1 INFO-VAX	Tue, 05 Nov 2002	Volume 2002 : Issue 611       Contents:
 Re: <None>
 Re: <None>3 Re: A valid point why VMS is superior to other OS's  alpha generation AST Restriction??  Re: AST Restriction?? + Re: can vms C access the keyboard directly? + Re: can vms C access the keyboard directly? + Re: can vms C access the keyboard directly?  Re: CHARON-ALPHA - soon ???? Re: CHARON-ALPHA - soon ????5 Re: Configuring IP (was: Re: Configuring OpenVMS-6.2) ! Re: Configuring VMS as DNS server ! Re: Configuring VMS as DNS server ! Re: Configuring VMS as DNS server ! Re: Configuring VMS as DNS server D DECtalk A Guide To Voice 1985 First Printing EB 26375 56/06 23 100.0H Re: DECtalk A Guide To Voice 1985 First Printing EB 26375 56/06 23 100.0 Re: Digital Press discount Re: Disk size specification  Re: Disk size specification  Re: Disk size specification  Re: Disk size specification ' Re: Does anyone use P7 in DCPS$STARTUP? ' Re: Does anyone use P7 in DCPS$STARTUP? ' RE: Does anyone use P7 in DCPS$STARTUP? ' Re: Does anyone use P7 in DCPS$STARTUP? ( Re: DS10L and backups on an IDE CDRW ...( Re: DS10L and backups on an IDE CDRW ...( Re: DS10L and backups on an IDE CDRW ... Re: errno equivalentH FA: OLD DEC VAX Circuit Card - cheap - NO RESERVE - ebay ends 11/08/2002 Re: HP Advocacy Site1 HP World News says You Can't Crack It or Crash It 1 HP World News says You Can't Crack It or Crash It 5 Re: HP World News says You Can't Crack It or Crash It 5 RE: HP World News says You Can't Crack It or Crash It 5 Re: HP World News says You Can't Crack It or Crash It 5 Re: HP World News says You Can't Crack It or Crash It 5 RE: HP World News says You Can't Crack It or Crash It 5 Re: HP World News says You Can't Crack It or Crash It 5 Re: HP World News says You Can't Crack It or Crash It  Re: HP-ETS 2002 CD Re: HP-ETS 2002 CD Re: HP-ETS 2002 CD0 Re: import/export between oracle 8.1.7 and 8.0.37 Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek 7 Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek 7 Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek 7 Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek 7 Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek 7 Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek  Karma   Microsoft strikes (itself) again$ Re: Microsoft strikes (itself) again$ Re: Microsoft strikes (itself) again$ RE: Microsoft strikes (itself) again$ Re: Microsoft strikes (itself) again Monitoring NIC utilization Re: Monitoring NIC utilization Re: Monitoring NIC utilization OpenVMS & XDMCP  Re: OpenVMS & XDMCP  Re: OpenVMS & XDMCP  RE: OpenVMS & XDMCP  Re: OpenVMS & XDMCP 0 Re: OpenVMS printing to printer via windows 2000 Re: Palladium article ( Problem with /AUTOSTART_ON on VMS V7.3-1 Re: Remote Console access  Re: Remote Console access  Setting NIC to 100mb/Full  Re: Setting NIC to 100mb/Full  Re: Setting NIC to 100mb/Full   Re: Slow performance with BACKUP  Re: Slow performance with BACKUP  Re: Slow performance with BACKUP  Re: Slow performance with BACKUP smp license per card?  TCP/IP Services NFS tuning The Leper colony at ETS  Re: The Leper colony at ETS ( Re: VMS on Itanium not mission critical?( Re: VMS on Itanium not mission critical?( Re: VMS on Itanium not mission critical?( Re: VMS on Itanium not mission critical?( Re: VMS on Itanium not mission critical?( Re: VMS on Itanium not mission critical?( Re: VMS on Itanium not mission critical?7 Re: VMS to support KDE over CDE in which future version 7 Re: VMS to support KDE over CDE in which future version $ Re: Was OpenVMS left out on purpose?$ Re: Was OpenVMS left out on purpose?$ Re: Was OpenVMS left out on purpose?$ Re: Was OpenVMS left out on purpose?$ Re: Was OpenVMS left out on purpose? What Ethertype do I use? Re: What Ethertype do I use? Re: What Ethertype do I use?# Re: [very OT] OS X, the (bad?) turn   F ----------------------------------------------------------------------   Date: 4 Nov 02 20:51:56 +0100 ) From: p_sture@elias.decus.ch (Paul Sture)  Subject: Re: <None> ) Message-ID: <S9RcJ9BSfub4@elias.decus.ch>   g In article <3DC5A058.C2E0C7AE@blueyonder.co.uk>, Tim Llewellyn <tim.llewellyn@blueyonder.co.uk> writes:  >  >  > Paul Sture wrote:  >>  a >> In article <OFF8CE8C7A.5FE268DA-ON85256C62.006C3047@metso.com>, norm.raphael@metso.com writes:  >> > >> > Good catch!J >> > That might be a useful line to add at the end of the procedure  ;-) . >>  A >> Given that VUPOMETER has a date prior to Alphas, I don't think  >> it's valid any more.  >>  L >> Here is another I picked up a couple of years ago. It was pretty accurateG >> when applied to my Vaxstation 3100, and seems reasonably good on the G >> Alphas I have tried. I modified it then to do an Alpha/VAX test, but 4 >> I'll warn you now it is not Itanium compliant :-) >>   > B > It says my VMS7.2 running on ts10 on a PII 300 MHz is 1 VUP :-). > H > I've been thinking to myself "this is about as fast as a MicroVAX II",3 > good to have some semi-quantitative confirmation.  > A Now I'm not sure about the difference in the code between VAX and ? Alpha in the example I posted. The newsgroup posting I got that < procedure from is dated 1996, so maybe the figures were more& accurate for models current back then.   --  
 Paul Sture Switzerland    ------------------------------  $ Date: Mon, 4 Nov 2002 16:03:50 -0500 From: norm.raphael@metso.com Subject: Re: <None> ? Message-ID: <OFE7D5206E.0F2F121E-ON85256C67.0073705C@metso.com>   6 As this thread well-confirms, no figures are accurate.H I found that running on an unloaded system (Only one user logged in to a priviledged G account and nothing else using cycles in the background) gives a larger 
 value than on an active system.H I was a customer of a software company that tried to build a sizing tool some years back just to K figure out which VAX to recommend with it's product.  They gave up, because 	 it always ( recommended the "biggest" VAX available.        = p_sture@elias.decus.ch (Paul Sture) on 11/04/2002 02:51:56 PM   5 Please respond to p_sture@elias.decus.ch (Paul Sture)    To:    Info-VAX@Mvb.Saic.Com cc:  Subject:    Re: <None>    > In article <3DC5A058.C2E0C7AE@blueyonder.co.uk>, Tim Llewellyn( <tim.llewellyn@blueyonder.co.uk> writes: >  >  > Paul Sture wrote:  >>B >> In article <OFF8CE8C7A.5FE268DA-ON85256C62.006C3047@metso.com>, norm.raphael@metso.com writes: >> > >> > Good catch!J >> > That might be a useful line to add at the end of the procedure  ;-) . >>A >> Given that VUPOMETER has a date prior to Alphas, I don't think  >> it's valid any more.  >>C >> Here is another I picked up a couple of years ago. It was pretty  accurateG >> when applied to my Vaxstation 3100, and seems reasonably good on the G >> Alphas I have tried. I modified it then to do an Alpha/VAX test, but 4 >> I'll warn you now it is not Itanium compliant :-) >> > B > It says my VMS7.2 running on ts10 on a PII 300 MHz is 1 VUP :-). > H > I've been thinking to myself "this is about as fast as a MicroVAX II",3 > good to have some semi-quantitative confirmation.  > A Now I'm not sure about the difference in the code between VAX and ? Alpha in the example I posted. The newsgroup posting I got that < procedure from is dated 1996, so maybe the figures were more& accurate for models current back then.   --
 Paul Sture Switzerland    ------------------------------  % Date: Mon, 04 Nov 2002 21:50:59 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> < Subject: Re: A valid point why VMS is superior to other OS's' Message-ID: <3DC74023.26359F4F@fsi.net>    Bob Ceculski wrote:  > E > I was arguing VMS superiority to linux on their board when a poster B > made the following excellent point, except he gave a link at theC > end that I can't find ... anyone know what that is all about "why D > vms wasn't ported to intel before", and wouldn't this be a problem? > for the itanium vms port? Itanium cannot be mission critical?   F Well stated as simply as my peanut-brain will allow, IA32 and the mostH common IA32 mobo architectures (if you can call 'em that) lacks\ much ofB what VMS requires in terms of interrupts and lots of other stuff I cannot hope to understand.  F Prior to the IPF edict, it was never worth DEC's while to work out theA kinks and make it work, like they did with PALcode to make up for  inherent shortfalls in Alpha.   A Now that HP has a degree of control over IPF design, it can cause E suitable design features to find their way into IPF so Intel does the  work instead of OVMS Engr.  , Mobo.'s, of course, remain a separate issue.  H Terribly inaccurate, but in a nutshell, that encompasses the bulk of the- "not on IA32" excuses^H^H^H^H^H^H^Harguments.    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  $ Date: Tue, 5 Nov 2002 03:02:18 +01001 From: "T.R." <tr303@notmail.com.replace.n.with.h>  Subject: alpha generation 3 Message-ID: <JwFx9.34447$I6.3407450@zwoll1.home.nl>   I Right now I'm assembling an Alpha system based on the AXPPci 33 board. It H will be housed in a generic clone AT minitower. To make it look a littleL better, I'd like to put an Alpha sticker on it. Is there a place where I can? get 'Alpha generation' or other Digital/Alpha related stickers?    :-)  -- T    ------------------------------  % Date: Mon, 04 Nov 2002 20:26:27 +0000 - From: Gerald Marsh <gerald@cyfer.demon.co.uk>  Subject: AST Restriction??8 Message-ID: <u0ldsuk5nj1sobka01h9f68ub9g3jf6l7q@4ax.com>   Heelllppppp!  D We are having real hassle with a Compaq BASIC program calling an AST routine written in BLISS.   E The following ACCVIO is produced when another routine is being called E from within the AST routine. The instruction suggests that either R27 $ or the stack pointer is corrupted...  D 11:45:59.48-BALGV-AFTMQO--%SYSTEM-F-ACCVIO, access violation, reason? mask=04, virtual address=00000000029C19A0, PC=000000000003ACB8,  PS=0000001B D 11:45:59.51-BALGV-AFTMQO--  Improperly handled condition, image exit forced. : 11:45:59.54-BALGV-AFTMQO--    Signal arguments:   Number = 0000000000000005: 11:45:59.58-BALGV-AFTMQO--                        Name   = 000000000000000C 11:45:59.61-BALGV-AFTMQO-- 0000000000000004 11:45:59.64-BALGV-AFTMQO-- 00000000029C19A0 11:45:59.68-BALGV-AFTMQO-- 000000000003ACB8 11:45:59.71-BALGV-AFTMQO-- 000000000000001B, 11:45:59.76-BALGV-AFTMQO--    Register dump:; 11:45:59.79-BALGV-AFTMQO--    R0  = 0000000000000065  R1  =  819764C010000FF0 R2   = 0000000000011660; 11:45:59.82-BALGV-AFTMQO--    R3  = 00000000029A00A0  R4  =  00000000029CBB40 R5   = 0000000000000000; 11:45:59.85-BALGV-AFTMQO--    R6  = FFFFFFFFFFFFFFFF  R7  =  00000000029A0110 R8   = 0000000000000000; 11:45:59.89-BALGV-AFTMQO--    R9  = 0000000000000000  R10 =  00000000029A0128 R1 1 = 000000007BC01110; 11:45:59.92-BALGV-AFTMQO--    R12 = 0000000000000001  R13 =  FFFFFFFF8D44FCD0 R1 4 = FFFFFFFF81ACB240; 11:45:59.95-BALGV-AFTMQO--    R15 = 000000007BC010E0  R16 =  00000000029C5A68 R1 7 = 00000000029C5A68; 11:45:59.99-BALGV-AFTMQO--    R18 = 000000000000002C  R19 =  0000000000000001 R2 0 = FFFFFFFF8E1DA000; 11:46:00.02-BALGV-AFTMQO--    R21 = FFFFFFFDFF7FE000  R22 =  FFFFFFFF80109240 R2 3 = 000000007FFA1FC0; 11:46:00.05-BALGV-AFTMQO--    R24 = 0000000000000000  R25 =  000000000000002F R2 6 = 000000000003B034; 11:46:00.09-BALGV-AFTMQO--    R27 = 0000000000011620  R28 =  FFFFFFFF800C6328 R2 9 = FFFFFFFF8003BEC8; 11:46:00.13-BALGV-AFTMQO--    SP  = 000000007AF12000  PC  =  000000000003ACB8 PS   = 000000000000001BA =================================================================   F It is interesting to note that the ACCVIO occurs as the called routine is being entered.   F The following is the AST routine. The ACCVIO occurs at the line marked with "++++"   & ROUTINE SUBSERVER_MAILBOX_AST:NOVALUE= BEGIN  LOCAL   STATUS,          GETLIST:VECTOR[4],,         SENDER_PROCESS_NAME:VECTOR[16,BYTE],@         SUBSERVER_DESC:VECTOR[2] INITIAL(0,SENDER_PROCESS_NAME),         SUBSERVER_INDEX1,          SUBSERVER_INDEX,         SUBSERVER_INDEX2, (         SUBSERVER_MESSAGE_DESC:VECTOR[2]% INITIAL(0,SUBSERVER_MAILBOX_MESSAGE),          CONNECTION_INDEX, $         QUEUE_INFO:REF BLOCK[,BYTE];@ BIND    SUBSERVER_MESSAGE_LENGTH=SUBSERVER_MAILBOX_IOSB[1]:WORD,0         SUBSERVER_PID=SUBSERVER_MAILBOX_IOSB[2],+         SENDER_SIZE=SUBSERVER_DESC[0]:WORD;  EXTERNAL         ROUTINE DELAY;3 WRITE_MESSAGE ('Subserver mailbox ast  delivered'); & IF NOT .SUBSERVER_MAILBOX_IOSB[0] THEN
         BEGIN 6         WRITE_MESSAGE('SUBSERVER MAILBOX READ ERROR');,         MESSAGE(.SUBSERVER_MAILBOX_IOSB[0]);         RETURN;          END;5 GETLIST[0]=16+JPI$_PRCNAM^16;           !PROCESS NAME  GETLIST[1]=SENDER_PROCESS_NAME;  GETLIST[2]=SENDER_SIZE; 
 GETLIST[3]=0; 5 STATUS=$GETJPIW(PIDADR=SUBSERVER_PID,ITMLST=GETLIST); 
 CHECK_STATUS; 4 WRITE_MESSAGE('***SUBSERVER_MAILBOX_AST STEP 1***'); INCR I FROM 1 TO 100 DO          DUMMY_ROUTINE(); !SUBSERVER_INDEX=0; 8 ++++SUBSERVER_INDEX=GET_SUBSERVER_INDEX(SUBSERVER_DESC);4 WRITE_MESSAGE('***SUBSERVER_MAILBOX_AST STEP 2***');3 SUBSERVER_MESSAGE_DESC[0]=.SUBSERVER_MESSAGE_LENGTH     -SUBSERVER_MAILBOX_HEADER_SIZE; - CONNECTION_INDEX=.SUBSERVER_CONNECTION_INDEX; 3 QUEUE_INFO=SUBSERVER_QUEUE_INFO[.CONNECTION_INDEX]; . WRITE_VALUE('SUB MBX AST', .CONNECTION_INDEX);4 WRITE_DESC ('SUB MBX AST',  SUBSERVER_MESSAGE_DESC);9 ENQUEUE_COMM_MESSAGE(.QUEUE_INFO,SUBSERVER_MESSAGE_DESC); C ENQUEUE_COMM_MESSAGE(.QUEUE_INFO,SUBSERVER_LIST[.SUBSERVER_INDEX]);  STATUS=$SETEF(EFN=19);
 CHECK_STATUS; $ STATUS=$QIO(CHAN=.SUBSERVER_MAILBOX,         FUNC=IO$_READVBLK,%         ASTADR=SUBSERVER_MAILBOX_AST, $         IOSB=SUBSERVER_MAILBOX_IOSB,$         P1=SUBSERVER_MAILBOX_BUFFER,$         P2=SUBSERVER_MAILBOX_LIMIT);
 CHECK_STATUS; ' WRITE_MESSAGE('SUBSERVER_MAILBIX_AST');  $waitfr(efn=17); END;    !SUBSERVER_MAILBOX_AST  T ====================================================================================  @ Any other routine called from within the AST code results in the ACCVIO.   B (The original worked fine for years before we upgraded MQueue from3 2.2.1.1 to 5.1 but no MQ routine seems implicated!)   D If the code for GET_SUBSERVER_INDEX is placed "in-line", the routine works...    + ROUTINE SUBSERVER_MAILBOX_AST_AFTS:NOVALUE=  BEGIN  LOCAL   STATUS,          GETLIST:VECTOR[4],,         SENDER_PROCESS_NAME:VECTOR[16,BYTE],@         SUBSERVER_DESC:VECTOR[2] INITIAL(0,SENDER_PROCESS_NAME),         SUBSERVER_INDEX,(         SUBSERVER_MESSAGE_DESC:VECTOR[2]% INITIAL(0,SUBSERVER_MAILBOX_MESSAGE),          CONNECTION_INDEX, $         QUEUE_INFO:REF BLOCK[,BYTE];@ BIND    SUBSERVER_MESSAGE_LENGTH=SUBSERVER_MAILBOX_IOSB[1]:WORD,0         SUBSERVER_PID=SUBSERVER_MAILBOX_IOSB[2],+         SENDER_SIZE=SUBSERVER_DESC[0]:WORD; 3 WRITE_MESSAGE ('Subserver mailbox ast  delivered'); & IF NOT .SUBSERVER_MAILBOX_IOSB[0] THEN
         BEGIN 6         WRITE_MESSAGE('SUBSERVER MAILBOX READ ERROR');,         MESSAGE(.SUBSERVER_MAILBOX_IOSB[0]);         RETURN;          END;5 GETLIST[0]=16+JPI$_PRCNAM^16;           !PROCESS NAME  GETLIST[1]=SENDER_PROCESS_NAME;  GETLIST[2]=SENDER_SIZE; 
 GETLIST[3]=0; 5 STATUS=$GETJPIW(PIDADR=SUBSERVER_PID,ITMLST=GETLIST); 
 CHECK_STATUS; 7 SUBSERVER_INDEX=(INCR I FROM 0 TO .SUBSERVER_COUNT-1 DO D         IF STR$COMPARE(SUBSERVER_DESC,SUBSERVER_LIST[.I]) EQL 0 THEN                 EXITLOOP .I);  IF .SUBSERVER_INDEX LSS 0 THEN BEGIN 2         IF .SUBSERVER_COUNT GEQ MAX_SUBSERVER THEN
         BEGIN 4                 WRITE_VALUE('MAX SUBSERVER EXCEEDED: ',.SUBSERVER_COUNT);         END;)         SUBSERVER_INDEX=.SUBSERVER_COUNT; E         STR$COPY_DX(SUBSERVER_LIST[.SUBSERVER_COUNT],SUBSERVER_DESC); +         SUBSERVER_COUNT=.SUBSERVER_COUNT+1;  END;: !SUBSERVER_INDEX=GET_SUBSERVER_INDEX_AFTS(SUBSERVER_DESC); !SUBSERVER_INDEX=0; > !STR$COPY_DX(SUBSERVER_LIST[.SUBSERVER_COUNT],SUBSERVER_DESC);3 SUBSERVER_MESSAGE_DESC[0]=.SUBSERVER_MESSAGE_LENGTH     -SUBSERVER_MAILBOX_HEADER_SIZE; - CONNECTION_INDEX=.SUBSERVER_CONNECTION_INDEX; 3 QUEUE_INFO=SUBSERVER_QUEUE_INFO[.CONNECTION_INDEX]; . WRITE_VALUE('SUB MBX AST', .CONNECTION_INDEX);4 WRITE_DESC ('SUB MBX AST',  SUBSERVER_MESSAGE_DESC);9 ENQUEUE_COMM_MESSAGE(.QUEUE_INFO,SUBSERVER_MESSAGE_DESC); C ENQUEUE_COMM_MESSAGE(.QUEUE_INFO,SUBSERVER_LIST[.SUBSERVER_INDEX]);  STATUS=$SETEF(EFN=19);
 CHECK_STATUS; $ STATUS=$QIO(CHAN=.SUBSERVER_MAILBOX,         FUNC=IO$_READVBLK,*         ASTADR=SUBSERVER_MAILBOX_AST_AFTS,$         IOSB=SUBSERVER_MAILBOX_IOSB,$         P1=SUBSERVER_MAILBOX_BUFFER,$         P2=SUBSERVER_MAILBOX_LIMIT);
 CHECK_STATUS;  END;    !SUBSERVER_MAILBOX_AST  F What concerns me is that there is still an underlying problem and it's" going to come up and bite us soon!  F Can anyone see if there is anything suspect regarding the AST routine?  @ Note that, to overcome a known hang problem, the PTHREADS rtl is= explicitly linked in the LINK command. (This version of MQ is 1 multi-threaded but the application image is not.)   B We have even tried to upgrade to the latest Bliss compiler (1.9, I2 think) to no avail. The OpenVMS version is V7.2-2.  @ IBM have suggested a quota problem but massive increases haven't helped.     E Any help to keep this high-availability system highly available would  be gratefully received.   C In the meantime, keep up the good work  - The threads in this group % are much better than the rtl threads!    Bye for now,     Gerald.  Gerald Marsh  / gerald -at- cyfer -dot- demon -dot- co -dot- uk    ------------------------------  % Date: Mon, 04 Nov 2002 21:29:46 -0500 2 From: rdeininger@mindspring.com (Robert Deininger) Subject: Re: AST Restriction??K Message-ID: <rdeininger-0411022129460001@1cust216.tnt2.nashua.nh.da.uu.net>   8 In article <u0ldsuk5nj1sobka01h9f68ub9g3jf6l7q@4ax.com>, gerald@cyfer.demon.co.uk wrote:   
 >Heelllppppp!  > E >We are having real hassle with a Compaq BASIC program calling an AST  >routine written in BLISS.   > F >The following ACCVIO is produced when another routine is being calledF >from within the AST routine. The instruction suggests that either R27% >or the stack pointer is corrupted...   G Why do you suspect this has to do with AST?  Does the same code work if I it's in the main thread of execution?   (Not to say it's NOT AST-related; ! just curious about the evidence.)   I The BASIC code isn't shown.  Of course you will need to be careful around 3 synchronization of any data shared between threads.   : I don't like to depend on register dumps if I can help it.  F Can you compile the module(s) /DEBUG/NOOPT, and link with /DEBUG/MAP ?  ? Even if you run /NODEBUG, you should get a symbolic stack dump.   J Even better would be to step through the program with the debugger and see  exactly where you are ACCVIOing.  F Note that the problem might "go away" when you run with the debugger. H Uninitialized stack variables _change_ when the debugger takes over, andJ symptoms can change as well.  There are other classes of bug that try hard to evade the debugger as well.     > E >11:45:59.48-BALGV-AFTMQO--%SYSTEM-F-ACCVIO, access violation, reason @ >mask=04, virtual address=00000000029C19A0, PC=000000000003ACB8, >PS=0000001B  ' You are trying to read address 29C19A0.    (register dump removed)    > G >It is interesting to note that the ACCVIO occurs as the called routine  >is being entered. > G >The following is the AST routine. The ACCVIO occurs at the line marked  >with "++++"  H How do you know where the ACCVIO occurs?  From the reported PC, with theJ compiler listings and link map?  Or from the embedded WRITE_MESSAGE calls?  G I don't see the called GET_SUBSERVER_INDEX routine anywhere.  Maybe you $ are ACCVIOing _inside_ that routine?  H It takes me more than a few minutes to digest a chunk of BLISS.  I don't see the cause at first glance.  C I would approach this with the VMS debugger.  If the symptom evades D detection under debugger control, an image dump (SET PROCESS/DUMP inF recent versions of VMS) can be analyzed with the debugger, bug you can! only see the moment of the error.        > ' >ROUTINE SUBSERVER_MAILBOX_AST:NOVALUE=  >BEGIN >LOCAL   STATUS, >        GETLIST:VECTOR[4], - >        SENDER_PROCESS_NAME:VECTOR[16,BYTE], A >        SUBSERVER_DESC:VECTOR[2] INITIAL(0,SENDER_PROCESS_NAME),    ...   9 >++++SUBSERVER_INDEX=GET_SUBSERVER_INDEX(SUBSERVER_DESC);   H You are passing the address of SUBSERVER_DESC to the routine.  What doesJ it do with it?  How does it declare the parameter?  What does it return?    A >Any other routine called from within the AST code results in the  >ACCVIO.  D Strange.  Even a dummy routine with no parameters that just returns?    C >(The original worked fine for years before we upgraded MQueue from 4 >2.2.1.1 to 5.1 but no MQ routine seems implicated!)  G I recently found a bug is a chunk of BLISS that has been working for 25 A years (literally!).  An uninited string descriptor on the stack.  J Apparently I was the first person to trip over it.  BLISS is almost as bad- as C for hiding latent errors in plain sight.   E >If the code for GET_SUBSERVER_INDEX is placed "in-line", the routine 	 >works...   : So maybe there's something wrong in the calling mechanism?    G >What concerns me is that there is still an underlying problem and it's # >going to come up and bite us soon!   	 Probably.   G >Can anyone see if there is anything suspect regarding the AST routine?    Not so far.   A >Note that, to overcome a known hang problem, the PTHREADS rtl is > >explicitly linked in the LINK command. (This version of MQ is2 >multi-threaded but the application image is not.) > C >We have even tried to upgrade to the latest Bliss compiler (1.9, I 3 >think) to no avail. The OpenVMS version is V7.2-2.     A >IBM have suggested a quota problem but massive increases haven't  >helped.  0 ACCVIO doesn't sound like a quota problem to me.  D Maybe the debugger will help, or someone with sharper eyes will spot something...   ------------------------------   Date: 4 Nov 02 20:37:49 +0100 ) From: p_sture@elias.decus.ch (Paul Sture) 4 Subject: Re: can vms C access the keyboard directly?) Message-ID: <HlYOtTW0IP+D@elias.decus.ch>   R In article <us8o5lfggq965@corp.supernews.com>, Z  <zarlenga@conan.ids.net> writes:, > Paul Sture <p_sture@elias.decus.ch> wrote:I > :> On Unix you set the tt control with ioctl() and then call getchar().  > :> That's hard?  > / > : For someone who hasn't done it before, yes.  > H > : I've just had a swift look at "man ioctl" and while I could probablyF > : figure it out eventually, it would be swifter to scour the net for# > : an example, or ask a newsgroup.  > H > IMO, the following is MUCH easier reading than the 1 or 2 chapters oneM > must read in the I/O User's Guide to properly use $QIO for single-character 	 > input :  >  > P > IOCTL(2)                  FreeBSD System Calls Manual                 IOCTL(2)   <snip>  @ And, I'd personallly go for the manual every time. It's just the> way I like to work. Indeed when I am doing something I haven't5 done before, I prefer to work with hard copy manuals.    --  
 Paul Sture Switzerland    ------------------------------   Date: 4 Nov 02 21:49:03 +0100 ) From: p_sture@elias.decus.ch (Paul Sture) 4 Subject: Re: can vms C access the keyboard directly?) Message-ID: <rBFz+oyYSLNE@elias.decus.ch>   T In article <5P1ujMua5X2y@eisner.encompasserve.org>, briggs@encompasserve.org writes:T > In article <us8o5lfggq965@corp.supernews.com>, Z  <zarlenga@conan.ids.net> writes:- >> Paul Sture <p_sture@elias.decus.ch> wrote: J >> :> On Unix you set the tt control with ioctl() and then call getchar(). >> :> That's hard? >>  0 >> : For someone who hasn't done it before, yes. >>  I >> : I've just had a swift look at "man ioctl" and while I could probably G >> : figure it out eventually, it would be swifter to scour the net for $ >> : an example, or ask a newsgroup. >>  I >> IMO, the following is MUCH easier reading than the 1 or 2 chapters one N >> must read in the I/O User's Guide to properly use $QIO for single-character
 >> input : >  > ROFL > I > You expect someone to be able to write an ioctl call based on _THAT_!?!  >     That was rather my reaction too.   F > At least have the grace to quote a page from termio (7I) if you want@ > to show something that someone has a prayer of coding against. > / > What I got from the quoted man page was that:  > E > The previous poster already knows what to do, how to do it and what H > side effects to expect in the context of $QIO.  And that's nine tenths > of the battle. >   E In actual fact, I recall the whole exercise took me 3 or so days, not D as a dedicated task (my earlier reference to finding somewhere quiet; to read the treeware was a serious factor on that project).   C The hard part came about by doing it from COBOL rather than FORTRAN 9 or Macro, _not_ by the complexity of the manuals or $QIO.    B > Yes, I'm sure that a competent Unix programmer could whip out an? > ioctl for single character I/O quite easily and would find it E > quite difficult to slog through the $QIO documentation to the point A > that he or she could perform a similar task under VMS.  Indeed, @ > that's rather the previous poster's point:  It's easier to useC > what you're used to than to get yourself up to speed on something  > new.  = That was indeed the intention. The other point is that once I ? had invested the time to "slog through the $QIO documentation", @ that knowledge stayed put, and I could do it again in a fraction? of the time, (and have done so). It was an excellent investment  of my time.    --  
 Paul Sture Switzerland    ------------------------------   Date: 4 Nov 2002 22:05:46 GMT # From: hoff@hp.nospam (Hoff Hoffman) 4 Subject: Re: can vms C access the keyboard directly?* Message-ID: <aq6qvq$k3k$1@web1.cup.hp.com>  ` In article <80d4088b.0210310921.76a9b7fb@posting.google.com>, aronoffs@hotmail.com (stu) writes:  D :I need to be able to access the vax vms keyboard directly so that IA :can capture key presses as they occur, including the arrow keys.   K   There are several hidden assumptions lurking within this innocent-looking I   and apparently simple statement, in terms of the keyboard response, the I   processing of the input record terminator, and that various keys return    multiple characters.  $   I'll work some of these through...  E   Different keyboards can and do provide different data for seemingly E   the same keypress.  While this has certainly gotten better over the H   years, you can and do still see disagreements over certain characters.G   For example, the escape key and the delete/backspace key.  The former I   is usually assumed to indicate a control sequence on OpenVMS, but there H   are also applications (qv: emacs) that assume and even require that itJ   occur in isolation.  In the case of the delete and backspace processing,B   UNIX systems tend to map this key differently than does OpenVMS.  J   SMG provides a mechanism -- please see the FAQ -- to look up the controlH   sequences that can be generated.  I prefer to use this for processing I   these sequences if for no other reason than to avoid having to rebuild  J   if I have a different terminal -- and folks that manage the SMG databaseH   correctly can "fix" the application themselves.  The whole of SMG doesE   not need to be hauled in to use the SMG escape and control sequence 7   dictionary.  But yes, you can use SMG for this too...   H   Now as for display and keyboard management, others have pointed to theI   ability of the SMG RTL to handle the display for you.  C and curses can H   also do this for you as well, but the two curses APIs that are presentG   in OpenVMS are rather more limited than is that of SMG.  (I'd tend to G   switch to or to use the BSD curses starting with OpenVMS V7.0, vs the '   older OpenVMS curses implementation.)   H   And as for curses, if you really want to do single character I/O usingJ   getchar or getc, you can use [no]crmode and the curses getch call -- perH   the C documentation, single character input on UNIX (without the input0   of a return) is enabled via crmode and cbreak.  D   As for key-presses, the numeric keypad has differing responses forI   keypresses depending on the terminal mode, and various of the keys can  J   return `n' or (usually?) `n+1' characters depending on when the keyboardL   is operating in its eight-bit or in seven bit modes.  Eight-bit operations1   usually mean fewer characters in the sequences.   I   As for I/O, UNIX tends to be character-oriented, while OpenVMS operates I   with record-oriented assumptions.  Put another way, UNIX usually passes K   the characters along largely unmolested and leaves most of the processing K   to the upper layers.  With OpenVMS, the original design attempts to clump G   data together into groups of characters -- records -- for reasons of  H   performance.  (This is where the terminator stuff creeps in -- OpenVMSJ   hardware and software tends to buffer input until it sees a terminator.)G   The OpenVMS theory is that blocks or groups or packets of data can be H   processed more quickly and more efficiently by the I/O system than canI   single character I/O.  DMA, for instance, has less system overhead than J   does character I/O.  (Now does the difference in I/O processing overheadG   actually matter to you?  Donno.  That's up to the amount of processor F   cycles and the amount of I/O hardware you can throw at the problem.   G   Character I/O chews up the host processor, while smarter I/O hardware J   is more expensive but can offload the host -- dumb I/O widgets are oftenI   far cheaper (to design, build, and buy), but the relative `dumbness' of F   these widgets usually comes at a cost of host processor cycles.  ButH   I digress....  For examples of the two approaches, see the discussionsJ   of clustering over a LAN or Memory Channel controller vs clustering overJ   a CI controller -- the former are "stupid" and fast I/O widgets, but theG   host processor is definitely involved in the I/O.  The CI widgets are L   far "smarter", and can accordingly more easily offload the host processor.  H   I've been involved in both ends of this hardware discussion, and thereJ   are be trade-offs due to available board real-estate, part costs, designG   costs, and the target price point (expected selling price).  There is J   seldom One True and Right Answer here, though the target price point is    usually the winning criteria.   I   As for the portability of this stuff, I've seen systems with character  G   sets requiring five, six, seven, eight, and nine bits for characters. G   And then there are the differences in endian-ness, a valued absorbant #   of superfluous network bandwidth.   E :When you use the getchar() library routine (which advertises that it E :gets a character from stdin), it actually doesn't get any characters G :until the user has pressed the ENTER key.  And, any arrow keys pressed  :are filtered out of stdin.   D   My local LK421 keyboard has no "enter" key, that key is overloadedC   onto the return key -- the LK46W does have both the enter and the F   return.  (See what I mean? :-)  But seriously, yes, this is expectedF   behaviour, and due to the OpenVMS interest in chunking the I/O data.    2 :Is there a way for a C program to do what I need?  A   Sure.  I've got code on the Freeware that does exactly this at:   +     http://www.openvms.compaq.com/freeware/ 1       freeware50/srh_examples/qio_extended_read.c   D   and I've added a discussion and some pointers into the OpenVMS FAQA   on Escape sequences (control sequences) and related processing. B   The SMG materials referenced in the FAQ and the OpenVMS terminalA   driver documentation in the I/O User's Reference Manual are two A   of the resources you will want to become familiar with.  Things =   get rather more interesting if you want to grab ^Y, ^C, or  A   (depending on the serial line) ^P, as DCL, DCL, and the console *   hardware can get involved, respectively.  C   Single-character I/O has been discussed before, and I expect that D   a search of the newsgroup archives will identify previous threads.    N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com0   ------------------------------  % Date: Mon, 04 Nov 2002 21:06:22 -0600h1 From: "David J. Dachtera" <djesys.nospam@fsi.net> % Subject: Re: CHARON-ALPHA - soon ?????' Message-ID: <3DC735AE.2FAEB6EF@fsi.net>e   Robert Boers wrote:c > ; > We are building the Alpha to IPF binary code translator. i  $ Any chance of a VAX->IPF translator?   -- R David J. Dachtera? dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/S   ------------------------------  % Date: Mon, 04 Nov 2002 23:30:40 -0500r* From: "Stanley F. Quayle" <stan@stanq.com>% Subject: Re: CHARON-ALPHA - soon ????t/ Message-ID: <3DC70320.17107.24831F87@localhost>   0 On 4 Nov 2002 at 21:06, David J. Dachtera wrote:& > Any chance of a VAX->IPF translator?  @ One of the specs is that the Alpha->IPF translator can translate VAX->Alpha translated images.C  
 --Stan Quayle  Quayle Consulting Inc.  
 ----------C Stanley F. Quayle, P.E. N8SQ  +1 614-868-1363  Fax: +1 614 868-1671C1 8572 North Spring Ct. NW, Pickerington, OH  43147.= Preferred address:  stan@stanq.com       http://www.stanq.comk   ------------------------------   Date: 4 Nov 2002 23:07:41 GMTi# From: hoff@hp.nospam (Hoff Hoffman)d> Subject: Re: Configuring IP (was: Re: Configuring OpenVMS-6.2)* Message-ID: <aq6ujt$k3k$2@web1.cup.hp.com>  T In article <3DC2879E.955BF9B9@127.0.0.1>, Nic Clews <sendspamhere@127.0.0.1> writes: :Jurren Bouman wrote:YH :> We have moved a box with OpenVMS version 6.2 to another location. ButG :> now we need to change the IP-address, Gateway address and the Subnet G :> mask address. I've never done and I'd like to know how this is done.s( :> Does anyone here know how to do this?  C   Please move to a current release of OpenVMS, there have been many E   changes and many enhancements to OpenVMS made since 1995; since the1G   release of V6.2.  The current release of OpenVMS VAX is V7.3 (shippedeG   2001) and the current OpenVMS Alpha release is V7.3-1 (shipped 2002).n  F   As for TCP/IP Services, information on setting the gateway is listed   in the OpenVMS FAQ.   A   If you are changing the whole IP environment, I'd likely simplyoE   reconfigure the network using the UCX$CONFIG tool.  Details on this C   are in the TCP/IP Services product documentation, and pointers tot3   the current documentation are in the OpenVMS FAQ.a  H   With current TCP/IP Services and current OpenVMS versions, you can useJ   the DHCP client capabilities and avoid most of this configuration stuff.  I :> Then I'd like to know the mac-address of the ethernet card. How can weM :> find this under OpenVMS 6.2?t  E   The multicast address varies, depending on what networking softwareRG   is loaded.  DECnet Phase IV will set the controller physical address  D   based on the DECnet address (AA-00-04-00-xx-yy), for instance.  IfF   (as I would expect) you really want to know the hardware address forF   the Ethernet or IEEE 802.3 network controller (the address stored inD   non-volatile controller memory) , then please see the OpenVMS FAQ.  @ :What IP package, UCX, TCPIP, Multinet, TCPware, something else?  H   Correct.  Please see the OpenVMS FAQ for the details -- product names,G   versions, etc -- that can be required when answering the question(s).aF   Providing the necessary information is to your benefit, as it makes D   for a far quicker answer to the questions you have posed.  Without7   the details, the answer to your questions is delayed.s    N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comd   ------------------------------  $ Date: Mon, 4 Nov 2002 21:42:51 +0100" From: "Hans Vlems" <hvlems@iae.nl>* Subject: Re: Configuring VMS as DNS server5 Message-ID: <aq6m4j$6uo6g$1@ID-143435.news.dfncis.de>R  = "Lyndon Bartels" <lbartels@pressenter.com> schreef in berichtc( news:3DC535E7.15BC22DC@pressenter.com... > Hello all, >.
 > VMS v7.3-1,o) > TCPIP 5.3-xx? (Whatever was on the CD.)w >e >  > F > I'm in the process of becoming Linux free. I need to replace the DNS > server on my home LAN. > 4 > How do I configure my VMS machine to do this task? >t1 The configuration is in @sys$manager:tcpip$configGE Run the utility and select Core environments and next [1] - domain  :  that'll configure the domainH Return to the main menu and select Server environments and next [1] BINDH That'll configure the server. You have to start the server manually with TCPIP ENABLE SERVICE BINDi  K  My local Domain Name System server runs on a VAXstation 3100/GPX under VMSi 7.2 and TCPIP V5.1. ; I find the following commands particularly useful in TCPIP:    sho service bind /full sho config binde sho name  L Don't forget to alter the serialno in the .db files after each change and to" enable it with TCPIP SET BIND/INIT    
 Good luck, Hans   ------------------------------  % Date: Mon, 04 Nov 2002 21:23:53 -0600g1 From: "David J. Dachtera" <djesys.nospam@fsi.net>e* Subject: Re: Configuring VMS as DNS server' Message-ID: <3DC739C9.29E6C50D@fsi.net>w   Brian Tillman wrote: > A > >I'd going into TCPIP (there may be an alias verb UCX) (invokesmG > >TCPIP$UCP, I think) and look at the HELP for SET HOST and SET CONFIGE > >HOST or SET HOST/PERM > K > Neither SET CONFIG HOST nor SET HOST/PERM exist.  SET HOST is a permanentn > setting command. > % > What does this have to so with DNS?l  @ See the on-line help and/or doc. for TCP/IP services and/or UCX.   -- r David J. Dachteran dba DJE Systems. http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Mon, 04 Nov 2002 22:59:54 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca>i* Subject: Re: Configuring VMS as DNS server, Message-ID: <3DC74230.C54C1880@videotron.ca>  M > > Neither SET CONFIG HOST nor SET HOST/PERM exist.  SET HOST is a permanentc > > setting command. > > ' > > What does this have to so with DNS?d  H SET HOST populates the local hosts database. From that database, you canM create a DNS hosts database with the CONVERT/UNIX BIND  command. (you need to7M use that command twice, once for the hosts database, and once for the reverse  lookup one).  H It is must better to just edit the bind databases (the domain.db and the0 reverse lookp table x.y.z.IN-ADDR_ARPA.DB files)    H You are also better off editing the .db files and not bothering with theN TCPIP> SET HOST and SET MX commands and having a "real" configuration for both hosts and MX records.    ------------------------------  % Date: Mon, 04 Nov 2002 23:48:18 -0500>. From: Lyndon Bartels <lbartels@pressenter.com>* Subject: Re: Configuring VMS as DNS server. Message-ID: <3DC70742.3CD66CEF@pressenter.com>  
 Thanks JF.    8 Looking at the pointers you gave me, I figured it out...   The kicker is:  C TCPIP v5.3 uses BIND version 9. There are some new stuff there thatt& isn't covered in the documentation....    7 But it's working now. Now onto the web server software.     
 Thanks again,    Lyndon     -- sG My opinions are mine and mine alone. They seldom align with those of myt	 employer.     H The only good thing about putting the cart before the horse is you don't have to look at the horse's butt.   ------------------------------   Date: 4 Nov 2002 12:18:26 -0800S% From: yolydeltoro@yahoo.com (yolanda)nM Subject: DECtalk A Guide To Voice 1985 First Printing EB 26375 56/06 23 100.0h= Message-ID: <fd84de93.0211041218.358ad740@posting.google.com>   B I'm looking to purchase or copy the DECtalk A Guide To Voice First8 Printing 1985 EB 26375 56/06 23 100.0.  Thanks,  Yolanda yolydeltoro@yahoo.comt   ------------------------------   Date: 5 Nov 2002 01:37:08 GMTs# From: hoff@hp.nospam (Hoff Hoffman)lQ Subject: Re: DECtalk A Guide To Voice 1985 First Printing EB 26375 56/06 23 100.0(* Message-ID: <aq77c4$k3k$4@web1.cup.hp.com>  e In article <fd84de93.0211041218.358ad740@posting.google.com>, yolydeltoro@yahoo.com (yolanda) writes:tC :I'm looking to purchase or copy the DECtalk A Guide To Voice Firstn9 :Printing 1985 EB 26375 56/06 23 100.0.  Thanks,  Yolanda  :yolydeltoro@yahoo.com  G   AFAICR, The "DECtalk a Guide to Voice" book was an old sales-oriented 9   publication, and not particularly technically detailed.t  I   All of the on-line DECtalk documentation and supporting materials that rH   I have been able to locate are available on the OpenVMS Freeware -- asD   I've prised the materials loose -- and this cache has most of the E   available technical materials, the support documentation, examples,i   and other technical stuff:  >     http://www.openvms.compaq.com/freeware/freeware50/dectalk/>     http://www.openvms.compaq.com/freeware/freeware40/dectalk/  >   The contents of these two directories are (quite) different!  K   Most of the DECtalk materials (greatly!) pre-date the web, and were never >   particularly ported forward into HTML or other such formats.  H   If there is a scanned copy of this manual around (and I don't know of K   one), it would probably/eventually be available at the following website:t  .     http://www.montagar.com/~patj/dec/hcps.htm  C   If you do find a copy of this sales document in the wild, please 2   let Pat (see URL above) know.r  G   The folks over at Fonix (http://www.fonix.com/) might well be of somel@   assistance here, as well, as Fonix is where DECtalk now lives.  N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comE   ------------------------------  % Date: Mon, 04 Nov 2002 15:10:41 -05001! From: Jim Agnew <jpagnew@vcu.edu>0# Subject: Re: Digital Press discountr' Message-ID: <3DC6D441.5673277B@vcu.edu>c  G sorry guys... I didn't see that orange writing, was looking too fast in0 the 2 main menus..  & "pass the salt, I gotta eat my hat..."   David D Miller wrote:1 >  > Folks: >  > http://www.openvms.compaq.com9 > K > really does work.  Look for the Digital Press logo on the bottom right of86 > the first page.  It's orange on my Netscape browser. >  > dave.M   ------------------------------  # Date: Tue, 05 Nov 2002 01:13:09 GMT0$ From: winston@SSRL.SLAC.STANFORD.EDU$ Subject: Re: Disk size specification8 Message-ID: <00A167AD.E9AB4834@SSRL04.SLAC.STANFORD.EDU>  @ In article <3dc6bc91$1@NUK.Uni-Lj.Si>, rok@nuk.uni-lj.si writes:   > OpenVMS V7.3, Alpha and VAX.E > I have two RA82 lookalike disks, but alas, they differ: one has one0; >cylinder less than the other, they differ for 1120 blocks.1A > I there a way to tell VMS both disks are of the same size, as IF >would like to shadow them.F  I I seem to recall that disks no longer had to be the same size or geometry0L to be shadowed, that the shadowing stuff will just use as much of the largerL one as will cover the smaller one.  (I think this was a 7.3 change - haven't( gone back to the release notes to look.)  1 I'm not using HBVS so I don't know this for sure.    -- Alan-   ------------------------------  % Date: Mon, 04 Nov 2002 19:59:47 -0600 7 From: "Craig A. Berry" <craigberry@nospam.telocity.com>0$ Subject: Re: Disk size specificationG Message-ID: <craigberry-67A9D4.19594704112002@news.directvinternet.com>A  8 In article <00A167AD.E9AB4834@SSRL04.SLAC.STANFORD.EDU>,&  winston@SSRL.SLAC.STANFORD.EDU wrote:  B > In article <3dc6bc91$1@NUK.Uni-Lj.Si>, rok@nuk.uni-lj.si writes:  C > > I there a way to tell VMS both disks are of the same size, as It > >would like to shadow them.r > K > I seem to recall that disks no longer had to be the same size or geometryI > to be shadowed  B Not quite.  In 7.2 the type and geometry could be different.  The E number of physical blocks still has to be the same as of 7.3-1.  The , requirements are stated here:   I <http://www.openvms.compaq.com/doc/731FINAL/5423/5423pro.html#compatible__ disks_sect>]  G A knowledgeable abuser of the bad block locator utility could probably  G convince a disk it had some bad blocks that it didn't really have, but R that's over my head.   ------------------------------   Date: 5 Nov 2002 02:29:42 GMTS# From: hoff@hp.nospam (Hoff Hoffman)L$ Subject: Re: Disk size specification* Message-ID: <aq7aem$sag$1@web1.cup.hp.com>  _ In article <00A167AD.E9AB4834@SSRL04.SLAC.STANFORD.EDU>, winston@SSRL.SLAC.STANFORD.EDU writes: A :In article <3dc6bc91$1@NUK.Uni-Lj.Si>, rok@nuk.uni-lj.si writes:U :  :> OpenVMS V7.3, Alpha and VAX.$     Which Alpha, and which VAX?M  =   How are these disks connected to the OpenVMS VAX and to theE   OpenVMS Alpha systems?  F :> I have two RA82 lookalike disks, but alas, they differ: one has one< :>cylinder less than the other, they differ for 1120 blocks.B :> I there a way to tell VMS both disks are of the same size, as I :>would like to shadow them.  A   The RA82-class disks are ancient and physically giant UDA-classSE   disk drives.  If these two lookalike disks are of a similar vintageXF   as the RA82 series, then I'm mildly surprised these widgets are even   still working.  E   But you will also want to check with the vendor or the organization;A   supporting your hardware, as the details of how these disks areMD   configured are obviously known to the particular vendor's support.  J :I seem to recall that disks no longer had to be the same size or geometryM :to be shadowed, that the shadowing stuff will just use as much of the larger M :one as will cover the smaller one.  (I think this was a 7.3 change - haven't ) :gone back to the release notes to look.)S  E   As currently configured, these two disks cannot be shadowed by the XF   OpenVMS host-based shadowing package -- the block counts must match.  F   While disk geometry no longer matters for shadowing -- and there areE   shadowing ECO kits with this capability available for older OpenVMSCF   releases -- the overall disk block counts must be identical for the D   volumes to be shadowed using host-based volume shadowing software.  F   There is work going on in this area for future OpenVMS releases, but+   this information is current as of V7.3-1.   F   And yes, I would encourage you to replace these RA82-lookalike disks    with some more modern storage.    N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comS   ------------------------------  % Date: Mon, 04 Nov 2002 21:42:44 -0600:1 From: "David J. Dachtera" <djesys.nospam@fsi.net>v$ Subject: Re: Disk size specification' Message-ID: <3DC73E34.C2C4C134@fsi.net>_   rok@nuk.uni-lj.si wrote: >  >  OpenVMS V7.3, Alpha and VAX.VF >  I have two RA82 lookalike disks, but alas, they differ: one has one< > cylinder less than the other, they differ for 1120 blocks.B >  I there a way to tell VMS both disks are of the same size, as I > would like to shadow them./ >  INITIALIZE/SIZE=1312155 would be so handy :(   B Well, it's kinda bogus, but on the larger disk you could use LD toH create a container the size of the small disk and use that to shadow the
 smaller disk.    REALLY reaching here....   -- E David J. DachteraS dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Mon, 04 Nov 2002 14:57:04 -0500T& From: David M Smith <dsmit115@csc.com>0 Subject: Re: Does anyone use P7 in DCPS$STARTUP?8 Message-ID: <kvjdsusrgd8h6t3gui9ci9eeb4o01ulbsj@4ax.com>  P On Sat, 02 Nov 2002 02:54:08 GMT, "John E. Malmberg" <wb8tyw@qsl.network> wrote:  G >I found that a data file driven approach was too confusing to those I IJ >tried to distribute system management tasks to.  Also maintenance on the I >command files was a pain, as a mis-edit in a data file would cause many O
 >problems.  N We have found that data-file driven procedure work very well for us, and allowO leveraging of system managers across many, many systems because the environmentLN "looks the same" on all the systems, except for the specific details contained in the various data files.  L Here is an example of a simple DCPS queue definition in our environment (for tutorial purposes only):   /queue=xxx_na_dcps -         /type=lps_dcps -1         /description="XXXXXXLexmark Optra T614" -S'         /autostart_on=(xxxvx1,xxxal1) -i/         /server="""ip_rawtcp/20.x.y.z:9100""" -t         /processor=dcps$smb -sO         /dcps_parameter="number_up=1,sides=one_sided_simplex,layup=noborders" - F         /initialize="define/system/exec dcps$xxx_na_dcps_no_sync true"  J where I've munged some of the identifying data to avoid spreading internalK information. It is easy to understand the key variables in this queue setupsD (IMHO) and easy for non-expert managers / operators to make changes.  I -------------------------------------------------------------------------eI David M. Smith 302.391.8533                       dsmit115 at csc dot com I Computer Sciences Corporation     (Opinions are those of the writer only) I -------------------------------------------------------------------------    ------------------------------  # Date: Mon, 04 Nov 2002 20:54:04 GMTt* From: Paul Anderson <paul.anderson@hp.com>0 Subject: Re: Does anyone use P7 in DCPS$STARTUP?5 Message-ID: <041120021553401041%paul.anderson@hp.com>c  F In article <kvjdsusrgd8h6t3gui9ci9eeb4o01ulbsj@4ax.com>, David M Smith <dsmit115@csc.com> wrote:   F > We have found that data-file driven procedure work very well for us,C > and allow leveraging of system managers across many, many systemseE > because the environment "looks the same" on all the systems, except.? > for the specific details contained in the various data files.   A This is a good discussion.  I've noted each suggestion for futuretF reference and discussion with my peers who maintain the queue manager.  D For now, we're trying to add AUTOSTART support to DCPS but think theC syntax of DCPS$STARTUP, especially the P2 parameter, is getting tooc complex.  = Long term, I think a data-driven approach is a good one.  ThewF possibility for syntax errors exists in both a command procedure and a+ data file, but I think less in a data file.i   Paul   --    Paul Anderson   OpenVMS Engineeringu   Hewlett-Packard Companyk   ------------------------------  $ Date: Mon, 4 Nov 2002 16:26:02 -0500! From: VAXVMS <bounce@notmail.com>e0 Subject: RE: Does anyone use P7 in DCPS$STARTUP?K Message-ID: <BA52530E3149734A9BAABDBBFA808E4903027B95@rlghncst964.usps.gov>y    I concur with David's sentiment.  C We consolidated 85 district VAXen by migrating them to a cluster ofdE ES40s where group context takes the place of what used to be discrete  hardware differentiation.   J There are literally thousands of queues that we manage, and not only do weH have data-file driven queue management but we periodically run a commandJ procedure that surveys each district context and recreates the queue setup andmK queue logical creation files so we don't keep setting up queues that aren't  needed any more.   WWWebb    F In article <kvjdsusrgd8h6t3gui9ci9eeb4o01ulbsj@4ax.com>, David M Smith <dsmit115@csc.com> wrote:y  F > We have found that data-file driven procedure work very well for us,C > and allow leveraging of system managers across many, many systems E > because the environment "looks the same" on all the systems, except ? > for the specific details contained in the various data files.   A This is a good discussion.  I've noted each suggestion for futurerF reference and discussion with my peers who maintain the queue manager.  D For now, we're trying to add AUTOSTART support to DCPS but think theC syntax of DCPS$STARTUP, especially the P2 parameter, is getting toob complex.  = Long term, I think a data-driven approach is a good one.  TheRF possibility for syntax errors exists in both a command procedure and a+ data file, but I think less in a data file.6   Paul   --  Paul Anderson   OpenVMS Engineering,   Hewlett-Packard Company.   ========================  William W. Webb / DSSC/RLM, USPS OpenVMS Support Services& 4924 Green Road Raleigh, NC 27616-2800: 919.874.3043 <FirstInitialDotLastNameAtEmailDotUSPSDotGov>   ------------------------------  % Date: Mon, 04 Nov 2002 20:42:36 -0500S- From: JF Mezei <jfmezei.spamnot@videotron.ca>v0 Subject: Re: Does anyone use P7 in DCPS$STARTUP?* Message-ID: <3DC7220B.676A6F@videotron.ca>  
 VAXVMS wrote:gL > There are literally thousands of queues that we manage, and not only do weJ > have data-file driven queue management but we periodically run a commandL > procedure that surveys each district context and recreates the queue setup > andf  I The question becomes: should Digital change the queue manager to be "dataeL driven" or should it let each customer design their own customized data file0 format and procedures that fit their own needs ?  H I would think that the needs are diverse enough that no single data-file driven solution would fit all.  R But enabling the quaue maneger to be more easily dara driven would be a good idea.  J Perhaps a SHOW/QUEUE/MANAGER/CONFIG=myfile.dat in a documented file formatB would go a long way towards helping customize their queue configs.   ------------------------------  $ Date: Mon, 4 Nov 2002 14:45:52 -0500% From: "John Vottero" <John@mvpsi.com> 1 Subject: Re: DS10L and backups on an IDE CDRW ...s/ Message-ID: <usdjjgpmp6eiec@news.supernews.com>V  K I backup to CD-R all the time.  We've completely eliminated tape.  You needrJ to backup to an intermediate device (or virtual device) and then burn that device to the CD-R.r  5 "Bob Ceculski" <bob@instantwhip.com> wrote in messageh7 news:d7791aa1.0211041018.545a7be6@posting.google.com...29 > does the cdwrite program mentioned on this board before ; > allow you to install and IDE CDRW drive and do consistentg3 > backups (daily) on rewritable cd's on a ds10l ...s0 > would this work with an automated dcl routine? > Any success stories?   ------------------------------  $ Date: Mon, 4 Nov 2002 16:12:38 -0500$ From: "rob kas" <rob@netcarrier.net>1 Subject: Re: DS10L and backups on an IDE CDRW ...a3 Message-ID: <3dc6e244$0$1392$8e9e3842@news.atx.net>w  &    Backing up to CD , works real well.:    Just make sure you back up to a tempfile then burn that    to CD  $                                  Rob    5 "Bob Ceculski" <bob@instantwhip.com> wrote in messageh7 news:d7791aa1.0211041018.545a7be6@posting.google.com... 9 > does the cdwrite program mentioned on this board before ; > allow you to install and IDE CDRW drive and do consistentn3 > backups (daily) on rewritable cd's on a ds10l ...r0 > would this work with an automated dcl routine? > Any success stories?   ------------------------------  * Date: Mon, 4 Nov 2002 22:09:14 +0000 (UTC), From: lewis@mazda.mitre.org (Keith A. Lewis)1 Subject: Re: DS10L and backups on an IDE CDRW ... . Message-ID: <aq6r6a$ge9$1@newslocal.mitre.org>  { "John Vottero" <John@mvpsi.com> writes in article <usdjjgpmp6eiec@news.supernews.com> dated Mon, 4 Nov 2002 14:45:52 -0500:cL >I backup to CD-R all the time.  We've completely eliminated tape.  You needK >to backup to an intermediate device (or virtual device) and then burn thatw >device to the CD-R. >s6 >"Bob Ceculski" <bob@instantwhip.com> wrote in message8 >news:d7791aa1.0211041018.545a7be6@posting.google.com...: >> does the cdwrite program mentioned on this board before< >> allow you to install and IDE CDRW drive and do consistent4 >> backups (daily) on rewritable cd's on a ds10l ...1 >> would this work with an automated dcl routine?: >> Any success stories?s  L I'm not familiar with cdwrite, but I know cdrecord + VD (virtual disk, newerH version known as LD) works great with VMS for creating disk images.  And" yes, you can automate it with DCL.   Here's what I don't know:t  D 1.  Will cdrecord (or cdwrite) work with IDE CD burners on an Alpha?  > 2.  Will those programs work in CD-RW mode?  I'm guessing yes.  H 3.  Can you jam a full-size CD-RW drive into a DS10L if you toss out the CD/floppy combo?  I One thing though -- A CD doesn't give you all that much backup capacity. pI As long as you're writing your backup data to IDE anyway, consider an IDExI hard drive.  It's officially unsupported, but so is the IDE CDRW I think.   + --Keith Lewis              klewis$mitre.orgo> The above may not (yet) represent the opinions of my employer.   ------------------------------   Date: 4 Nov 2002 23:30 CST' From: carl@gerg.tamu.edu (Carl Perkins)p Subject: Re: errno equivalent , Message-ID: <4NOV200223301738@gerg.tamu.edu>  " briggs@encompasserve.org writes...s }In article <c3c6388a.0211010652.2d24328e@posting.google.com>, johan.nilsson@esrange.ssc.se (Johan Nilsson) writes:sB }> are there any native errno (*nix) or Get/SetLastError() (Win32)E }> equivalent under OpenVMS, or do you always have to store conditions* }> values returned by the system services? }  }You have to store it. } B }The notion that it could be done any other way has always boggled> }my mind.  The idea of going out and sampling a global cell toB }see what error last occurred and blindly assuming that this error9 }is relevant to the problem at hand in the current threadu? }of execution seems patently absurd.  (Though I'm sure it works 8 }out fairly well in practice.  Please take my comment as1 }honest bewilderment rather than as O/S bashing).- } B }Under VMS, you call a system service and you get a return status.@ }And, if the system service is potentially asynchronous you also> }get the opportunity to provide an I/O status block in which aE }completion status code will be written.  If you have 50 asynchronous A }calls outstanding, you can have 50 I/O status blocks assigned toI& }pick up the various completion codes. } G }The return status (or completion status) is itself fairly descriptive.gC }There is no need to refer to an external "errno" to get a detailed  }problem description.e } = }For RMS services, you get an RMS error code and you can also 6 }retrieve an underlying error code from, for instance, }your-fab.FAB$L_STV. } G }For RTL calls, you generally just receive the RTL return status.  ThislC }status is generally detailed enough that the underlying O/S return  }status would be redundant.  } ? }A few RTL routines will signal their errors, thus providing ane= }"exception vector" so that registered exception handlers (orlF }debuggers or traceback dump printers) can provide detailed diagnosticC }information.  If I remember correctly, it's only the STR$ and CLI$  }routines that do this.h } 2 }For C RTL calls, there is an errno and it is set. } 
 }	John Briggst  B Onthe other hand, in Fortran there are some values stored that you" can get via the ERRSNS subroutine:  
     ERRSNS  ;        ERRSNS ([io-err] [,sys-err] [,stat] [,unit] [,cond])e          Class:  Subroutinef  F        Returns information about the last Fortran error that occurred.I        The arguments are all return values and must be defined as integerx#        variables or array elements:o  ?           io-err    Stores the most recent Fortran error numberd;                     that occurred during program execution.o?                     The value is zero if no error has occurred.t  ?           sys-err   Stores the most recent RMS STS status code.s  @           stat      Stores the most recent RMS STV status value.@                     This status value provides additional status                      information.  ?           unit      Stores the logical unit number (if the laste5                     the last error was an I/O error).o  <           cond      Stores the actual processor value.  This)                     value is always zero.n     --- Carl   ------------------------------   Date: 4 Nov 2002 21:26:48 -0800r" From: all_pinball@yahoo.com (Mike)Q Subject: FA: OLD DEC VAX Circuit Card - cheap - NO RESERVE - ebay ends 11/08/2002u= Message-ID: <49d19389.0211042126.5a402552@posting.google.com>g  C For Auction: Miscellaneous, likely non-operational, DEC VAX Circuitu  Cards.  Auctions end 11/08/2002.  2 ALL ARE LISTED WITH NO RESERVE AND NO MINIMUM BID!  p http://cgi6.ebay.com/ws/eBayISAPI.dll?ViewSellersOtherItems&userid=all_pinball&include=0&since=-1&sort=3&rows=25   <A HREF="http://cgi6.ebay.com/ws/eBayISAPI.dll?ViewSellersOtherItems&userid=all_pinball&include=0&since=-1&sort=3&rows=25">Auction Listings<A>r  # Part numbers offered are as folows:v 59-L0119-YAe L0108 K.SAI / K.STIe 59-L0111-00.' M8971 5015050F1 STI INTERFACE BUS LOGICp$ M8959 5013186D WRITE MICROCONTROLLER
 M8950 5012673r  D These items are offered entirely AS-IS and are best suited for parts! or for collector's purposes only.   7 A cheap and easy way to add to your DEC VAX collection.   5 Email questions through the ebay member mail service.o  p http://cgi6.ebay.com/ws/eBayISAPI.dll?ViewSellersOtherItems&userid=all_pinball&include=0&since=-1&sort=3&rows=25   <A HREF="http://cgi6.ebay.com/ws/eBayISAPI.dll?ViewSellersOtherItems&userid=all_pinball&include=0&since=-1&sort=3&rows=25">Auction Listings<A>u   ------------------------------  % Date: Mon, 04 Nov 2002 21:04:11 -0600s1 From: "David J. Dachtera" <djesys.nospam@fsi.net>e Subject: Re: HP Advocacy Siteh' Message-ID: <3DC7352B.2AD38122@fsi.net>I   "David J. Dachtera" wrote: >  > Syltrem wrote: > >t > > It's already there!o > > J > > TITLE: Market OpenVMS just like Tru64 - not only for so-called "niche" > > marketst > >a& > > There's only one me too right now. > ) > I don't see that one. The ones I see ateQ > http://www.hpuseradvocacy.org/Advocacy/Issues/MeToo/Index.cfm?IssueCategoryID=0c >  > are: >  > Universal SYSLOG Capabilities- > CISC COMPILER SUPPORTa# > Documentation on Legacy Equipmentp< > roadmap to migrate CA's Advantage:Gen to HP OpenVMS on IPF > NAPSE North American Support > Embeded Wireless LAN in Ipaq > : > What page are you looking at, and how did you get there?  G Forgive my self-reply, but I found out that the "CategoryId" at the enda. of the URL makes the difference. Try 64 or 71.   -- r David J. Dachtera  dba DJE SystemsT http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/i   ------------------------------   Date: 4 Nov 2002 13:59:49 -0800 1 From: keithparris_NOSPAM@yahoo.com (Keith Parris)x: Subject: HP World News says You Can't Crack It or Crash It= Message-ID: <cf15391e.0211041359.7f0a052d@posting.google.com>   > HP World News, published by Interex, included a pointer to theB following article in the Oct. 31 issue of their e-mail newsletter.  7 You can find this article on the Interex.org website atuF http://www.interex.org/hpworldnews/hpw211/news9.jsp (while one must beF a member of Interex to read it, there is a level of membership that is8 free -- see http://www.interex.org/member/benefit.html). ---e You Can't Crack It or Crash It! OpenVMS Inspires Loyal Following i  l  by Susan B. Shore  ? For 25 years, OpenVMS and its predecessor VMS have been runningIF unhacked and with extremely infrequent if not non-existent reboots, or" so say their very loyal followers.  E "When you port to VMS, you know that the operating system is going to < be stable and just about bulletproof," said Champ Clark, whoC administers the DeathRow OpenVMS cluster (www.deathrow.vistech.net)EE under HP's hobbyist license program. A common refrain, said Clark and = others, is "OpenVMS uptime is measured in years, not months."v  > Clark started working on VMS as a teen. "I've always loved the? handling of VMS and the security model it uses," he said. "It's E incredibly stable, even under heavy loads. It's just a load of fun toiD use." He said he has also worked on HP-UX and other UNIX releases asE well as Linux. His company, Vistech Communications (www.vistech.net),s does networking and security.    Back to DEC   D VMS has its roots in Digital Equipment Corp., where it ran on 32-bit@ VAX (Virtual Address eXtension) systems. DEC changed the name toB OpenVMS when the operating system was ported to Alpha servers. TheF name is unrelated to open source as we think of it today. OpenVMS is aD proprietary OS. HP is now porting OpenVMS to Itanium using the AlphaA source code so that Alpha users will get the same improvements as 5 those who upgrade to the 64-bit Itanium architecture.e  : "OpenVMS has many, many extremely loyal users," Bob Blatz,> Hewlett-Packard's director of OpenVMS marketing, told HP WorldC Magazine. "It simplifies their life, simplifies their job. We're at 4 the 25th anniversary, and it still has a long life."  C Through the hobbyist program, individuals can get free licenses forrE the UNIX-based OpenVMS and the layered software products for VMS, but-F they must purchase the distribution kit to get the binary code to run.D The license allows non-commercial use of OpenVMS and the software. AC similar program allows educational institutions to obtain a license0F and use the OS and software for anything other than revenue-generatingD functions. "It's nice to get back information around the cool things@ they've done," said Blatz, "but it also provides a way of having? people on [their] own initiative learn about the system, and it8' provides a talent pool for businesses."S  B The best estimate of OpenVMS's install base is about 411,000, withA much of that in government, education, financial services, health E care, manufacturing and telecommunications, according to Blatz. Those E industries all require high availability and security. OpenVMS allows-C 96 nodes per cluster and has featured clustering-now catching on in2C Linux environments-since the mid-1980s. It can also handle disasterwD variant clustering, in which nodes can be as far apart as 500 miles.   A Homegrown Community   @ Admirers of OpenVMS, like admirers of other OSes, have their ownD community that helps solve problems and generate ideas. The DeathRowC cluster, for instance, gives new users access to the DCL prompt andlD compilers as well as a bulletin board. Most in the OpenVMS communityD swear by the system's security, going so far as to say that it can't
 be hacked.  @ According to an entry on alt.ph.uk, a discussion group set up toB discuss hacking and cracking, "you'd be hard pushed to find anyone7 these days who has successfully hacked a fully secured,e: enterprise-scale VMS machine. Which is not to say it's not@ possible-far from it. Indeed, shell code for VMS does exist in a@ Digital [sic] form, out there... somewhere :-)." The VMS section( concludes, "Want a challenge? Hack VMS."  @ Others say the reason OpenVMS has not been hacked is its lack ofD bandwidth. Hackers want to go where the notoriety is, where they canD affect the most people. Blatz said, somewhat jokingly, that he likedF to keep it quiet that VMS was so secure in order to avoid waving a red flag at the hacker community.   C Blatz believes part of the reason OpenVMS was ahead of the curve on D issues such as clustering was just plain good business practice. "We8 spend a lot of time talking to our customers about their? requirements," he said. HP has an OpenVMS executive council andaE several technical advisory forums, but loyalty is born of more than aeE quarter decade of familiarity, Blatz said. "In addition to growing up D with it, you find out how useful it is, and you're not fighting it."   ------------------------------   Date: 4 Nov 2002 14:03:16 -0800i1 From: keithparris_NOSPAM@yahoo.com (Keith Parris)e: Subject: HP World News says You Can't Crack It or Crash It= Message-ID: <cf15391e.0211041403.685c06eb@posting.google.com>   B Oh, and I've already given feedback to the editor about mistakenlyC classifying OpenVMS as "UNIX-based", about saying "disaster variant-F clustering" when "disaster-tolerant" was obviously intended, and aboutB how hobbyists can get media without paying megabucks, and I'm toldD that info will be included in a letter to the editor in the DecemberA issue, so let's not give them any grief on those issues, because,f? those minor mistakes aside, this is a very positive article forv OpenVMS.   ------------------------------  % Date: Mon, 04 Nov 2002 14:08:10 -0800S% From: Dean Woodward <deanw@rdrop.com>6> Subject: Re: HP World News says You Can't Crack It or Crash It( Message-ID: <3DC6EFCA.9000409@rdrop.com>   Keith Parris wrote:e@ > HP World News, published by Interex, included a pointer to theD > following article in the Oct. 31 issue of their e-mail newsletter. > 9 > You can find this article on the Interex.org website at H > http://www.interex.org/hpworldnews/hpw211/news9.jsp (while one must beH > a member of Interex to read it, there is a level of membership that is: > free -- see http://www.interex.org/member/benefit.html).  D I don't want to read more than I have.  Does anyone know how to get < ahold of Ms. Shor, and a level enough temper to educate her?    > You Can't Crack It or Crash It# > OpenVMS Inspires Loyal Following y >  s >  by Susan B. Shora   [Snip Happens]  E > Through the hobbyist program, individuals can get free licenses foruG > the UNIX-based OpenVMS and the layered software products for VMS, but         ^^^^^^^^^^^^^^^^^^e4 AAAAAAAAAAAAAAHHHHHHHHHHHHHHHHHH!!!!!!!!!!!!!!!!!!!!   ------------------------------  $ Date: Mon, 4 Nov 2002 14:02:12 -0800$ From: Shane Smith <ssmith@icius.com>> Subject: RE: HP World News says You Can't Crack It or Crash It0 Message-ID: <01C2840A.D75ED150@sulfer.icius.com>  E My initial reaction to this could have been mistaken for Jay's rap atIH the beginning of "Jay and Silent Bob Strike Back". Then I noticed the HPG in "HP World News" and I calmed down a little. I thought for a moment arF pro-VMS article had actually made it into the press. Surprisingly good5 article though, except the "Unix based OpenVMS" line.t  3 Anybody know what the distribution's like for this?p   Shaneo   -----Original Message-----H From: keithparris_NOSPAM@yahoo.com [mailto:keithparris_NOSPAM@yahoo.com]' Sent: Monday, November 04, 2002 2:00 PMg To: Info-VAX@Mvb.Saic.Comm: Subject: HP World News says You Can't Crack It or Crash It    > HP World News, published by Interex, included a pointer to theB following article in the Oct. 31 issue of their e-mail newsletter.  7 You can find this article on the Interex.org website atoF http://www.interex.org/hpworldnews/hpw211/news9.jsp (while one must beF a member of Interex to read it, there is a level of membership that is8 free -- see http://www.interex.org/member/benefit.html). ---S You Can't Crack It or Crash It! OpenVMS Inspires Loyal Following       by Susan B. Shor   ? For 25 years, OpenVMS and its predecessor VMS have been running.F unhacked and with extremely infrequent if not non-existent reboots, or" so say their very loyal followers.  E "When you port to VMS, you know that the operating system is going to < be stable and just about bulletproof," said Champ Clark, whoC administers the DeathRow OpenVMS cluster (www.deathrow.vistech.net)-E under HP's hobbyist license program. A common refrain, said Clark and = others, is "OpenVMS uptime is measured in years, not months."n  > Clark started working on VMS as a teen. "I've always loved the? handling of VMS and the security model it uses," he said. "It'sIE incredibly stable, even under heavy loads. It's just a load of fun to D use." He said he has also worked on HP-UX and other UNIX releases asE well as Linux. His company, Vistech Communications (www.vistech.net),  does networking and security.    Back to DEC   D VMS has its roots in Digital Equipment Corp., where it ran on 32-bit@ VAX (Virtual Address eXtension) systems. DEC changed the name toB OpenVMS when the operating system was ported to Alpha servers. TheF name is unrelated to open source as we think of it today. OpenVMS is aD proprietary OS. HP is now porting OpenVMS to Itanium using the AlphaA source code so that Alpha users will get the same improvements as85 those who upgrade to the 64-bit Itanium architecture.r  : "OpenVMS has many, many extremely loyal users," Bob Blatz,> Hewlett-Packard's director of OpenVMS marketing, told HP WorldC Magazine. "It simplifies their life, simplifies their job. We're atm4 the 25th anniversary, and it still has a long life."  C Through the hobbyist program, individuals can get free licenses for-E the UNIX-based OpenVMS and the layered software products for VMS, but5F they must purchase the distribution kit to get the binary code to run.D The license allows non-commercial use of OpenVMS and the software. AC similar program allows educational institutions to obtain a licenseoF and use the OS and software for anything other than revenue-generatingD functions. "It's nice to get back information around the cool things@ they've done," said Blatz, "but it also provides a way of having? people on [their] own initiative learn about the system, and ith' provides a talent pool for businesses."h  B The best estimate of OpenVMS's install base is about 411,000, withA much of that in government, education, financial services, health E care, manufacturing and telecommunications, according to Blatz. Those5E industries all require high availability and security. OpenVMS allowslC 96 nodes per cluster and has featured clustering-now catching on in C Linux environments-since the mid-1980s. It can also handle disasterrD variant clustering, in which nodes can be as far apart as 500 miles.   A Homegrown Community   @ Admirers of OpenVMS, like admirers of other OSes, have their ownD community that helps solve problems and generate ideas. The DeathRowC cluster, for instance, gives new users access to the DCL prompt and D compilers as well as a bulletin board. Most in the OpenVMS communityD swear by the system's security, going so far as to say that it can't
 be hacked.  @ According to an entry on alt.ph.uk, a discussion group set up toB discuss hacking and cracking, "you'd be hard pushed to find anyone7 these days who has successfully hacked a fully secured,V: enterprise-scale VMS machine. Which is not to say it's not@ possible-far from it. Indeed, shell code for VMS does exist in a@ Digital [sic] form, out there... somewhere :-)." The VMS section( concludes, "Want a challenge? Hack VMS."  @ Others say the reason OpenVMS has not been hacked is its lack ofD bandwidth. Hackers want to go where the notoriety is, where they canD affect the most people. Blatz said, somewhat jokingly, that he likedF to keep it quiet that VMS was so secure in order to avoid waving a red flag at the hacker community.   C Blatz believes part of the reason OpenVMS was ahead of the curve on D issues such as clustering was just plain good business practice. "We8 spend a lot of time talking to our customers about their? requirements," he said. HP has an OpenVMS executive council and0E several technical advisory forums, but loyalty is born of more than aeE quarter decade of familiarity, Blatz said. "In addition to growing up D with it, you find out how useful it is, and you're not fighting it."   ------------------------------  $ Date: Mon, 4 Nov 2002 17:45:39 -0500* From: "Bill Todd" <billtodd@metrocast.net>> Subject: Re: HP World News says You Can't Crack It or Crash It2 Message-ID: <OP2dnYSCGs8OZVugXTWcpw@metrocast.net>  > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message7 news:cf15391e.0211041359.7f0a052d@posting.google.com...h   ...s  > > The best estimate of OpenVMS's install base is about 411,000  E Now, that is truly amazing:  estimated VMS-related annual revenue has G dropped from $4 billion to $2 billion (perhaps even lower now), but its J estimated installed base is *exactly* the same as it was back at its peak.   - bill   ------------------------------  % Date: Mon, 04 Nov 2002 15:46:12 -0800j& From: Greg Cagle <gregc@gregcagle.com>> Subject: Re: HP World News says You Can't Crack It or Crash It/ Message-ID: <use1lut78e7311@corp.supernews.com>T  D As Keith noted, HP World News is an Interex publication. It goes outC to Interex members as well as other subscribers outside the InterexaE membership. It's a "controlled circulation" perodical like InfoWorld.s9 I can find out the exact numbers if anyone is interested.a   - Greg   Shane Smith wrote:G > My initial reaction to this could have been mistaken for Jay's rap at J > the beginning of "Jay and Silent Bob Strike Back". Then I noticed the HPI > in "HP World News" and I calmed down a little. I thought for a moment acH > pro-VMS article had actually made it into the press. Surprisingly good7 > article though, except the "Unix based OpenVMS" line.b > 5 > Anybody know what the distribution's like for this?  >  > Shaneb   -- h
 Greg Cagle gregc at gregcagle dot com   ------------------------------  $ Date: Mon, 4 Nov 2002 16:05:15 -0800$ From: Shane Smith <ssmith@icius.com>> Subject: RE: HP World News says You Can't Crack It or Crash It0 Message-ID: <01C2841C.06FE8750@sulfer.icius.com>  D I'd actually be as interested in the demographics as the numbers, if3 they're available. Techies? Managers? Sheepfarmers?V   Shanea   -----Original Message------ From: Greg Cagle [mailto:gregc@gregcagle.com]e' Sent: Monday, November 04, 2002 3:46 PMM To: Info-VAX@Mvb.Saic.Comv> Subject: Re: HP World News says You Can't Crack It or Crash It    D As Keith noted, HP World News is an Interex publication. It goes outC to Interex members as well as other subscribers outside the InterextE membership. It's a "controlled circulation" perodical like InfoWorld.n9 I can find out the exact numbers if anyone is interested.a   - Greg   Shane Smith wrote:G > My initial reaction to this could have been mistaken for Jay's rap at J > the beginning of "Jay and Silent Bob Strike Back". Then I noticed the HPI > in "HP World News" and I calmed down a little. I thought for a moment a H > pro-VMS article had actually made it into the press. Surprisingly good7 > article though, except the "Unix based OpenVMS" line.i > 5 > Anybody know what the distribution's like for this?i >  > Shanes   -- f
 Greg Cagle gregc at gregcagle dot com   ------------------------------  % Date: Mon, 04 Nov 2002 22:03:37 -0600e1 From: "David J. Dachtera" <djesys.nospam@fsi.net>t> Subject: Re: HP World News says You Can't Crack It or Crash It' Message-ID: <3DC74319.9C7BA16A@fsi.net>s   Bill Todd wrote: > @ > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message9 > news:cf15391e.0211041359.7f0a052d@posting.google.com...n >  > ...e > @ > > The best estimate of OpenVMS's install base is about 411,000 > G > Now, that is truly amazing:  estimated VMS-related annual revenue hassI > dropped from $4 billion to $2 billion (perhaps even lower now), but itsaL > estimated installed base is *exactly* the same as it was back at its peak.  H ...but that's good! If revenue is down and the user base the same, maybeH that means there have been price cuts that somehow escaped us - "stealth marketing", y'know.a  E In such case, we need to get the word out far and wide and OpenVMS is-6 now affordable, even if the hardware it runs on isn't.  7 1st step: increase the user base (revenues rise again).[  = 2nd step: sell more support contracts (revenues rise higher).   E 3rd step: make VMS-capable hardware that is affordable (revenues rise 
 even higher).s  E 4th step: use the reveue increases to advertise the living SPIT outta." VMS! (Revenues rise higher still).  F 5th step: when the user base reaches a sustainable level, pull back onE marketing, but don't relapse back into "stealth" mode (revenues levelbA off at a much higher, but much more sustainable level: due to theeF greater market penetration, loss of any one large customer or group of( smaller ones does not kill the product).  + Baby steps - slow and steady wins the race!a   -- n David J. Dachterat dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/J   ------------------------------  % Date: Mon, 04 Nov 2002 23:28:04 -0500g- From: JF Mezei <jfmezei.spamnot@videotron.ca>-> Subject: Re: HP World News says You Can't Crack It or Crash It, Message-ID: <3DC748C8.39EC2AFE@videotron.ca>   "David J. Dachtera" wrote:I > > Now, that is truly amazing:  estimated VMS-related annual revenue has K > > dropped from $4 billion to $2 billion (perhaps even lower now), but itstN > > estimated installed base is *exactly* the same as it was back at its peak. > J > ...but that's good! If revenue is down and the user base the same, maybe@ > that means there have been price cuts that somehow escaped us     J I don't personally believe the 411,000 number for installed base.  I don'tL think that Digital , even less Compaq had the means to measure the installedK base. My guess is that the number represents the number of licences issued,pG minus only the ones known to have been de-activated. So in reality, thee4 current installed base is probably MUCH MUCH lower.   J Having said this, the 411,000 number is important for the little amount ofJ marketing that is done as it gives VMS some impressive number.  Since thisM "statistic" is to VMS' advantage, we should shut up and let the VMS group use  it as much as they want/can.  J However, if the 50% drop in revenus is true, that is significant. With theF murder of Alpha, the whole merger, and Stallard's memo, as well as theL uncertainty about IA64, it should be no surprise to anyone. But it does giveO the wintel camp at HP ammunition to carry through their dream to eliminate VMS.)   ------------------------------   Date: 4 Nov 2002 14:47:29 -0600-; From: kaplow_r@eisner.encompasserve.org.TRABoD (Bob Kaplow)- Subject: Re: HP-ETS 2002 CD33 Message-ID: <rPLm0lftDM2r@eisner.encompasserve.org>   a In article <usd8s6i82u16a3@corp.supernews.com>, "Encompass - HP-ETS" <KilleenJ@toast.net> writes:-K > It is likely that any member of Encompass will be able to request a CD of-J > the conference presentations at no additional charge.  Details are being: > worked out and I will have them in about a week or so...  I Thank you very much! As I said, I'd be glad to pay a reasonable price forv  one, but I can live with free...  1 	26-October, 2001: A day that will live in infamy 4 	Support Freedom: http://www.indefenseoffreedom.org/  J         Guard with jealous attention the public liberty. Suspect every oneJ         who approaches that jewel. Unfortunately, nothing will preserve itE         but downright force. Whenever you give up that force, you are C         inevitably ruined. -- Patrick Henry, speech of June 5 1788     ------------------------------  $ Date: Mon, 4 Nov 2002 17:15:18 -0500/ From: "Encompass - HP-ETS" <KilleenJ@toast.net>  Subject: Re: HP-ETS 2002 CD / Message-ID: <usdsbs8gm2l745@corp.supernews.com>t  K > Thank you very much! As I said, I'd be glad to pay a reasonable price for " > one, but I can live with free...   You do have to be a member   --   Jeff Killeen   All Info: http://www.Killeen.cc   ? ---------------------------------------------------------------a  H "Bob Kaplow" <kaplow_r@eisner.encompasserve.org.TRABoD> wrote in message- news:rPLm0lftDM2r@eisner.encompasserve.org...eF > In article <usd8s6i82u16a3@corp.supernews.com>, "Encompass - HP-ETS" <KilleenJ@toast.net> writes:J > > It is likely that any member of Encompass will be able to request a CD ofL > > the conference presentations at no additional charge.  Details are being< > > worked out and I will have them in about a week or so... >eK > Thank you very much! As I said, I'd be glad to pay a reasonable price for " > one, but I can live with free... > 2 > 26-October, 2001: A day that will live in infamy5 > Support Freedom: http://www.indefenseoffreedom.org/e >eL >         Guard with jealous attention the public liberty. Suspect every oneL >         who approaches that jewel. Unfortunately, nothing will preserve itG >         but downright force. Whenever you give up that force, you are D >         inevitably ruined. -- Patrick Henry, speech of June 5 1788   ------------------------------  % Date: Mon, 04 Nov 2002 21:38:59 -0600a1 From: "David J. Dachtera" <djesys.nospam@fsi.net>t Subject: Re: HP-ETS 2002 CDf& Message-ID: <3DC73D53.FBBF1A9@fsi.net>   Encompass - HP-ETS wrote:s > M > > Thank you very much! As I said, I'd be glad to pay a reasonable price forp$ > > one, but I can live with free...    Strictly my opinion (follows)...   > You do have to be a member  & Well, that gets us up to the B+ mark.   F Anyone wanna try for an A-? How 'bout free to members, non-members pay, $mumble (not more than two digits) plus S&H?   Notes to make for next year:  * Attendees (member or not): free + free S&H' non-attendees (members): $19.95US + S&H + non-attendees (non-members): $49.95US + S&Hl  H Again, strictly my opinion. Your mileage may vary considerably, possibly* even to the point of diametric opposition.   -- a David J. Dachtera0 dba DJE Systemsi http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/2   ------------------------------  # Date: Tue, 05 Nov 2002 03:27:04 GMT:1 From: Michael Austin <maustin@firstdbasource.com>D9 Subject: Re: import/export between oracle 8.1.7 and 8.0.3 2 Message-ID: <3DC7391E.895BFAAD@firstdbasource.com>  
 Manser wrote:  >  > Hi oracle users  > E > i don't know if it is the right place to post but as my systems ares > running alpha VMSt > i will try it.+ > i have the following system configuration  > 1 > node 1 and node 2 are connected via TCPIP (wan)r > G > sqlnet are configured on both sides, tnsnames.ora match in both nodeso' > so connection via sqlplus is possiblee > B > +--------------+                                +--------------+B > !    node1     !                                !   node2      !B > ! vms 7.3      !                                !  vms 7.2-1   !B > ! oracle SRV   !                                ! oracle SRV   !B > !   8.1.7      !                                !   8.0.3      !B > +--------------+                                +--------------+ > ; > catexp.sql and catexp7 were run both at node 1 and node 2L > . > i want to transfer Data from node1 to node 2 > H > 1) i run on node 1 the export utility to create a dump file expdat.dmp: >    i transfer this file via ftp in binary mode to node 2. >    at node 2 i run import to import the file" > NODE2> imp <username>/<password> > A > Import: Release 8.0.3.2.0 - Production on Fri Nov 1 0:8:40 20025 > > > (c) Copyright 1997 Oracle Corporation.  All rights reserved. > > > Connected to: Oracle8 Enterprise Edition Release 8.0.3.2.0 - > Production+ > With the Partitioning and Objects options ' > PL/SQL Release 8.0.3.2.0 - Productiono > 7 > Import file: DISK$ORACLE:[000000.IMPORTS]EXPDAT.DMP >-3 > Enter insert buffer size (minimum is 4096) 30720>0@ > IMP-00010: not a valid export file, header failed verificationA > IMP-00021: operating system error - error code (dec 2, hex 0x2)s- > IMP-00000: Import terminated unsuccessfullyw > > > 2) on node 2 i run the export utility against 8.1.7 (node 1) > 8 > NODE2> exp <username>/<password>@<db_name>.<db_domain> > B > Export: Release 8.0.3.2.0 - Production on Fri Nov 1 0:31:49 2002 > > > (c) Copyright 1997 Oracle Corporation.  All rights reserved. > = > %SYSTEM-F-ACCVIO, access violation, reason mask=00, virtualn< > address=77774500000000E0, PC=0000000000668F40, PS=0000001B1 > %TRACE-F-TRACEBACK, symbolic stack dump followso; >   image    module    routine             line      rel PC- > abs PC) >  ORACLIENT64_V803  NTTUTIL  nttbnd2addrm@ >                                         84536 0000000000000AF0 > 0000000000668F40@ >  ORACLIENT64_V803  NTAC  ntacbnd2addr   80395 0000000000000384 > 000000000056FAC4@ >  ORACLIENT64_V803  NSC2  nsc2addr       91273 0000000000000AA8 > 00000000005CF5B8@ >  ORACLIENT64_V803  NSCALL  nscall1      91477 0000000000000D88 > 00000000005D0468@ >  ORACLIENT64_V803  NSCALL  nscall       91254 000000000000059C > 00000000005CFC7C@ >  ORACLIENT64_V803  NRICALL  nricdt      92178 0000000000001554 > 00000000005C02A4@ >  ORACLIENT64_V803  NRICALL  nricall     91909 0000000000000CC8 > 00000000005BFA18@ >  ORACLIENT64_V803  NIO  niotns         100006 0000000000005E24 > 000000000059D284@ >  ORACLIENT64_V803  NIG  nigcall         95089 0000000000000054 > 0000000000596954@ >  ORACLIENT64_V803  NIGCON  osncon       96370 0000000000000CEC > 000000000057897C@ >  ORACLIENT64_V803  UPIPRV  upiini      194027 0000000000000FCC > 00000000004A351C@ >  ORACLIENT64_V803  UPICDC  upiah0      169170 00000000000000C8 > 00000000004A3CA8@ >  ORACLIENT64_V803  UPICDC  upiahm      169222 0000000000000154 > 00000000004A3D34@ >  ORACLIENT64_V803  KPU  kpuatch        268352 0000000000000848 > 00000000004795D8* >  ORACLIENT64_V803  OCI8  OCIServerAttach@ >                                        162404 0000000000000194 > 0000000000493384@ >  EXP  EXULON  exulon                   259471 0000000000000684 > 00000000000C5584@ >  EXP  EXUDRV  main                     259952 00000000000007E8 > 00000000000B07E8@ >  EXP  EXUDRV  __main                        0 000000000000006C > 00000000000B006C@ >  PTHREAD$RTL                                0 0000000000030E7C > 000000007BC04E7C@ >  PTHREAD$RTL                                0 0000000000012B48 > 000000007BBE6B48@ >                                             0 FFFFFFFF84CBD3D4 > FFFFFFFF84CBD3D4 > + > 3) i try the same with the import utilityj > 8 > NODE2> imp <username>/<password>@<db_name>.<db_domain> > A > Import: Release 8.0.3.2.0 - Production on Fri Nov 1 0:32:3 2002n > > > (c) Copyright 1997 Oracle Corporation.  All rights reserved. > = > %SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual < > address=77774500000000E0, PC=0000000000648F40, PS=0000001B1 > %TRACE-F-TRACEBACK, symbolic stack dump follows ; >   image    module    routine             line      rel PCs > abs PC) >  ORACLIENT64_V803  NTTUTIL  nttbnd2addrt@ >                                         84536 0000000000000AF0 > 0000000000648F40@ >  ORACLIENT64_V803  NTAC  ntacbnd2addr   80395 0000000000000384 > 000000000054FAC4@ >  ORACLIENT64_V803  NSC2  nsc2addr       91273 0000000000000AA8 > 00000000005AF5B8@ >  ORACLIENT64_V803  NSCALL  nscall1      91477 0000000000000D88 > 00000000005B0468@ >  ORACLIENT64_V803  NSCALL  nscall       91254 000000000000059C > 00000000005AFC7C@ >  ORACLIENT64_V803  NRICALL  nricdt      92178 0000000000001554 > 00000000005A02A4@ >  ORACLIENT64_V803  NRICALL  nricall     91909 0000000000000CC8 > 000000000059FA18@ >  ORACLIENT64_V803  NIO  niotns         100006 0000000000005E24 > 000000000057D284@ >  ORACLIENT64_V803  NIG  nigcall         95089 0000000000000054 > 0000000000576954@ >  ORACLIENT64_V803  NIGCON  osncon       96370 0000000000000CEC > 000000000055897C@ >  ORACLIENT64_V803  UPIPRV  upiini      194027 0000000000000FCC > 000000000048351C@ >  ORACLIENT64_V803  UPICDC  upiah0      169170 00000000000000C8 > 0000000000483CA8@ >  ORACLIENT64_V803  UPICDC  upiahm      169222 0000000000000154 > 0000000000483D34@ >  ORACLIENT64_V803  KPU  kpuatch        268352 0000000000000848 > 00000000004595D8* >  ORACLIENT64_V803  OCI8  OCIServerAttach@ >                                        162404 0000000000000194 > 0000000000473384@ >  IMP  IMPLON  implon                   177779 0000000000000664 > 00000000000B2974@ >  IMP  IMPDRV  main                     178285 000000000000075C > 00000000000A075C@ >  IMP  IMPDRV  __main                        0 000000000000006C > 00000000000A006C@ >  PTHREAD$RTL                                0 0000000000030E7C > 000000007BC04E7C@ >  PTHREAD$RTL                                0 0000000000012B48 > 000000007BBE6B48@ >                                             0 FFFFFFFF84CBD3D4 > FFFFFFFF84CBD3D4 > " > 4) on node 1 i try the following > 8 > NODE1> exp <username>/<password>@<db_name>.<db_domain> > C > Export: Release 8.1.7.0.0 - Production on Fri Nov 1 00:07:12 2002s > > > (c) Copyright 2000 Oracle Corporation.  All rights reserved. > ) > EXP-00008: ORACLE error 942 encounteredt) > ORA-00942: table or view does not existu? > EXP-00024: Export views not installed, please notify your DBAa- > EXP-00000: Export terminated unsuccessfully  > ; > 5) NODE1> imp <username>/<password>@<db_name>.<db_domain>  > C > Import: Release 8.1.7.0.0 - Production on Fri Nov 1 00:02:16 2002  > > > (c) Copyright 2000 Oracle Corporation.  All rights reserved. > ) > IMP-00003: ORACLE error 942 encountered-) > ORA-00942: table or view does not exist6? > IMP-00023: Import views not installed, please notify your DBAD- > IMP-00000: Import terminated unsuccessfully  > * > can someone tell me what's going wrong ?" > any help is greatly appreciated. > 
 > reguards >  > Nazim Manser    ; why not just create a database link and copy the data over..  , insert into node2.table select * from table;  = Exp on the 8.1.7 may not be backward compatable with 8.0.3.  t   -- r Regards,  6 Michael Austin            OpenVMS User since June 19847 First DBA Source, Inc.    Registered Linux User #261163i7 Sr. Consultant            http://www.firstdbasource.com    ------------------------------  # Date: Mon, 04 Nov 2002 20:23:05 GMT # From: "John Smith" <a@nonymous.com>0@ Subject: Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeekJ Message-ID: <JGAx9.172573$Q3S.139129@news01.bloor.is.net.cable.rogers.com>  4 "David Mathog" <mathog@caltech.edu> wrote in message2 news:20021104101052.6a2e48ea.mathog@caltech.edu..." > On Fri, 01 Nov 2002 23:04:42 GMT& > "John Smith" <a@nonymous.com> wrote: >bJ > > The old adage, "Where there's smoke, there's fire" probably holds true on	 > > this.r >=A > Naw, it's the new version that should hold true for the companye > stockholders:_ >bE > "What have Carly and Curly been smoking?  Shouldn't they be fired?"-      J Well if they've been to Canada recently, marijuana is practically legal. I; can see the trailers now, "Carly and Curly Go Up In Smoke".S   ------------------------------   Date: 4 Nov 2002 13:47:54 -0800 1 From: keithparris_NOSPAM@yahoo.com (Keith Parris)y@ Subject: Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek= Message-ID: <cf15391e.0211041347.4b7b3031@posting.google.com>o  \ Shane Smith <ssmith@icius.com> wrote in message news:<01C281A6.6053E3E0@sulfer.icius.com>...I > So far we've seen two versions of Itanium not so much in the wild as in  > wildlife parks.p   The Inquirer says:  E "A glitch on the Web site meant that the ZX6000 wasn't easily visibleeB while the ZX 2000, listed as on back order, is actually available,? said Larry Mahoney, Itanium Product Manager at HP's workstationk division in the US today.   @ In fact, said Mahoney, although the HP site showed the ZX2000 onE backorder, it and the ZX8000 were readily available and could ship in  four to five days.  ' 'Both machines are shipping,' he said."   ( http://www.theinquirer.net/?article=6025   ------------------------------   Date: 4 Nov 2002 22:54:40 -0000a= From: Doc.Cypher <Use-Author-Supplied-Address-Header@[127.1]>c@ Subject: Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek4 Message-ID: <20021104225440.577.qmail@gacracker.org>  9 On Mon, 04 Nov 2002, "John Smith" <a@nonymous.com> wrote:,5 >"David Mathog" <mathog@caltech.edu> wrote in messaget3 >news:20021104101052.6a2e48ea.mathog@caltech.edu...d# >> On Fri, 01 Nov 2002 23:04:42 GMTi' >> "John Smith" <a@nonymous.com> wrote:s >>K >> > The old adage, "Where there's smoke, there's fire" probably holds truel >ono
 >> > this. >>B >> Naw, it's the new version that should hold true for the company >> stockholders: >>F >> "What have Carly and Curly been smoking?  Shouldn't they be fired?" >o >5 >aK >Well if they've been to Canada recently, marijuana is practically legal. Ir< >can see the trailers now, "Carly and Curly Go Up In Smoke".  ) Say, isn't that where all the money goes?M  : Of course... why would C&C complain? It isn't their money.       Doc. -- l6 The bigger the humbug, the better people will like it.K ~ Phineas Taylor Barnum.                             https://vmsbox.cjb.netr   ------------------------------  $ Date: Mon, 4 Nov 2002 17:39:24 -0500* From: "Bill Todd" <billtodd@metrocast.net>@ Subject: Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek2 Message-ID: <LmKdnWssyoeBalugXTWc2Q@metrocast.net>  > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message7 news:cf15391e.0211041347.4b7b3031@posting.google.com...e1 > Shane Smith <ssmith@icius.com> wrote in messaget, news:<01C281A6.6053E3E0@sulfer.icius.com>...K > > So far we've seen two versions of Itanium not so much in the wild as inp > > wildlife parks.r >o > The Inquirer says: >dG > "A glitch on the Web site meant that the ZX6000 wasn't easily visibleiD > while the ZX 2000, listed as on back order, is actually available,A > said Larry Mahoney, Itanium Product Manager at HP's workstation  > division in the US today.e >eB > In fact, said Mahoney, although the HP site showed the ZX2000 onG > backorder, it and the ZX8000 were readily available and could ship in  > four to five days. >s) > 'Both machines are shipping,' he said."y >e* > http://www.theinquirer.net/?article=6025  A You should, however, note that all the HP statements in L'Inq are B conspicuously in the present tense.  In other words, systems *are*> available, but there's no indication that they were yesterday.  I This is consistent with a discussion I recently had elsewhere:  HP's July F 8th press releases spoke glowingly about their new Itanic2 toys in theF present tense, but I happened to notice that HP's SPEC submissions allI listed *September* as the hardware availability date.  This is consistentuH with the earliest actual delivery date I've heard of (latish September),H which seems to suggest that the July launch was indeed a 'paper' one butF that at least some systems starting making it out the door a couple of
 months later..   - bill   ------------------------------   Date: 4 Nov 2002 22:56:49 GMT-& From: Rick Jones <foo@bar.baz.invalid>@ Subject: Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek* Message-ID: <aq6tvh$nat$1@web1.cup.hp.com>  ) Bill Todd <billtodd@metrocast.net> wrote:l? > No, it doesn't.  Published benchmarks have been selected withoA > extreme care to avoid head-to-head comparisons, save for SPECfpe@ > (where Itanic indeed excels, with the aid of feedback-directedE > optimization in the base tests - which might bring the newest Alphae8 > within striking distance if it had used them as well)   C For baseline, one can only use a limited number of flags, so adding / feedback would mean taking something else away.t  B Have any of the SPECfp2000 results for Alpha, going back to 1999 -D which appears to be the earliest SPECcpu2000 results - used feedbackC directed optimizations in baseline?  I've made a cursory check, andcD while not terribly familiar with the syntax for compilers for Alpha,@ didn't see any options in the baseline flags that appeared to be invoking such an optimization.  B > and SPECint (where McKinley doesn't knock anyone's socks off butF > does more or less equal the best of the current 64-bit competition -: > though, again, only by virtue of using feedback-directed2 > optimizations that the competition doesn't use).  > Again, I am no compiler flag expert, but looking at a bunch ofC SPECint2000 results as returned by the SPEC search engine, it seemstD that the ones using x86 compilers on Xeon (yes, that is only 32 bit) and such were using feedback.n   The one for the ES45:-  M http://www.spec.org/osg/cpu2000/results/res2002q3/cpu2000-20020801-01512.html   A did seem to use a feedback flag (+CFB). It looks like the SPARC64 ; results from Fujitsu also used feedback, as did the POWER4,I? UltraSPARC-III and PA-RISC figures. And of course, the Itanium2c results.  A > But as for other benchmarks: SPECweb99_SSL is nowhere nearly as # > popular a benchmark as SPECweb99,   < It would be good to define how you are measuring popularity.  F SPECweb99_SSL was announced on the 2nd of April, 2002, or seven monthsD of life thusfar. Since that time, 28 SPECweb99_SSL results have been@ published on www.spec.org. In that same time, there have been 25C SPECweb99 results published. That is a total of 53 between the two.r  D In the first seven months of SPECweb99's life there were indeed moreE results published - 59 if I take results published before April, 2001hE (the original SPECweb99 announcement was mid-month in August of 2000,DC and the public SPEC search engine only does "whole months".)  ThosehF results are across the same broad families of CPUs as the current crop= of SPECweb99_SSL results - x86, SPARC, PA-RISC, POWER, Alpha.f  @ However, I think it is interesting that the sum of SPECweb99 and7 SPECweb99_SSL submissions for the first seven months of A SPECweb99_SSL's life is so close to that of the first 7 months of)A SPECweb99's life. My experience thusfar is that each benchmark isnD about the same degree of effort to run. (for completeness, I checkedA for SPECweb96 benchmarks published after the announcement date ofn6 SPECweb99 (August 2000), and there appear to be none.)  E > hence does not expose Itanic to competition from, e.g., other chipsyC > with large on-chip caches (the performance of PA-RISC and Xeon inPC > SPECweb99_SSL suggests that they may be a very important factor --  A Seems one would need cache miss-rate per SPECweb99_SSL figures tos validate that assertion.  B How small can on-chip cache be before it is not large? Would it beF correct to assume that you believe that the Xeons with a 512KB on-chip@ are not large? Is PA-RISC in the form of the PA-8700 used in theD SPECweb99_SSL results a chip with a large on-chip cache? If so wouldD then the POWER4 chip used in the IBM p630 SPECweb99_SSL results with7 its 1440KB on-chip cache also be a large on-chip cache?t  
 rick jones -- r= denial, anger, bargaining, depression, acceptance, rebirth...e where do you want to be today?F these opinions are mine, all mine; HP might not want them anyway... :)A feel free to post, OR email to raj in cup.hp.com  but NOT BOTH...    ------------------------------  $ Date: Mon, 4 Nov 2002 20:00:29 -0500* From: "Bill Todd" <billtodd@metrocast.net>@ Subject: Re: Intel Conceding 64-Bit Desktops to IBM, AMD - eWeek2 Message-ID: <ItycnWYTmcWwhVqgXTWcpw@metrocast.net>  3 "Rick Jones" <foo@bar.baz.invalid> wrote in messagee$ news:aq6tvh$nat$1@web1.cup.hp.com...+ > Bill Todd <billtodd@metrocast.net> wrote:oA > > No, it doesn't.  Published benchmarks have been selected withiC > > extreme care to avoid head-to-head comparisons, save for SPECfpiB > > (where Itanic indeed excels, with the aid of feedback-directedG > > optimization in the base tests - which might bring the newest Alpha 9 > > within striking distance if it had used them as well)r >.E > For baseline, one can only use a limited number of flags, so adding 1 > feedback would mean taking something else away.i > D > Have any of the SPECfp2000 results for Alpha, going back to 1999 -F > which appears to be the earliest SPECcpu2000 results - used feedbackE > directed optimizations in baseline?  I've made a cursory check, andeF > while not terribly familiar with the syntax for compilers for Alpha,B > didn't see any options in the baseline flags that appeared to be  > invoking such an optimization.  E I only checked the most recent one(s), but I didn't find them either.b   >iD > > and SPECint (where McKinley doesn't knock anyone's socks off butH > > does more or less equal the best of the current 64-bit competition -< > > though, again, only by virtue of using feedback-directed4 > > optimizations that the competition doesn't use). >=@ > Again, I am no compiler flag expert, but looking at a bunch ofE > SPECint2000 results as returned by the SPEC search engine, it seemsOF > that the ones using x86 compilers on Xeon (yes, that is only 32 bit) > and such were using feedback.  >0 > The one for the ES45:E >4 >0L http://www.spec.org/osg/cpu2000/results/res2002q3/cpu2000-20020801-01512.htm l4 >0) > did seem to use a feedback flag (+CFB).A  < I did notice that one, but the other two flags (which may beK SPIKE-specific - my knowledge in that area is minimal) weren't present.  OfcJ course, the fact that Alpha uses 3 feedback flags rather than Itanic's oneK (IIRC) would penalize it for its additionally fine-grained compiler control4J (guess compiler flag definitions should start being optimized for SPEC :).    It looks like the SPARC64= > results from Fujitsu also used feedback, as did the POWER4,TA > UltraSPARC-III and PA-RISC figures. And of course, the Itanium29
 > results.  I Blargh - mea culpa.  Shouldn't have repeated hearsay without checking theC sources myself:  thanks.   >uC > > But as for other benchmarks: SPECweb99_SSL is nowhere nearly as4% > > popular a benchmark as SPECweb99,0 >0> > It would be good to define how you are measuring popularity.  L By the number and range of submissions, thus allowing comparisons to be madeJ against a considerably wider spectrum of competition (and, in the process,G providing more information from which to draw inferences about possible0> correlations between implementation features and performance).   ...   G > > hence does not expose Itanic to competition from, e.g., other chips E > > with large on-chip caches (the performance of PA-RISC and Xeon inDE > > SPECweb99_SSL suggests that they may be a very important factor -C > C > Seems one would need cache miss-rate per SPECweb99_SSL figures to0 > validate that assertion.  L But not to raise it as a reasonable possibility, which you will note is what I did above.   > D > How small can on-chip cache be before it is not large? Would it beH > correct to assume that you believe that the Xeons with a 512KB on-chip > are not large?  K No, it would not - not by comparison with Alpha's 64 KB instruction plus 64 I KB data on-chip cache:  unless the benchmark's performance were carefullySK balanced between I-cache and D-cache size, 512 KB would be close to 8 times0J the size of the relevant Alpha on-chip cache, while only 1/6th the size of, the largest on-chip cache tested (Itanic's).  E Furthermore, since the 4-processor Xeon result is obtained using 1 MBLK on-chip cache, you seem to be pointing out the 2-processor result.  If thatAI is to suggest that its high per-processor performance with less cache may0A indicate a lack of correlation with cache size, note  1) that thedH 2-processor Xeon box is using 2.4 GHz Xeons vs. the 1.6 GHz Xeons in theF 4-processor box and  2) that Xeon scaling is also far less than linearJ between 4- and 8-processor boxes (I can't say I know why, but it does makeI comparisons between Xeon boxes and anything else with different processor0K counts at least somewhat suspect, even though some of the HP results indeed * scale very linearly with processor count).  2  Is PA-RISC in the form of the PA-8700 used in the: > SPECweb99_SSL results a chip with a large on-chip cache?  I With 1.5 MB of data cache plus another 768 KB of I-cache?  Definitely, at6 least by comparison with Alpha.0    If so wouldF > then the POWER4 chip used in the IBM p630 SPECweb99_SSL results with9 > its 1440KB on-chip cache also be a large on-chip cache?0   Yup.   - bill   ------------------------------  $ Date: Mon, 4 Nov 2002 17:55:54 -0500* From: "Bill Todd" <billtodd@metrocast.net> Subject: Karma2 Message-ID: <NaydnXHoCp1kZ1ugXTWcpQ@metrocast.net>  I L'Inq reports that Avalanche (the server solution favored by Dave Fenwick L which he claimed Alpha wasn't particularly well-suited for) has been dumped:  ( http://www.theinquirer.net/?article=6043   - bill   ------------------------------   Date: 4 Nov 2002 13:16:20 -0600C- From: koehler@encompasserve.org (Bob Koehler) ) Subject: Microsoft strikes (itself) again03 Message-ID: <IvfVfu5dWmbR@eisner.encompasserve.org>   =    The following page would not open from Netscape 6.2 today:          www.microsoft.com   7    maybe they don't want those kind of people any more?    ------------------------------  % Date: Mon, 04 Nov 2002 13:46:18 -0800o% From: Dean Woodward <deanw@rdrop.com>E- Subject: Re: Microsoft strikes (itself) againi( Message-ID: <3DC6EAAA.8020204@rdrop.com>   Bob Koehler wrote:? >    The following page would not open from Netscape 6.2 today:i >  >       www.microsoft.com: > 9 >    maybe they don't want those kind of people any more?   > I never got along with NS 6.x.  The above page works for NS 7.   ------------------------------  # Date: Tue, 05 Nov 2002 01:57:14 GMT18 From: "Jerome H. Fine" <jhfineb9rv@b9rvnospamcompsys.to>- Subject: Re: Microsoft strikes (itself) againo4 Message-ID: <3DC7256F.C23D6903@b9rvnospamcompsys.to>   >Dean Woodward wrote:s   > > Bob Koehler wrote:> > > The following page would not open from Netscape 6.2 today: > >  www.microsoft.com; > >    maybe they don't want those kind of people any more?0@ > I never got along with NS 6.x.  The above page works for NS 7.   Jerome Fine replies:  = I just tried with Netscape 4.78 - while it works, the page isr= so difficult to read that it might as well have been blocked.r; I used the mouse and clicked on one of the icons in the topt9 right hand corner and for some reason it all disappeared.-9 I guess that is what I was supposed to do with a web sitee* that can't be used with standard software.   http://www.microsoft.com/    Sincerely yours,   Jerome Finer --4 To obtain the original e-mail address, please remove5 the ten characters which immediately follow the 'at'.@8 If you attempted to send a reply and the original e-mail7 address has been discontinued due a high volume of junk15 e-mail, then the semi-permanent e-mail address can beo7 obtained by replacing the four characters preceding the1. 'at' with the four digits of the current year.   ------------------------------  $ Date: Mon, 4 Nov 2002 17:57:20 -0800$ From: Shane Smith <ssmith@icius.com>- Subject: RE: Microsoft strikes (itself) again 0 Message-ID: <01C2842B.AC0DC940@sulfer.icius.com>  H It's fine with Mozilla 1.2b, even while rejecting cookies and javascript off. Interesting.'   Shane    -----Original Message-----= From: Jerome H. Fine [mailto:jhfineb9rv@b9rvnospamcompsys.to]a' Sent: Monday, November 04, 2002 5:57 PMG To: Info-VAX@Mvb.Saic.Com-- Subject: Re: Microsoft strikes (itself) again7     >Dean Woodward wrote:i   > > Bob Koehler wrote:> > > The following page would not open from Netscape 6.2 today: > >  www.microsoft.com; > >    maybe they don't want those kind of people any more?m@ > I never got along with NS 6.x.  The above page works for NS 7.   Jerome Fine replies:  = I just tried with Netscape 4.78 - while it works, the page iss= so difficult to read that it might as well have been blocked.t; I used the mouse and clicked on one of the icons in the topl9 right hand corner and for some reason it all disappeared.l9 I guess that is what I was supposed to do with a web sitek* that can't be used with standard software.   http://www.microsoft.com/h   Sincerely yours,   Jerome Finek --4 To obtain the original e-mail address, please remove5 the ten characters which immediately follow the 'at'.p8 If you attempted to send a reply and the original e-mail7 address has been discontinued due a high volume of junk25 e-mail, then the semi-permanent e-mail address can bed7 obtained by replacing the four characters preceding thei. 'at' with the four digits of the current year.   ------------------------------  % Date: Mon, 04 Nov 2002 21:41:24 -0500a- From: JF Mezei <jfmezei.spamnot@videotron.ca>@- Subject: Re: Microsoft strikes (itself) again0, Message-ID: <3DC72FCF.EBCF156C@videotron.ca>   "Jerome H. Fine" wrote:i? > I just tried with Netscape 4.78 - while it works, the page ise? > so difficult to read that it might as well have been blocked.   M MAC, Netscape 4.76, Javascript turned off, it renders fine and readable, eveny with style sheets enabled.  L But there are many wintel-powered sites that are unreadable due to the smallI font they spefify. This si why I generally go with style-sheets disabled.r   ------------------------------  % Date: Mon, 04 Nov 2002 23:57:45 -0500t. From: Lyndon Bartels <lbartels@pressenter.com># Subject: Monitoring NIC utilizationi. Message-ID: <3DC70979.52948A85@pressenter.com>  
 Hello all,  F Today, I got asked "Is there a slowdown?" And I couldn't answer. I can@ see the number of packets going across the interface (TCPIP SHOW
 INTERFACE)  G But can I monitor, in real-time, or at least really close to real time,R the load on a given NIC? e   If so, how?A     Thanks in advance,   Lyndon     -- WG My opinions are mine and mine alone. They seldom align with those of my 	 employer.     H The only good thing about putting the cart before the horse is you don't have to look at the horse's butt.   ------------------------------  % Date: Tue, 05 Nov 2002 00:38:02 -0500s- From: JF Mezei <jfmezei.spamnot@videotron.ca> ' Subject: Re: Monitoring NIC utilization:+ Message-ID: <3DC7592A.3F76AB3@videotron.ca>t   Lyndon Bartels wrote:iI > But can I monitor, in real-time, or at least really close to real time,, > the load on a given NIC?  K Look in the freeware or decus archives for ethermon.  Gives you a "monitor"e@ style output, as well as ability to trace the etherenet traffic.  L And at one point, one could patch MONITOR.EXE to enable the MONITOR ETHERNET" class. But I think it is gone now.   ------------------------------  $ Date: Tue, 5 Nov 2002 17:03:33 +1100L From: forSALTsythm@optushome.com.aSALTu (Mark(desalinate for e-mail)Forsyth)' Subject: Re: Monitoring NIC utilizationi9 Message-ID: <slrnasenpj.63.forSALTsythm@plague.bogus.com>r  Z On Mon, 04 Nov 2002 23:57:45 -0500, Lyndon Bartels <lbartels@pressenter.com> gushed forth: >Hello all,e >eG >Today, I got asked "Is there a slowdown?" And I couldn't answer. I cancA >see the number of packets going across the interface (TCPIP SHOWs >INTERFACE)i >iH >But can I monitor, in real-time, or at least really close to real time, >the load on a given NIC?  >c >If so, how?    H $ ucx help sho interf /cont may well be where the answer is. If you need) more detail there's always ethermon from p8 ftp://ftp.process.com/vms-freeware/fileserv/ethermon.zip   -- v Ooroot	 Mark F...f  $ Another Optus Cable Traffic Monitor.3 http://www.members.optushome.com.au/forsythm/traff/o  H Today is Prickle-Prickle, the 17th day of The Aftermath in the YOLD 3168   ------------------------------  % Date: Mon, 04 Nov 2002 23:10:35 +0100q) From: Georges A. Tomazi <gt@diapason.com>S Subject: OpenVMS & XDMCP8 Message-ID: <9srdsu0g94dfeja6j5qn6qk1p4ksnuaeqg@4ax.com>   Hi -  , I just installed OpenVMS 7.2 on a DEC Alpha.  8 TCP/IP is up and running, as well as CDE on the console.  E Now, I'm trying to find out how can I use a  X remote display, such a & X11 server running on a PC with XDMCP.   Any hints ?h   Thanks,p   Georges    --$ Georges A. Tomazi - gt@sunwizard.net   ------------------------------  # Date: Mon, 04 Nov 2002 23:23:54 GMT . From: peter@langstoeger.at (Peter LANGSTOEGER) Subject: Re: OpenVMS & XDMCP5 Message-ID: <ekDx9.141720$aa2.1671022@news.chello.at>   d In article <9srdsu0g94dfeja6j5qn6qk1p4ksnuaeqg@4ax.com>, Georges A. Tomazi <gt@diapason.com> writes:- >I just installed OpenVMS 7.2 on a DEC Alpha.h >s9 >TCP/IP is up and running, as well as CDE on the console.m >iF >Now, I'm trying to find out how can I use a  X remote display, such a' >X11 server running on a PC with XDMCP.  >m >Any hints ?  B Enable XDM (item 21 in server menue) with SYS$STARTUP:TCPIP$CONFIG   --   Peter "EPLAN" LANGSTOEGERM% Network and OpenVMS system specialista E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  % Date: Tue, 05 Nov 2002 01:44:24 +0100 ) From: Georges A. Tomazi <gt@diapason.com>  Subject: Re: OpenVMS & XDMCP8 Message-ID: <805esu8c8c9s6dohdi215sum2vabgk90at@4ax.com>   Hi -  = On Mon, 04 Nov 2002 23:23:54 GMT, peter@langstoeger.at (Peter  LANGSTOEGER) wrote:a  : >>TCP/IP is up and running, as well as CDE on the console. >>G >>Now, I'm trying to find out how can I use a  X remote display, such at( >>X11 server running on a PC with XDMCP. >>
 >>Any hints ?t >BC >Enable XDM (item 21 in server menue) with SYS$STARTUP:TCPIP$CONFIGf   I don't have that choice :  =         DIGITAL TCP/IP Services for OpenVMS SERVER Components: Configuration Menu  t         Configuration options:.                  1  -  BIND           Disabled.                  2  -  BOOTP          Disabled.                  3  -  TFTP           Disabled-                  4  -  FTP            Enabled .                  5  -  LPR/LPD        Disabled.                  6  -  NFS            Disabled.                  7  -  PC-NFS         Disabled.                  8  -  PORTMAPPER     Disabled-                  9  -  TELNET/RLOGIN  Enabled .                 10  -  SNMP           Disabled.                 11  -  NTP            Disabled.                 12  -  METRIC         Disabled.                 13  -  POP            Disabled.                 14  -  FINGER         Disabled.                 15  -  RMT            Disabled.                 16  -  LBROKER        Disabled.                 17  -  DHCP           Disabled/                  A  -  Configure options 1 - 17w                  [E] -  Exit menu    Enter configuration option:    Georgesl   --$ Georges A. Tomazi - gt@sunwizard.net   ------------------------------  $ Date: Mon, 4 Nov 2002 17:13:30 -0800$ From: Shane Smith <ssmith@icius.com> Subject: RE: OpenVMS & XDMCP0 Message-ID: <01C28425.9E5C7130@sulfer.icius.com>  F XDCMP is a relatively recent addition to TCP/IP Services. You may needF to update your stack. Alternatively, set up applications on the PC end* that go directly to specific IP addresses.  D If it's any help, in eXcursion you can create an application to run  SYS$SYSTEM:DECW$SESSIONsF and get a DECWindows/Motif session started. I assume there's something6 similar for CDE, but I didn't look very hard for that.   Shaneu   -----Original Message-----0 From: Georges A. Tomazi [mailto:gt@diapason.com]' Sent: Monday, November 04, 2002 4:44 PMH To: Info-VAX@Mvb.Saic.Com! Subject: Re: OpenVMS & XDMCP       Hi -  = On Mon, 04 Nov 2002 23:23:54 GMT, peter@langstoeger.at (Peter@ LANGSTOEGER) wrote:R  : >>TCP/IP is up and running, as well as CDE on the console. >>G >>Now, I'm trying to find out how can I use a  X remote display, such a ( >>X11 server running on a PC with XDMCP. >>
 >>Any hints ?n >iC >Enable XDM (item 21 in server menue) with SYS$STARTUP:TCPIP$CONFIGI   I don't have that choice :  =         DIGITAL TCP/IP Services for OpenVMS SERVER Components  Configuration Menu  d         Configuration options:.                  1  -  BIND           Disabled.                  2  -  BOOTP          Disabled.                  3  -  TFTP           Disabled-                  4  -  FTP            Enabledo.                  5  -  LPR/LPD        Disabled.                  6  -  NFS            Disabled.                  7  -  PC-NFS         Disabled.                  8  -  PORTMAPPER     Disabled-                  9  -  TELNET/RLOGIN  Enablede.                 10  -  SNMP           Disabled.                 11  -  NTP            Disabled.                 12  -  METRIC         Disabled.                 13  -  POP            Disabled.                 14  -  FINGER         Disabled.                 15  -  RMT            Disabled.                 16  -  LBROKER        Disabled.                 17  -  DHCP           Disabled/                  A  -  Configure options 1 - 17                   [E] -  Exit menu  w Enter configuration option:    Georgese   --$ Georges A. Tomazi - gt@sunwizard.net   ------------------------------  # Date: Tue, 05 Nov 2002 01:19:07 GMTi. From: peter@langstoeger.at (Peter LANGSTOEGER) Subject: Re: OpenVMS & XDMCP5 Message-ID: <f0Fx9.142929$aa2.1679960@news.chello.at>n  d In article <805esu8c8c9s6dohdi215sum2vabgk90at@4ax.com>, Georges A. Tomazi <gt@diapason.com> writes:> >On Mon, 04 Nov 2002 23:23:54 GMT, peter@langstoeger.at (Peter >LANGSTOEGER) wrote: >a; >>>TCP/IP is up and running, as well as CDE on the console." >>>aH >>>Now, I'm trying to find out how can I use a  X remote display, such a) >>>X11 server running on a PC with XDMCP.o >>>U >>>Any hints ? >>D >>Enable XDM (item 21 in server menue) with SYS$STARTUP:TCPIP$CONFIG >. >I don't have that choice :e   Then upgrade TCPIP.    --   Peter "EPLAN" LANGSTOEGERu% Network and OpenVMS system specialistA E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------   Date: 5 Nov 2002 00:31:24 GMTe# From: hoff@hp.nospam (Hoff Hoffman)p9 Subject: Re: OpenVMS printing to printer via windows 2000A* Message-ID: <aq73gs$k3k$3@web1.cup.hp.com>  e In article <1d08b916.0211010814.41bdc271@posting.google.com>, mb301@hotmail.com (Mark Bowman) writes:,E :Does anyone know how to print to local attached printer from OpenVMS  :which is running Windows 2000?   L   Yes, you can do this, but having this pass-through printing configuration J   is (IMNSHO) more effort and more complexity than the benefit -- you willI   particularly want to ask for help configuring the Microsoft Windows LPDrI   package from the Microsoft folks, too.  Troubleshooting this problem -- G   LPR jobs disappearing -- will involve at least some rummaging around OH   inside the Microsoft Windows LPD configuration and logging mechanisms.  C   I would get a NIC for the printer, either internal or external.  t  J   This approach will be easier, simpler, more reliable, and more flexible.  G   Various Ask The Wizard topics are related to this discussion, but I'daF   start with topic (1020).  For some peripherally related discussions E   (and pointers to Microsoft), see topics (2507), (3971), (4032), andSG   (6316) -- the OpenVMS configuration is documented in the manuals, buttJ   you'll have to sort out the Microsoft LPD configuration and connectivityI   with folks that are more knowledgeable in Microsoft Windows and its LPDs9   implementation, with the Microsoft knowledge base, etc.u  I   Since you will undoubtedly not listen to me here, you will end up usingtE   telnet to connect to the LPD port on the Microsoft box, then send asF   formfeed or some such -- send some dreck to the daemon that makes itG   think you are an incoming print job -- and see what happens.  The useeG   of ping and (if any -- doesn't everyone use ping? :-) other availablen9   IP connectivity-testing tools, of course, also applies.s  G   It is far easier (again, IMNSHO) to hook the printer to a NIC, and touH   print directly from all of the platforms -- the direct-attached schemeJ   works nicely in very small networks, but the scheme does not scale well.F   Standalone options (HP JetDirect, various third-party NICs, etc) areE   available quite inexpensively from HP and various vendors, and mosttG   (all?) newer mid-range (and higher) printers offer integrated NICs ass   an add-on or option.    N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comq   ------------------------------   Date: 4 Nov 2002 13:14:46 -0600h- From: koehler@encompasserve.org (Bob Koehler)t Subject: Re: Palladium article3 Message-ID: <JlOCQ+XJPyuA@eisner.encompasserve.org>n  b In article <3dc6c1dd$1@news.si.com>, "Brian Tillman" <tillman_brian@notnoone.notnohow.com> writes:G >>   You ever install Exceed?  It puts an icon on your desktop which isi% >>   not in the file system anywhere.e > 0 > But Microsoft's TweakUI might let you hide it.  B    The TweakUI that I could find was for W95.  I'm running WNT and    W2K.n   ------------------------------  % Date: Mon, 04 Nov 2002 21:48:49 -0500 + From: John Johnstone <jj_usenet2@yahoo.com>e1 Subject: Problem with /AUTOSTART_ON on VMS V7.3-1a( Message-ID: <3DC6EB40.CB51FEC@yahoo.com>  E There seems to be a problem with the /AUTOSTART_ON qualifier with VMSyE V7.3-1 and TCPIP V5.3 ECO 1.  The following command works on a systemo& running VMS V7.3 and TCPIP V5.3 ECO 1:  8 $ INITIALIZE/QUEUE/START/AUTOSTART_ON=node2::"pr1:9100"-,    /RECORD_BLOCKING/PROCESSOR=UCX$TELNETSYM-    /SCHEDULE=(NOSIZE) pr1m  E The same command issued on a system running VMS V7.3-1 and TCPIP V5.3n ECO 1 results in:a  *    %SYSTEM-F-IVDEVNAM, invalid device name   and OPCOM messages:-  9    TCPIP$TELNETSYM - (PR1) Invalid /ON="host:port" formatr  D    %QMAN-I-QUENOTSTART, queue PR1 could not be started on node NODE2  :    -QMAN-I-QUEAUTOOFF, queue PR1 is now autostart inactive  *    -SYSTEM-F-IVDEVNAM, invalid device name  F The initialize command also works on VMS V7.3-1 if you use /ON instead of /AUTOSTART_ON.u  % Has anyone seen this with VMS V7.3-1?o   ------------------------------   Date: 4 Nov 02 20:16:31 +0100e) From: p_sture@elias.decus.ch (Paul Sture) " Subject: Re: Remote Console access) Message-ID: <47QynHIxt$9y@elias.decus.ch>t   In article <rdeininger-0211020935120001@1cust151.tnt1.nashua.nh.da.uu.net>, rdeininger@mindspring.com (Robert Deininger) writes: > In articleK > <BE56C50EA024184DAF48F0B9A47F5CF402660B0E@kaoexc01.americas.cpqcorp.net>, * > "Main, Kerry" <Kerry.Main@hp.com> wrote: > 	 >>Robert,x >>G >>>> IMHO, the Consoleworks/PC/Windows package that is foisted with thehA >>Alphaserver GS series is horrid.  What were they thinking?  The H >>Billy-ware has a seizure, and your GS-320 either hangs or crashes as a
 >>result. <<<t >>F >>If Billy-ware hangs or crashes, then it should not impact OpenVMS at >>all.   > K > If the steenking billy-box yanks on the alpha's console port hard enough, $ > it can leave VMS in a lot of pain. > C >>If it does, then imho, OpenVMS (or Tru64) has a bug. ConsoleWorksOI >>is simply collecting, analyzing and storing the ascii output being sent , >>to it (from any system/device it supports) > I > I don't know the names of the various components and layers that are intK > the typical "console" billy-box.  I agree that collecting, analyzing, andoL > storing should not effect VMS.  But transmitting is a different story, andF > something in that evil box is able to capture keystrokes from the PCH > keyboard and transmit them (modified) to the alpha's console port.  IfJ > that component falls over, or becomes possessed by a billy-demon, it can# > do damage to the VMS environment.e > J > When a PC "hangs", it's usually not really dead.  The processor is stillL > executing code from hell.  Sometimes that code seems to concentrate on the > console port.S > K > I've seen the PC take out GS-series systems several times.  I've seen the E > PC keep the GS from powering on completely.  I've seen irate system-A > managers form a posse, with wire cutters, and go after the PC. -L > Disconnecting it from the GS box and installing a VT terminal rendered theL > GS box operational.  I know of customers who have logged calls complainingK > about the PC console management.  One root cause I heard about was simple L > -- a corner of a notebook was accidentally leaning on the space bar of theJ > PC, and the GS system (this one was running Tru64) _went away_ until theD > notebook was found and removed.  (After which the PC "went away".) > I > Oh, and did I mention that the billy-ware on the PC is a perfect little L > petri-dish for hosting all sorts of viruses?  Totally harmless, of course,L > unless one of them types a CTRL-P while your alpha is trying to process an > important transaction. > D > I've never heard of a single system manager who was happy with theH > PC-based console on a GS-series alphaserver.  But then I don't get out > much.t >w  B For me, it's simply a matter of having an extra piece of kit which@ is not under my control. Give me a good old fashioned VT please.  r? > I don't think billy-ware should be permitted in a data centeri0 > environment.  I guess that makes me a radical. > I >>Btw - ConsoleWorks server can run natively on a VMS box (VAX or Alpha).e6 >>It does not need Wintel if Cust does not want it.=20 > B > I've heard this rumor, but I've never seen it in non-vapor form.  B I've been looking at a demo version of ConsoleWorks and indeed theC server does run on VMS. There is a browser interface which actuallyu@ works with Netscape 3.03 (doesn't look as pretty as on a PC, but2 it appears to work). There's also a DCL interface.   --  
 Paul Sture Switzerlands   ------------------------------  # Date: Tue, 05 Nov 2002 03:22:10 GMTJ1 From: Michael Austin <maustin@firstdbasource.com>f" Subject: Re: Remote Console access2 Message-ID: <3DC737F5.C818D932@firstdbasource.com>   Bart Zorn wrote: >  > Main, Kerry wrote: > > Bart, Robert - > >eK > > ConsoleWorks started as an OpenVMS based product (VAX and Alpha), so ito@ > > is a very good VMS based product for managing many consoles. > >aL > > If Customers or Prod Mgrs :-) decide to implement PC platforms to manageK > > important systems, then that is their decision, but it can certainly beo > > done in other ways.e > >t > > Reference:; > > http://www.tditx.com/services_support_faqs.html#consolea > >oG > > (scroll down and look at the OS listed first under supported servere	 > > OS's)n > >e > > :-)t > >h > > Regards  > >  > > Kerry Main > > Senior Consultantc  > > Hewlett-Packard (Canada) Co.% > > Consulting & Integration Servicesi > > Voice: 613-592-4660d > > Fax   : 613-591-4477  > > Email: kerryDOTmain@hpDOTcom1 > >     (remove the DOT's and replace with "."'s)t > J > Fair enough! However, as a consultant who has not been involved with theF > initial configuration of my current client's Wildfire systems, I seeG > that they are configured with billyboxes in such a way that those arewI > definitely a single point of failure with a miserable user interface onhE > top of it (you can't even use cut and paste in the console window!)a > H > The only way to make them usable is through the ConsoleWorks client onF > OpenVMS. But there I have not found a way to do something similar asG > Console Manager's CONSOLE CONNECT <node>. You have to do it in a muchi% > more complex way (logging in, e.a.)t >  > Bart Zornc    F I work for a company that uses consoleworks and I absolutely love it. G It connects using a TCPIP-based DECSERVER700 (16 ports). We have over a-G hundred VMS nodes plus 2 ports for each of 28 HSG80's (soon to be 36). .G The logging facility is a great tracking tool. It can identify everyoneeF that connects to the consoles.  You can even <ctrl>P the systems.  YouE can do everything but power it on - unless of course it is a GS160 ore? 320 to which you are connected (and the main breakers are on ofs	 course)..S  H I can even use a VPN connection to the corporate network and use the WebH or VMS interface to talk to any console in the data center.  Now that isH really cool especially when you don't want to go to the office to in theH middle of the night. Another neat feature is that you can have more thanF one person looking at this console at the same time.  I can "watch" as9 Compaq Support connects to the console to do some seriousSE troubleshooting.  It is funny sometimes to have dueling keyboards. :)    -- a Regards,  6 Michael Austin            OpenVMS User since June 19847 First DBA Source, Inc.    Registered Linux User #261163 7 Sr. Consultant            http://www.firstdbasource.comm   ------------------------------  % Date: Mon, 04 Nov 2002 13:18:28 -0600b& From: tolkien <limer1don@netscape.net>" Subject: Setting NIC to 100mb/Full8 Message-ID: <fthdsu8jl5271obub25mnk6ir2c9jvrbml@4ax.com>  D I have a DS20E Alpha server running OpenVMS 7.2-1 and DIGITAL TCP/IP) Services for OpenVMS Alpha Version V5.0A.   B How do I go about setting the NIC to 100mb Full from 10mb Half and* have it stay that way even after a reboot.     Thanks In Advance    ------------------------------  * Date: Mon, 4 Nov 2002 20:43:03 +0000 (UTC), From: lewis@mazda.mitre.org (Keith A. Lewis)& Subject: Re: Setting NIC to 100mb/Full. Message-ID: <aq6m4n$g8j$1@newslocal.mitre.org>   tolkien <limer1don@netscape.net> writes in article <fthdsu8jl5271obub25mnk6ir2c9jvrbml@4ax.com> dated Mon, 04 Nov 2002 13:18:28 -0600:E >I have a DS20E Alpha server running OpenVMS 7.2-1 and DIGITAL TCP/IP * >Services for OpenVMS Alpha Version V5.0A. >dC >How do I go about setting the NIC to 100mb Full from 10mb Half and + >have it stay that way even after a reboot.e       >show *_mode*     [list of NIC card modes currently set]4     [assuming you want to change the value for ewa0]     >set ewa0_mode fastfd5  K These settings are supposed to survive a reboot, power-cycle, or just abouteL anything else.  However, the hardware that supports VMS booting isn't always  up to VMS runtime standards...    G Yesterday I power-cycled my 500au (the Unix video card stays off unlessiM there's a monitor connected during bootstrap) and it booted the ARC console! )E It took some archive searching to find out how to go from ARC to SRM.r  I (Just so it's written down in one more place: "F2 - Setup", "CMOS Setup",oN "F6 - Advanced", and then it's obvious -- "boot a different OS" or something.)  + --Keith Lewis              klewis$mitre.orge> The above may not (yet) represent the opinions of my employer.   ------------------------------  $ Date: Mon, 4 Nov 2002 14:46:52 -0600* From: "Dave Gudewicz" <dgudewicz@core.com>& Subject: Re: Setting NIC to 100mb/Full9 Message-ID: <3dc6dcbd$0$17651$1dc6e903@news.corecomm.net>g  	 $mc lancpdK lancp>set dev ewa0: /speed=100/full_dup    (check the /perm qualifiers too)r lancp>show dev ewa0 /chari# lancp>help    (for further details)n
 lancp>exit     -- Dave...   G It is noble to teach oneself, but still nobler to teach others-and less  trouble. -----Mark Twaina  3 "tolkien" <limer1don@netscape.net> wrote in messageo2 news:fthdsu8jl5271obub25mnk6ir2c9jvrbml@4ax.com...F > I have a DS20E Alpha server running OpenVMS 7.2-1 and DIGITAL TCP/IP+ > Services for OpenVMS Alpha Version V5.0A.S >nD > How do I go about setting the NIC to 100mb Full from 10mb Half and, > have it stay that way even after a reboot. >W >  > Thanks In Advancep   ------------------------------   Date: 4 Nov 2002 12:01:47 -0800r( From: bob@instantwhip.com (Bob Ceculski)) Subject: Re: Slow performance with BACKUPa= Message-ID: <d7791aa1.0211041201.1d9758d5@posting.google.com>s  d Rudolf Wingert <win@fom.fgan.de> wrote in message news:<200211040952.KAA13417@sinet1.fom.fgan.de>... > Hello, > : > I did see a very slow performance with BACKUP. Scenario: > < > 	Backup node = dual 500Mhz processor DS20 with 4GB memory.6 > 	OpenVMS 7.1-2 AXP, backup media 2x73GB Seagate disk3 > 	parallel BACKUP of 10 Alphas a two disk. Evererys3 > 	disk is MSCP mounted on the backup node. The ten 8 > 	Alpas are connected via 100BaseTX and the backup node8 > 	with 1000BaseSC. For the backup we used the followingB > 	command BACK/IMA/VER/REC DISKn: DISKm:[BACKUP]DISKn.IBCK/SAVE -B > 	/noCRC/GROUP=0. The disk's on board write cache was enabled -->+ > 	18.5MB/s possible throuput to the disks.d > K > I started the backup last monday. Two backups will be processed till now.iJ > The complete backup will have less then 180GB --> less then 16GB/day -->A > less then 0.8GB/h. I did compute the following max. throughput:uH > min of (10*100Mb/s, 1000Mb/s, 2*18.5MB/s) = 37MB/s = 133MB/h --> afterH > three hour BACKUP must be finished. Does anybody have an idea why not?E > Is anything wrong with my calculation or anything else? Any help ise
 > welcome. > % > TIA and best regards Rudolf Wingerty  > /verify will slow it down, and using compression will help ... /media=compact   ------------------------------  # Date: Mon, 04 Nov 2002 23:32:20 GMT . From: "Tom Simpson" <simpsont@attbi.com.fubar>) Subject: Re: Slow performance with BACKUPn. Message-ID: <8sDx9.10330$ar4.107416@sccrnsc03>  K Or not...  Compression very well may make the situation worse if the system- can not keep enough dataD ready to maintain the tape in "streaming mode".  My experience is 3X worse...   Regards, Tom   5 "Bob Ceculski" <bob@instantwhip.com> wrote in messagec7 news:d7791aa1.0211041201.1d9758d5@posting.google.com...h3 > Rudolf Wingert <win@fom.fgan.de> wrote in messager2 news:<200211040952.KAA13417@sinet1.fom.fgan.de>...
 > > Hello, > > < > > I did see a very slow performance with BACKUP. Scenario: > >2= > > Backup node = dual 500Mhz processor DS20 with 4GB memory. 7 > > OpenVMS 7.1-2 AXP, backup media 2x73GB Seagate disk 4 > > parallel BACKUP of 10 Alphas a two disk. Everery4 > > disk is MSCP mounted on the backup node. The ten9 > > Alpas are connected via 100BaseTX and the backup nodel9 > > with 1000BaseSC. For the backup we used the followingiC > > command BACK/IMA/VER/REC DISKn: DISKm:[BACKUP]DISKn.IBCK/SAVE -dC > > /noCRC/GROUP=0. The disk's on board write cache was enabled -->e, > > 18.5MB/s possible throuput to the disks. > >bH > > I started the backup last monday. Two backups will be processed till now.L > > The complete backup will have less then 180GB --> less then 16GB/day -->C > > less then 0.8GB/h. I did compute the following max. throughput:fJ > > min of (10*100Mb/s, 1000Mb/s, 2*18.5MB/s) = 37MB/s = 133MB/h --> afterJ > > three hour BACKUP must be finished. Does anybody have an idea why not?G > > Is anything wrong with my calculation or anything else? Any help isp > > welcome. > > ' > > TIA and best regards Rudolf Wingertc >e@ > /verify will slow it down, and using compression will help ... > /media=compact   ------------------------------  $ Date: Mon, 4 Nov 2002 21:58:49 -0500  From: John Santos <JOHN@egh.com>) Subject: Re: Slow performance with BACKUP.4 Message-ID: <1021104213627.400A-100000@Ives.egh.com>  & On Mon, 4 Nov 2002, Tom Simpson wrote:  M > Or not...  Compression very well may make the situation worse if the systemo > can not keep enough dataF > ready to maintain the tape in "streaming mode".  My experience is 3X
 > worse... > 
 > Regards, > Tomw > 7 > "Bob Ceculski" <bob@instantwhip.com> wrote in messagea9 > news:d7791aa1.0211041201.1d9758d5@posting.google.com...o5 > > Rudolf Wingert <win@fom.fgan.de> wrote in message54 > news:<200211040952.KAA13417@sinet1.fom.fgan.de>... > > > Hello, > > > > > > > I did see a very slow performance with BACKUP. Scenario: > > >b? > > > Backup node = dual 500Mhz processor DS20 with 4GB memory.a9 > > > OpenVMS 7.1-2 AXP, backup media 2x73GB Seagate diske6 > > > parallel BACKUP of 10 Alphas a two disk. Everery6 > > > disk is MSCP mounted on the backup node. The ten; > > > Alpas are connected via 100BaseTX and the backup noded; > > > with 1000BaseSC. For the backup we used the following2E > > > command BACK/IMA/VER/REC DISKn: DISKm:[BACKUP]DISKn.IBCK/SAVE - E > > > /noCRC/GROUP=0. The disk's on board write cache was enabled -->a. > > > 18.5MB/s possible throuput to the disks.  B What tape?  The output is to a DISK.  Therefor no compaction.  :-)  ? (Compaction and keeping the tape streaming are *very* important-< for tape.  Usually a big win.  The compaction is done by the> tape drive or controller, so it puts no additional load on the< CPU.  Only an extremely slow (or very heavily loaded) system? wouldn't be able to keep up with the tape drive with compactionr@ enabled, but would be able to with it disabled.  The main reason= for not enabling compaction is if there is a chance you mighth@ need to read the tape on an older drive that doesn't support it.; For example, when making a 4mm DAT tape on a TLZ07 that youn might want to read on a TLZ04.)p  < 1) Tune the backup account as per the recommendations in the system manager's manual.  < 2) Use a large blocksize.  /BLOCK=32256 is the largest legal block size for a disk saveset.    : 3) Use a large default extend quantity on the output disk.  8 Easiest way is to use $ set rms/extend=LARGE_NUMBER.  If< you know how big the saveset is going to be, then try to get8 it all in one extent.  Otherwise, the fewer extents, the: better.  (Less than 7 doesn't matter much.)  I think RMS's6 handling when extend quantity is zero (the default) is7 much better than it used to be, but you should at least.9 check that someone hasn't set an explicit smallish extendi8 quantity, either for you process or as a system default. (View both with $ SHOW RMS.)   Hope one of these things helps!    > > >wJ > > > I started the backup last monday. Two backups will be processed till > now.N > > > The complete backup will have less then 180GB --> less then 16GB/day -->E > > > less then 0.8GB/h. I did compute the following max. throughput:gL > > > min of (10*100Mb/s, 1000Mb/s, 2*18.5MB/s) = 37MB/s = 133MB/h --> afterL > > > three hour BACKUP must be finished. Does anybody have an idea why not?I > > > Is anything wrong with my calculation or anything else? Any help is  > > > welcome. > > >h) > > > TIA and best regards Rudolf Wingert  > >-B > > /verify will slow it down, and using compression will help ... > > /media=compact >  >  >  >    -- c John Santosu Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Mon, 04 Nov 2002 21:29:30 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>d) Subject: Re: Slow performance with BACKUPw' Message-ID: <3DC73B1A.A9CED562@fsi.net>a   Rudolf Wingert wrote:n >  > Hello, > : > I did see a very slow performance with BACKUP. Scenario: > C >         Backup node = dual 500Mhz processor DS20 with 4GB memory. = >         OpenVMS 7.1-2 AXP, backup media 2x73GB Seagate diske: >         parallel BACKUP of 10 Alphas a two disk. Everery: >         disk is MSCP mounted on the backup node. The ten? >         Alpas are connected via 100BaseTX and the backup nodew? >         with 1000BaseSC. For the backup we used the followingwI >         command BACK/IMA/VER/REC DISKn: DISKm:[BACKUP]DISKn.IBCK/SAVE -gI >         /noCRC/GROUP=0. The disk's on board write cache was enabled -->l2 >         18.5MB/s possible throuput to the disks. > K > I started the backup last monday. Two backups will be processed till now.-J > The complete backup will have less then 180GB --> less then 16GB/day -->A > less then 0.8GB/h. I did compute the following max. throughput:eH > min of (10*100Mb/s, 1000Mb/s, 2*18.5MB/s) = 37MB/s = 133MB/h --> afterH > three hour BACKUP must be finished. Does anybody have an idea why not?E > Is anything wrong with my calculation or anything else? Any help is.
 > welcome.  G The only comment I would add is that such calculations assume exclusiveI- access to the wire, which is rarely the case.-   -- r David J. Dachtera[ dba DJE Systems@ http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/:   ------------------------------  $ Date: Mon, 4 Nov 2002 18:58:40 -0500# From: "Hal Kuff" <kuff@comcast.net>  Subject: smp license per card?O Message-ID: <09AECF17AD8A2EB8.85053A913299723C.49CDEE1C31A896C4@lp.airnews.net>p  ? How do I tell if we have loaded enough licenses (OPENVMS-ALPHA)w  ( to run all of the smp cards in a system?  = Is there a way in SDa or what-have-you to tell that all cardsd  ; are running... a show cpu is a hardware list not loaded o/ss   licenses is it not?d   ------------------------------  * Date: Mon, 4 Nov 2002 14:59:33 -0700 (MST)" From: John Nebel <nebel@csdco.com># Subject: TCP/IP Services NFS tuning F Message-ID: <Pine.OSF.4.21.0211041425440.2452-100000@athena.csdco.com>  H The object is to use the VMS NFS server for a network of Macs running OSE X, but I don't think the client operating system has a lot to do with' it.f  J The server is TCP/IP 5.3 eco 1 on VMS 7.2-2.  NFS has a container file set9 up for testing.  The disk is on fibre-channel and an HSG.e  I When copying files, typically image files which are about 400mb each, thecJ server runs about 700 disk i/o's /sec but according to the HSG only aroundH 1.8 mb/sec of data are moving off(??) and on the disk, Rd% 13  Wr% 87.  J Why should NFS be reading anything ??  The HSG read cache hit rate is 80%.  G There must be a way to get a higher data transfer rate out of all thoseeK i/o's.  Changing the default "TCPIP$CFS_WRITESIZE" = "512" to 8192 doesn't f
 seem to help.t  H I did try this on VMS 7.3-1 to see if the XFC made any difference.  ThatF test machine doesn't have a fibre-channel interface, just a local 10K,G 72GB disk and the i/o rate was a lot worse, around 100/sec and the datatC transfer rate was similarly worse.  The read/write mixture without e( HSG-caching probably kills a local disk.  E It looks like larger block sizes and fewer RMS extends should improvebG things, but playing with RMS_DFMBC and RMS_EXTEND_SIZE didn't appear to  do much.  H I also have played around with the parameters on the client to make sureH the transfer size was 8192 and tried TCP rather than UDP and that didn't do much good either.  A Getting this to work well will mean a having supported way of VMSa& serving Macs which would be very good.   Thanks.e  
 John Nebel   ------------------------------   Date: 4 Nov 2002 14:35:43 -0800a1 From: susan_skonetski@hotmail.com (Sue Skonetski)   Subject: The Leper colony at ETS= Message-ID: <857e9e41.0211041435.4142d4cf@posting.google.com>o  F A few of you have sent me mail asking if the Leper Colony got together at ETS.  Yes we did.  A We had tea and cookies and spoke about world peace and the higherM things of life.l   Anyone buy that?  ? We did get together and had a wonderful time.  Its funny how if E someone is a VMS Zealot you already know you are going to have a good F time.  We went to dinner a couple of times and they either put us in aF different room or outside so we were pretty much free to laugh as loudF as we wanted. And then went places for a few drinks.  The big bang was: great, I will leave that story to one of the other lepers.  F I finally got to talk to Mrs VAXman on the cell phone, who is one heck of a great person.  F Granted much of the humor was at the expense of poor quality software,@ which is our favorite topic, followed up by embarasing moments. @ Unfortunatly I seem to have the most amount of them or the other lepers are not fessing up.  E Mike Stewart, one of the enginers wore a kilt, I have to tell you its E a babe magnet, and it sure provides a few good opening lines as well.c  E Just so you know they did not announce the next ETS as soon as I findI out I will let you know.  
 Warm Regards,c Suet   ------------------------------  % Date: Mon, 04 Nov 2002 22:07:38 -0600s1 From: "David J. Dachtera" <djesys.nospam@fsi.net> $ Subject: Re: The Leper colony at ETS' Message-ID: <3DC7440A.6D293074@fsi.net>-   Sue Skonetski wrote: > G > Just so you know they did not announce the next ETS as soon as I findt > out I will let you know.  @ Yes! Please, especially the date for submissions of sessions andD seminars! Don't know how to convince them that speakers need time toG prepare, but I've already started working on the presentations I intendnF to submit for (whatever)-2003. Hopefully, the close of submissions andG session/seminar selections will be more than two months before the show.3 (four months would great, six would be ideal, IMO).o   --   David J. Dachterao dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/p   ------------------------------   Date: 4 Nov 2002 12:09:41 -0600-+ From: young_r@encompasserve.org (Rob Young)21 Subject: Re: VMS on Itanium not mission critical?A3 Message-ID: <qjpLcN5ZTeZs@eisner.encompasserve.org>   h In article <d7791aa1.0211041041.494279da@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes: > B > I am not sure I would like to see a message in my browser when IC > transfer my money to the bank that "sorry, our rtlX ethernet cardcA > misbehaved, you lost your money". See the OpenBSD kernel sourceeA > for what they write about the rtl cards (redefines low end). Iny >   = 	But in the real world, it doesn't work that way.  You simplyI; 	don't lose transactions.  You need robust software to maken8 	sure you don't lose transactions (assuming transactions? 	are very important to you) and you should also make sure thosetA 	transactions hit disk platters many miles apart (if transactions2D 	are very very important).  A good solution is VMS Volume Shadowing  	coupled with RTR. 	D& 	Keith has good links for an overview:    	Comparing VMS Clusters and RTR:  S http://www.geocities.com/keithparris/decus_presentations/f99_rtr_cf_vmsclusters.pptd  ? 	Smart (or even marginal) software can overcome cruddy hardwarefA 	and/or cruddy OSes.  As an example, many of us are surrounded byn; 	laughable "solutions" running on numerous Intel boxes with D 	certain OSes.  Failovers for each and every system.  It sorta works 	well.   	Very good software (RTR):  9 http://www.compaq.com/products/software/ntenterprise/rtr/e  A 	Can make broken OSes/clusters/hardware quite reliable to the end  	user.  @ 	If a ndoe in your Itanium cluster crashes, hopefully the entireB 	application environment is designed such that it isn't the end ofC 	the world or doesn't have major user community / financial impact.    	Nodes do crash ya know!   				Robl   ------------------------------  # Date: Mon, 04 Nov 2002 20:19:16 GMT # From: "John Smith" <a@nonymous.com> 1 Subject: Re: VMS on Itanium not mission critical? J Message-ID: <8DAx9.188416$%h2.169451@news02.bloor.is.net.cable.rogers.com>  5 "Bob Ceculski" <bob@instantwhip.com> wrote in message 7 news:d7791aa1.0211041041.494279da@posting.google.com... E > I was arguing VMS superiority to linux on their board when a posteriB > made the following excellent point, except he gave a link at theC > end that I can't find ... anyone know what that is all about "whyoD > vms wasn't ported to intel before", and wouldn't this be a problem? > for the itanium vms port? Itanium cannot be mission critical?n    @ VMS on Itanic will be a low-volume solution for several reasons:  A 1) In order for Itanic to even come close to acceptable levels ofXL performance, it will require HP's custom (read proprietary) ZX-1 chipset (orI whatever it's called). HP is not likely to sell/license this to any otherlI Itanic integrator/OEM. Ergo, the only vendor a VMS-capable Itanic will bec available from will be HP.  J 2) Without any competition in the VMS-capable Itanic market, there will beL no low price 'commodity' hardware to run VMS on. It will be EXACTLY the same@ situations as today vis-a-vis Alpha. How many ways can you spell 'proprietary'.  L 3) Itanic will forever be a lower perfoming solution than Alpha at any point5 in time up until Alpha's are no longer actively sold.   # 4) Itanic is no cheaper than Alpha.-  J 5) By the time 3) above rolls around, many current VMS customers will have# bolted to a unix solution from IBM.i  J 6:5 odds HP will be bankrupt in 5 years by continuing to be a high cost PCI producer, having its margins eroded by Dell's competition in the printing.I business, and by slinging the Itanic Albatross around the neck of all itsAL enterprise platforms. HP's position across all enterprise lines will be thatF of high-cost, low-volume producer of systems nobody wants to purchase.   ------------------------------  % Date: Mon, 04 Nov 2002 21:41:39 +0100 6 From: Jan van Mastbergen <jan.vanmastbergen@planet.nl>1 Subject: Re: VMS on Itanium not mission critical?4( Message-ID: <3DC6DB83.3030605@planet.nl>  I I vaguely recall from earlier discussions about this subject that (apart aB from commercial arguments) differences in the interrupt mechanism E between the i386 architecture and the VAX cpu would not allow you to  G realise the same sort of real-time reliability with a ported VMS/86 as   with VMS/VAX /AXP etc.E I am not very solid on this but, on i386, you would have to take the  H interrupt and analyze it to decide its less important than what the cpu G is currently doing while on VAX the hardware would hold an outstanding  C interrupt until the cpu's IPL drops below the IPL of the interrupt.a   Jan    Bob Ceculski wrote:n  E > I was arguing VMS superiority to linux on their board when a postergB > made the following excellent point, except he gave a link at theC > end that I can't find ... anyone know what that is all about "why9D > vms wasn't ported to intel before", and wouldn't this be a problem? > for the itanium vms port? Itanium cannot be mission critical?l >  >  > C > The problem with the "widely" available and manufactured hardwareaB > is that there are a lot of compatibility issues. Also, as we seeC > on this ng, there are a lot of quality issues. Since VMS will runnE > only on their hardware, they have complete control over the working B > of the system. While I might not want someone completely controlB > my box when it is just an apache/ftp/quake server, or it is justB > my desktop, I am not sure if I want to see these hardware piecesD > in some mission critical devices like medical computers, computersB > aboard airplanes, etc. In those situations, I absolutely accept,F > even demand, that someone (reputable, and trustable, and I think DECF > satisfied both) have total control over the soft/hardware. Since VMSF > is targeted at such "industries", besides finance, I see no problem. > B > I am not sure I would like to see a message in my browser when IC > transfer my money to the bank that "sorry, our rtlX ethernet card>A > misbehaved, you lost your money". See the OpenBSD kernel sourceoA > for what they write about the rtl cards (redefines low end). IniB > such situations, I want, better said demand, the best. HopefullyC > such things are not present in VMS systems and in turn in missiontC > critical systems. However, for my home network, I don't mind whenoC > the network card is not the fastest or it drops more packets thanl& > it should. It is just a home system. > D > I don't mind if the gov't pays a 10x surcharge in mission critical@ > situations and no US warship needs to be towed to harbour. OK,E > that was a software issue, but the point is that they used a systemx< > which was not designed for such a mission critical system. >  > B >>You might get away with a "foreign" SCSI disk drive, but they'llC >>pester you every time they come in to do provocative maintenance,p6 >>blaming all problems on not buying /their/ hardware. >> > @ > I do expect my bank/hospital/military to have the funds to buy4 > quality hard disks for *MISSION CRITICAL* systems. >  > G >>By the way, oops, the only supported video hardware is obsolete stufftG >>four years old that is no longer available except at truly exorbitanta >>pricing... >> > E > So why do you need a 3d card for a box which controls the radiationmC > dosage in a hospital? A simple video card which provides the usern > with a 80x25 display is fine.  > D > The page http://www.kjsl.com/vmsfaq/vms.htm talks about why it was > not ported to Intel before.  >  > Vilmos >    ------------------------------  $ Date: Mon, 4 Nov 2002 16:04:22 -0500* From: "Bill Todd" <billtodd@metrocast.net>1 Subject: Re: VMS on Itanium not mission critical? 2 Message-ID: <L4WdneUXI7RIfVugXTWcqQ@metrocast.net>  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:qjpLcN5ZTeZs@eisner.encompasserve.org...s   ...t  @ > Smart (or even marginal) software can overcome cruddy hardware > and/or cruddy OSes.r   Only within certain limits.a  -   As an example, many of us are surrounded byr< > laughable "solutions" running on numerous Intel boxes withE > certain OSes.  Failovers for each and every system.  It sorta works- > well.-  J Where it doesn't work well at all is where the hardware or software is notH 'fail-fast'.  If you don't use ECC memory, then unless you use lock-stepB hardware-comparator-monitored solutions no software will help you:I corrupted memory can cause your system to corrupt data without batting an-K eye (i.e., no fail-over will occur, because no one knows anything's wrong).eF Same problem if your buses aren't parity- or ECC-protected, or if someL random piece of unqualified hardware (or, for that matter, your OS) corrupts data without notice.  D A high-quality hardware/software combination like VMS on Alpha can'tK guarantee that no such corruption will occur (only lock-step systems can doeB that), but it makes the chances of it a lot smaller than with lessL controlled hardware environments.  HP could certainly qualify certain ItanicH boxes as thoroughly as it does Alpha boxes - but don't expect them to be( priced the same as generic Itanic boxes.   - bill   ------------------------------  % Date: Mon, 04 Nov 2002 16:45:42 -0500.- From: JF Mezei <jfmezei.spamnot@videotron.ca>U1 Subject: Re: VMS on Itanium not mission critical?., Message-ID: <3DC6EA77.72EDE89A@videotron.ca>  C If Intel added the stuff needed for Tandem to run on IA64, then theTN architecture should have the necessary design to allow mission critical stuff.  J If the chips turns out to have many bugs, it may delay the use of IA64 forJ real applications though. And perhaps for real applications, the chip will: need to be accompanied by anciliary chips to help it cope.   ------------------------------   Date: 4 Nov 2002 20:23:21 -0800u1 From: keithparris_NOSPAM@yahoo.com (Keith Parris)N1 Subject: Re: VMS on Itanium not mission critical? = Message-ID: <cf15391e.0211042023.35437987@posting.google.com>c  m bob@instantwhip.com (Bob Ceculski) wrote in message news:<d7791aa1.0211041041.494279da@posting.google.com>...1D > The page http://www.kjsl.com/vmsfaq/vms.htm talks about why it was > not ported to Intel before.(  D Based on looking at the Wayback Machine (http://www.archive.org), itB appears this page merely contained a copy of the OpenVMS FAQ as ofC some date in 2000 or so.  The same section of the OpenVMS FAQ todayn contains this text:a  ?   2.9  Why hasn't OpenVMS been ported to Intel (IA-32) systems?a  #            Why? Business reasons...y  :            There is presently a belief that there would be;            insufficient market to justify the cost involveds<            in porting OpenVMS to systems using the Intel IA-;            32 architecture. In addition to the direct costss=            involved in any port, each maintainer of a product >            or a package for OpenVMS has to justify the port to?            "OpenVMS Pentium" or to OpenVMS Itanium, akin to the B            required justifications for a product port from OpenVMS             VAX to OpenVMS Alpha.  -            But yes, it would be nice to have.t  <            And yes, OpenVMS Engineering is well aware of AMD            Opteron/Hammer, too.n  A If you're really interested in what this used to say, do a Google 
 search forB   OpenVMS FAQ VMS11 Why hasn't OpenVMS been ported to Intel (IA32) systemsnF and you'll find lots of older versions of various vintages archived at various locations on the net.t   ------------------------------   Date: 4 Nov 2002 20:48:31 -0800o1 From: keithparris_NOSPAM@yahoo.com (Keith Parris)l1 Subject: Re: VMS on Itanium not mission critical?c= Message-ID: <cf15391e.0211042048.46be5830@posting.google.com>s  * billtodd@metrocast.com (Bill Todd) wrote: L > Where it doesn't work well at all is where the hardware or software is notJ > 'fail-fast'.  If you don't use ECC memory, then unless you use lock-stepD > hardware-comparator-monitored solutions no software will help you:K > corrupted memory can cause your system to corrupt data without batting aneM > eye (i.e., no fail-over will occur, because no one knows anything's wrong). H > Same problem if your buses aren't parity- or ECC-protected, or if someN > random piece of unqualified hardware (or, for that matter, your OS) corrupts > data without notice.  F In light of the above, it's interesting to note that a report from theD Microprocessor Forum indicates AMD's Opteron apparently doesn't evenB bother to protect general register contents with parity, much less ECC.  @ "Opteron implements ECC data protection mostly everywhere but itE doesn't yet implement parity for internal registers.  Thus Opteron iso4 not good enough yet for mission-critical servers" --( http://www.theinquirer.net/?article=5815  B AMD would have to do another Hammer chip to correct this omission.  E Itanium was designed with the intent of supporting NonStop (which, as D things turned out, it will), so its design is deterministic and will support lockstep operation.x   ------------------------------  $ Date: Mon, 4 Nov 2002 14:10:10 -0500% From: "John Vottero" <John@mvpsi.com>l@ Subject: Re: VMS to support KDE over CDE in which future version/ Message-ID: <usdhgjar9ac218@news.supernews.com>W  % Dang!  I was really hoping to get Qt.   @ "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote in message+ news:olyx9.49$DO.989288@news.cpqcorp.net...e> > Please point the ambassador to me.  There are no such plans. >o >f! > C.W.Holeman II wrote in messager6 > <77555df7.0211022058.535989ff@posting.google.com>...F > >Anyone have any info on the rumor that a future release of VMS willF > >support KDE? I was told by a VMS Ambassador the CDE was going to be > >replaced by KDE.- > >- > >--- > >C.W.Holeman IIe? > >cwhii5@Julian5Locals.com                http://also.as/cwhiii > >remove the fivesu& > >Send spam to junkmail@earthlink.net >D >3   ------------------------------  % Date: Mon, 04 Nov 2002 21:15:56 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>s@ Subject: Re: VMS to support KDE over CDE in which future version' Message-ID: <3DC737EC.378548F9@fsi.net>n   Simon Clubley wrote: > ] > In article <3DC57CF1.22128165@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:, > > Simon Clubley wrote: > >>? > >> I sincerely hope that it will be GTK/Gnome instead of KDE.e > >>I > >> Amongst other things, GTK has been designed to support many languagerC > >> bindings from day one, and has a wider range of them than KDE.  > >nL > > Then you better hope that KOffice gets ported to Gnome (GOffice?). ThereL > > may be a similar package for Gnome, but it's never come to my attention. > >  > ? > There is a gnome-office, but I don't know how advanced it is.i > , > Of course, there's always Open Office. :-) > H > Seriously though, I suppose it comes down to what a VMS port of such aN > toolkit would be used for. Would it be used mainly for in house development,F > (ie: to allow the same applications to be developed on VMS and otherJ > platforms), or would there suddenly be a large increase in the number of1 > VMS workstations sold for office applications ?a > L > I know (from your public comments) that your wish is for the latter, but I7 > suspect unfortunately that the former is more likely.  > J > BTW, I like GTK. I've used it quite a bit on Linux, and more recently on@ > Win98. (The languages that I have used it with are C and Ada).  H I can see this descending into another "religious" war. KDE seems to havE mroe software already but the super-techies seem to prefer Gnome. Not(E sure why in either case, except that "out of the box", KDE looks very2H WhineBloze-like, although the "K" and the foot-print substitutes for the2 "Start" button are bit too "cutsie" for my taste.   . What's wrong with a key "Menu" button instead?  C I mean, cosmetically that's all they "improved" in W/9x and later - C WhineBloze really gives you little more than an "organized" menuingrH system with a central GUI API. Having individual programs on the desktop: rather than program groups was a vast improvement, but the! under-pinnings are still bogus...    -- l David J. Dachteraa dba DJE Systemsf http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/l   ------------------------------  $ Date: Mon, 4 Nov 2002 14:37:28 -0600* From: "Dave Gudewicz" <dgudewicz@core.com>- Subject: Re: Was OpenVMS left out on purpose?v9 Message-ID: <3dc6da8b$0$17644$1dc6e903@news.corecomm.net>n  K Got a reply from Stallard.  Message said VMS should have been on the surveyLH and it will be fixed.  BTW, I think the survey came from SBA, the peopleB that manage Encompass (nee DECUS).  They also emailed the (paying)K membership today regarding on-line voting for BOD candidates.  Didn't work.06 I sent them an e-mail and after a while, it was fixed.   -- Dave...   G It is noble to teach oneself, but still nobler to teach others-and lesss trouble. -----Mark TwainV  1 "Dave Gudewicz" <k9jdk@arrl.net> wrote in messageF3 news:3dc43688$0$17642$1dc6e903@news.corecomm.net...tH > I was a bit harsh, sorry Ken.  Thought I softened the message with the > spider and fly closing.: >+G > But I will admit I was angered, once again, by the absence of VMS.  It wroteCF > to Stallard and Jessica about VMS being AWOL.  Hope someone learns a lesson
 > by this. >e	 > Dave...n >t. > "Scratch" <scratch@aol.com> wrote in message: > news:D3Cw9.61438$dn3.2014117@twister.southeast.rr.com... > >e9 > > "Dave Gudewicz" <dgudewicz@core.com> wrote in messageo7 > > news:3dc2b758$0$17648$1dc6e903@news.corecomm.net...tG > > > Explains what?  That LINUX and Windows are no considered non-stope > > platforms,! > > > said the spider to the fly.f > >cH > > I was just passing the message along.  I guess by "explains" I meant that > itK > > was the Nonstop group and not anyone related to OpenVMS.  Heck, I don'tr > know > > why they left it out.s > >o > >, > > Kenw > >h > > -- > > Kenneth Farmer > > http://www.OpenVMS.org > >a > >. > >  > >r9 > > "Dave Gudewicz" <dgudewicz@core.com> wrote in messageu7 > > news:3dc2b758$0$17648$1dc6e903@news.corecomm.net... G > > > Explains what?  That LINUX and Windows are no considered non-stoph > > platforms,! > > > said the spider to the fly.  > > >  > > > --
 > > > Dave...u > > >OH > > > It is noble to teach oneself, but still nobler to teach others-and less > > > trouble. > > > -----Mark Twain  > > >$2 > > > "Scratch" <scratch@aol.com> wrote in message> > > > news:k8xw9.61154$dn3.1979508@twister.southeast.rr.com...' > > > > I guess this kinda explains it.D > > > >i > >d > > > >c > > > >  > > > >R$ > > > > ----- Original Message -----7 > > > > From: "Moser, Jessica " <Jessica_Moser@sba.com> 4 > > > > To: "'Kenneth Farmer'" <kfarmer@openvms.org>0 > > > > Sent: Friday, November 01, 2002 10:28 AM' > > > > Subject: RE: Instapoll questioni > > > >c > > > >  > > > > > Kennth -	 > > > > >oA > > > > >    This survey was sponsored and written by the NonStop. DevelopmentwG > > > > > department within HP so it's scope was limited to the NonStop8
 > > platform.6 > > > WeL > > > > > apologize if this wasn't communicated properly. Thank you for your > > > > interestH > > > > > in the advocacy program. Please continue to check our web siteK > > > > > (www.hpuseradvocacy.org) for future advocacy related initiatives.o	 > > > > > $ > > > > > -----Original Message-----; > > > > > From: Kenneth Farmer [mailto:kfarmer@openvms.org]o3 > > > > > Sent: Thursday, October 31, 2002 12:12 PMe& > > > > > To: admin@hpuseradvocacy.org% > > > > > Subject: Instapoll questione	 > > > > >u	 > > > > >t* > > > > > Was OpenVMS left out on purpose?	 > > > > >KE > > > > > http://www.hpuseradvocacy.com/Surveys/2002October/index.cfme	 > > > > >i	 > > > > >SK > > > > > For which platform(s) are you developing enterprise applications?  > > > > > NonStop Kernel > > > > > HP-UXg > > > > > Tru64 UNIX > > > > > Windows  > > > > > Linuxr" > > > > > Other (please elaborate) > > > >e > > > >  > > > >s > > > >e > > > >1 > > >X > > >e > >u > >e >  >t   ------------------------------   Date: 4 Nov 2002 13:59:38 -0600M+ From: young_r@encompasserve.org (Rob Young)E- Subject: Re: Was OpenVMS left out on purpose?e3 Message-ID: <foTyX7xaUm0b@eisner.encompasserve.org>   f In article <3dc6da8b$0$17644$1dc6e903@news.corecomm.net>, "Dave Gudewicz" <dgudewicz@core.com> writes:   > M > Got a reply from Stallard.  Message said VMS should have been on the survey/J > and it will be fixed.  BTW, I think the survey came from SBA, the peopleD > that manage Encompass (nee DECUS).  They also emailed the (paying)M > membership today regarding on-line voting for BOD candidates.  Didn't work. 8 > I sent them an e-mail and after a while, it was fixed. >   = 	Which goes to show if your tin hat isn't blocking the proper,@ 	frequencies, you believe things can be changed.  Otherwise, you@ 	have to visit Area 51 to see what is really going on.  Prior toC 	the visit, make sure you smoke the same thing everyone else at thea& 	top is smoking.  Not so unreasonable.   				Rob1  cN "The reasonable man adapts himself to the world; the unreasonable one persistsJ  in trying to adapt the world to himself.  Therefore all progress depends 5   on the unreasonable man." --George Bernard Shaw        ------------------------------  % Date: Mon, 04 Nov 2002 20:38:53 -05002- From: JF Mezei <jfmezei.spamnot@videotron.ca>o- Subject: Re: Was OpenVMS left out on purpose? , Message-ID: <3DC7212C.75211466@videotron.ca>   Rob Young wrote:F >         Which goes to show if your tin hat isn't blocking the proper: >         frequencies, you believe things can be changed.     L Is it the customer's jobs to find all the omissions and errors that HP makes! and then beg HP to correct them ?   G The fact that VMS was omitted to begin with shows that there is still a  problem at HP.    I Has this been a TV commercial that mentioned all of HP's OS's except VMS, O would that Stallard guy have gone to bat to pull the ad and have it corrected ?G   ------------------------------  % Date: Mon, 04 Nov 2002 21:09:49 -0600i1 From: "David J. Dachtera" <djesys.nospam@fsi.net> - Subject: Re: Was OpenVMS left out on purpose?e' Message-ID: <3DC7367D.604A8373@fsi.net>l   JF Mezei wrote:G >  > Rob Young wrote:H > >         Which goes to show if your tin hat isn't blocking the proper; > >         frequencies, you believe things can be changed.3 > N > Is it the customer's jobs to find all the omissions and errors that HP makes# > and then beg HP to correct them ?    Hey  - we work cheap... WTF?   -- i David J. Dachtera  dba DJE Systemse http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/m   ------------------------------   Date: 4 Nov 2002 20:51:01 -06000+ From: young_r@encompasserve.org (Rob Young).- Subject: Re: Was OpenVMS left out on purpose?.3 Message-ID: <9WEdXCJGrw+L@eisner.encompasserve.org>>  \ In article <3DC7212C.75211466@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: > Rob Young wrote:G >>         Which goes to show if your tin hat isn't blocking the properr; >>         frequencies, you believe things can be changed. , >  > N > Is it the customer's jobs to find all the omissions and errors that HP makes# > and then beg HP to correct them ?m >   ! 	Beg?  That's one way to spin it.>  1 	I got an e-mail four days ago.  It said in part:e   	Hot Backup up's  1 	I replied kindly and mentioned that should read:f   	hot backups    	It was appreciated being found.  E 	Was it my job to send email highlighting that error?  Of course not. D 	But let me tell you about being an asset.  An asset is someone thatE 	does things of their own volition.  Someday in this networked world,.= 	it pays off.  Maybe not always.  But it does give the personp? 	pause on the other end when they say to themselves "I remembern; 	how many times he helped me."  Sort of like when I give up., 	my seat on the train.  Part of being alive.   				Robr   ------------------------------  % Date: Mon, 04 Nov 2002 23:53:56 -0500 . From: Lyndon Bartels <lbartels@pressenter.com>! Subject: What Ethertype do I use? . Message-ID: <3DC70894.2C5763FC@pressenter.com>  , I was asked that by our network guys lately.    # I told them "DECnet, LAT, TCP/IP." l  D But I want to give them a better answer than that. Given that I haveH clusters at multiple sites. I'm positioning the hardware to be ready forG a Wide Area, Fault Tolerant Cluster. What am I using now, and what do IT. definitely not want to give up for the future.     I know it's vauge...     Thanks in advance,   Lyndon       --  G My opinions are mine and mine alone. They seldom align with those of myZ	 employer.n    H The only good thing about putting the cart before the horse is you don't have to look at the horse's butt.   ------------------------------  % Date: Tue, 05 Nov 2002 00:26:02 -0600 7 From: "Craig A. Berry" <craigberry@nospam.telocity.com>t% Subject: Re: What Ethertype do I use? @ Message-ID: <craigberry-729F92.00255705112002@news.telocity.com>  . In article <3DC70894.2C5763FC@pressenter.com>,0  Lyndon Bartels <lbartels@pressenter.com> wrote:  . > I was asked that by our network guys lately. >  > % > I told them "DECnet, LAT, TCP/IP." t  G Yes, that is vague.  Something like MCR LANCP SHO DEV/CHAR EW might be d a place to start.h   ------------------------------  % Date: Tue, 05 Nov 2002 01:49:27 -0500u- From: JF Mezei <jfmezei.spamnot@videotron.ca>t% Subject: Re: What Ethertype do I use?n, Message-ID: <3DC769E2.57F5FA9D@videotron.ca>   Lyndon Bartels wrote:p > . > I was asked that by our network guys lately. > $ > I told them "DECnet, LAT, TCP/IP."    K You should have asked them "whats is an ethertype, and what does it eat ?".o  > Was the question "what protocols to you use on the ethernet" ?   If so, in a cluster, excpect:   ( SCS (cluster), DECNET, LAT, MOP, TCPIP.   M SCS and LAT are non-routable as they are based on hardware ethernet address. e  I Note that for that IA64 thing coming in the future, MOP may or may not begL available, and another protocol to allow nodes to boot from another node may be used.   ------------------------------   Date: 4 Nov 02 21:20:43 +0100 ) From: p_sture@elias.decus.ch (Paul Sture)-, Subject: Re: [very OT] OS X, the (bad?) turn) Message-ID: <POfZSaWZ3obZ@elias.decus.ch>a  c In article <s2cUn0qDPirs@eisner.encompasserve.org>, koehler@encompasserve.org (Bob Koehler) writes:f   <snip>  ? >> 5. Why does the Classic indexation process not work anymore?  > I >    You can always boot back to OS 9 for a while to get some old featurecB >    you really need.  Most Classic apps to run under OS X ClassicG >    environment, but a lot of functions built using Clasic extensions s# >    just don't exist yet for OS X.u > I >    I don't miss indexing because I never had enough time for it to run.n >hH There is a command line  alternative in the locate and updatedb command.  4 updatedb builds an index of all files on the system.2 locate is used to search that index for filenames.  H The Linux installations I have tried come set up to run updatedb nightly0 (via cron), but OS X doesn't do that by default.  " To do it manually (idiot's guide):   o - launch a Terminal sessione o - man locate+ o - hit the space bar to scroll down a page>; o - select and copy the script name under the FILES sectiong%     it's /usr/libexec/locate.updatedb  o - hit q to exit man1 o - execute as root:%     sudo /usr/libexec/locate.updatedba= (sudo asks for your privileged password and then executes thel' following command as a privileged user)e   That was easy wasn't it? :-)  9 I haven't got around to trying to set up a cron job to do  the same, but it's on the list.e    -- r
 Paul Sture Switzerland    ------------------------------   End of INFO-VAX 2002.611 ************************ww.djesys.com/vms/soho/:   ------------------------------  $ Date: Mon, 4 Nov 2002 18:58:40 -0500# From: "Hal Kuff" <kuff@comcast.net>  Subject: smp license per card?O Message-ID: <09AECF17AD8A2EB8.85053A913299723C.49CDEE1C31A896C4@lp.airnews.net>p  ? How do I tell if we have loaded enough licenses (OPENVMS-ALPHA)w  ( to run all of the smp cards in a system?  = Is there a way in SDa or what-have:    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :    :     ;    ;    ;    ;    ;    ;    ;    ;    ;    	;    
;    ;    ;    
;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;     ;    !;    ";    #;    $;    %;    &;    ';    (;    );    *;    +;    ,;    -;    .;    /;    0;    1;    2;    3;    4;    5;    6;    7;    8;    9;    :;    ;;    <;    =;    >;    ?;    @;    A;    B;    C;    D;    E;    F;    G;    H;    I;    J;    K;    L;    M;    N;    O;    P;    Q;    R;    S;    T;    U;    V;    W;    X;    Y;    Z;    [;    \;    ];    ^;    _;    `;    a;    b;    c;    d;    e;    f;    g;    h;    i;    j;    k;    l;    m;    n;    o;    p;    q;    r;    s;    t;    u;    v;    w;    x;    y;    z;    {;    |;    };    ~;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    ;    