                  [apologies to DeForest Kelly, 1920-1999]y3 <A Href="http://perdita.lcs.psu.edu">home page</a> yJ <a Href="http://perdita.lcs.psu.edu/junkdec.htm">junk mail declaration</a>   ------------------------------  % Date: Thu, 15 Jun 2000 09:46:11 -0400 ( From: Ed Vogel <edward.vogel@compaq.com>8 Subject: Re: C bashing (was Re: VMS File Caching Futures* Message-ID: <3948DE23.2DB50150@compaq.com>   Wayne Sewell wrote:s   > M > I assume you already saw this one, since it appeared earlier in the thread:r >m > >  > >    for (i=0; i<100; i++);p > >       {  > >       do that(); > >       }t > >s > Q > I don't remember ever making this particular mistake, though obviously somebodynQ > must have for it to make the list.  I'm sure I will make this error someday, so K > I certainly won't mind the compiler pointing it out when it happens.  :-)E  &     Yes.  I did see this one.  Thanks!                               Ed   ------------------------------  % Date: Thu, 15 Jun 2000 07:56:29 +0100 4 From: John Laird <john@laird-towers.freeserve.co.uk>G Subject: Re: Canadian Association of Compaq Users Symposium rescheduledj8 Message-ID: <99pfks8teoj5inu0jtoi7rjujhe0iqh2ir@4ax.com>  2 On Mon, 12 Jun 2000 15:37:13 -0400, "Peter Weaver" <peter.weaver@stelco.ca> wrote:3  D >As a side note, are there any DECUS groups in the world that charge >money for membership?   DECUS UK do.   -- H
 John Laird   ------------------------------  % Date: Wed, 14 Jun 2000 18:42:18 -0400   From: kuff@tessco.com (Hal Kuff) Subject: Cisco Configs??O Message-ID: <EC14A7B88FBBA478.93303B6167563098.D6A453C0D7290563@lp.airnews.net>h  @    Looking for DECNET/LAT Transparent Bridging configs for CISCOH routers.... anyone out there have a working config they can sanitize and post?R     HK   ------------------------------  + Date: Thu, 15 Jun 2000 06:52:12 -0600 (MDT)0) From: John Nebel <nebel@athena.csdco.com>d Subject: Re: Cisco Configs? F Message-ID: <Pine.OSF.4.21.0006150633001.3270-100000@athena.csdco.com>  F Basically just use bridge-group on the interfaces for bridging and theG bridge command to specify protocol.  Use DEC protocol for compatibilityoI with older DEC bridges.  Bridge priority controls which devices will end rF up as spanning tree roots and in a situation with only a couple or few$ bridges the default will usually do.  J On the below router, DECnet and LAT are being bridged and IP and AppletalkD routed.  You may want to consider routing DECnet which Cisco routers8 with the right firmware (ie Enterprise) are happy to do.   appletalk routing    interface Ethernet0/0 '  ip address x.y.128.129 255.255.255.128f  appletalk cable-range 6-6 6.55I  appletalk zone ana   bridge-group 1w   interface Serial3/1 %  ip address a.b.3.154 255.255.255.252a   appletalk cable-range 5-5 5.176  appletalk zone router  bridge-group 1-   bridge 1 protocol ieee bridge 1 priority 200   
 John Nebel    $ On Wed, 14 Jun 2000, Hal Kuff wrote:   >  > B >    Looking for DECNET/LAT Transparent Bridging configs for CISCOJ > routers.... anyone out there have a working config they can sanitize and > post?a >  >  > HK >  >    ------------------------------  % Date: Thu, 15 Jun 2000 10:33:41 +0100e5 From: "Adrian Lumsden" <A.Lumsden@spamtrap.xdt.co.uk>i+ Subject: Re: CMQ/Dec with a sense of Humor?t/ Message-ID: <8ia9c2$a79$2@newsg1.svr.pol.co.uk>w  4 I remember being told a story about a user trying to4 interrupt some on-going tape operation with repeated3 CTL-Ys. After a few had been typed in to no effect,N9 she got the message: "I'm stopping as quickly as I can!".e   Probably VMS 4.n'ish.*   Adrian   --( Adrian Lumsden, XDT Computer Systems, UK" A dot Lumsden at xdt dot co dot uk      : Terry Marosites <TMarosites@unitedad.com> wrote in messageG news:1137A4A23A51D311B2D600105A1D5213019AEE9B@seantexch.unitedad.com...VF >    My first VMS class was in Orlando Fl. (Near Disney World), Dec at thatH > time had developed a two node cluster with the nodes as snow and white (they B > also had 7 terminal servers, can you name the terminal servers).A > I am sure with info-vax group you can collect some very strange[ cluster nameC > and node name configuration. Even just looking at some of the oldx
 orange and8 > gray wall references there were some fun name in them. >T > -----Original Message-----H > From: Shane.F.Smith@Healthnet.com [mailto:Shane.F.Smith@Healthnet.com]) > Sent: Wednesday, June 14, 2000 10:30 AMI > To: Info-VAX@Mvb.Saic.ComR- > Subject: Re: CMQ/Dec with a sense of Humor?tF > Ok, so it's not a part of VMS, but I have a server here called Doris whosenB > shutdown packet consists of the string "die screaming with sharp	 things in D > your head". It's a reference to a movie title mentioned in the Red Dwarf TV	 > series.l >l > Shanea >a >u >i >d >uA > Phil Tregoning <ptregoni@esoc.esa.de> on 06/14/2000 09:09:56 AMa >y > To:   Info-VAX@Mvb.Saic.Comt > cc:  > . > Subject:  Re: CMQ/Dec with a sense of Humor? >j   ------------------------------  % Date: Thu, 15 Jun 2000 07:34:25 -0400 & From: Ken Robinson <ksrobin@erenj.com>+ Subject: Re: CMQ/Dec with a sense of Humor?m, Message-ID: <39488701.2259.14D557@localhost>  - On 14 Jun 2000, at 19:42, Dan O'Reilly wrote:   H > I don't know if VMS Datatrieve had this message, but Datatrieve-11 hadD > an error message, in response to a user trying "AT MIDDLE OF PAGE  > <dosomething>"L > that was "AT MIDDLE OF PAGE is innovative, but, alas, illegal".  One of myI > all-time faves.  Along with, of course, "LINK-F-Illegal Error" from ther > RT-11 linker...r  B Speaking of Datatrieve... Does anybody remember what you got when A you typed "Help wombat" when you were in Datatrieve and you were  J using a terminal that recognized REGIS graphics???  The last time I tried < the command, at a recent Decus Symposium, it was gone... :-(   Ken Robinson ksrobin@erenj.comM   ------------------------------  % Date: Thu, 15 Jun 2000 14:14:29 +0200l( From: Etienne Vogt <vogt@alice.obspm.fr>+ Subject: Re: CMQ/Dec with a sense of Humor? . Message-ID: <3948C8A5.C1915694@alice.obspm.fr>   Roy Omond wrote:S > I can't remember which system it was (ICL ? Emas ? mid 1970's), but uninitialisedsR > memory was set to xF00l  (spells "fool" !).  Used to make reading dumps a little > bit . > easier.  Must admit, xDEADBEEF is better :-)  O There's a debugging tool for AmigaOS (called MungWall) that fills uninitializedaF memory, free momory and the likes with funny longwords like $DEADBEEF,
 $DEADF00D, $ABADCAFE or $C0DEDBAD.a   -- b# 		Etienne Vogt (vogt @ na.astro.it) ) 		Osservatorio Astronomico di Capodimontee 		Napoli, Italia   ------------------------------  % Date: Thu, 15 Jun 2000 09:16:03 -0400o# From: Jim Agnew <agnew@hsc.vcu.edu>a+ Subject: Re: CMQ/Dec with a sense of Humor? + Message-ID: <3948D713.7D408A80@hsc.vcu.edu>m  D I know, I know!!!  you got a picture of the Datatrieve Wombat... ;-)   jim.   Ken Robinson wrote:  > / > On 14 Jun 2000, at 19:42, Dan O'Reilly wrote:  > J > > I don't know if VMS Datatrieve had this message, but Datatrieve-11 hadE > > an error message, in response to a user trying "AT MIDDLE OF PAGE  > > <dosomething>"N > > that was "AT MIDDLE OF PAGE is innovative, but, alas, illegal".  One of myK > > all-time faves.  Along with, of course, "LINK-F-Illegal Error" from the: > > RT-11 linker..._ > C > Speaking of Datatrieve... Does anybody remember what you got whentB > you typed "Help wombat" when you were in Datatrieve and you wereK > using a terminal that recognized REGIS graphics???  The last time I tried > > the command, at a recent Decus Symposium, it was gone... :-( >  > Ken Robinson > ksrobin@erenj.com-   ------------------------------  # Date: Thu, 15 Jun 2000 13:22:22 GMT^/ From: "Tom Wade" <t.wade@vms.eurokom.ie.nospam>t+ Subject: Re: CMQ/Dec with a sense of Humor? . Message-ID: <iM425.2368$QT5.27051@news.iol.ie>  4 My favorite compiler humor was a C compiler warning:  H "Call me paranoid, but when I see '/*' in a comment block I get worried"   Toma   ------------------------------  + Date: Thu, 15 Jun 2000 08:31:25 -0500 (CDT)i From: JBUTLER@utmem.edu + Subject: Re: CMQ/Dec with a sense of Humor? 1 Message-ID: <01JQMJ2YLJGY9M1E0M@UTMEM1.UTMEM.EDU>e  C 	On the first version of the VAX/VMS DBMS database software, if you09 got a deadlock the error message was something like this:c   	"Have you tried whitefish?"  H I don't know if the error made it from the final field test version into& the production version of DBMS or not.   ------------------------------   Date: 15 Jun 2000 12:55:51 GMTF From: lederman@star.enet.dec.DISABLE-JUNK-EMAIL.com (Bart Z. Lederman)+ Subject: Re: CMQ/Dec with a sense of Humor?n, Message-ID: <8iajon$at2$1@sniff.shr.dec.com>  U In article <39488701.2259.14D557@localhost>, Ken Robinson <ksrobin@erenj.com> writes: C >Speaking of Datatrieve... Does anybody remember what you got when  B >you typed "Help wombat" when you were in Datatrieve and you were K >using a terminal that recognized REGIS graphics???  The last time I tried ,= >the command, at a recent Decus Symposium, it was gone... :-(o  A HELP WOMBAT is still included.  However, I think you're confusingd= it with PLOT WOMBAT, which is what requires a ReGIS terminal.E  0 They're both still present, and both still work.   -- t(  B. Z. Lederman   Personal Opinions Only  8  Posting to a News group does NOT give anyone permission8  to send me advertising by E-mail or put me on a mailing  list of any kind.  5  Please remove the "DISABLE-JUNK-EMAIL" if you have ab5  legitimate reason to E-mail a response to this post.p   ------------------------------    Date: 15 Jun 2000 12:17:04 -05001 From: kaplow_r@eisner.decus.org.mars (Bob Kaplow)a+ Subject: Re: CMQ/Dec with a sense of Humor?o+ Message-ID: <DuvCVSfadwwJ@eisner.decus.org>   u In article <8iajon$at2$1@sniff.shr.dec.com>, lederman@star.enet.dec.DISABLE-JUNK-EMAIL.com (Bart Z. Lederman) writes:0 > W > In article <39488701.2259.14D557@localhost>, Ken Robinson <ksrobin@erenj.com> writes:eD >>Speaking of Datatrieve... Does anybody remember what you got when C >>you typed "Help wombat" when you were in Datatrieve and you were eL >>using a terminal that recognized REGIS graphics???  The last time I tried > >>the command, at a recent Decus Symposium, it was gone... :-( > C > HELP WOMBAT is still included.  However, I think you're confusing ? > it with PLOT WOMBAT, which is what requires a ReGIS terminal.s  ' Wasn't there also HELP WOMBAT ADVANCED?s   	Bob Kaplow	  E SPAM:	spamrecycle@ChooseYourmail.com	uce@ftc.gov	postmaster@127.0.0.1    ------------------------------  % Date: Thu, 15 Jun 2000 11:53:25 -0400D/ From: "Joe H. Gallagher" <dtrwiz@ix.netcom.com>g+ Subject: Re: CMQ/Dec with a sense of Humor?t) Message-ID: <3948FBEC.4F74@ix.netcom.com>c   Ken Robinson wrote:e ><SNIP> C > Speaking of Datatrieve... Does anybody remember what you got whenoB > you typed "Help wombat" when you were in Datatrieve and you wereK > using a terminal that recognized REGIS graphics???  The last time I triedl> > the command, at a recent Decus Symposium, it was gone... :-( >  > Ken Robinson > ksrobin@erenj.com2  5 Perhaps you were thinking of DATATRIEVE's response ton  
 	DTR> help mes   	You are beyond help.    andb   	DTR> help me advanced   	You are not advanced.  A My memory is a little fuzzy on this, but I think Jim Starkey (the C creator of DATATRIEVE) said that Digital management was not amused    by this "feature" of DATATREIVE.   Joe H. Gallagher, Ph. D.$ Former SIG Chair & Newsletter Editor DATATRIEVE/4GL SIG of DECUSm dtrwiz@ix dot netcom dot com" See "The DATATRIEVE Programmer" at2 http://www.geocities.com/SiliconValley/Pines/8958/   ------------------------------  + Date: Thu, 15 Jun 2000 04:38:55 -0700 (PDT)  From: bethomas@qwestonline.com Subject: Computer Prosperity= Message-ID: <200006151141.HAA14834@betelgeuse.concentric.net>h  9 This message was brought to you by Email Platinum v.3.1b.aC For your own copy of Email Platinum v.3.1b contact:1-909-797-5425. c Bruce Thomas (Associate) SkyBiz 2000  Carlsbad, California  USA  Phone  760-438-1067s REPLY TO:  b-thomas@skybiz.com   Dear Business Entrepreneur:C3 You can make a lot of money in your spare-time withO8 your computer.  Would you like to retire this time next 8 year?   SkyBiz 2000 is creating millionaires faster and 9 easier than any other company in history!  Find out how,  ; visit my website listed below..........                    c       /          http://www.companyontheweb.com/be-rich     n   ------------------------------  # Date: Thu, 15 Jun 2000 15:39:44 GMT 9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)sG Subject: DEC Pascal 5.7 Alignment warnings (was: AXP Volatile Warnings)8+ Message-ID: <YfpmBsixB4gz@eisner.decus.org>>  I In article <8iakfk$obe$1@nnrp1.deja.com>, paul_hallam@hotmail.com writes:4G > When compiling code on a V7.1 ALPHA, Pascal 5.7 I am getting warnings  > about unaligned variablesmI > BUT as far as I can see all references are VOLATILE and I don't believes% > I should be getting these warnings.a  1 VOLATILE has nothing at all to do with UNALIGNED.   ) > xmi_rec_type        = [VOLATILE] RECORDr' >                       CASE BOOLEAN OF = >                            TRUE  : (data   : xmi_buf_type); I >                            FALSE : (tpl_no : [POS (0)] xmi_tpl_no_type;tG >                                     rtl    : [POS (8)] xmi_rtl_type);o >                       END; > 1 > xmi_rec_ptr         = [VOLATILE] ^xmi_rec_type;e > $ > reqst_record        : XMI_rec_ptr;$ > rtl                 : xmi_rtl_type >  > ****** > H > WHY, WHEN I COMPILE ON THE V7.1 ALPHA WITH THE PASCAL-V57-073 COMPILER! > DO I GET THE FOLLOWING WARNING.- >  >      reqst_record^.rtl := 1; > .....^E > %PASCAL-W-UNAVOLACC, volatile access appears unaligned, but must be > > aligned at run-time to ensure atomicity and byte granularity > at line number 6421d  F I see nothing to assure the compiler that xmi_rec_type will be aligned at runtime.    ------------------------------  % Date: Thu, 15 Jun 2000 16:40:29 +0100:* From: "Richard Brodie" <R.Brodie@rl.ac.uk>K Subject: Re: DEC Pascal 5.7 Alignment warnings (was: AXP Volatile Warnings)m+ Message-ID: <8iatdb$ttk@newton.cc.rl.ac.uk>o  F "Larry Kilgallen" <Kilgallen@eisner.decus.org.nospam> wrote in message% news:YfpmBsixB4gz@eisner.decus.org...m  3 > VOLATILE has nothing at all to do with UNALIGNED.t  I It seems to me that there might be an interaction: if you have a volatilegL location, for example a status register, that is incorrectly aligned it willE be impossible to read/write it atomically. There is no guarantee that"E reading the top byte then the bottom byte (or vice versa) is the samemK as reading it all at once (which you can't do). Even if it's just plain oldw1 shared memory, similar considerations will apply.    ------------------------------  # Date: Thu, 15 Jun 2000 18:39:42 GMTs9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)tK Subject: Re: DEC Pascal 5.7 Alignment warnings (was: AXP Volatile Warnings)e+ Message-ID: <3gVQQGp5CIgb@eisner.decus.org>i  X In article <8iatdb$ttk@newton.cc.rl.ac.uk>, "Richard Brodie" <R.Brodie@rl.ac.uk> writes: > H > "Larry Kilgallen" <Kilgallen@eisner.decus.org.nospam> wrote in message' > news:YfpmBsixB4gz@eisner.decus.org...c > 4 >> VOLATILE has nothing at all to do with UNALIGNED. > K > It seems to me that there might be an interaction: if you have a volatile)N > location, for example a status register, that is incorrectly aligned it willG > be impossible to read/write it atomically. There is no guarantee that G > reading the top byte then the bottom byte (or vice versa) is the same:M > as reading it all at once (which you can't do). Even if it's just plain old 3 > shared memory, similar considerations will apply.y  F Sorry, what I meant was, the fact that VOLATILE is sprinkled liberallyE through the source code provided will not cure any alignment problemsv* (which I understood to be the assumption).  F I saw _lots_ of VOLATILE attributes and no use of the ALIGN attribute.   ------------------------------  % Date: Thu, 15 Jun 2000 12:40:11 -1300,- From: John Reagan <reagan@hiyall.zko.dec.com>"K Subject: Re: DEC Pascal 5.7 Alignment warnings (was: AXP Volatile Warnings) 2 Message-ID: <3948CEAB.21488B88@hiyall.zko.dec.com>   Richard Brodie wrote:. > H > "Larry Kilgallen" <Kilgallen@eisner.decus.org.nospam> wrote in message' > news:YfpmBsixB4gz@eisner.decus.org...- > 5 > > VOLATILE has nothing at all to do with UNALIGNED.  > K > It seems to me that there might be an interaction: if you have a volatileiN > location, for example a status register, that is incorrectly aligned it willG > be impossible to read/write it atomically. There is no guarantee thatiG > reading the top byte then the bottom byte (or vice versa) is the same M > as reading it all at once (which you can't do). Even if it's just plain oldt3 > shared memory, similar considerations will apply.   D What Larry was trying to say (at least what I think he was trying toC say), is that the VOLATILE attribute does not change the compiler'si@ rules for positioning fields in records.  Since the code used anB explicit POS(8) attribute to position the word field at bit 8 in aE volatile record, the compiler was just saying that it cannot generate.C code to do an unaligned volatile word read/write operation.  If theyH POS(8) was left off, the compiler would have positioned the field at bitG 16 to ensure natural alignment (and then would have able to perform the@! atomic word update as requested).t   -- t John ReaganS Compaq Pascal Project Leader   ------------------------------  # Date: Thu, 15 Jun 2000 19:42:07 GMT-9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)0K Subject: Re: DEC Pascal 5.7 Alignment warnings (was: AXP Volatile Warnings) + Message-ID: <+3o7MO7Su1rS@eisner.decus.org>2  b In article <3948CEAB.21488B88@hiyall.zko.dec.com>, John Reagan <reagan@hiyall.zko.dec.com> writes: > Richard Brodie wrote:  >> nI >> "Larry Kilgallen" <Kilgallen@eisner.decus.org.nospam> wrote in message-( >> news:YfpmBsixB4gz@eisner.decus.org... >> M6 >> > VOLATILE has nothing at all to do with UNALIGNED. >> :L >> It seems to me that there might be an interaction: if you have a volatileO >> location, for example a status register, that is incorrectly aligned it willtH >> be impossible to read/write it atomically. There is no guarantee thatH >> reading the top byte then the bottom byte (or vice versa) is the sameN >> as reading it all at once (which you can't do). Even if it's just plain old4 >> shared memory, similar considerations will apply. > F > What Larry was trying to say (at least what I think he was trying toE > say), is that the VOLATILE attribute does not change the compiler'srB > rules for positioning fields in records.  Since the code used anD > explicit POS(8) attribute to position the word field at bit 8 in aG > volatile record, the compiler was just saying that it cannot generatefE > code to do an unaligned volatile word read/write operation.  If theeJ > POS(8) was left off, the compiler would have positioned the field at bitI > 16 to ensure natural alignment (and then would have able to perform thel# > atomic word update as requested).f  C John is being too kind.  I missed the POS(8).  I deserve credit foroH only one thing -- adding "Pascal" to the title to attract the expert :-)   ------------------------------    Date: 15 Jun 2000 02:50:27 -0500/ From: jlauret@?.chem.sunysb.edu (Jerome LAURET)u" Subject: DPW433 and graphic's card. Message-ID: <39487cb3_2@dilbert.ic.sunysb.edu>  B 	I tried to turn a DPW433 worstation into an OpenVMS Alpha stationK (making join our cluster). It had WindowsNT on it before and we do not haved any use for it so ... H 	It works perfectly as far as I can tell  except that I have a hard timeN to find a graphic's card which works. I have tried the PBGXGC-AA but this cardM cuases an immediate system crash whenever DECwindows starts. I also tried an  P older card (PMAGB-BA) and it is not recongnized by the hardware (i.e. at consoleN level, it tells me that an uknown card is plugged to slot #bla and that I have
 to remove it.l  E 	Does anyboddy know or have some experience with a DPW433 station andaP what kind of graphic's card it can use ? This is not an Open3D issue !!! (I haveD a rather new version of it OPEN3DA049 which includes ZLXp-E* cards).  H 	BTW : I do realize that the DPW433 is an "unsuported" model for VMS butN it's trying hard or having a machine doing nothing and resting on a desk ... I think the choice is clear.    F 	I would appreciate any advices (please, copy your answer to the Email: above) since I am struggling with this for some time now.  	n     -- b6                   Jerome LAURET S.U.N.Y. @ Stony Brook$        ,,,,,      Dept. of Chemistry+       ( o o )     Stony Brook NY 11794-3400 ;   ---m---U---m---------------------------------------------s&   E-mail: jlauret@mail.chem.sunysb.edu<   URL   : http://nucwww.chem.sunysb.edu/jlauret/jlauret.html   ------------------------------  % Date: Thu, 15 Jun 2000 08:29:44 -0400s+ From: Tim Shoppa <shoppa@trailing-edge.com> 3 Subject: Re: Files-11 ODS-2 Readability for FreeBSDM1 Message-ID: <394893F8.6E34ED76@trailing-edge.com>-   Terry Kennedy wrote: >  > mccrobie@my-deja.com writes: > > UnixH > > would be the issue since "container file systems" seem to be a "new" > > idea there.8 > D >   Not really. The vnode device driver has been around for a while: > 
 > DESCRIPTIONwO >      The vnd driver provides block and raw disk interfaces to a regular file.nP >      A vnode disk special file is associated with a regular file with the vnd-P >      config(8) utility.  Once configured, the regular file can be accessed viaN >      the appropriate block and raw special files in ``/dev'' allowing tradi-I >      tional disk operations such as disksetup(8),  newfs(8),  mount(8),  >      fsck(8),  and swapon(8).t > N >   The above is from BSD/OS. The FreeBSD version has the following SCCS identP > string: "@(#)vnconfig.c  8.1 (Berkeley) 12/15/93". So I'd say it's been around > for a while.  C It's all relative :-).  There are other operating systems where the0E idea of a "container file system" has been around since the mid-70's.RB And I'm sure there are much earlier examples that I haven't worked with.s   Tim.   ------------------------------  % Date: Thu, 15 Jun 2000 08:43:40 +0100d5 From: "Chris Casey" <Chris.Casey@HendersonNOSPAM.com>  Subject: Re: Fun VMS Facts?m) Message-ID: <8ia1fi$9i4$1@soap.pipex.net>r   >s  >Since when is "downtime" cheap? >h  H It's very cheap to the consultants who recommended the approach and thenB have to come back in to redesign for resilience and then again for1 performance and then again for upgrades, etc.....b Nice work if you can get it.   ------------------------------  % Date: Thu, 15 Jun 2000 09:30:49 +0100-B From: Andrew Harrison SUNUK Consultancy <andrew.nospam@uk.sun.com> Subject: Re: Fun VMS Facts?c* Message-ID: <39489439.62EB6857@uk.sun.com>   Arne Vajhj wrote:  * > Andrew Harrison SUNUK Consultancy wrote: > > Arne Vajhj wrote:. > > > Andrew Harrison SUNUK Consultancy wrote:A > > > > Why is it fun, OpenVMS vunerabilities to a number of CERTy? > > > > advisories have been fixed by Compaq/Digital in patches,< > > > > released for OpenVMS but no response has been posted > > > > to CERT. > > > >P? > > > > The poster asked for VMS fact's, you don't want fictionv; > > > > in this list and including any CERT/OpenVMS relatedAC > > > > material in the list is going to make it a work of fiction.k > > > 
 > > > ???? > > >a! > > > I can not see the logic in:h > > >gL > > > VMS engineering does not report back to CERT about fixes => using CERT > > > VMS advisories is fictiond > >S: > > So using CERT advisories or the lack of them in a list8 > > interesting "facts" about OpenVMS is pointless isn't > > it.o >@ > ???? > B > You claimed that. I argued that your argumentation does not make0 > any sense. And you think I agree with you ????  8 I claimed that because it is true and demonstrably true.  6 There are advisories for violations like SMURF and POD4 for which there are no OpenVMS responses from Compaq% in the CERT vendor response sections.l  9 This would be OK but scanning the patch reports published 4 by Compaq reveals that OpenVMS was vunerable to some; of these attacks and that Compaq supplied patches for them.r  ; Since the patches were generated in response to the attacksh= any specifically refer to the attacks by name I think that we ; can safely assume that Compaq didn't anticipate that attack + and rush out the patch prior to the attack.i  8 In this light using CERT lists as a basis of any "facts"5 relating to OpenVMS security is pointless because the-/ OpenVMS CERT responses are incomplete and thatsT putting a nice gloss on it.c  @ > > The fact that OpenVMS is vunerable to these attacks and thatA > > you require patches for some of them has now been lost to you 8 > > and you may or may not be vunerable depending on how > > up-to-date your system is. >e+ > Now you have changed your argumentation !w  7 No I havn't, the fact that the CERT reports from CompaqiB are incomplete is verifiable by reading the OpenVMS patch reports.   >i   > ? > First you argued that VMS engineering did not tell CERT, whenu > a bug was fixed. > : > Now you claim that problems on VMS does not get reported > in the first place ! >A  5 Correct on both counts, how many times do you need too; read this. Compaq for a major series of CERT vunerabilitiesa@ has posted no response at all for OpenVMS. They havn't confirmed: that it is vunerable and they havn't then posted that fact. that there is a patch to fix the vunerability.  9 To discover if it is vunerable you have to read the patch 7 reports which then present you with a fix for a problem-5 that if you were using CERT as a guide you would have  though did not exist.e   >r; > That may be true or not true. It does not matter. This ism8 > an official list. SUN employees posting to news-groups7 > about non-registered VMS vulnerabilities is just FUD.e; > If you have the facts, then report it, or keep your mouth  > shot.c ><  4 Look this has been covered exhaustively two or three6 times in the past. For what ever reason Compaq/Digital6 do not routinely post responses to CERT vunerabilities. for OpenVMS that may apply to OpenVMS. In some0 cases OpenVMS is not vunerable, in some it is as9 illustrated by the fact that Compaq have produced patchess for some of the vunerabilities.   ; You can argue the merits of this approach, security throughf7 obscurity vs security through public scrutiny. What youe4 cannot argue is that CERT is a usefull basis for any "facts" about OpenVMS security.m  : CERT is also not an "official list" is simply represents a6 list of vendor responses to known security violations.  5 What is so laughable about your knee jerk response toe7 my posting is that I have heard almost exactly the same 5 responses from NT bigot's. They say, look at the CERTs1 advisories there arn't any NT ones. Guess what MSo3 don't post responses to CERT, they arn't actually aH member.t   Regardsw Andrew Harrison  Enterprise IT Architecte   ------------------------------  # Date: Thu, 15 Jun 2000 09:19:37 GMTo( From: Terry Kennedy <terry@gate.tmk.com> Subject: Re: Fun VMS Facts?s' Message-ID: <Fw6v8p.K3K@spcuna.spc.edu>e  D Andrew Harrison SUNUK Consultancy <andrew.nospam@uk.sun.com> writes:8 > There are advisories for violations like SMURF and POD6 > for which there are no OpenVMS responses from Compaq' > in the CERT vendor response sections.D >T; > This would be OK but scanning the patch reports published 6 > by Compaq reveals that OpenVMS was vunerable to some= > of these attacks and that Compaq supplied patches for them.m  M   Note that those patches are not to OpenVMS, but to a different product. TheoJ VMS operating system does not include a TCP/IP package, though certainly aO large number of sites install Compaq's TCP/IP product. Other sites use a TCP/IPc package from another vendor.  O   Whether or not VMS should include a TCP/IP product as part of the VMS licenset is a different issue...   - 	Terry Kennedy             http://www.tmk.com.5         terry@tmk.com             Jersey City, NJ USA    ------------------------------  % Date: Thu, 15 Jun 2000 11:53:20 +0100b/ From: Nigel Arnot <sysmgr@maxwell.ph.kcl.ac.uk>C Subject: re: Fun VMS Facts? 7 Message-ID: <009EBA48.C3BCB401.24@maxwell.ph.kcl.ac.uk>t  @ > Wayne Sewell <wayne@tachysoft.xxx.044962.killspam.0138> wrote:O > : It contains a section called Ten Ways to Tell if You're an OpenVMS fanatic, Q > : contining items like "You know what happened on November 17, 1858"  "You know.N > : why the 'sho system' display had to have the system uptime field increased! > : beyond 999 days."  And so on.  >  > Here's an example: >  > $ show sysO > VAX/VMS V5.5-2  on node ZEUS  14-JUN-2000 10:03:12.32   Uptime  1312 05:11:508 > " > That's over 3.5 years of uptime. > --    @ Even more routinely long-lived are VMS clusters. To find out, do   $ SHOW CLUSTER/CONTINg   and then ADD FORMEDP  # to show the cluster formation date.u  F Those that don't show 4-JAN-2000 because of Y2K-related organisationalO powerdowns not infrequently  show the date when VMScluster was first installed nO at the site, meaning that the cluster has been continuously available -- acrossr. software and hardware upgrades  -- since then.  O (Irish Railways still takes the prize - clusters didn't *exist*  13 years ago!)    ---------------o  L Not really a VMS fact. but I've heard that microVAX-II chips had the RussianG for "Digital VAX - for those who know to steal the very best" etched in E the silicon! (This was of course when exporting them to the USSR was eA strictly illegal, and the communist bloc's best(?) computers wereH cloned VAX 11/780s)T  V 	Yours, 
 		Nigel Arnot-- 		NRA@MAXWELL.PH.KCL.AC.UK                   I  7 		"In the beginning there was nothing, which exploded."A   ------------------------------  % Date: Thu, 15 Jun 2000 07:41:03 -0400t& From: Ken Robinson <ksrobin@erenj.com> Subject: re: Fun VMS Facts?6+ Message-ID: <3948888F.529.1AE87B@localhost>t  6 On 15 Jun 2000, at 11:53, Nigel Arnot wrote (in part):  -K > (Irish Railways still takes the prize - clusters didn't *exist*  13 yearsB > ago!).  H Actually clusters did exist 13 years ago. I helped install a cluster in . December 1984 that consisted of 3 VAX 11/785s!   Ken Robinson ksrobin@erenj.comn   ------------------------------  # Date: Thu, 15 Jun 2000 13:15:11 GMT@* From: Art Rice <arice.NOcSPAM@ue.itug.org> Subject: Re: Fun VMS Facts?D8 Message-ID: <uglhks0c7kg2ioban32o009h3bb39a382d@4ax.com>  C On 14 Jun 2000 22:57:50 +0200, eplan@kapsch.net (Peter LANGSTOEGER)s wrote:  X >In article <3947f06c$1@news.kapsch.co.at>, eplan@kapsch.net (Peter LANGSTOEGER) writes:I >>Since only the purchase costs are counted. And see below, nobody had tosL >>prove their statements (with numbers, eg. total costs). It's a shame, whatM >>bad marketing (and "everyone does it this way") make to the mind of people:-
 >>INSANITY !!, > I >Oops. I forgot. Just now, our Exchange server (cluster) is down (for theaF >third of fourth time this year) because the "database" got corrupted.M >Restoration normaly requires 5 to 8 hours. No Mailboxes for over 1000 users,bJ >for many hours and not for the first time this year And everyone seems toH >accept this as "it's still the best way" (except me, but I'm one of theJ >few having the mailbox on the VMScluster, so I'm lucky). Downtime costs ?  6 Just proves the old adage, "You get what you pay for."     --  
 Art Rice   ** # Special Data Processing Corporation & --------------------------------------' All opinions expressed are mine and do o% not reflect the views of my employer.g   ------------------------------  # Date: Thu, 15 Jun 2000 13:19:30 GMTa* From: Art Rice <arice.NOcSPAM@ue.itug.org> Subject: Re: Fun VMS Facts? 8 Message-ID: <eilhksgv6u46tmsalkq08hp3bu1cpe4vp2@4ax.com>  3 On Wed, 14 Jun 2000 16:37:22 -0700, Terry Marositesh  <TMarosites@unitedad.com> wrote:   >Bill,M >    You're  preaching to the choir here. We all know that $ for $ VMS is thetM >best buy for a world-class system. I am a VMS bigot and tell all the praises I >of VMS.  But as my managers have stated  "I can buy a dozens of desk top L >system and some m$ software with the swipe of the pen. But a new VMS systemH >has to be approved for capital expenditures and approved by the board."I >This is the fight most of us have, we can tell them that they are saving J >money in the long run and getting a better system. But when a manager canH >stay under budget and gets a fat bonus check because of it, what do youK >think he is going to do. Yep he is going to put money in his pocket. UntilaG >VMS is shouted from the top offices as the OS of choice we will have a, >battle.   r     Been through that before.pE A large S-series Himalaya with all the bells and whistles and NonStopd? SQL is pricey when compared to the "initial" cost of HP systems > running Oracle.  I say "initial" because I have seen the V2500D rebooted three times this week already.  I suppose it'l happen about 15:30 today also.t   Oh, the Tandem K2000?       uptimew&  9:20am  up 325 days, 20:44,  77 users     -- e
 Art Rice   **e# Special Data Processing Corporationn& --------------------------------------' All opinions expressed are mine and do o% not reflect the views of my employer.e   ------------------------------  # Date: Thu, 15 Jun 2000 13:31:48 GMTo/ From: "John Nixon" <jorlnixon@worldnet.att.net>  Subject: Re: Fun VMS Facts?,F Message-ID: <8V425.3211$Xx5.197110@bgtnsc06-news.ops.worldnet.att.net>  C Can you point me to any kind of link that I can use to support thata
 statement.J It may help tip the scales in a very large decision regarding OVMS that we have coming up real soon.h? "Martin Vorlaender" <martin@radiogaga.harz.de> wrote in messages5 news:3947e3fe.524144494f47414741@radiogaga.harz.de...g > jbecker@ui.urban.org wrote:l8 > : I'm putting together an orientation presentation for4 > : internal end users who are new to OpenVMS Alpha. > ...e: > : Here's where you come in. What fun VMS facts would you; > : include in such an orientation? I'm looking for tidbitsg: > : that can be stated briefly and effectively to (in this: > : situation) a general but educated audience. I want the1 > : items to show that VMS is a good place to be.t > J > Rich Marcello just pointed out today in the OpenVMS Directions TeleforumI > that Kevin Mitnick stated the only OS he couldn't break into easily wase > VMS. >l > cu,r
 >   Martin > --F >                        |  Martin Vorlaender  |  VMS & WNT programmer3 >   OpenVMS: When you    |  work: mv@pdv-systeme.de J >   KNOW where you want  |        http://www.pdv-systeme.de/users/martinv/: >   to go today.         |  home: martin@radiogaga.harz.de   ------------------------------   Date: 15 Jun 2000 14:42:46 GMT2 From: mathog@seqaxp.bio.caltech.edu (David Mathog) Subject: Re: Fun VMS Facts? , Message-ID: <8iaq16$87j@gap.cco.caltech.edu>  p In article <3947e3fe.524144494f47414741@radiogaga.harz.de>, martin@radiogaga.harz.de (Martin Vorlaender) writes: >oI >Rich Marcello just pointed out today in the OpenVMS Directions Teleforum)H >that Kevin Mitnick stated the only OS he couldn't break into easily was >VMS.e  I Well duh, pay the man some bucks for the right to the quote and run it aspK an ad.  That's pretty high praise, and more people know who Mitnick is thanr have ever heard of OpenVMS.    David Mathog mathog@seqaxp.bio.caltech.edu ? Manager, sequence analysis facility, biology division, Caltech     ------------------------------  # Date: Thu, 15 Jun 2000 14:44:56 GMT: From: d.webb@mdx.ac.uk Subject: Re: Fun VMS Facts?i) Message-ID: <8iaq52$t1s$1@nnrp1.deja.com>D  * In article <3947cb0e$1@news.kapsch.co.at>,   eplan@kapsch.net wrote:.B > In article <8i8hap$obf$1@nntp9.atl.mindspring.net>, "Randy Park"' <rjpark@mindspring.com.nospaam> writes:a >e > 	*) everyone does it this waye >o    * Eat Sh*t 60 billion flies can't be wrong !    
 David Webb VMS and Unix team leader CCSS Middlesex University    & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------  # Date: Thu, 15 Jun 2000 16:21:57 GMTl= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)e Subject: Re: Fun VMS Facts?b0 Message-ID: <009EBA4C.C2F35647@SendSpamHere.ORG>  B In article <8iaq52$t1s$1@nnrp1.deja.com>, d.webb@mdx.ac.uk writes:+ >In article <3947cb0e$1@news.kapsch.co.at>,o >  eplan@kapsch.net wrote:C >> In article <8i8hap$obf$1@nntp9.atl.mindspring.net>, "Randy Park"p( ><rjpark@mindspring.com.nospaam> writes: >>  >> 	*) everyone does it this way >> >g > + >Eat Sh*t 60 billion flies can't be wrong !   / Hey, now you're treading upon my argument... ;)a   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM    ------------------------------    Date: 15 Jun 2000 11:17:06 -05001 From: kaplow_r@eisner.decus.org.mars (Bob Kaplow)n Subject: Re: GS320/GS160+ Message-ID: <xLhws5+64sML@eisner.decus.org>-  Y In article <FuvE0x.3I2@world.std.com>, "Terry C. Shannon" <shannon@world.std.com> writes:/ > <snip>> >> Work continues apace on Marvel (the 256 CPU follow with EV7E >> technology) and other servers on the Roadmap.  EV7 has interestinguF >> technologies to make for a much strong server.  On-chip L2, on-chipF >> memory controller and on-chip network controller.  Marvel is a much& >> more powerful server than Wildfire. >>E >> By the way, Wildfire is almost 2 years late and under constructioneC >> for 5 so it falls somewhere halfway in your analysis above.  One G >> other tidbit, 39 AlphaServer 8400s ran simulation/verification for 2wE >> years straight prior ot rollout and are still running full tilt oni >> the next box.  G And it looks like it's slipped again. We're one of those 200+ first dayhL orders that were mentioned in the announcement. We're supposed to be gettingJ a GS160-16, the first model to be shipped. It was supposed to be here thisG month. Now we hear next month. Has everything slipped, or the GS160-16s J being pushed back with the GS160-08s that were scheduled to go out a month later?  6 BTW, the Q FINALLY updated the web page at            J http://www.compaq.com/alphaserver/performance/perf_by_name.html to includeL some of the new Wildfire boxes, in response to my note to the webmaster. ButL the first version had messed up values for the GS320. Turns out they overdidL the cut-and-paste, and had the CPU counts as 1, 2, 4, 6, and 8 instead of 1,C 4, 8, 16, and 32. Would be nice if an 8 CPU GS320 had 4.4 times theeJ performance of my GS140, instead of 1.4 :-) Ironically, that GS140 that we@ got last summer, should have been a GS160, but for the delays...   	Bob Kaplow	  E SPAM:	spamrecycle@ChooseYourmail.com	uce@ftc.gov	postmaster@127.0.0.1n   ------------------------------  # Date: Thu, 15 Jun 2000 12:10:00 GMTn0 From: Timothy Stark <sword7@grace.speakeasy.org>5 Subject: Help! VAX emulator on Intel with Hobbyist CD 8 Message-ID: <sI325.186068$MB.3451983@news6.giganews.com>   Hello Folks:  J I have a demo version of Charon-VAX emulator.  I was told by Charon that II can install with OpenVMS Hobbyist CD.  I tried to follow instructions but F something is wrong.  Now I have OpenVMS 7.2 software when I ordered it	 recently.   G It said that check 'show devices' on VAX console prompt (>>>) and typediE 'show devices' but got an error message - ~"?18 INV DGT".  I tried tof4 enter 'help' but it resulted the same message.  :-(   G Does anyone install OpenVMS Hobbyist CD on Charon-VAX successfully?  IfiE so, how do I install it?  Or I will have wait for Hobbyist version of I Charon-VAX, etc?  I hope that Hobbyist version should support Ethernet or H useless product.  Yes, I am waiting for the Hobbyist version for Eternet support right now.  
 Thank you!   -- Tim Stark   -- sC Timothy Stark	<><	Inet: sword7@speakeasy.org, sword7@firesword7.netdJ --------------------------------------------------------------------------F "For God so loved the world, that he gave his only begotten Son, that H whosoever believeth in him should not perish, but have everlasting life.. Amen." -- John 3:16 (King James Version Bible)   ------------------------------    Date: 15 Jun 2000 12:30:16 -05001 From: kaplow_r@eisner.decus.org.mars (Bob Kaplow)rP Subject: Re: Large amount of files in one directory causes very slow performance+ Message-ID: <m2PlOCmdMksB@eisner.decus.org>C  F In article <8hudks$ced$1@nnrp1.deja.com>, sabolich@my-deja.com writes:L > In article <8holi1$vuu$1@rubidium.news.lion-access.net>,"salvi@freeler.nl"M > <s.schrijen@libertel.nl> wrote:> Hi there,>> At this time we have al lot ofhN > files, about 25,000 all in onedirectory on a> VMS 6.2-1H3 2node DSSI clusterH > with Alpha 4100 machines with each 2Gb of> memory and two CPU's.> Both ...   L Breaking things up into smaller directories is the best solution. If left onE one directory, be sure that the unique part of the filename is at the 6 BEGINNING, not the end, to aid the hashing algorithms.  E Also not that upgrading from VMS V6.2-1H3 to the current V7.2-1 gives > significant enhancements in processing very large directories.   	Bob Kaplow	  E SPAM:	spamrecycle@ChooseYourmail.com	uce@ftc.gov	postmaster@127.0.0.1h   ------------------------------  % Date: Thu, 15 Jun 2000 15:27:28 +0000i$ From: Steve.Spires@yellowpages.co.uk Subject: Lock Statistics/ Message-ID: <002568FF.0055017F.00@quegw01.btyp>,   cc:  bcc:= Contact:   Tel: 3063  -  VSSG, 1st Floor, Bridge Street Plaza    Lock Statisticst  K I know it's difficult to answer the sort of question I'm about to pose, but 	 anyway...C  K I have a two-node cluster, both Alpha 8400s with 2 CPUs in each, 4Gb memory K each, all data disks in an ESA10000 with HSZ70s, and all disks RZ1CBs, some 7 JBOD, some RAID5, and the system is loaded with Oracle.b  J Locks. Is there any sort of rule of thumb I can use to say whether lockingM activity is excessive? Is there any kind of upper limit I should be trying tos keep activity below?  L Vague questions I know, but any help, pointers, URLs to relevant information would be a help.  F And yes, I AM working my way through the Tuning manual and Performance
 manuals...   Steve Spires VMS System Manager BT/Yellow Pagest     [Information] -- PostMaster:D This transmission is intended solely for the addressee(s) and may beL confidential. If you are not the named addressee, or if the message has beenP addressed to you in error, you must not read, disclose, reproduce, distribute or use this transmission.  L Delivery of this message to any person other than the named addressee is notH intended in any way to waive confidentiality.  If you have received thisF transmission in error please contact the sender or delete the message.  
 Thank you.   ------------------------------  # Date: Thu, 15 Jun 2000 12:13:59 GMTp0 From: Timothy Stark <sword7@grace.speakeasy.org>" Subject: Looking for Jnet software8 Message-ID: <bM325.186069$MB.3451983@news6.giganews.com>   Hello Folks:  H Since I was Bitnet user at Gallaudet University, I know Jnet software soE well.  I am looking for Jnet software to allow communications between H IBM mainframes (i.e. Hercules-390) and VAX system?  If so, do they offerE hobbyist version?  Yes, I have IBM mainframe emulator (Hercules-390).   
 Thank you!   -- Tim Stark   -- -C Timothy Stark	<><	Inet: sword7@speakeasy.org, sword7@firesword7.neteJ --------------------------------------------------------------------------F "For God so loved the world, that he gave his only begotten Son, that H whosoever believeth in him should not perish, but have everlasting life.. Amen." -- John 3:16 (King James Version Bible)   ------------------------------  % Date: Thu, 15 Jun 2000 11:54:47 -0400w= From: "Stephane Paquin" <stephane.paquin@nospam.ispatnet.com>a Subject: Mail batch log file1 Message-ID: <c%625.4096$227.94893@nnrp1.uunet.ca>s  K Does anyone can suggest a way to mail the log file of a batch at the end of I the batch ? Is it possible to obtain the batch log file name while in theP batch procedure ?r   More...a  K Once in a while, I run a long batch and would like to receive the output in G mail when it finishes. Up to now, I have done it by creating a separate/I procedure submitted at the end of the batch whose sole purpose is to mailKG the log file of the first procedure back to me. I am looking for a more I elegant way. I would at least like to have a general procedure that could J find the name of the current batch log file and submit a general procedure  with the file name as parameter.  ! Any pointers ? Thanks in advance.k   Stephane Paquinn   ------------------------------  % Date: Thu, 15 Jun 2000 11:22:00 -0500 + From: Don Rogstad <Don.Rogstad@dalsemi.com>n  Subject: RE: Mail batch log fileH Message-ID: <EA5FE99F5DEAD311A6CB00805F199992BCA88A@misnts1.dalsemi.com>  (Getting the Log specification from a batch job can be tricky since it depends upon what is specified in the /LOG qualifier on SUBMIT.  If the /LOG switch is not specified, then the log file becomes SYS$LOGIN:<proc-name>.LOG, else parts given to the /LOG switch is used to replace the items above.  C    I have used the following to procedure to get the log file name:C  ) $    My_Proc = F$Environment("PROCEDURE") 6 $    My_Name = F$Parse(My_Proc,,,"NAME","SYNTAX_ONLY")F $    My_Job_Entry = F$GetQui("DISPLAY_JOB","ENTRY_NUMBER",,"THIS_JOB")O $    Log_File = F$GetQui("DISPLAY_ENTRY","LOG_SPECIFICATION","''My_job_entry'")h? $    Log_File = F$Parse(Log_File,"Sys$Login:''My_Name'.Log",,,)d   Hope this helps, Don Rogstad. Dallas Semiconductor   -----Original Message-----B From: Stephane Paquin [mailto:stephane.paquin@nospam.ispatnet.com]& Sent: Thursday, June 15, 2000 10:55 AM To: Info-VAX@Mvb.Saic.Como Subject: Mail batch log file    K Does anyone can suggest a way to mail the log file of a batch at the end ofxI the batch ? Is it possible to obtain the batch log file name while in theb batch procedure ?r   More...V  K Once in a while, I run a long batch and would like to receive the output in G mail when it finishes. Up to now, I have done it by creating a separate I procedure submitted at the end of the batch whose sole purpose is to mailrG the log file of the first procedure back to me. I am looking for a moregI elegant way. I would at least like to have a general procedure that couldrJ find the name of the current batch log file and submit a general procedure  with the file name as parameter.  ! Any pointers ? Thanks in advance.y   Stephane Paquini   ------------------------------  % Date: Thu, 15 Jun 2000 13:53:49 -0300 1 From: "Boyle, Darren" <boyledj@bankofbermuda.com>-  Subject: RE: Mail batch log fileK Message-ID: <9F664D538536D411BD3200508B6FF01A024761@bdant027.bda.bobda.com>n   > ----------B > From: 	Stephane Paquin[SMTP:stephane.paquin@nospam.ispatnet.com]) > Sent: 	Thursday, June 15, 2000 12:54 PMr > To: 	Info-VAX@Mvb.Saic.Com > Subject: 	Mail batch log fileo > J > Does anyone can suggest a way to mail the log file of a batch at the end > ofK > the batch ? Is it possible to obtain the batch log file name while in thee > batch procedure ?  >  > C 	Yes see  HELP LEXICAL F$GETQUI and look for LOG_SPECIFICATION, see ! Examples in help for correct use.,      	 > More...r > J > Once in a while, I run a long batch and would like to receive the output > inI > mail when it finishes. Up to now, I have done it by creating a separateiK > procedure submitted at the end of the batch whose sole purpose is to mail I > the log file of the first procedure back to me. I am looking for a moret > elegant way.   > L There are various ways of redirecting the output either by using the logical: SYS$OUTPUT or having the command in a shell which uses theG @COMMAND_PROCEDURE/OUTPUT=logfile_name.log, then simply mail this file. J It's not so easy to get the whole LOG file output to be mailed because theJ file is still open and the output not completely flushed, again definately not impossible  > > I would at least like to have a general procedure that couldL > find the name of the current batch log file and submit a general procedure" > with the file name as parameter. > I Yes use the F$GETQUI and then have a simple command file submitted as thet job finishes withY  K SUBMIT MAIL_ME_LOGS.COM  /PARAMETER=log_file_name  ! This obtained form the  F$GETQUI above   mail_me_logs.com $ WAIT 00:018 $ MAIL 'P1'  my_vms_account/SUBJECT="Log File For ''P1'" $ EXIT  # > Any pointers ? Thanks in advance.e >  > Stephane Paquinw >  >  - Darren      F **********************************************************************C This message and any files transmitted with it are confidential andtJ may be privileged and/or subject to the provisions of privacy legislation.M They are intended solely for the use of the individual or entity to whom theyeL are addressed. If the reader of this message is not the intended recipient, B please notify the sender immediately and then delete this message.I You are notified that reliance on, disclosure of, distribution or copyingp of this message is prohibited.   Bank of BermudawF **********************************************************************   ------------------------------   Date: 15 Jun 2000 07:38:13 GMT* From: helbig@astro.rug.nl (Phillip Helbig) Subject: patches, graphic cardso. Message-ID: <8ia155$nn9$1@info.service.rug.nl>   I got an email this morning:  % From:   SMTP%"ECO-Queries@compaq.com"eK To:     "OpenVMS Patch/ECO Announcements" <openvms@list.support.compaq.com>- CC:-D Subj:   [openvms] OpenVMS VAXF11X06_071 VAX V7.1 F11BXQP ECO Summary  4 Looks like the patches list is back up in some form.  C I notice there are a couple of new (ALPHA) patches (no, no email onaC this, at least not yet).  One involves a graphics card.  Now, I do 9E minimal graphics stuff and no little about this.  How can I find out d$ what graphics card I have installed?   ------------------------------  # Date: Thu, 15 Jun 2000 14:09:10 GMTd1 From: "Mark D. Jilson" <jilly@clarityconnect.com>i# Subject: Re: patches, graphic cardsd2 Message-ID: <3948E36C.63CEF181@clarityconnect.com>   $ ANA/SYSTEM CLUE CONFIGb  5 See if SDA lists a Gnn: device and what SDA calls it.r   Phillip Helbig wrote:r >  > I got an email this morning: > ' > From:   SMTP%"ECO-Queries@compaq.com"yM > To:     "OpenVMS Patch/ECO Announcements" <openvms@list.support.compaq.com>' > CC:nF > Subj:   [openvms] OpenVMS VAXF11X06_071 VAX V7.1 F11BXQP ECO Summary > 6 > Looks like the patches list is back up in some form. > E > I notice there are a couple of new (ALPHA) patches (no, no email on D > this, at least not yet).  One involves a graphics card.  Now, I doF > minimal graphics stuff and no little about this.  How can I find out& > what graphics card I have installed?   --  D Jilly	- Working from Home in the Chemung River Valley - Lockwood, NY0 	- jilly@clarityconnect.com			- Brett Bodine fan. 	- Mark.Jilson@Compaq.com			- since 1975 or so, 	- http://www.jilly.baka.com               -   ------------------------------   Date: 15 Jun 2000 15:26:45 GMT* From: helbig@astro.rug.nl (Phillip Helbig)# Subject: Re: patches, graphic cards). Message-ID: <8iasjl$2mi$1@info.service.rug.nl>  C In article <3948E36C.63CEF181@clarityconnect.com>, "Mark D. Jilson"@# <jilly@clarityconnect.com> writes: h  
 >$ ANA/SYSTEMr >CLUE CONFIG > 6 >See if SDA lists a Gnn: device and what SDA calls it.   No Gnn:  Closest is:    6000   14 GQA:   13 S3 Trio32/64   What's that?   ------------------------------    Date: 15 Jun 2000 10:10:34 -04004 From: "Robert Deininger" <rdeininger@mindspring.com>> Subject: Re: Proxy problem: Why does node:: work but 0:: fail?* Message-ID: <B56E5C1C-1B12A@165.247.46.84>  I On Wed, Jun 14, 2000 7:04 PM, Lawrence Bleau <bleau@umtof.umd.edu> wrote:d    I >If this is true, though, I should be able to make a copy of NETPROXY.DAT1 intoH >NET$PROXY.DAT .  Also, if this is the sole reason, this (multiple files with aJ >process logical) should not be the case on my other system.  Let me check >the latter...  F Don't forget, there are two commands in AUTHORIZE for looking at proxy info:@ SHOW /PROXYm SHOW /PROXY /OLD  J The first looks in your NET$PROXY file, the second looks in your NETPROXY.  J I don't think the files are identical.  On our cluster, NET$PROXY has nodeF names in the "new" format, with LOCAL:. prefix.  NETPROXY just has theJ "old" decnet-IV style node names.  At least that's the way AUTHORIZE showsD them.  (I can't DIFF the files to be sure, since NET$PROXY is always locked.)  C >On the UMTOF system, NET$PROXY.DAT is in SYS$SYSROOT:[SYSEXE], and< NETPROXY.DATH >is in SYS$COMMON:[SYSEXE], with the process logical NETPROXY defined as8 >SYS$SYSTEM:NETPROXY.DAT .  So, there *is* a difference.    D Ugh.  The files should be in sys$common.  And logicals NET$PROXY and NETPROXYI should be defined /EXEC/SYSTEM via SYLOGICALS.COM at boot time.  Process-bA private versions of these logicals will just cause you confusion,a particularly in the SYSTEM account.  > And you might as well define logicals SYSUAF and RIGHTSLIST in SYLOGICALS.COMC also. Once you do this, AUTHORIZE no longer cares what your defaulteH directory is when you run it.  Organizing these definitions consistentlyH really reduced our goof-ups here, where we have 2 or 3 people "managing" our cluster from time to time.   Just my two cents.   ---------------------------c Robert Deininger rdeininger@mindspring.comt   ------------------------------    Date: 15 Jun 2000 14:00:57 -0500 From: briggs@eisner.decus.orgh7 Subject: Re: Removing/Stripping RFC headers from emailsr+ Message-ID: <tzlpeeCtMfd9@eisner.decus.org>   [ In article <8i8q52$ce7$3@info.service.rug.nl>, helbig@astro.rug.nl (Phillip Helbig) writes:e7 > In article <3947E71D.3ACC1EAE@videotron.ca>, JF Mezeih) > <jfmezei.spamnot@videotron.ca> writes: h > N >>My problem is not processing the RFC header (I already have code for that inP >>another piece of software), but building code that will detect the presence orN >>not of such header in situations where the software gets messages in variousE >>formats. (header at top, no header at all or header at the bottom).6 > ' > Again, why not search for the string u > : >    ================== RFC 822 Headers ================== > 9 > I don't think anyone will type this in by mistake.  :-)1  F On the other hand, they might type it in on purpose.  As you just did.  D That's the kind of paranoia I get into when I write mail filters.  IC don't want to interfere with someone engaging in a discussion aboutxC e-mail which happens to contain embedded examples of e-mail syntax.a   	John Briggs   ------------------------------  % Date: Thu, 15 Jun 2000 10:18:25 -0300U1 From: "Boyle, Darren" <boyledj@bankofbermuda.com>6  Subject: RE: Secondary PasswordsK Message-ID: <9F664D538536D411BD3200508B6FF01A024759@bdant027.bda.bobda.com>c   From Help..0  D      o  To set only the first password and clear the second, specify         /PASSWORD=password.p  :      o  To set both the first and second password, specify)         /PASSWORD=(password1, password2).   B      o  To change the first password without affecting the second,)         specify /PASSWORD=(password, "").I  B      o  To change the second password without affecting the first,)         specify /PASSWORD=("", password).    - Darren   > ----------( > From: 	JHansen[SMTP:jhansen@fnbnd.com]) > Sent: 	Thursday, June 15, 2000 12:12 AMo > To: 	Info-VAX@Mvb.Saic.Com# > Subject: 	Re: Secondary Passwords  > L > No, I was able to enforce the secondary password on my own account ( usingJ > the " MODIFY username  /PASSWORD=("",secondarypwd)." )  I try and log on > andND > I am able to enter my first password, it then prompts for a second > password,aJ > but I have no idea what it is so I am unable to logon. I tried  " MODIFYK > username  /PASSWORD=("",secondarypwd=xxxxx)."  All this does is change mytI > first password.  I found the command to change the secondary password (c > DCL,L > command SET PASSWORD/SECONDARY ) but I can't logon to run this command.  IJ > hope this helps clear up the confusion and thanks for your help,time and > patience.s > JeffB > David J. Dachtera <djesys.nospam@earthlink.net> wrote in message) > news:394842B3.65941B38@earthlink.net...a > > JHansen wrote: > > > L > > > I am trying to set up secondary passwords for "high profile" accounts. > I.L > > > was able to impose a secondary password by using the " MODIFY usernameL > > > /PASSWORD=("",secondarypwd)."  This would enforce a secondary password > but IlI > > > don't know what the secondary password is  and thus I was unable top > login.G > > > I would like to know how to set a password for the secondary when1 > usingj > the & > > > above command if it is possible.K > > > If it is not possible to do this when enforcing a secondary password,i > howeJ > > > can I change this for each user so they can login and then be forced > to% > > > change the secondary password?? I > > > Any other tips, hints or suggestions to secondary passwords will be  > > > appreciated.$ > > > Thanks for your time and help.
 > > > Jeff > >eL > > Sorry, but your explanation appears to be contradictory. You changed the= > > secondary password but don't know what you changed it to?g > >r > > Non capisco. > >  > > -- > > David J. Dachterat > > dba DJE Systemse& > > http://home.earthlink.net/~djesys/ > >r> > > Unofficial Affordable OpenVMS Home Page and Message Board:/ > > http://home.earthlink.net/~djesys/vms/soho/  >  >  >     F **********************************************************************C This message and any files transmitted with it are confidential andFJ may be privileged and/or subject to the provisions of privacy legislation.M They are intended solely for the use of the individual or entity to whom theyaL are addressed. If the reader of this message is not the intended recipient, B please notify the sender immediately and then delete this message.I You are notified that reliance on, disclosure of, distribution or copying  of this message is prohibited.   Bank of BermudacF **********************************************************************   ------------------------------  % Date: Thu, 15 Jun 2000 08:37:00 -0600d= From: Arlen Williams <remove.arlen.williams@remove.sabre.com>   Subject: Re: Secondary Passwords0 Message-ID: <394895AC.15CFBD11@remove.sabre.com>   JHansen wrote: > K > I am trying to set up secondary passwords for "high profile" accounts.  ImH > was able to impose a secondary password by using the " MODIFY usernameN > /PASSWORD=("",secondarypwd)."  This would enforce a secondary password but IL > don't know what the secondary password is  and thus I was unable to login.M > I would like to know how to set a password for the secondary when using theo" > above command if it is possible.K > If it is not possible to do this when enforcing a secondary password, howcI > can I change this for each user so they can login and then be forced ton! > change the secondary password??gE > Any other tips, hints or suggestions to secondary passwords will be  > appreciated.  > Thanks for your time and help. > JeffD The qualifier "/PASSWORD=("",secondarypwd)" will enforce a secondaryH password and set it to "SECONDARYPWD" (or whatever you put as the second= parameter). To get rid of the secondary password you would doaE "/PASSWORD=("")". If you go into help in authorize and look at modifyr /password it will explain this.s   ------------------------------  % Date: Thu, 15 Jun 2000 11:32:22 -0500e# From: "JHansen" <jhansen@fnbnd.com>!  Subject: Re: Secondary Passwords3 Message-ID: <8x725.106$KK1.12294@news.corpcomm.net>d  K Thanks to all who responded.  I am off and running.  Thanks again the help.- JeffH Arlen Williams <remove.arlen.williams@remove.sabre.com> wrote in message* news:394895AC.15CFBD11@remove.sabre.com... > JHansen wrote: > >.J > > I am trying to set up secondary passwords for "high profile" accounts. I-J > > was able to impose a secondary password by using the " MODIFY usernameJ > > /PASSWORD=("",secondarypwd)."  This would enforce a secondary password but I4G > > don't know what the secondary password is  and thus I was unable toh login.K > > I would like to know how to set a password for the secondary when usingl thel$ > > above command if it is possible.I > > If it is not possible to do this when enforcing a secondary password,e howhK > > can I change this for each user so they can login and then be forced tos# > > change the secondary password??eG > > Any other tips, hints or suggestions to secondary passwords will be  > > appreciated." > > Thanks for your time and help. > > JeffF > The qualifier "/PASSWORD=("",secondarypwd)" will enforce a secondaryJ > password and set it to "SECONDARYPWD" (or whatever you put as the second? > parameter). To get rid of the secondary password you would dopG > "/PASSWORD=("")". If you go into help in authorize and look at modify1! > /password it will explain this.D   ------------------------------  % Date: Thu, 15 Jun 2000 11:22:58 +01005- From: Tim Llewellyn <tim.llewellyn@bbc.co.uk>s Subject: Re: Some HW questions) Message-ID: <3948AE82.575B95AC@bbc.co.uk>n   "David J. Dachtera" wrote:   > David A Froble wrote:r > >UQ > > I've recently gotten my hands on a DEC TZ87N.  What tapes does this use?  DLTo > > III? > >s  N Yup, the N variant won't write "older" densities like TZ85,TZ86 and won't read
 TK50/TK70, whereas the TZ87 will.    --l6 Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project0 MedAS at the BBC, Whiteladies Road, Bristol, UK.A Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.ukR  A I speak for myself only and my views in no way represent those ofC MedAS or the BBC.t   ------------------------------  % Date: Thu, 15 Jun 2000 11:51:57 -0500 * From: WILLIAM WEBB <WWEBB1@email.usps.gov>0 Subject: Stealing the best- (was Fun VMS Facts?)- Message-ID: <0033000024101135000002L052*@MHS>t  C =0A> Wayne Sewell <wayne@tachysoft.xxx.044962.killspam.0138> wrote:.H > : It contains a section called Ten Ways to Tell if You're an OpenVMS = fanatic,H > : contining items like "You know what happened on November 17, 1858" =  "YoulH know > : why the 'sho system' display had to have the system uptime fie= ld+ increased > : beyond 999 days."  And so on.  >  > Here's an example: >f > $ show sysH > VAX/VMS V5.5-2  on node ZEUS  14-JUN-2000 10:03:12.32   Uptime  1312 = 05:11:50 >c" > That's over 3.5 years of uptime. > --  @ Even more routinely long-lived are VMS clusters. To find out, do   $ SHOW CLUSTER/CONTINs   and then ADD FORMEDl  # to show the cluster formation date.f  F Those that don't show 4-JAN-2000 because of Y2K-related organisationalH powerdowns not infrequently  show the date when VMScluster was first in= stalled-H at the site, meaning that the cluster has been continuously available -= - across. software and hardware upgrades  -- since then.  H (Irish Railways still takes the prize - clusters didn't *exist*  13 yea= rs ago!)   ---------------   H Not really a VMS fact. but I've heard that microVAX-II chips had the Ru= ssianeH for "Digital VAX - for those who know to steal the very best" etched in=  D the silicon! (This was of course when exporting them to the USSR wasA strictly illegal, and the communist bloc's best(?) computers weres cloned VAX 11/780s)t        Yours,e           Nigel Arnot "           NRA@MAXWELL.PH.KCL.AC.UK  ?           "In the beginning there was nothing, which exploded."   H      There's a picture and blurb about this on page thirteen of the Ope= nVMS      at Twenty book-  8      "VAX...when you care enough to steal the very best-  H      During the Cold War, VAX systems could bot be sold behind the Iron=  H      Curtain.  Recognizing superior technology, technical people cloned=  VAXH      systems in Russia, Hungary and China.  After learning that VAX sys= temsH      were being cloned, DIGITAL had the following words etched on the C= VAXe@      chip, 'VAX...when you care enough to steal the very best.'"  H      And the accompanying picture indicates that they did it in Russian=        using Cyrillic characters.u  *      transliterated it goes something like  @      VAKS- vwui zabateetye dovoyno vorovats nastoyschee loochee.    H      I've still got the .PDF of this and it appears to be one of the ma= nyH      wonderful things that has disappeared in the "Compaq-tion" of the = olde      Digital websites.  @      I'll be glad to put it up on my website if anyone wants it.        WWWebb=   ------------------------------  % Date: Thu, 15 Jun 2000 09:10:22 -0700 % From: Rob Davis <rob-d@ix.netcom.com>A, Subject: System call to get CPU utilization?- Message-ID: <3948FFEE.BF5848D9@ix.netcom.com>?  G Anyone know if there's a way to get the % of CPU utilization for a node*H in a cluster (like what's displayed for MONITOR SYSTEM) via some sort of system call? Thanks.   Robd   ------------------------------  % Date: Thu, 15 Jun 2000 18:32:02 +0200v= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>e0 Subject: Re: System call to get CPU utilization?) Message-ID: <39490502.2F3F06A7@gtech.com>i   Rob Davis wrote:I > Anyone know if there's a way to get the % of CPU utilization for a nodegJ > in a cluster (like what's displayed for MONITOR SYSTEM) via some sort of > system call? Thanks.   The undocumented EXE$GETSPI !*   For examples see:*&   ftp://ftp.hhs.dk/pub/vms/spi/spi.zip   Arne   ------------------------------  % Date: Thu, 15 Jun 2000 12:28:49 -0400e" From: Dan Sugalski <dan@sidhe.org>0 Subject: Re: System call to get CPU utilization?: Message-ID: <4.3.2.7.0.20000615122746.00ea7b40@24.8.96.48>  + At 09:10 AM 6/15/00 -0700, Rob Davis wrote:oH >Anyone know if there's a way to get the % of CPU utilization for a nodeI >in a cluster (like what's displayed for MONITOR SYSTEM) via some sort ofy >system call? Thanks.h  K The only way to do it is to call the undocumented EXE$GETSPI service twice p@ and do the math. The wizard has docs on this somewhere, I think.   					Dan  I --------------------------------------"it's like this"------------------- 2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and even ;                                       teddy bears get drunkj   ------------------------------  # Date: Thu, 15 Jun 2000 16:42:53 GMTo= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)G0 Subject: Re: System call to get CPU utilization?0 Message-ID: <009EBA4F.AF9887EE@SendSpamHere.ORG>  U In article <3948FFEE.BF5848D9@ix.netcom.com>, Rob Davis <rob-d@ix.netcom.com> writes:oH >Anyone know if there's a way to get the % of CPU utilization for a nodeI >in a cluster (like what's displayed for MONITOR SYSTEM) via some sort ofs >system call? Thanks.   4 You can use the "undocumented" EXE$GETSPI routine.    K You need to call this twice to get the stats necessary to compute the % CPUN utilization (busy time).  & The item lists to pass will look like:  ) SPI_ITEM_LIST1: .WORD   8*4,GETSPI$_MODESt+                 .ADDRESS INITIAL_MODE_TICKSt                 .LONG   0[2]   INITIAL_MODE_TICKS:i                 .LONG   0i ICPU_ID:        .BYTE   0a IINTERRUPT:     .LONG   0  IMP_SYNCH:      .LONG   0m IKERNEL:        .LONG   0- IEXECUTIVE:     .LONG   0a ISUPERVISOR:    .LONG   0T IUSER:          .LONG   0  ICOMPATIBILITY: .LONG   0R IIDLE:          .LONG   0   ) SPI_ITEM_LIST2: .WORD   8*4,GETSPI$_MODES>#                 .ADDRESS MODE_TICKSr                 .LONG   0[2]   MODE_TICKS:a                 .LONG   0  CPU_ID:         .BYTE   0F INTERRUPT:      .LONG   0  MP_SYNCH:       .LONG   0l KERNEL:         .LONG   0i EXECUTIVE:      .LONG   03 SUPERVISOR:     .LONG   0  USER:           .LONG   0- COMPATIBILITY:  .LONG   00 IDLE:           .LONG   0     I Then you need to "correct" a few items by subtracting the IDLE times fromdI the INTERRUPT times.  Then, compute the differences of each item...  for dI example, SUBL2 IKERNEL,KERNEL.  Then compute the total time by adding theIH elements in the vector to get the total time.  You then compute the per-6 centages by dividing the item times by the total time.   Simple, Eh?e --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMc   ------------------------------  % Date: Thu, 15 Jun 2000 18:25:27 +0200o= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>h( Subject: Techwise report on availability) Message-ID: <39490377.99DC4FE5@gtech.com>h  6 Have you read the Techwise report on availability that is on www.openvms.compaq.com ?   It looks pretty good !  7 But what I wonder about is: how independent & reputable  is this company Techwise ?   Arne   ------------------------------  # Date: Thu, 15 Jun 2000 12:29:27 GMTo9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)R Subject: Re: VAX on Intel?+ Message-ID: <1tHuZ3PZVE2Z@eisner.decus.org>o  g In article <39483ACA.41D9DA15@earthlink.net>, "David J. Dachtera" <djesys.nospam@earthlink.net> writes:i > Larry Kilgallen wrote: > [snip]& >> So another tactic will be required. >  > Such as...  I It is your crusade (with which I disagree), so you will have to innovate.oA The stated problem is that you have (thus far) failed to convince D those in control at Compaq (although Steve Hoffman seems convinced).  A Since you also have not convinced me, it is clear I am not a goode( source of tactics for convincing Compaq.   ------------------------------  # Date: Thu, 15 Jun 2000 12:22:48 GMTg. From: Peter Moreton <petermoreton@my-deja.com> Subject: Re: VAX on Intel?) Message-ID: <8iahqb$mh4$1@nnrp1.deja.com>-  D I think it is interesting to consider that right now, the Charon VAXG emulator gives about 10 VUPS on a 1Ghz PC. One doesn't have to look toogC far along the performance curve to see a future Intel chip (itaniummC maybe) in an 8-way configuration, possibly offering maybe 160 VUP's<D upward, and serious I/O. It might not be long before the fastest VAX ever, is Charon!   -- Peter Moreton, Northamptonshire,J UK.e    - In article <394124EC.EF9BB5A3@tsoft-inc.com>, -   David A Froble <davef@tsoft-inc.com> wrote:I > Peter LANGSTOEGER wrote: > >>> > > So, this niche product surely wasn't made for me. Sorry... >pF > No, it wasn't.  You've got to expect that the original market was/is the veryG > old VAXs, probably embedded, that are still running out there.  Partst and6B > maintenance for such are probably getting real hard to find, and
 expensive. > C > Still, regardless of the actual value in terms of performance and  avoiding MS,B > there is some real possibilities with regard to exposure of VMS, marketing, andA > such.  Many people are impressed by technological inovation andI tricks.  This,E > if mentioned often enough, might help with the public perception ofp VMS.
 > Imagine: >dF > Trade rag guy:  What's all this talk about VMS?  It's DEAD!  Haven't
 you heard? >eD > VMS bigot with VMS running on shirt pocket PC>  Dead?  Then what's
 running on> > this? (as shirt pocket PC is shoved in trade rag guy's face) >1G > Trade rag guy, usually impressed by things like shirt pocket PC, getso a newtC > perspective on VMS, and possibly mentions it in his next article.W >=A > So, while shirt pocket PC running VMS is in most cases useless,e (maybe real ? > cheap/small replacement for embedded MVII), it still can be a  positive thing for/ > VMS, if only from a perception point of view.s >rE > Me, I'd like to see a copy on every PC Compaq delivers.  I'll never 
 use it forD > any real VMS work, I have more VAXs and Alphas than one person can
 reasonablyD > use.  But I'm thinking about a copy on my notebook, just for those
 times someoneA > tells me VMS is dead.: >e > Dave >  > --6 > David Froble                       Tel: 724-529-04506 > Dave Froble Enterprises, Inc.      Fax: 724-529-0596@ > DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com8 > T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486 >d    & Sent via Deja.com http://www.deja.com/ Before you buy.e   ------------------------------  % Date: Thu, 15 Jun 2000 11:13:14 -0400.* From: Chuck Chopp <ChuckChopp@rtfmcsi.com> Subject: Re: VAX on Intel?+ Message-ID: <3948F289.F711493F@rtfmcsi.com>    Peter Moreton wrote:  F > I think it is interesting to consider that right now, the Charon VAXI > emulator gives about 10 VUPS on a 1Ghz PC. One doesn't have to look too E > far along the performance curve to see a future Intel chip (itaniumeE > maybe) in an 8-way configuration, possibly offering maybe 160 VUP's F > upward, and serious I/O. It might not be long before the fastest VAX > ever, is Charon! >t > -- > Peter Moreton, > Northamptonshire,M > UK.s >a  F That brings up an interesting question.  Would the emulator need to beJ modified to present the multiple Intel CPU's as multiple VAX CPU's capableC of SMP as part of the hardware emulation?  Or, would it simply be aaF multithreaded application w/respect to consuming resources on multipleG Intel CPU's while only emulating a uniprocessor VAX system?  If SMP was J going to be emulated, what would be the best VAX hardware model to emulateF so that OpenVMS would properly handle the SMP features of the emulated	 hardware?n     -- Chuck Choppr  8 ChuckChopp@rtfmcsi.com            http://www.rtfmcsi.com0                                   ICQ # 22321532@ RTFM Consulting Services Inc.     864 801 2795 voice & voicemail2 103 Autumn Hill Road              864 801 2774 fax4 Greer, SC  29651                  800 400 4935 pagerC                                   8004004935@alphapage.airtouch.com    ------------------------------  % Date: Thu, 15 Jun 2000 02:01:30 -0400 ' From: "Bill Todd" <billtodd@foo.mv.com>hD Subject: Re: Which cpu does Sun server use: I386, Mip, Ppc or Alpha?( Message-ID: <8i9r96$slk$1@pyrite.mv.net>  E Andrew Harrison SUNUK Consultancy <andrew.nospam@uk.sun.com> wrote int, message news:3947B2A3.E7D4D027@uk.sun.com... > Bill Todd wrote: >cI > > Andrew Harrison SUNUK Consultancy <andrew.nospam@uk.sun.com> wrote inn0 > > message news:39477A39.7740A67E@uk.sun.com... > >s > > ...i > > L > > Or not.  VMS already supports locality ('RAD' may be the Wildfire term - I'd F > > have to check) to help keep accesses local to a single board in anG > > application-transparent manner.  Since a single board can support 4*K > > processors (each of which will enjoy the process and generational speed-H > > improvements of the Alpha road map over time, including extension toJ > > simultaneous multi-threading) and up to 32 GB of on-board memory, this isH > > not as significant a restriction as you suggest:  in fact, only very *few* H > > monolithic applications can really benefit much from more than 4-wayD > > multi-processing (especially if each processor supports multiple interleavedRE > > threads at the hardware level) and more than 32 GB of memory, and E > > non-monolithic applications are relatively easy to segregate ontoi multiple> > > boards that need to interact only relatively infrequently. > >h > : > Firstly, other NUMA OS's like Dynix for example have the@ > same sort of code in the OS to try and encourage locality, theG > Sequents also have local hardware assist to cache heavily used pages.tA > But even with this support many people using these systems have A > found that it is necessary to modify their applications to helpe@ > the OS. In Sequents case running Oracle Parallel Server in the9 > box, partitioning the DBMS across nodes within the sameD > OS instance has been tried.e  K I guess it in part depends on what you think systems like Wildfire are goodD for.  K They will be excellent at running large numbers of applications in a singleiL box ('server consolidation'), where almost every individual application willG run just fine in a 4-Alpha/32GB board (typically sharing the board withw$ other similarly-happy applications).  C They will also be excellent at running the occasional humongous butiA fairly-readily-partitionable application, where the definition oflE 'partitionable' is considerably less strict than with non-shared-disk.J databases (where the impact of accessing something in another partition isL at least 3+/- decimal orders of magnitude greater latency rather than a mere@ factor of 3).  Most such applications are already designed to beH partitioned, precisely because architectures that can successfully marryH more than at most dozen processors to shared UMA memory are, let us say,J rare (some would say non-existent - which may be how, looking ahead to SMTJ Alphas, the QBB designers decided that 4 was the right number of Alphas to place on each board).r  G Since with larger numbers of processors (especially SMT processors) thenJ aggregate processing bandwidth tends to swamp the shared memory bandwidth,L efficient use of both resources demands partitioning anyway, though having aG small off-board access penalty rather than a large one may skew optimal"G design enough to be worth some tuning at some future time.  So the realDL question is whether 32 GB of memory is sufficient to keep large applicationsF (and humongous but reasonably partitioned applications) from having toK access off-board memory frequently - though the 32GB/board limitation seems-I likely to be able to be increased as memory densities increase over time.i   >w? > Even with this level of tuning and with the support of the OS  > 80% locality is high.B  G Certainly not in the case of something like OPS, which was designed for H cluster-style environments in which off-node operations take hundreds ofH microseconds (or more) rather than hundreds of nanoseconds:  if anythingD like 20% of OPS data accesses were inter-node, it just wouldn't run.  K Which is why Oracle continues to pursue a hybrid strategy with OPS (and whytE the nay-sayers who assert that shared-disk clustering "doesn't scale"HH haven't a clue what they are talking about):  it *always* makes sense to> partition where it's efficient to do so (even in a shared-diskK architecture) - the times you have nodes cooperate in accessing common data K are precisely those times when a single node can't handle the load (and the)J data resists partitioning), making the two approaches complementary ratherB than competitive (but only VMS - and Parallel Sysplex - offer suchJ flexibility in their native, non-Oracle, environments, and one could argueK that of these two only the VMS lock manager really allows one to capitalizewI on the locking efficiency of partitionable operations, since the ParalleleA Sysplex 'coupling facility' is a far more centralized mechanism).    >h; > Secondly large DBMS applications will easily consume more3: > than 4 CPU's worth of resource making it necessary to go: > of node with things like the SGA being a chunk of memory3 > which all nodes may need to have equal access to.n  I I tend to agree that 4 EV67 Alphas may be less than the optimal number totG harness to the memory on a single board, though I question whether it'sfK anything like the disadvantage you seem to be suggesting - again, we'll let,L the benchmarks decide, rather than debate opinions.  Over time I suspect, asC I said above, that 4 SMT Alphas per board will be just about ideal.e   > 7 > Thirdly the latencies quoted by Compaq in their paper-6 > for the WildFire are "idle" latencies, I assume that5 > these are for an unloaded system and if that is theV8 > case then the latency of a node under load or off node9 > access under load is likely to be worse than the values  > that have been quoted. >l >n > > C > > So on-board latency will indeed likely dominate overall average  application K > > latency (80+% may well be a very conservative estimate for the on-boardeD > > fraction), and IIRC there's a copy of VMS on each board to avoid	 excessiveaI > > system off-board code (and presumably pure data) access.  Given these"L > > considerations, plus Alpha's respectable caching facilities, I'd suggest you H > > not get too complacent about Sun's next-generation competitiveness - evenJ > > ignoring the fact that creation of the Wildfire follow-on is also well underh > > way. > >r >gK > I can't really comment on Sun's next generation servers but consider this A > example. The HP N4000 which occupies the 1-8 CPU space which istF > also occupied by the GS80 has an Idle latency of 130ns, true it does@ > not have the same throughput as WildFire but it does have muchG > lower latency. There is very little actual technical detail availablev@ > on HP's SuperDome but I would suggest that they will be aimingE > for sub 200ns local idle access because that is what they currentlyg% > are able to support with the N4000.i  D As an external observer I haven't any information about why WildfireL latencies are what they are.  Had I to guess, I would guess that the need toJ support interaction with the 'Global Switch' (or simply with a second QBB,I in the two-QBB configuration) has a deleterious effect upon local latency H through the local on-board switch (since DEC's engineers weren't usuallyC slouches at hardware design), and that scaling to a second level ofaK switching in SuperDome might well encounter a similar penalty.  Bet someone-3 here knows, though they may not be willing to talk.1  L An interesting sidelight (which I have no idea how to quantify) is that, forC sufficiently parallel workloads (scientific and database - or other0D server-style - applications, for example) and a reasonable number ofI processors harnessed to a given chunk of memory (which is where SMT help,'J given the current QBB configuration), latency becomes far less relevant toJ overall system throughput and bandwidth becomes dominant (though you stillK want to maintain reasonable on-board locality if Global Switch bandwidth isu8 significantly lower than the aggregate local bandwidth).   >jJ > This is the sort of memory latency you should be looking at for WildFireG > otherwise you will have huge issues delivering the levels of ILP thateA > the CPU's are capable and moving forward Alpha will support SMTo1 > which will only increase the demand for memory.M  K Actually, my guess is that SMT *reduces* the demand for low-latency memory,iI at least in environments exhibiting any significant degree of parallelismTL (for the throughput reason noted above).  And Alpha's deep pre-fetch abilityJ tends to mesh well with such situations (though speculative pre-fetch doesJ place even more demands on memory bandwidth - again making it desirable to# keep accesses local when possible).    > E > > > Just for your information the Origin 2000 a 3-4 year old system-? > > > had at launch 310ns local and 950ns remote as illustrateda< > > > in the white paper SGI published for the origin intro. > > >t/ > > > http://www.sgi.com/origin/images/isca.pdf8 > > >nE > > > Secondly the bisectional bandwidth of the WildFire is about thei@ > > > same as the E10K, yes Compaq have published higher STREAMSB > > > numbers but these included local on-node memory access whichB > > > assumes that a process never goes off node. This is unlikelyB > > > except for a very small subset of applications that WildFire > > > will support.c > >gE > > As pointed out above, this is in fact likely for the overwhelming  majorityK > > of applications Wildfire will support, and the majority of the rest are L > > easily partitionable to meet this criterion.  Leaving aside the fact for forAA > > *really* monstrous applications, neither E10K nor Wildfire is  sufficient:TH > > you've got to expand into clustering, and Sun really doesn't want to take$ > > this competitive argument there. > >  >T> > Sure you can partition applictions and people using OPS in a< > single Sequent box are examples of this. Would you do this> > however if the alternative was a similarly sized system with: > a more uniform memory architecture which did not require= > you to make these changes. I think on the basis of Sequents 1 > experience in this are the answer has to be no.:  H I'm not familiar with Sequent's experience, so have no way to judge yourJ conclusions based on it (and being a software type, wouldn't have completeL confidence in my judgement in such areas anyway).  Perhaps someone else here* can weigh in with a knowledgeable opinion.  K I do suspect that monolithic applications that can saturate the performancenF capabilities of a single QBB (hence will either have less-than-optimalL performance or require partitioning work - beyond work already performed dueJ to the constraints of other environments - to achieve optimal performance)L will be present in a relatively small percentage Wildfire installations (notK to say that this small percentage won't include some important sites), justuI as they are likely present in a relatively small percentage of Sun ExxxxxhL installations and IBM S80 installations.  Such applications tend to push theL limits of commercial hardware (from any vendor - and hardware from differentH vendors often exhibits bottlenecks in different areas) but usually don'tG define its design center, and as a result tend to be pretty flexible ingE allowing configuration flexibility to get the best performance out ofn- whatever environment they find themselves in.g  A So I'm still not convinced that Wildfire suffers from any notableaI deficiencies of the sort you've suggested:  to my (relatively uninformed)]K eye, it appears to be a well-balanced design (given the expectation of SMT,aI anyway).  But benchmarks would convince me, if you can provide persuasivel numbers.   - bill   > 	 > regards  > Andrew Harrison  > Enterprise IT Architectu >S >e >f   ------------------------------  % Date: Thu, 15 Jun 2000 00:57:06 -0500 ' From: Bill Bradford <mrbill@mrbill.net>o/ Subject: WTB: VAX-bootable external SCSI CD-ROM-- Message-ID: <20000615005706.I6262@mrbill.net>   B I'm looking for a RRD42, RRD43, etc - whatever I can get cheapest,D fastest, and will let me install VMS on a VAXstation 4000-VLC, as it9 looks like my Plextor CD-ROM isnt going to be compatible.s   Thanks.U   Bill   -- .* +--------------------+-------------------+* |   Bill Bradford    |   Austin, Texas   |* +--------------------+-------------------+* | mrbill@sunhelp.org | mrbill@mrbill.net |* +--------------------+-------------------+   ------------------------------  # Date: Thu, 15 Jun 2000 08:44:54 GMTs( From: Terry Kennedy <terry@gate.tmk.com>3 Subject: Re: WTB: VAX-bootable external SCSI CD-ROM-' Message-ID: <Fw6tMu.ICM@spcuna.spc.edu>t  ) Bill Bradford <mrbill@mrbill.net> writes:aD > I'm looking for a RRD42, RRD43, etc - whatever I can get cheapest,F > fastest, and will let me install VMS on a VAXstation 4000-VLC, as it; > looks like my Plextor CD-ROM isnt going to be compatible.r  E   Which Plextor? I have the 32x tray version, and it lists VMS as onewI of the guaranteed compatible operating systems (you need to set the blocka1 size switch to configure it for 512-byte blocks).a  - 	Terry Kennedy             http://www.tmk.coma5         terry@tmk.com             Jersey City, NJ USAh   ------------------------------    Date: 15 Jun 2000 13:52:56 +0100T From: pmoreau@cenaath.cena.dgac.fr (Patrick MOREAU, CENA Athis, Tel: 01.69.57.64.40)3 Subject: Re: WTB: VAX-bootable external SCSI CD-ROMs! Message-ID: <ER96fspKJp3t@gaelic>s  6 In article <Fw6tMu.ICM@spcuna.spc.edu>, Terry Kennedy  <terry@gate.tmk.com> writes:  G >   Which Plextor? I have the 32x tray version, and it lists VMS as one K > of the guaranteed compatible operating systems (you need to set the blocka3 > size switch to configure it for 512-byte blocks).c  H Yes, there is no problems for booting VMS CD (at least on Alpha) with a > Plextor 32x (I've just installed VMS 7.1-2 with such a drive).   Patrick  -- cO ===============================================================================aO pmoreau@cena.dgac.fr  (CENA)     ______      ___   _           (Patrick MOREAU) 4 moreau_p@decus.fr (DECUS)       / /   /     / /|  /|J CENA/Athis-Mons FRANCE         / /___/     / / | / |   __   __   __   __  N BP 205                        / /         / /  |/  |  |  | |__| |__  |__| |  |N 94542 ORLY AEROGARE CEDEX    / /   ::    / /       |  |__| | \  |__  |  | |__|N http://www.ath.cena.fr/~pmoreau/            http://www.multimania.com/pmoreau/O ===============================================================================    ------------------------------   Date: 15 Jun 2000 16:07:52 GMT- From: ba600@FreeNet.Carleton.CA (Mike Kenzie)t! Subject: WTD: VAX parts in OTTAWAs/ Message-ID: <8iav0o$85o$1@freenet9.carleton.ca>   : I'm looking for a BCC08 cable to connect a serial terminal OR3 a vr290 display and a BC18Z-10 cable to connect it.e  > Also interested in SCSI adapter and Ethernet for VAXstation II  > I've checked with Computer recyclres and they didn't have any., I contacted Compac and they haven't replied.  C Does anyone have any OLD VAXen siting in storage they want removed?    ------------------------------   End of INFO-VAX 2000.333 ************************