1 INFO-VAX	Thu, 15 Nov 2001	Volume 2001 : Issue 636       Contents:P Alt group for VMS (was Re: Special IPF-Inside Issue of Shannon Knows Compaq at w5 Re: Anyone know of 100 MB NIC supported under VMS 7.1 5 re: Anyone know of 100 MB NIC supported under VMS 7.1  Auto CD Mounting Re: Auto CD Mounting? Re: check this out  Web hosting  On OpenVMS  (customer link...) ? Re: check this out  Web hosting  On OpenVMS  (customer link...) & CJL's NOTLIB_COPY.C (Was: Re: LIB$xxx)* CJL's NOTLIB_PAGINATE.C (Was: Re: LIB$xxx) Re: Clustering Textbook < Comdex Fall special report praises VMS but criticises CompaqG Re: Compaq and OSDN Create Clustering Foundry for Open Source Community  Re: Compaq guarantees? Re: Compaq guarantees?0 Re: Compaq/HP merger - let them know how to vote0 Re: Compaq/HP merger - let them know how to vote0 Re: Compaq/HP merger - let them know how to vote0 Re: Compaq/HP merger - let them know how to vote DEC 4000 300 has no halt?  Re: DEC 4000 300 has no halt?   DEC/X-Windows, VMS and eXcursion$ Re: DEC/X-Windows, VMS and eXcursionD Re: Editing DCL command lines longer than the current terminal widthD Re: Editing DCL command lines longer than the current terminal widthD Re: Editing DCL command lines longer than the current terminal width# Re: errlog, decevent, CA, and WEBES P Re: How to connect a DS20 to a Fibre  channel SAN switch an that one to a hsg80?P Re: How to connect a DS20 to a Fibre  channel SAN switch an that one to a hsg80?  Re: HP kills MPE. It's official.  Re: HP kills MPE. It's official.  Re: HP kills MPE. It's official.  Re: HP kills MPE. It's official.  Re: HP kills MPE. It's official.  Re: HP kills MPE. It's official.  Re: HP kills MPE. It's official.  Re: HP kills MPE. It's official.  Re: HP kills MPE. It's official.  Re: HP kills MPE. It's official.  Re: HP kills MPE. It's official.P Re: HP's Acquisition of Compaq: Beyond the Naysayers, the Promise and   the OppoP HP's Acquisition of Compaq: Beyond the Naysayers, the Promise and the OpportunitP Re: HP's Acquisition of Compaq: Beyond the Naysayers, the Promise and the OpportP Re: HP's Acquisition of Compaq: Beyond the Naysayers, the Promise and the OpportP Re: HP's Acquisition of Compaq: Beyond the Naysayers, the Promise and the OpportP Re: HP's Acquisition of Compaq: Beyond the Naysayers, the Promise and the Opport Re: HP/Compaq merger Re: IBM to bid for CPQ ??  Re: IBM to bid for CPQ ??  Re: IBM to bid for CPQ ??  Re: IBM to bid for CPQ ??  Re: IBM to bid for CPQ ??  Re: IBM to bid for CPQ ??  Re: IBM to bid for CPQ ??  Re: IBM to bid for CPQ ??  Re: IBM to bid for CPQ ??  Re: IBM to bid for CPQ ?? # Is it a DEC C problem, or is it me? ' Re: Is it a DEC C problem, or is it me? ' Re: Is it a DEC C problem, or is it me? ' Re: Is it a DEC C problem, or is it me? ' Re: Is it a DEC C problem, or is it me? ' Re: Is it a DEC C problem, or is it me? $ Itanium flunking Compaq server tests( Re: Itanium flunking Compaq server tests( Re: Itanium flunking Compaq server tests( Re: Itanium flunking Compaq server tests( Re: Itanium flunking Compaq server tests( Re: Itanium flunking Compaq server tests( Re: Itanium flunking Compaq server tests( Re: Itanium flunking Compaq server tests Re: Items for sale! A Re: Linking shareable library  using another shareable library !? A Re: Linking shareable library  using another shareable library !? * Re: Need help, was: Rob's British Champion1 Re: OpenVMS o/s cd-rom insufficient quotas/limits  P/OS anyone? Re: P/OS anyone?
 Re: Pathworks  Re: Proxy Removal Problem. Re: Proxy Removal Problem.' Re: Sources of class information needed E Re: Special IPF-Inside Issue of Shannon Knows Compaq at www.tru64.org E Re: Special IPF-Inside Issue of Shannon Knows Compaq at www.tru64.org E Re: Special IPF-Inside Issue of Shannon Knows Compaq at www.tru64.org E Re: Special IPF-Inside Issue of Shannon Knows Compaq at www.tru64.org E Re: Special IPF-Inside Issue of Shannon Knows Compaq at www.tru64.org E Re: Special IPF-Inside Issue of Shannon Knows Compaq at www.tru64.org C Re: Unix and Linux certs advisory - VMS remains only unhackable OS! C Re: Unix and Linux certs advisory - VMS remains only unhackable OS! C Re: Unix and Linux certs advisory - VMS remains only unhackable OS!   usage of new products on vms axp$ Re: usage of new products on vms axp? Re: Using INETACP_FUNC$C_GETHOSTBYNAME using fortran- examples? ? Re: Using INETACP_FUNC$C_GETHOSTBYNAME using fortran- examples? ? Re: Using INETACP_FUNC$C_GETHOSTBYNAME using fortran- examples? ? Re: Using INETACP_FUNC$C_GETHOSTBYNAME using fortran- examples? ? Re: Using INETACP_FUNC$C_GETHOSTBYNAME using fortran- examples?  Using vfork and exit
 Re: VMS Crash 8 VMS debug image runs OK but nocheck, check version crash< Re: VMS debug image runs OK but nocheck, check version crash< Re: VMS debug image runs OK but nocheck, check version crash< Re: VMS debug image runs OK but nocheck, check version crash VMS emulator needed  Re: VMS emulator needed  RE: VMS emulator needed  Re: VMS emulator needed 4 Re: VMS/RMS - destroyed sequential files - NEED HELP4 Re: VMS/RMS - destroyed sequential files - NEED HELP4 Re: VMS/RMS - destroyed sequential files - NEED HELP4 Re: VMS/RMS - destroyed sequential files - NEED HELP4 Re: VMS/RMS - destroyed sequential files - NEED HELP4 Re: VMS/RMS - destroyed sequential files - NEED HELP4 Re: VMS/RMS - destroyed sequential files - NEED HELP4 Re: VMS/RMS - destroyed sequential files - NEED HELP4 Re: VMS/RMS - destroyed sequential files - NEED HELP RE: Wanted: Small MicroVAX0 Re: Web Connecting from an Alpha (using TCPWARE)! Welcome To Microsoft's IT Academy   F ----------------------------------------------------------------------    Date: 15 Nov 2001 15:29:29 -00004 From: Doc.Cypher <Use-Author-Address-Header@[127.1]>Y Subject: Alt group for VMS (was Re: Special IPF-Inside Issue of Shannon Knows Compaq at w 6 Message-ID: <20011115152929.21245.qmail@gacracker.org>  ; On Wed, 14 Nov 2001, Alan Greig <a.greig@virgin.net> wrote:    <snip>  = >But we are going off topic again (at least for those reading F >c.o.v./info-vax) Maybe time for alt.sys.vms.chat but I'm not sure howE >many main stream servers honour random NNTP newgroup controls in alt ? >these days. Creating a  comp.os.vms.chat would be even harder.  >  >--  >Alan   J This isn't the first time the suggestion has been made that a new group isJ needed for VMS. As Alan notes, getting a new comp. group would be a littleJ difficult. All an alt group supposedly needs is discussion in the relevantK newsgroup(s) followed by a charter being drawn up and posted to alt.config. F (This is assuming there is a broad base of support for the new group.)I After that most servers will honour the newgroup or you can petition your  ISP to carry the group.      Doc. --  6 The bigger the humbug, the better people will like it.K ~ Phineas Taylor Barnum.                              http://vmsbox.cjb.net    ------------------------------  % Date: Thu, 15 Nov 2001 00:55:22 -0500 ( From: David Froble <davef@tsoft-inc.com>> Subject: Re: Anyone know of 100 MB NIC supported under VMS 7.1, Message-ID: <3BF358CA.4030501@tsoft-inc.com>   David Silvers wrote:  M > I've got a cluster which I need to backup across a 100Mbit ethernet - but I K > cannot upgrade to VMS 7.2 due to incompatible versions of VMS existing on A > the cluster (a VMS 5.5-2 system running WordPerfect for VMS, an M > application uses WP print routines - can't upgrade this machine) and when I   > upgrade the 7.1 system to 7.2,E > it bugachecks (incompatible s/w versions in cluster).  I know 5.5-2 0 > shouldn't be in a cluster w/6.x & 7.x systems. > N > However, this situation existed b4 I got here, and I've been told "It works,< > so quit griping".  Except 7.2 actually enforces the rules. >  > Any ideas? >  >  >   H The DE500 works as far back as 6.2 I believe.  I'd have to read the SPD I again to be sure.  It's a PCI card.  What type of system are you wanting  , to upgrade?  That's the first consideration.   Dave   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com6 T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486   ------------------------------  % Date: Thu, 15 Nov 2001 12:58:31 +0100 / From: Nigel Arnot <sysmgr@maxwell.ph.kcl.ac.uk> > Subject: re: Anyone know of 100 MB NIC supported under VMS 7.16 Message-ID: <00A0515D.B0A69825.5@maxwell.ph.kcl.ac.uk>   > J > The DE500 works as far back as 6.2 I believe.  I'd have to read the SPD K > again to be sure.  It's a PCI card.  What type of system are you wanting  . > to upgrade?  That's the first consideration. >  > Dave >   D You need 6.2-1H3, not vanilla 6.2, and only a particular rev. of theG DE500 works with it. (can't recall which, there are -AA -AB -BA amd -XA L variants of the DE500 and I *do* recall that Digital delivered the one that E 6.2-1H3 would not support and we had to hassle to get it "downgraded"  to one that worked).  L Should be a lot less fussy with 7.1, but check the SPD (which I just looked  for and can't find).   	Yours, 
 		Nigel Arnot - 		NRA@MAXWELL.PH.KCL.AC.UK                      7 		"In the beginning there was nothing, which exploded."    ------------------------------  % Date: Thu, 15 Nov 2001 08:40:06 -0800 . From: Jack Trachtman <Jack.Trachtman@vmmc.org> Subject: Auto CD Mounting ( Message-ID: <3BF3EFE6.9D00A24F@vmmc.org>  E I'm trying to write a DCL script which will mount a CD no matter what  formatF it's in (ODS, ISO, etc).  So far I'm taking the simplistic approach of mountingG the CD /FOREIGN and looking for a lable.  If  I see one, I presume it's  inE ODS format, else not.  Can anyone give me other suggestions?  Thanks.    ------------------------------   Date: 15 Nov 2001 18:05:51 GMT+ From: Dale Dellutri <ddellutr@enteract.com>  Subject: Re: Auto CD Mounting + Message-ID: <9t105v$o03$1@bob.news.rcn.net>   S On Thu, 15 Nov 2001 08:40:06 -0800, Jack Trachtman <Jack.Trachtman@vmmc.org> wrote: G > I'm trying to write a DCL script which will mount a CD no matter what  > formatH > it's in (ODS, ISO, etc).  So far I'm taking the simplistic approach of
 > mountingI > the CD /FOREIGN and looking for a lable.  If  I see one, I presume it's  > inG > ODS format, else not.  Can anyone give me other suggestions?  Thanks.   E Here's what I use.  MOUNT_ANY.COM tries to mount the CD with /SYSTEM.   7 $! MOUNT_ANY.COM - Mount whatever disk is in the device  $! Dale Dellutri 1997-05-29  $ on severe_error then continue  $ say = "WRITE SYS$OUTPUT"" $ if p1 .eqs. "" .or. p1 .eqs. "?" $ then< $  say "MOUNT_ANY.COM - Mount whatever disk is in a device." $  say "(Dale 1997-05-29)" $  say " P1 = device"  $  exit  $ endif 	 $ say " " ' $ say "MOUNT /SYSTEM ",p1," <whatever>" 	 $ say " "   $ if .not. f$getdvi(p1,"EXISTS") $ then' $  say "*** Error: No such device: ",p1  $  exit %X10000002 $ endif # $ if f$getdvi(p1,"DEVCLASS") .ne. 1  $ then* $  say "*** Error: Device not a disk: ",p1 $  exit %X10000002 $ endif ! $ assign mount_any.tmp sys$output ! $ mount /noassist /system 'p1' ""  $ code = $status $ deassign sys$output  $ show symbol code' $! %X107201A4 MEDOFL, medium is offline  $!   No disk in device. 1 $! %X1072006C DEVMOUNT, device is already mounted 1 $! %X1072010C INCVOLLABEL, incorrect volume label 9 $!   Files-11 volume, get correct label and re-try mount. @ $! %X107208E4 NOHOMEBLK, Files-11 home block not found on volumeF $!   ISO 9660 volume?  Get correct label and re-try with /MEDIA=CDROM.) $! %X10720E0C DEADLOCK, deadlock detected B $!   ISO 9660 volume?  But cannot mount /SYSTEM since label is "".4 $ if code .ne. %X1072010C .and. code .ne. %X107208E4 $ then! $  say "Mount not done, see code"  $  delete mount_any.tmp.0  $  exit code $ endif   $ open /read lu_in mount_any.tmp $ read lu_in a $ read lu_in a
 $ close lu_in  $ delete mount_any.tmp.0, $ if code .eq. %X1072010C  ! Files-11 format $ then $  i = f$locate("label = '",a) $  if i .ne. f$length(a) $  then ( $   b = f$extract(i+9,f$length(a)-i-9,a) $   j = f$locate("'",b) 6 $   if j .eq. f$length(b) then goto bad_mount_response $   lbl = f$extract(0,j,b)& $   mount /noassist /system 'p1' 'lbl' $   exit $status $  endif $  i = f$locate("CDROM_ISO",a) $  if i .ne. f$length(a) $  then * $   b = f$extract(i+11,f$length(a)-i-11,a) $   j = f$locate(":",b) 6 $   if j .eq. f$length(b) then goto bad_mount_response $   lbl = f$extract(0,j,b): $   mount /noassist /system /media_format=cdrom 'p1' 'lbl' $   exit $status $  endif $  goto bad_mount_response $ endif 7 $ if code .eq. %X107208E4  ! ISO 9660 format, hopefully  $ then $  i = f$locate("CDROM_ISO",a)5 $  if i .eq. f$length(a) then goto bad_mount_response ) $  b = f$extract(i+11,f$length(a)-i-11,a)  $  j = f$locate(":",b)5 $  if j .eq. f$length(b) then goto bad_mount_response  $  lbl = f$extract(0,j,b) 9 $  mount /noassist /system /media_format=cdrom 'p1' 'lbl'  $  exit $status  $ endif  $bad_mount_response:> $ say "*** Error: Unknown response from MOUNT for device: ",p1 $ exit 0   --  & Dale Dellutri -- ddellutr@enteract.com   ------------------------------    Date: 15 Nov 2001 09:41:28 -0600- From: koehler@encompasserve.org (Bob Koehler) H Subject: Re: check this out  Web hosting  On OpenVMS  (customer link...)3 Message-ID: <m9B1NCFe4NS8@eisner.encompasserve.org>   b In article <3bf2cf50$1@news.si.com>, "Brian Tillman" <tillman_brian@notnoone.notnohow.com> writes: > H > Sorry, Sue, but there is no "Services" ,emu and no "Web Hosting" menu.  B    It's not visiable from Explorer or Lynx, but it is visible fromE    Netscape or Mozilla if you get past about 1/2 dozen framing error  3    complaints.  But I don't see any mention of VMS.    ------------------------------  % Date: Thu, 15 Nov 2001 11:56:10 -0500 2 From: rdeininger@mindspring.com (Robert Deininger)H Subject: Re: check this out  Web hosting  On OpenVMS  (customer link...)L Message-ID: <rdeininger-1511011156110001@user-2iveb9o.dialup.mindspring.com>  C In article <WJBI7.1577$RL6.51306@news.cpqcorp.net>, "Sue Skonetski" # <susan.skonetski@compaq.com> wrote:   N > Well I just tried it again and it worked did you click on service at the top > of the first screen?   > > >http://www.eurokom.ie/   - Well, my browser shows a lovely blank window.   0 Second-guessing from the "html" source, I tried &    http://www.eurokom.ie/v1/index.htmlG which is a bit better.  It does indeed have a "web hosting" link, which 	 points to $     http://www.eurokom.ie/v1/web.htm  $ which has some nice words about VMS.  H Editorial question:  What's so important about the main page that it had3 to be obscured behind some hokey javascript?  Sigh.   9 ( http://validator.w3.org/ doesn't like the page either.)    --   Robert Deininger rdeininger@mindspring.com    ------------------------------  % Date: Thu, 15 Nov 2001 10:40:14 -0800 < From: "Kenneth H. Fairfield" <Kenneth.H.Fairfield@intel.com>/ Subject: CJL's NOTLIB_COPY.C (Was: Re: LIB$xxx) ) Message-ID: <3BF40C0E.CF379DCF@intel.com>    "David J. Dachtera" wrote:   > "Kenneth H. Fairfield" wrote:  > ; > >     I'd be happy to repost if anyone wants to see them.  >   > Register one Yes vote, please.  .     I'm easy, one vote does it for me! :-) :-)  ;     I don't have the original comp.os.vms posting, only the 7 extracted code.  Hopefully, the attribution is obvious.        -KenK ===========================================================================   P /******************************************************************************\O  *  NOTLIB_COPY:  A function to copy an arbitrary file in block mode          * O  *  Copyright 1992 by the Caltech Odd Hack Committee.  No rights reserved     * O  *  Author:  Carl J Lydick (carl@sol1.gps.caltech.edu)                        * O  *      Arguments:                                                            * O  *              inp_file                                                      * O  *                                                                            * O  *              VMS usage: input_filespec                                     * O  *              type: character-coded text string                             * O  *              access: read only                                             * O  *              mechanism: by descriptor--fixed                               *fO  *                  length string descriptor                                  * O  *                                                                            *uO  *              Name of the file to be copied.                                * O  *              ---------------------------------                             *PO  *              out_file                                                      *gO  *                                                                            *pO  *              VMS usage: output_filespec                                    *aO  *              type: character-coded text string                             * O  *              access: read only                                             * O  *              mechanism: by descriptor--fixed                               *hO  *                  length string descriptor                                  *eO  *                                                                            * O  *              Name of the destination file.                                 *mO  *              ---------------------------------                             * O  *              file_id                                                       *eO  *                                                                            *eO  *              VMS usage: input_file_id                                      *lO  *              type: character-coded text string                             *oO  *              access: read only                                             * O  *              mechansim: by descriptor--fixed                               * O  *                  length string descriptor                                  *.O  *                                                                            *oO  *              NAM$T_DVI, NAM$W_FID, and NAM$W_DID fields associated         *tO  *              with the input file                                           *tP \******************************************************************************/  5 long int notlib_copy(struct dsc$descriptor *inp_file,hF     struct dsc$descriptor *out_file, struct dsc$descriptor *file_id) {   struct FAB  inp_fab, out_fab;t   struct RAB  inp_rab, out_rab;o   struct NAM  inp_nam;   struct XABSUM  xabsum;   struct XABKEY  *xabkey = 0;    struct XABALL  *xaball = 0;b   char  buffer[32256];
   long  stat;?
   char  *ptr;b
   int   i;     xabsum = cc$rms_xabsum;      inp_fab = cc$rms_fab; ,   inp_fab.fab$b_fac = FAB$M_BRO | FAB$M_GET;.   inp_fab.fab$l_fna = inp_file->dsc$a_pointer;=   inp_fab.fab$b_fns = (unsigned char) inp_file->dsc$w_length;s    inp_fab.fab$l_fop = FAB$M_SQO;/   inp_fab.fab$b_shr = FAB$M_SHRPUT | FAB$M_UPI;s&   inp_fab.fab$l_xab = (char *)&xabsum;   if (file_id != 0) {v     inp_nam = cc$rms_nam;i'     ptr = (char *)&(inp_nam.nam$t_dvi);i<     for(i = 0; i < (unsigned int)file_id->dsc$w_length; ++i))       ptr[i] = file_id->dsc$a_pointer[i];s#     inp_fab.fab$l_fop |= FAB$M_NAM;t!     inp_fab.fab$l_nam = &inp_nam;e   }t8   if (!$VMS_STATUS_SUCCESS(stat = SYS$OPEN(&inp_fab))) {*     Error("sys$open failed, rc:%d", stat);     return stat;   }l  +   if ((unsigned int)xabsum.xab$b_nok > 0) {g
     int i;  K     if ((xabkey = (struct XABKEY *) malloc((unsigned int)xabsum.xab$b_nok * '         sizeof(struct XABKEY))) == 0) {a&       Error("could not alloc xabkey");       return SS$_INSFMEM;.     }r  9     for(i = 0; i < (unsigned int)xabsum.xab$b_nok; ++i) {         xabkey[i] = cc$rms_xabkey;5       xabkey[i].xab$l_nxt = (char *)(&xabkey[i + 1]);-.       xabkey[i].xab$b_ref = (unsigned char) i;     }   ;     xabkey[(unsigned int)xabsum.xab$b_nok-1].xab$l_nxt = 0;6&     xabsum.xab$l_nxt = (char *)xabkey;   }o  +   if ((unsigned int)xabsum.xab$b_noa > 0) {P
     int i;  K     if ((xaball = (struct XABALL *) malloc((unsigned int)xabsum.xab$b_noa *h'         sizeof(struct XABALL))) == 0) { &       Error("Could not alloc xaball");       if (xabkey != 0) {         free(xabkey);s         xabkey = 0;n       }h       return SS$_INSFMEM;L     }r  :     for (i = 0; i < (unsigned int)xabsum.xab$b_noa; ++i) {        xaball[i] = cc$rms_xaball;5       xaball[i].xab$l_nxt = (char *)(&xaball[i + 1]);T.       xaball[i].xab$b_aid = (unsigned char) i;     }CH     xabkey[(unsigned int)xabsum.xab$b_nok-1].xab$l_nxt = (char *)xaball;;     xaball[(unsigned int)xabsum.xab$b_noa-1].xab$l_nxt = 0;G   }T  ;   if (!$VMS_STATUS_SUCCESS(stat = SYS$DISPLAY(&inp_fab))) {O-     Error("sys$display failed, rc:%d", stat);      if (xabkey != 0) {       free(xabkey);s       xabkey = 0;e     }i     if (xaball != 0) {       free(xaball);h       xaball = 0;c     }e     return stat;   }O     out_fab = inp_fab;,   out_fab.fab$b_fac = FAB$M_BRO | FAB$M_PUT;.   out_fab.fab$l_fna = out_file->dsc$a_pointer;=   out_fab.fab$b_fns = (unsigned char) out_file->dsc$w_length;m   out_fab.fab$w_ifi = 0;/   out_fab.fab$b_shr = FAB$M_SHRPUT | FAB$M_UPI;M"   out_fab.fab$l_fop &= ~FAB$M_NAM;   out_fab.fab$l_nam = 0;  :   if (!$VMS_STATUS_SUCCESS(stat = SYS$CREATE(&out_fab))) {,     Error("sys$create failed, rc:%d", stat);     SYS$CLOSE(&inp_fab);     if (xabkey != 0) {       free(xabkey);i       xabkey = 0;P     }M     if (xaball != 0) {       free(xaball);P       xaball = 0;t     }e     return stat;   }E     inp_rab = cc$rms_rab;    inp_rab.rab$l_fab = &inp_fab;     inp_rab.rab$l_rop = RAB$M_BIO;     out_rab = inp_rab;   out_rab.rab$l_fab = &out_fab;o   out_rab.rab$w_isi = 0;  ;   if (!$VMS_STATUS_SUCCESS(stat = SYS$CONNECT(&inp_rab)) ||-;       !$VMS_STATUS_SUCCESS(stat = SYS$CONNECT(&out_rab))) {u-     Error("sys$connect failed, rc:%d", stat);p     SYS$CLOSE(&inp_fab);     SYS$CLOSE(&out_fab);     if (xabkey != 0) {       free(xabkey);2       xabkey = 0;r     }r     if (xaball != 0) {       free(xaball);g       xaball = 0;      }i     return stat;   }      inp_rab.rab$l_ubf = buffer;s   inp_rab.rab$w_usz = 32256;   out_rab.rab$l_rbf = buffer;      while (1==1) {L     if (!$VMS_STATUS_SUCCESS(stat = SYS$READ(&inp_rab)) && stat != RMS$_EOF) {d,       Error("sys$read failed, rc:%d", stat);       SYS$CLOSE(&inp_fab);       SYS$CLOSE(&out_fab);       if (xabkey != 0) {         free(xabkey);s         xabkey = 0;A       },       if (xaball != 0) {         free(xaball);f         xaball = 0;u       }d       return stat;!     } else if(stat == RMS$_EOF) {l       SYS$CLOSE(&inp_fab);       SYS$CLOSE(&out_fab);       if (xabkey != 0) {         free(xabkey);s         xabkey = 0;o       }e       if (xaball != 0) {         free(xaball);o         xaball = 0;        }r       return RMS$_NORMAL;-     } else {,       out_rab.rab$w_rsz = inp_rab.rab$w_rsz;=       if (!$VMS_STATUS_SUCCESS(stat = SYS$WRITE(&out_rab))) {s/         Error("sys$write failed, rc:%d", stat);          SYS$CLOSE(&inp_fab);         SYS$CLOSE(&out_fab);         if (xabkey != 0) {           free(xabkey);            xabkey = 0;V	         }s         if (xaball != 0) {           free(xaball);v           xaball = 0;t	         }w           return stat;       }0     }h   }      return stat; }g  L ============================================================================   ------------------------------  % Date: Thu, 15 Nov 2001 10:43:48 -0800 < From: "Kenneth H. Fairfield" <Kenneth.H.Fairfield@intel.com>3 Subject: CJL's NOTLIB_PAGINATE.C (Was: Re: LIB$xxx)a) Message-ID: <3BF40CE4.3DB3D183@intel.com>-  D    By request (well _one_ request :-) here's a copy of Carl Lydick'sC original comp.os.vms post showing how to make LIB$PUT_OUTPUT "page"o its output.  Enjoy!p       -KenG -----------------------------------------------------------------------s    B Someone recently complained about the fact that applications usingO LBR$OUTPUT_HELP to display help don't paginate the results.  Well, I've writtenhM a small set of routines that allow one to do such pagination.  They won't, of O course, help at all with utilities already out there, but if you're planning totM write an application that uses LBR$OUTPUT_HELP (or any other application thataL writes to SYS$OUTPUT and for which you'd like to have the results paginated, you might find them useful.   L NOTLIB_PAGINATE_OUTPUT has the same semantics as LIB$PUT_OUTPUT, but, if theM output device is a terminal, paginates the output.  NOTLIB_PAGINATE_INPUT hasuH the same semantics as LIB$GET_INPUT, but resets the line counter used byO NOTLIB_PAGINATE_OUTPUT.  NOTLIB_PAGINATE_RESET resets the static variables usedn1 by NOTLIB_PAGINATE_OUTPUT to their initial state.o   #include dvidef  #include descrip
 long stat;8 #define ckstat(x) if (((stat = x) & 7) != 1) return stat5 notlib_paginate_output(struct dsc$descriptor *string) * {       return notlib_paginate(0, string); }D  J notlib_paginate_input(struct dsc$descriptor *result, struct dsc$descriptor(         *prompt, unsigned short *reslen) {       notlib_paginate(1, 0);5         return LIB$GET_INPUT(result, prompt, reslen);o }n   notlib_paginate_reset() % {       return notlib_paginate(2, 0);. }   : notlib_paginate(int action, struct dsc$descriptor *string)G {       static int paglen = 0, initialized = 0, curline = 0, istty = 0;01         static $DESCRIPTOR(devnam, "SYS$OUTPUT");h         static char buffer[80];M+         static $DESCRIPTOR(result, buffer);.=         static $DESCRIPTOR(prompt, "Hit return to continue");n           if(action == 2)s          {       initialized = 0;                 return 1;c	         }i         else if (action == 1)                  curline = 0;         else if (action == 0) %         {       if (initialized == 0)g(                 {       initialized = 1;J                         ckstat(LIB$GETDVI(&DVI$_TRM, 0, &devnam, &istty));O                         ckstat(LIB$GETDVI(&DVI$_TT_PAGE, 0, &devnam, &paglen));s$                         curline = 0;                 }a5                 if (istty && (curline >= paglen - 1))t@                 {       ckstat(LIB$GET_INPUT(&result, &prompt));$                         curline = 0;                 }p                 curline++;.                 return LIB$PUT_OUTPUT(string);	         }n }IP --------------------------------------------------------------------------------L Carl J Lydick | INTERnet: CARL@SOL1.GPS.CALTECH.EDU | NSI/HEPnet: SOL1::CARL  N Disclaimer:  Hey, I understand VAXen and VMS.  That's what I get paid for.  MyI understanding of astronomy is purely at the amateur level (or below).  SoUJ unless what I'm saying is directly related to VAX/VMS, don't hold me or myM organization responsible for it.  If it IS related to VAX/VMS, you can try toiJ hold me responsible for it, but my organization had nothing to do with it.   ------------------------------  % Date: Thu, 15 Nov 2001 07:21:38 -0800:* From: James Gessling <jgessling@yahoo.com>  Subject: Re: Clustering Textbook) Message-ID: <3BF3DD82.ED509647@yahoo.com>v  ' This might be what you are looking for:   5 http://www.aplawrence.com/Books/searchofclusters.htmls  J I haven't read it, but I understand it concludes that the "shared nothing" model!( is to be preferred.  (Not the VMS model)   Jimi   David D Miller wrote:0   > Folks: >eH > I'm under the impression that someone (several someones?) wrote a bookG > recently that compared the clustering techniques supported by various0L > operating systems.  But it didn't pop out when I searched Amazon.com.  Any > recommendations? >r > thanks, dave.=   ------------------------------  % Date: Thu, 15 Nov 2001 13:03:59 +0000B% From: Alan Greig <a.greig@virgin.net>"E Subject: Comdex Fall special report praises VMS but criticises Compaqy8 Message-ID: <s3f7vtk79huof8ci8ofqikiapsfrc646md@4ax.com>  C http://www.zdnet.com/special/stories/report/0,13518,2824020,00.htmla  
 Sample quote:   > Many IT decision makers, particularly those not well-versed inD technology, are leery of OpenVMS' shrinking talent pool, and believeB Microsoft's solutions are a good substitute. But Windows currentlyA lacks industrial strength compared to Compaq's operating systems._D Smith speaks fondly of an old button that reads, "Microsoft's answer@ to OpenVMS is NT 8.0." "That tells you how long it will take forC Windows to be as revered as OpenVMS," he says. "I think we are justlF about halfway there." To date, Windows server versions have had littleD success finding their way into datacenters. The result, according toC Smith, is that these decision makers rule out OpenVMS as an option.d  A Smith says that Compaq had a hand in creating this problem in the!E first place. In 1999, Compaq decertified its only authorized training A outfit, Global Knowledge. According to Smith, shortly thereafter,a< Compaq authorized four companies, including Parsec, to offerD certifications. But Compaq did a poor job communicating where to getF the training. As a result, OpenVMS lost some of its momentum. AlthoughA the market continues to grow, that growth is mostly from existingwA OpenVMS shops that are upgrading their OS, adding more systems touD existing clusters, or expanding their infrastructure. "You don't see= new companies coming online and buying OpenVMS," notes Smith.B   -- Alan   ------------------------------    Date: 15 Nov 2001 09:02:44 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris)eP Subject: Re: Compaq and OSDN Create Clustering Foundry for Open Source Community= Message-ID: <cf15391e.0111150902.4e3d0fc3@posting.google.com>S  W Alan Greig <a.greig@virgin.net> wrote in message news:<3BF16B3C.1838D0DC@virgin.net>... N > But Compaq have just open sourced the Tru64 clustering code  as far as I canM > see. In case you think the announcement just refers to a subset look at the-M > 50 odd powerpoint slides where it becomes obvious everything is going.  Them; > Single System Image code from Tru64 is going as well. See  > www.opensource.compaq.coms  3 I think this is unlikely.  Both sourceforge.net and1@ www.opensource.compaq.com say the technology for both the SingleF System Image (SSI) for Linux and Cluster Infrastructure (CI) for LinuxE projects comes from Compaq's NonStop Cluster for Unixware.  That came A out of the Tandem clustering effort, not the Digital (TruCluster)o effort.   F I suspect the features listed on the PowerPoint slides look similar toE those of TruClusters because they have similar goals, to meet similar ? customer needs, rather than because any Tru64 code is involved.   B If TruCluster code were in the public domain, I doubt Oracle would= have paid big bucks to license TruCluster technology for RAC.t  B If Compaq/HP end up doing to Tru64 what HP just did to HPE, then ID think it would be much more likely to see TruCluster code show up in Linux.C --------------------------------------------------------------------C Keith Parris | parris at encompasserve dot org | VMS consulting on:KC Clusters, Disaster Tolerance, Internals, Performance, Storage & I/ON   ------------------------------    Date: 15 Nov 2001 09:36:51 -0600- From: koehler@encompasserve.org (Bob Koehler)e Subject: Re: Compaq guarantees? 3 Message-ID: <NDAS2ziYuigW@eisner.encompasserve.org>m  q In article <cf15391e.0111140836.79ee16f5@posting.google.com>, keithparris_NOSPAM@yahoo.com (Keith Parris) writes:i > C > I've been wondering about this lately, because it occurred to me:r? > What's to prevent Compaq from providing a _different_ DII-COE=G > compliant system (e.g. Tru64) in place of VMS to existing VMS DII-COE*H > customers to meet the 15- or 20-year committments under the contracts,% > instead of continuing to offer VMS?   E    I don't know anywhere near enough about DII-COE to know if this istD    a possibility.  But I do know just enough about some customers toC    know they aren't buying VMS systems just because they're DII-COE I    compliant.  Even Compaq doesn't want to upset certain large customers._   ------------------------------    Date: 15 Nov 2001 10:50:41 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)  Subject: Re: Compaq guarantees? 3 Message-ID: <oTjn$guua71a@eisner.encompasserve.org>   c In article <NDAS2ziYuigW@eisner.encompasserve.org>, koehler@encompasserve.org (Bob Koehler) writes:*s > In article <cf15391e.0111140836.79ee16f5@posting.google.com>, keithparris_NOSPAM@yahoo.com (Keith Parris) writes:t >> iD >> I've been wondering about this lately, because it occurred to me:@ >> What's to prevent Compaq from providing a _different_ DII-COEH >> compliant system (e.g. Tru64) in place of VMS to existing VMS DII-COEI >> customers to meet the 15- or 20-year committments under the contracts,*& >> instead of continuing to offer VMS? > G >    I don't know anywhere near enough about DII-COE to know if this is F >    a possibility.  But I do know just enough about some customers toE >    know they aren't buying VMS systems just because they're DII-COEMK >    compliant.  Even Compaq doesn't want to upset certain large customers.   B The purpose of the DII-COE support is not to get sales from people@ who really wanted Solaris.  The purpose is to be able to sell toB government operations that really want VMS but are prohibited from+ buying systems that do not support DII-COE.   B That is the mid-term revenue opportunity.  In the long run some ofC the Solaris fans may come around, but that is not the justification  for the project.   ------------------------------  # Date: Thu, 15 Nov 2001 14:23:38 GMT # From: "John Smith" <a@nonymous.com>-9 Subject: Re: Compaq/HP merger - let them know how to votee> Message-ID: <KdQI7.193463$YL3.58599197@news3.rdc1.on.home.com>  = > Better write fast since Enron is being purchased by Dynegy:  > F >   http://www.chron.com/content/chronicle/special/01/enron/index.html+ >   HoustonChronicle.com - Hot Topic: Enronr >eB > Mr. Lay isn't accepting his $ 60.6 million severance pay. He did0 > sell $ 1.2 million of Compaq stock last month. >  > --Jerry Leslie   Jerry,  K Lay fouled up big-time, but at least he did the honorable thing in the end. G Who knows what he'd do if approached with the right arguments about the 
 CPQ/HP thing.   H I feel a bit like Diogenes wandering about with his lamp in search of an honest man.M   ------------------------------  % Date: Thu, 15 Nov 2001 16:35:36 +0100e= From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <noone@dummy.com>*9 Subject: Re: Compaq/HP merger - let them know how to vote*) Message-ID: <3BF3E0C7.DD528F43@dummy.com>c  2 Hm, it was "...in search for a human", wasn't it ?   Jan-Erik Sderholm.$   John Smith wrote:{ > J > I feel a bit like Diogenes wandering about with his lamp in search of an
 > honest man.    ------------------------------  % Date: Thu, 15 Nov 2001 09:11:24 -0800a' From: David Mathog <mathog@caltech.edu>g9 Subject: Re: Compaq/HP merger - let them know how to votea+ Message-ID: <3BF3F73C.34F15D54@caltech.edu>.   John Smith wrote:  > J > > Ben Rosen is the Chairman of our Board of Trustees.  So in our case it > > sounds liker. > > rather a waste of paper, ink, and postage. > B > Rosen is only one person on the Board, and he has only one vote.  H Somehow I don't see the other trustee members standing up to him on this issue - theyH have their own minds but the issue is not likely to be very important to them, and itB is important that they get along to some extent with the chairman.A I don't know that Caltech owns any Compaq stock, although that is  probableH since trustees often give a lot to the school - so BR might have donatedF Q stock.  But however much Q stock Caltech has it's a safe bet that BR has a lot more.    Regards,   David Mathog mathog@caltech.edu   ------------------------------  % Date: Thu, 15 Nov 2001 19:06:06 +0100=& From: John McLean <mcleanj@dplanet.ch>9 Subject: Re: Compaq/HP merger - let them know how to vote;* Message-ID: <3BF4040E.F61D4C23@dplanet.ch>   David Mathog wrote:r >  > John Smith wrote:i > >$' > > How to influence the Compaq/HP deal . > > ------------------------------------------ > > N > > If you are a university alumnus, write the Board of Trustees or Regents of. > > your alma mater and try and convince them. > T > Ben Rosen is the Chairman of our Board of Trustees.  So in our case it sounds like, > rather a waste of paper, ink, and postage.  B Hmmm .... Sounds like complaining to your mother-in-law about your
 wife   :-)     John   ------------------------------    Date: 15 Nov 2001 09:12:19 -0600- From: koehler@encompasserve.org (Bob Koehler)g" Subject: DEC 4000 300 has no halt?3 Message-ID: <I1PVSz3i7mr1@eisner.encompasserve.org>_  C    My DEC 4000 300 workstation doesn't seem to have a halt button.  A    There's just any empty hole.  This is a pain because VMS isn'tlD    properly tuned yet (taking hours to boot after I enabled DECnet VI    with only 32 MB RAM) and I can't get a chevron prompt to do a minimum fI    system boot until it comes up and I shut it down (boot on power on is  *    enabled, I WILL turn that off tonight).  D    Were these manufactured that way, or is this one just broke?  HowF    hard can I expect it to be to find the contacts to solder on my own%    (momentary contact switch, right)?u  B    And what's that "Secure system" jumper next to the "ROM update"E    jumper?  If I move it is there a keyboard combination that will doa
    a halt?   ------------------------------    Date: 16 Nov 2001 02:24:45 +0800, From: Paul Repacholi <prep@prep.synonet.com>& Subject: Re: DEC 4000 300 has no halt?- Message-ID: <874rnv6geq.fsf@prep.synonet.com>   / koehler@encompasserve.org (Bob Koehler) writes:   D >    My DEC 4000 300 workstation doesn't seem to have a halt button.! >    There's just any empty hole.t  F This is an Alpha?, not a Vax right? The switch has a cap on the switchF shaft, and it can go astray :( But they are a pretty standard unit, soA getting one from your favorite electronics pirate and replace the 1 cap. PITA isn't it; my M600 has lost its as well.p  F >    Were these manufactured that way, or is this one just broke?  HowD >    hard can I expect it to be to find the contacts to solder on my+ >    own (momentary contact switch, right)?!  D I know who had the right switch in stock, if your caught let me know  and I get a part number for you.   -- a< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.$@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Thu, 15 Nov 2001 12:11:38 -0000x+ From: "Tim Jackson" <tim.jackson@amsjv.com> ) Subject: DEC/X-Windows, VMS and eXcursionf& Message-ID: <3bf3ae41$1@pull.gecm.com>  G I am running a 2 node Alpha VMScluster VMS V7.3, TCPIP V5.1, DECwindows;C 1.2-6, Advanced Server V7.3 (as BDC to a WinNT 4 PDC) and ExcursionxF V7.2.177 (on Windows 95).  I have XDM enabled, configured and running,G userids hostmapped to NT usernames and with XDMCP enabled (in eXcursionME on the PC) I see the VMS nodes okay.  When I try to logon with a user H name of the form "xxx.yyyyy" it always fails, but user names of the formF "xxxxxxx" (i.e.. no period) work fine. Am I missing something obvious?8 Could it be that the period in a username is messing up?  * Any help or pointers would be appreciated?   TIA D ------------------ Purely Personal Opinion -------------------------D Tim Jackson                                    tim.jackson@amsjv.com Air Systems Group  Alenia Marconi Systems Ltd.=   ------------------------------  # Date: Thu, 15 Nov 2001 12:48:19 GMT-% From: Milton <mbhewitt@optonline.net>3- Subject: Re: DEC/X-Windows, VMS and eXcursiont8 Message-ID: <t2e7vtcc76d6bikqrjcisflpcfv9l0ocd0@4ax.com>  1 On Thu, 15 Nov 2001 12:11:38 -0000, "Tim Jackson"i <tim.jackson@amsjv.com> wrote:  H >I am running a 2 node Alpha VMScluster VMS V7.3, TCPIP V5.1, DECwindowsD >1.2-6, Advanced Server V7.3 (as BDC to a WinNT 4 PDC) and ExcursionG >V7.2.177 (on Windows 95).  I have XDM enabled, configured and running,-H >userids hostmapped to NT usernames and with XDMCP enabled (in eXcursionF >on the PC) I see the VMS nodes okay.  When I try to logon with a userI >name of the form "xxx.yyyyy" it always fails, but user names of the formiG >"xxxxxxx" (i.e.. no period) work fine. Am I missing something obvious?e  9 >Could it be that the period in a username is messing up?L  B It's not conforming to the 8.3 naming convention that I believe NTD uses to store the username to the registry or possibly the harddisk. Does xxxxxxxx.yyy work?P, If it does then, then you have your answer.   6 You may be able to change this behavior by editing the" NtfsDisable8dot3NameCreation value http://support.microsoft.com/support/kb/articles/Q201/1/29.ASP?LN=EN-US&SD=gn&FR=0&qry=8.3&rnk=2&src=DHCS_MSPSS_gn_SRCH&SPR=NTS40r [be aware of word wrapping]  n    + >Any help or pointers would be appreciated?e  
 Just a guess.i   Cheers,u Milton   ------------------------------  ! Date: Thu, 15 Nov 01 09:14:47 GMTt From: jmfbahciv@aol.comsM Subject: Re: Editing DCL command lines longer than the current terminal width + Message-ID: <9t081b$9gm$1@bob.news.rcn.net>   8 In article <hd55vtssjks56v7k3qpuv5duopg4t541mv@4ax.com>,)    Alan Greig <a.greig@virgin.net> wrote:;9 >On Wed, 14 Nov 01 13:38:39 GMT, jmfbahciv@aol.com wrote:n >  >l4 >>>   And having passwords on a per-directory basis. >>> >>For what?  Writing?  That happens to be a good design.  Most? >>directories could be read, unless somebody was really, really,< >>paranoid or there existed a user who liked to mess around. > F >I'm guessing he means that unlike VMS TOPS-20 did not have a distinct@ >user authorization file but instead held details in a directoryD >header. A normal directory would be "files only" but a user accountA >would be "not files only". Don't recall this ever causing me any B >problems though. TOPS-20 directory numbers mapped to TOPS-10 PPNs# >(UICs) for compatibility purposes.   8 The only people I encountered who found the architecture4 irritating were ones who were trying to fuck around.  9 In-house, most of us had POKE privs and we simply PIVOTed 6 to the PPN we needed to access as a user.  TOPS-10 and, TOPS-20 were not single-user system designs.   /BAH  ' Subtract a hundred and four for e-mail.    ------------------------------    Date: 15 Nov 2001 09:20:41 -0600- From: koehler@encompasserve.org (Bob Koehler) M Subject: Re: Editing DCL command lines longer than the current terminal width 3 Message-ID: <REiwMdQP5hH4@eisner.encompasserve.org>-  E In article <9su33p$t16$2@bob.news.rcn.net>, jmfbahciv@aol.com writes:C   re Macro-10:  8 > Then don't use it.  I suspect you're confused, though.  > Not an option since the compiler's couldn't do what we needed.  > > I find that to be proper.  What system call did you need forE > an application that wasn't available through the compiler language?r  H I don't recallw hich ones anymore, but we used several JSYS that neitherE Fortran not Cobol had a clue about.  Those were the only compilers we-G had until I loaded BLISS-10 off the freeware tape.  Coming from callingmC everything from Fortran on our RSX and VMS systems I found it to beo
 quite a pain.   > > For what?  Writing?  That happens to be a good design.  Most? > directories could be read, unless somebody was really, reallys< > paranoid or there existed a user who liked to mess around.  D Writing, reading, ...  No centralize UAF file, and up until the lastC couple years no option to encrypt the passwords.  The whole conceptiD of subdirectories seems to be solely for the admins to organize the + users, not for the users to organize files.a  & > It sounds more like you didn't RTFM.  D I had my head stuck in them all the time.  Coming from a VMS and RSXD background I "knew" DEC would have good manuals.  Something I didn't get let down on until ULTRIX.G   ------------------------------  % Date: Thu, 15 Nov 2001 16:49:12 +0000 % From: Alan Greig <a.greig@virgin.net>fM Subject: Re: Editing DCL command lines longer than the current terminal widths8 Message-ID: <d1s7vtk0auh5nqld9mfgcqqfh870aajt9a@4ax.com>  F On 15 Nov 2001 09:20:41 -0600, koehler@encompasserve.org (Bob Koehler) wrote:  F >In article <9su33p$t16$2@bob.news.rcn.net>, jmfbahciv@aol.com writes: >o
 >re Macro-10:B >W9 >> Then don't use it.  I suspect you're confused, though.p >s? >Not an option since the compiler's couldn't do what we needed.e >,? >> I find that to be proper.  What system call did you need forlF >> an application that wasn't available through the compiler language? >nI >I don't recallw hich ones anymore, but we used several JSYS that neither F >Fortran not Cobol had a clue about.  Those were the only compilers weH >had until I loaded BLISS-10 off the freeware tape.  Coming from callingD >everything from Fortran on our RSX and VMS systems I found it to be >quite a pain.  B You ran into the problem of trying to use a native TOPS-20 featureF from Fortran or Cobol which were designed to support TOPS-10 so had noB interface to things such as COMND.  A lot of sites wrote their ownE package but there were also (usually non DEC) provided languages suchrC as Pascal, Simula and SAIL which provided a high level interface touD TOPS-20 calls. DEC should have provided a general calling conventionA form HLLs to avoid everyone having to re-invent the wheel (no puno
 intended :-))i  D I am fairly certain that the PASCAL implementation was callable fromA any other TOPS-20 HLL including DEC's own. So you just called the-B PASCAL RTL COMND interface routines from FORTRAN. You had to go anF grab Rutgers Pascal first though and DEC should have done a better job
 here I agree.   D I can also recall a student who wrote a generally callable interfaceE from HLLs in Macro as part of a project. But he shouldn't really have6
 had to do so.r  C Sometimes DEC targeted layered products at TOPS-10 and assumed thatc6 not supporting the additional TOPS-20 features was ok.  ? >> For what?  Writing?  That happens to be a good design.  Mostl@ >> directories could be read, unless somebody was really, really= >> paranoid or there existed a user who liked to mess around.n >rE >Writing, reading, ...  No centralize UAF file, and up until the last D >couple years no option to encrypt the passwords.  The whole conceptE >of subdirectories seems to be solely for the admins to organize the j, >users, not for the users to organize files. >w' >> It sounds more like you didn't RTFM.e >gE >I had my head stuck in them all the time.  Coming from a VMS and RSXrE >background I "knew" DEC would have good manuals.  Something I didn'ts >get let down on until ULTRIX.   -- Alan   ------------------------------  % Date: Thu, 15 Nov 2001 09:51:12 -0500 - From: "Peter Weaver" <peter.weaver@stelco.ca> , Subject: Re: errlog, decevent, CA, and WEBES2 Message-ID: <NDQI7.59230$Z2.878638@nnrp1.uunet.ca>  2 "Jim Strehlow" <jims@data911.com> wrote in message* news:9sur3c$761@dispatch.concentric.net... >...( > Anyone familiar with "CA" and "WEBES"? >...  D I was going to send a link to the Deja cache of the notes I wrote onJ November 8th and 9th, but when I went looking for it I saw that Deja has aL gap and is missing all newsgroup postings from November 7th thru the 9th. So1 instead of a simple URL I am reposting both notesn  
 From the 8th;B8 "Dale Dellutri" <ddellutr@enteract.com> wrote in message% news:9se6ts$i82$1@bob.news.rcn.net...g> > Is TCP/IP really needed to install WEBES and Compaq Analyze? > C > I've got some ES40's, and the techs always ask if I've got Compaq D > Analyze installed when they dial in to look at a hardware problem.E > I've balked at installing CA (and WEBES) because of all the patchesm >...  J You'll be sorry... tell them that if they want to run that garbage to lookI at the error log then they can install it at their site and you will mailfK them a copy of the ERRLOG.SYS through DSN or some other means. That is whatsJ Compaq suggested we do if DIAGNOSE can not display the error correctly. SoI far DIAGNOSE has been able to handle every (2) error that we had with them ES40.   8 The only thing Compaq got right with CA is the initials.  L That said, when we were unfortunate enough to have that garbage installed onK one of our ES40's (we only put it on the test machine, I would not allow itsL to go on the production machine) we had TCPWare installed not UCX and we hadK no IP related problems. I do not recall seeing anything that looked like it  need an IP stack.e    
 From the 9th;58 "Dale Dellutri" <ddellutr@enteract.com> wrote in message% news:9seu3v$aoq$1@bob.news.rcn.net...n >...E > Can you give me some details on why you think it's [Compaq Analyze]m garbage? >...  J I'm sure that more than a few people saw this question and said "No! Don't8 ask him that! He'll go off into one of his rants again."   $ SET RANT/MODE=CA  K I no longer have CA installed on any running system here, so all of this is I from previous comp.os.vms postings and from e-mail exchanges with Compaq. I You may want to use www.deja.com (now known as Google Groups) to see whatuF others have said about CA. Sylvain Robitaille in particular has postedK several notes on problems using CA on TRU64, and we exchanged several notesIK comparing the problems between the two platforms. If you really search deja E enough you will find one of the original programmers who worked on CAtK stating "I neither like nor am particularly good at writing algorithm-based 0 classes," which he had to do for Compaq Analyze.  L The last version I tried was V4, (it was the current shipping version, not a0 beta) I do not know what the current version is.    L 1. It is a resource hog that is capable of bringing an ES40 to its knees CPUF wise. CA was great for demonstrating how a Galaxy can be configured toI migrate a CPU to the node that needed it most, but when we ran CA on bothfK Galaxy nodes at the same time the CPU kept switching back and forth. CompaqcK tells me that this is not an issue since V4, I never got far enough into V4T to see for myself.  G 2. It makes no use of standard VMS techniques. This means that a syntax1J error on the command line that should be caught by the CLI takes 6 secondsF to process and spawns at least one sub-process. If you have accountingF turned on for sub-processes you will see that the process fails with a
 INCHAN error.i  	 $ ca/zerrl@    Error During Command Parsing: Unrecognized CA command '/zerr' $ show symbol $status-      $STATUS == "%X1000013A"- $ write sys$output "''f$message("%X1000013A")-(    %SYSTEM-E-IVCHAN, invalid I/O channel  L I do not know about you, but when I see IVCHAN in an accounting record, I goF looking for the programmer to tell him that his application is broken.  L Notice that the error format of FACILITY-SEVERITY-IDENTIFICATION-TEXT is not used by CA.n  F 3. After installing V4 I booted the system, I then used the ACCOUNTINGK utility to check on what the processes did during the boot. I noticed threeiJ entries that had "File Not Found" as the exit status. I never did find out what caused these errors.   I 4. The RCM logs had "%DCL-W-UNDSYM, undefined symbol - check validity andtH spelling" errors in them, I never bothered to find out what caused this.  C 5. I had an unresolved problem with seeing sub-processes dying withnB "%SYSTEM-I-GSDFULL, global section descriptor table is full" in my accounting records.   L 6. My first experience with CA V1 had me waste a lot of time trying to trackJ down an non-existing TRU64 node on the network. For some reason CA decidedH to read through the example database and presented me with an error fromJ node "sabl28," I spent a long time trying to track down this node before IK found out that this was an example error. Look for "sabl28" in deja for thehL details. Compaq tells me that this bug is fixed, but they could not offer me any proof about it.c  I 7. If you look at the various .COM files that get installed, you will see0L that some (most?? I do not recall how many now) had the template information still there. i.e.o   $! Facility:
 $! {~tbs~} $! $! Abstract:
 $! {~tbs~} $!
 $! Author:
 $! {~tbs~}  L There was one .COM that I found that was nothing but a template, i.e. it hadG places to put standard comments, but it did nothing. This tells me thatmK there was no code review before this shipped, so I do not know what else isn hiding in the code.   F 8. ... I had more, but I stopped looking at it when some senior peopleJ started viewing my comments on CA as more reasons to get rid of VMS. Now IL have to convince them that VMS itself is still a stable, quality product, it@ is just CA (and DSN V2... and MIME...) that are severely broken.   ------------------------------    Date: 15 Nov 2001 05:22:13 -0800 From: richv@mail4u.nl (Rich)Y Subject: Re: How to connect a DS20 to a Fibre  channel SAN switch an that one to a hsg80? = Message-ID: <a38c070e.0111150522.73edc590@posting.google.com>i   "Main, Kerry" <Kerry.Main@compaq.com> wrote in message news:<BE56C50EA024184DAF48F0B9A47F5CF4D562B5@kaoexc01.americas.cpqcorp.net>...v > Rich,t > ? > As previous replies stated, if you do not have any experiencewJ > connectiing a host system (on any platform) to a SAN, it would likely beJ > a good idea to contract someone for a few days of to assist you with theA > initial setup, configuration, tuning and backup considerations.o > G > More additional information specific to OpenVMS and SANs can be found  > at: 8 > http://www.openvms.compaq.com/openvms/fibre/index.html > H > Reference Table 1 at the end of the page for updated VMS configuration
 > doc's .. > 
 > Regards, >  > Kerry Main > Senior Consultant. > Compaq Canada Corp.R > Professional Servicesl > Voice: 613-592-46606 > Fax  :  819-772-7036 > Email: Kerry.Main@Compaq.com >  >  > -----Original Message-----% > From: Rich [mailto:richv@mail4u.nl]v! > Sent: November 2, 2001 10:51 AM  > To: Info-VAX@Mvb.Saic.ComnJ > Subject: How to connect a DS20 to a Fibre channel SAN switch an that one > to a hsg80?? >  > 	 > Hi all,t > G > I recieved some beautifull equipment over here, an Alpha server DS20,eG > a hsg80 including 256Mb cache and a fibrechannel San. I do now how torC > connect the hsg to the server. But i haven't a single clue how toh1 > connect the SAN between them. Can anybody help?o >  > Youre help is apriciated,o >  > Rich    
 Gentleman,  F I Just did the job, in fact it wasn't that difficult, only a couple ofC hours work. That's what make me think: if you need a concultant for 7 this, mabey it is better you find yourself a new job...    ------------------------------  % Date: Thu, 15 Nov 2001 14:37:40 +0000s% From: Alan Greig <a.greig@virgin.net>oY Subject: Re: How to connect a DS20 to a Fibre  channel SAN switch an that one to a hsg80?l8 Message-ID: <rek7vts3le56sm3nor9vftk85jb3tmnv58@4ax.com>  < On 15 Nov 2001 05:22:13 -0800, richv@mail4u.nl (Rich) wrote:     >uG >I Just did the job, in fact it wasn't that difficult, only a couple ofcD >hours work. That's what make me think: if you need a concultant for8 >this, mabey it is better you find yourself a new job...  C You didn't say in your initial post Kerry responded to that you had,E previous experience configuring HSZ controllers. If you've configuredDC (say) an HSZ80 pair then configuring an HSG80 SAN should be do-ablen' with a bit of advice. As you've proved!   E Kerry is a consultant/ambassador for Compaq on VMS so he was probablyiE assuming you were starting from scratch and hoping Compaq could get a E little bit of business setting it up for you. If you really had neveruD configured any kind of external controllers and were unfamiliar withD some of the concepts then Kerry's advice was definitely good imho...     -- Alan   ------------------------------  % Date: Thu, 15 Nov 2001 09:44:24 +0000-% From: Alan Greig <a.greig@virgin.net>o) Subject: Re: HP kills MPE. It's official.t8 Message-ID: <d937vtsasckbgs6l48tkvlbtipefs8g93m@4ax.com>  , On Wed, 14 Nov 2001 20:36:12 -0500, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:   K >Oje thing that worries me is the timing of this announcement. Why would HP  >announce this NOW ?  A Very good question. If you read the MPE newsgroup HP was recentlysC giving an upbeat message about the future of MPE and several peopleaD have stated that they had just made new purchases based on seeminglyF cast iron guarantees about the future (new JAVA support, IA64 support,B updated SQL database, POSIX  etc). Then all of a sudden out of theC blue "BANG". Seems to have taken HP's MPE staff by surprise as they.E had been assured that MPE would move forward, Some of them apparentlye; now feel sick that they were (ab)used to mislead customers.n -- Alan   ------------------------------  % Date: Thu, 15 Nov 2001 09:50:24 +0000 % From: Alan Greig <a.greig@virgin.net> ) Subject: Re: HP kills MPE. It's official.t8 Message-ID: <kk37vt8p7an4jchh3tg4irl4q224k9alhg@4ax.com>  , On Wed, 14 Nov 2001 20:19:43 -0500, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:t   >s >o >kE >Question: Had  HP made any "commitments" about porting MPE to IA64 ?S  F Yes they had committed to supporting MPE on  HP IA64 servers. AlthoughD most of the porting work was to be eased by using the IA-64 hardwareB features designed in by HP to provide a fast emulation of PA-RISC.  D I was at the CA MANMAN Europe user group conference (CAMUS) the weekC of the Compaq/Intel announcement. MANMAN runs on only two platformssE MPE and VMS. The MPE crowd were in extremely confident mood as they'd F all just received wonderful glossies and presentations from HP telling6 of MPEs glorious future. Today they are spitting fire. -- Alan   ------------------------------  % Date: Thu, 15 Nov 2001 05:39:24 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca>,) Subject: Re: HP kills MPE. It's official.a, Message-ID: <3BF39B5B.A3394E4F@videotron.ca>   Alan Greig wrote:eH > Yes they had committed to supporting MPE on  HP IA64 servers. AlthoughF > most of the porting work was to be eased by using the IA-64 hardwareD > features designed in by HP to provide a fast emulation of PA-RISC.  I That looks pretty bad for VMS then. The commitment of Compaq to VMS is as , valid/believeable as HP's commitment to MPE.  L The one metric that remains to compare is the amount of profits generated byM MPE versus those generated by VMS. Is it possible that MPE was just a drop in N th ebucket and thus no loss for HP wheras for Compaq, VMS's profits are needed) to subsidize their core wintel business ?s  M Also, you mentioned that MANMAN runs only on MPE or VMS. Could the killing of K MPE be a way to migrate MPE customers over to VMS ? (which would give VMS'sh survival more chances).b  H Or is it likely that those customers will just go to SAP on IBM or SUN ?   ------------------------------  % Date: Thu, 15 Nov 2001 11:45:47 +0000a% From: Alan Greig <a.greig@virgin.net> ) Subject: Re: HP kills MPE. It's official. 8 Message-ID: <ml97vt45a2lr34p29g7khpkim6ioc9vjij@4ax.com>  , On Thu, 15 Nov 2001 05:39:24 -0500, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:b    N >Also, you mentioned that MANMAN runs only on MPE or VMS. Could the killing ofL >MPE be a way to migrate MPE customers over to VMS ? (which would give VMS's >survival more chances).  D I don't think there are anything like the number of MPE MANMAN usersE worldwide to make this an issue. Maybe a few hundred VMS system salesaA or a few thousand at most if all migrated Being an ERP system youiE don't exactly have one copy per desktop. Maybe one machine or cluster / serving hundreds or perhaps thousands of users.u  C At CAMUS Europe a major presentation was given by the Ford CosworthwE team who use MANMAN on MPE to control their specialist Cosworth partslD production line (including high performance engines). Last year theyA reviewed all options including switch to VMS and switch away fromtB MANMAN completely. They upgraded to new faster MPE systems feelingE that was the safest option. I don't know how many other parts of FordnD still use MANMAN but there could be a number of other divisions alsoE caught out. MANMAN was re-certified by Ford just last year as meetingL) corporate requirements for certain units.    > I >Or is it likely that those customers will just go to SAP on IBM or SUN ?u  < Probably. Absolutely anyone but HP judging by the newsgroup. -- Alan   ------------------------------    Date: 15 Nov 2001 07:22:10 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris)() Subject: Re: HP kills MPE. It's official.$= Message-ID: <cf15391e.0111150722.42070677@posting.google.com>r  a JF Mezei <jfmezei.spamnot@videotron.ca> wrote in message news:<3BF31C0A.1F82C74E@videotron.ca>...cL > One thing that worries me is the timing of this announcement. Why would HP > announce this NOW ?e  C HP had just announced its Q4 quarterly earnings earlier on the samenE day.  Net income was down 89% from a year ago; revenues were down 18%S from a year ago.B (Bad as that sounds, it was actually much better than analysts had predicted.)I  D Most likely this announcement was intended to make HP look better on Wall Street.C -------------------------------------------------------------------aC Keith Parris | parris at encompasserve dot org | VMS consulting on:rC Clusters, Disaster Tolerance, Internals, Performance, Storage & I/Of   ------------------------------  # Date: Thu, 15 Nov 2001 15:22:16 GMT . From: "Duane Sand" <duane.sand@mindspring.com>) Subject: Re: HP kills MPE. It's official.h> Message-ID: <I4RI7.15680$RG1.8277412@news1.rdc1.sfba.home.com>   "JF Mezei" wroteL > ... Could the killing of MPE be a way to migrate MPE customers over to VMS ? 5 > (which would give VMS's survival more chances). ...e  D MPE software expects a big-endian memory address space; VMS suppliesH the opposite.  That dooms the prospect of any kind of binary object codeD emulator or translator for MPE applications on VMS.  Moving onto theE big-endian OS's for Sun or Power would be technically easier.  Within ? the HP & Compaq spheres, HP/UX and NonStop (someday) on IPF arecC both big-endian.  With the completed port of HP/UX onto IPF and therF completion of the Aries PA-RISC binary emulators & translators on IPF,E HP had already completed well over half of the work needed to support?D MPE on IPF.  But sometime this year HP choose to go real slow on the; MPE port to reduce headcount.  And now they've said, never.    ------------------------------  % Date: Thu, 15 Nov 2001 11:35:49 -0500.- From: JF Mezei <jfmezei.spamnot@videotron.ca>s) Subject: Re: HP kills MPE. It's official.p, Message-ID: <3BF3EEE3.A5F1B420@videotron.ca>   Keith Parris wrote:oF > Most likely this announcement was intended to make HP look better on > Wall Street.  N I cannot understand why corporations have this idea that they will be rewardedJ by Wall Street if they cancell their most profitable products in favour of- their more popular but money losing products.l  M I cannot understand why Wall Street would push companies to embrace Microsoft0H products when they know that those products are not profitable, and that+ Microsoft itself isn't actually profitable.D (has MS declared dividends ?)u   ------------------------------  % Date: Thu, 15 Nov 2001 09:00:55 -0800$' From: David Mathog <mathog@caltech.edu>n) Subject: Re: HP kills MPE. It's official. + Message-ID: <3BF3F4C7.C3321593@caltech.edu>e   Alan Greig wrote:   F > I was at the CA MANMAN Europe user group conference (CAMUS) the weekE > of the Compaq/Intel announcement. MANMAN runs on only two platformstG > MPE and VMS. The MPE crowd were in extremely confident mood as they'deH > all just received wonderful glossies and presentations from HP telling8 > of MPEs glorious future. Today they are spitting fire.  ? Which would be a glorious chance for VMS to pick up the kind ofn customers who spendkD a lot of $$$ and stick with a platform through thick and thin, if it were not forF the multiple poison pills Compaq has been swallowing of late. The only	 chance of.H those customers moving now will be if the merger tanks (and quickly) and	 somethingeA miraculous happens at Compaq, like they split the company into anf enterprise companyG and a PC company, or IBM buys them (which would be pretty much the samet	 thing).     G So, unfortunately, rather than HP MANMAN users moving to VMS MANMAN, ite seems a loteG more likely that MANMAN will be ported to Solaris, AIX, or even Linux. i
 That's what It" would do if MANMAN was my product.   Regards,   David Mathog mathog@caltech.edu   ------------------------------   Date: 15 Nov 2001 11:32 CSTI' From: carl@gerg.tamu.edu (Carl Perkins)t) Subject: Re: HP kills MPE. It's official.C- Message-ID: <15NOV200111322057@gerg.tamu.edu>r  1 JF Mezei <jfmezei.spamnot@videotron.ca> writes...P }Keith Parris wrote:G }> Most likely this announcement was intended to make HP look better one }> Wall Street.s } O }I cannot understand why corporations have this idea that they will be rewardedyK }by Wall Street if they cancell their most profitable products in favour ofn. }their more popular but money losing products.  # Because a lot of people are stupid.   N }I cannot understand why Wall Street would push companies to embrace MicrosoftI }products when they know that those products are not profitable, and thatt  
 See above.  , }Microsoft itself isn't actually profitable. }(has MS declared dividends ?)  G You are mistaken. Microsoft is *hugely* profitable - its overall profitoF margin, after everything, is about 25% (some $6.4 billion in profit inF the last 4 quarters reported - they also have over $36 billion in cashF accumulated by being profitable and zero debt). There aren't very manyJ companies that make that much profit, or have that high of a profit marginG (GE outdoes them by a factor of 2 in quantity, but have a profit marginIJ that is less than half Microsoft's and under a quarter as much cash layingF around and, for some reason that is unclear to me, considerable debt).C Dividends have nothing to do with it. Some companies pay them, somemD don't - it has nothing to do with profitablilty (other than that youC can't pay them for too long if you don't make at least some money).h   --- Carl   ------------------------------    Date: 15 Nov 2001 10:01:20 -0800( From: bob@instantwhip.com (Bob Ceculski)) Subject: Re: HP kills MPE. It's official.a= Message-ID: <d7791aa1.0111151001.66783cb7@posting.google.com>1  a JF Mezei <jfmezei.spamnot@videotron.ca> wrote in message news:<3BF39B5B.A3394E4F@videotron.ca>...t > Alan Greig wrote:sJ > > Yes they had committed to supporting MPE on  HP IA64 servers. AlthoughH > > most of the porting work was to be eased by using the IA-64 hardwareF > > features designed in by HP to provide a fast emulation of PA-RISC. > K > That looks pretty bad for VMS then. The commitment of Compaq to VMS is asT. > valid/believeable as HP's commitment to MPE. > N > The one metric that remains to compare is the amount of profits generated byO > MPE versus those generated by VMS. Is it possible that MPE was just a drop inmP > th ebucket and thus no loss for HP wheras for Compaq, VMS's profits are needed+ > to subsidize their core wintel business ?  > O > Also, you mentioned that MANMAN runs only on MPE or VMS. Could the killing ofeM > MPE be a way to migrate MPE customers over to VMS ? (which would give VMS's  > survival more chances).  > J > Or is it likely that those customers will just go to SAP on IBM or SUN ?    J if anything saves vms, it will be the military and the DOE commitments ...   ------------------------------    Date: 15 Nov 2001 12:25:08 -0600+ From: young_r@encompasserve.org (Rob Young) ) Subject: Re: HP kills MPE. It's official.M3 Message-ID: <wiPxg$CpTEmv@eisner.encompasserve.org>i  U In article <3BF3F4C7.C3321593@caltech.edu>, David Mathog <mathog@caltech.edu> writes:c > Alan Greig wrote:  >    > I > So, unfortunately, rather than HP MANMAN users moving to VMS MANMAN, itn
 > seems a lotoI > more likely that MANMAN will be ported to Solaris, AIX, or even Linux. o > That's what In$ > would do if MANMAN was my product. >   
 	Yep.  But...i  > 	No more than you port DSM to Solaris or AIX.  You don't.  You= 	create a new Mumps that's kinda sorta like DSM but lacks thee? 	goodness DSM customers come to expect.... and call it ISM.  Ita@ 	doesn't support "true" clustering, etc. but who needs that!?!!!D 	Why you just stick in a failover box... yeah ... that's the ticket!  = 	So.. maybe Alan can comment.  Does MANMAN take advantage of	 E 	VMS in a big way?  i.e. uses its DLM, After Image Journalling, etc.?m   	If so... no easy port at all.  D 	MANMAN may be taking advantage of the strengths of MPE and VMS. . .> 	i.e. their native filesystems.  The Unix answer of course is > 	"we don't need no stinking filesystem features... why we have> 	all the tools you need to build what you need..."  And so theC 	mindset of least common denominator coding comes into play (unless)A 	of course you port it to Oracle as a backend and skip/substitutee6 	native filesystem features.  Get out your paychecks!)  < 	Gotta love clashing paradigms!  Doesn't get any better than 	that!   				Robo   ------------------------------  % Date: Thu, 15 Nov 2001 14:43:40 +0000 % From: Alan Greig <a.greig@virgin.net> Y Subject: Re: HP's Acquisition of Compaq: Beyond the Naysayers, the Promise and   the Oppo>8 Message-ID: <10l7vtcgriu5iqagfl1bilgqh5af7f7iti@4ax.com>  , On Thu, 15 Nov 2001 09:07:11 -0500, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:w  A >> >http://www.aberdeen.com/ab_company/freeresearch/hp_compaq.htmo >s ed as Hollywood sitcom writers., >nK >I can't believe that such reports would be written and PUBLISHED by a firmt- >that is regarded as a reputable serious one.   E I can't believe that such reports should be praised by Carly and theneD FILED WITH THE SEC by a firm that is regarded as a reputable serious one such as Compaq.a  @ Err maybe that last bit about Compaq being reputable and serious doesn't work :-( -- Alan   ------------------------------  # Date: Thu, 15 Nov 2001 11:50:34 GMTq% From: Milton <mbhewitt@optonline.net>-Y Subject: HP's Acquisition of Compaq: Beyond the Naysayers, the Promise and the OpportunitV8 Message-ID: <gaa7vt4lf05j73j3clf1osvo8u21oaheu4@4ax.com>  = http://www.aberdeen.com/ab_company/freeresearch/hp_compaq.htma  ? Readers of this newsgroup may be particularly interested in thea  A Unix Perspective: Creating Order in HP/Compaq's High-End and Unixn Platform Strategy    Cheers,h Milton   ------------------------------  % Date: Thu, 15 Nov 2001 12:28:39 +0000 % From: Alan Greig <a.greig@virgin.net>eY Subject: Re: HP's Acquisition of Compaq: Beyond the Naysayers, the Promise and the Opportt8 Message-ID: <75d7vt8cmpmir2vjcqkljlig55o3p07pb1@4ax.com>  A On Thu, 15 Nov 2001 11:50:34 GMT, Milton <mbhewitt@optonline.net>u wrote:  > >http://www.aberdeen.com/ab_company/freeresearch/hp_compaq.htm >1@ >Readers of this newsgroup may be particularly interested in the >uB >Unix Perspective: Creating Order in HP/Compaq's High-End and Unix >Platform Strategy  F Not only have we read it we've had the key author Peter Kastner try toD justify his "drop VMS port" advice in this newsgroup. He didn't stay	 for long.    >Cheers, >Miltoni   -- Alan   ------------------------------  # Date: Thu, 15 Nov 2001 13:03:45 GMTc% From: Milton <mbhewitt@optonline.net>oY Subject: Re: HP's Acquisition of Compaq: Beyond the Naysayers, the Promise and the Opport 8 Message-ID: <3pe7vtsa5he9bof18g9iq5kf9iv6q93r1a@4ax.com>  C On Thu, 15 Nov 2001 12:28:39 +0000, Alan Greig <a.greig@virgin.net>A wrote:  B >On Thu, 15 Nov 2001 11:50:34 GMT, Milton <mbhewitt@optonline.net> >wrote:/ >a? >>http://www.aberdeen.com/ab_company/freeresearch/hp_compaq.htmt >>A >>Readers of this newsgroup may be particularly interested in thee >>C >>Unix Perspective: Creating Order in HP/Compaq's High-End and Unixf >>Platform Strategy  > G >Not only have we read it we've had the key author Peter Kastner try totE >justify his "drop VMS port" advice in this newsgroup. He didn't staye
 >for long.  : Oh, that must have been interesting to have witnessed.  ;)  ? What migration strategy did he have to offer, other than a EOL?s   Cheers,v Milton   ------------------------------  % Date: Thu, 15 Nov 2001 16:17:09 -0000a- From: wspencer@ap.nospam.org (Warren Spencer)oY Subject: Re: HP's Acquisition of Compaq: Beyond the Naysayers, the Promise and the Opporto7 Message-ID: <915A72338warrenspencer1977@207.126.101.97>s  ) mbhewitt@optonline.net (Milton) wrote in f- <gaa7vt4lf05j73j3clf1osvo8u21oaheu4@4ax.com>:   > >http://www.aberdeen.com/ab_company/freeresearch/hp_compaq.htm >a@ >Readers of this newsgroup may be particularly interested in the >eB >Unix Perspective: Creating Order in HP/Compaq's High-End and Unix >Platform Strategy >t >Cheers, >Miltont   Here's quote from the report:o  D "To succeed, HP/ Compaq must focus on customer outreach, clarity of E messages, and marketing execution excellence, all of which have been -% largely absent for nearly two years."   H I believe this says both companies sucked at marketing for the last two L years.  So in the throes of a mega-merger, they need to become excellent at  it?  Bloody unlikely, it seems.   F Assuming HPaq won't become overnight marketing mavens (safe bet), and G applying the quote from above, Aberdeen is predicting they will fail.  hB 'Bout time Aberdeen woke up - just have to read between the lines.  F In fact, Aberdeen makes numerous such statements, setting the bar for K performance and execution so high that it seems unlikely any company could AC reach it. Good thing it's all couched in seemingly positive terms, e> otherwise it wouldn't have credibility with senior management.   ws   -- d   Warren Spencer' Senior Software Engineer (not a writer)i The Associated Press  L ** My employer does not necessarily agree with my statements - neither do I  **   ------------------------------  % Date: Thu, 15 Nov 2001 09:07:11 -05001- From: JF Mezei <jfmezei.spamnot@videotron.ca>oY Subject: Re: HP's Acquisition of Compaq: Beyond the Naysayers, the Promise and the OpportF, Message-ID: <3BF3CC0D.9C2F82B4@videotron.ca>  @ > >http://www.aberdeen.com/ab_company/freeresearch/hp_compaq.htm    Q They pay 2 guys to write a 2 page report ?  My digestive system is having spasms.e  L I like the "Compaq's profits all come from the non-wintel stuff therefore weM recomment that HP focus on Compaq's wintel business" argument they put forth.m  M I also liked the "Digital had planned to migrate VMS used to Windows but that N strategy was scapped when Windows on Alpha was scrapped". These guys should be" hired as Hollywood sitcom writers.  J I can't believe that such reports would be written and PUBLISHED by a firm, that is regarded as a reputable serious one.   ------------------------------  % Date: Thu, 15 Nov 2001 17:09:12 +0000 # From: John McNulty <knode@jmtl.com>w Subject: Re: HP/Compaq mergerl/ Message-ID: <tv7tmsjppr2l11@corp.supernews.com>s   M. Ranjit Mathews wrote: >BF > Many notebook computers are already made in Asia with only the labelF > being slapped on by a US company. Consider Qanta computer in Taiwan.= > It puts together notebook computers by the truckload for USk< > "manufacturers" who are really only reselling (rather thanF > manufacturing) them. It's just as easy for a US company to outsourceF > manufacturing to Asia as it is for an Asian company, so why would US# > companies not be able to compete?S  D The only problem with this model is what happens in low demand.  An H outsourced deal usually comes with some kind of estimate about how many B units are going to be sold by the vendor.  In order to be as cost H competative as possible, or even profitable, the outsource manafacturer H will tune their manufacturing process to meet that need, and the vendor H usually agrees to purchase a fixed quantity for sale.  If they suddenly L can't sell that number they are in trouble because they are still obligated L to buy the products from the manufacturer.  Product which then sit on their J shelves as surplus inventory waiting to be sold.  This can kill a vendors  profitability.  I Outsourcing 'Just In Time' manufacturing can be very profitable when the i6 going is good, and very dangerous when times are hard.   ------------------------------  % Date: Thu, 15 Nov 2001 08:38:06 +0100A) From: "Pio Baettig" <baettig@hotmail.com>r" Subject: Re: IBM to bid for CPQ ??+ Message-ID: <3bf37015@siufuxsun02.unifr.ch>t  @ Could it be possible, that at some point, IF both architectures,= Hammer & IA64 prove equally succesful (in sold quantity) that ? Features would be cross licensed between Intel and AMD, or that8@ one firm would be forced to make its processor compatible to the? other one (as AMD didn't enhance 3Dnow but took ISSE into their2 "XP" processors)?oB I don't know if software vendors want to support their apps on two: mainstream architectures and put two boxes on the shelves.    5 "Bill Todd" <billtodd@metrocast.net> wrote in message 9 news:L3CI7.739768$Lw3.46273245@news2.aus1.giganews.com...qJ > See:  when you can get Fred away from Compaq and Afghanistan, he becomes > positively rational: >w@ > Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote in message/ > news:aZAI7.1573$RL6.51017@news.cpqcorp.net...l >l > ...a >uC > > My crystal ball.  Hammer is a good idea with no vendor support.- >-H > Unfortunately, likely true:  it will have to prove itself before being" > embraced, very much unlike IA64. >n >   It will be3 > > used by 97% of all consumers only in IA32 mode.t >gI > Well, yuh - since soon it will perform double duty as AMD's *only* IA321L > offering, and IA32 is likely to vastly outsell *all* (non-embedded) 64-bit' > platforms for the foreseeable future.1 >0 >   There will be a LINUXaC > > version using the 64-bitness, but nobody will use the features,a >oL > That one's not so clear:  it could easily dominate the Linux 64-bit serverA > (and 'server appliance') space, and Linux servers are already as > non-negligible market. >o >  and mosto' > > Linux users won't use the features.j > L > At the end-user level, probably true - though it'll make one mean high-end- > workstation for a small percentage of them.a > ( >   Windows will only run IA32 mode, and+ > > never support the extended 64 features.h >gC > Sightings of Windows code with x86-64-specific features have beeni	 reported,rJ > but as I said elsewhere you can't count on it until you can buy it (look at > Win64 on Alpha). >  >   MS will embrace Itanium. >BI > Absolutely (unless Itanium tanks before they have the opportunity to doh so).H > The only question is whether they'll do so to the exclusion of Hammer:H > unless Hammer encounters some unanticipated disaster, it will offer anJ > *extremely* attractive platform on which to build low-end-to-medium-sizeJ > 64-bit servers, and since Linux (and *BSD, etc.) *will* be in that spaceG > Microsoft will have to decide whether they want to compete there moreu* > effectively by running 64-bit on Hammer. >t > >iG > > At some point in time, the Itanium Processor Family will exceed theu! > > performance of the IA32 code,r >t9 > Not until long after IA32 processor development ceases.r > 0 >  MS will write some new cool applications that > > require IA64,h >tH > If they can marginalize Hammer, that could happen.  But in the processI > they'll have to give up a lot to Linux, as described above - and in the-G > process would run the risk of marginalizing themselves.  Hard to knowg whichhJ > way they'll jump, and it may well not be clear until Hammer has had real. > exposure rather than promising expectations. >a > - bill >A >N >r   ------------------------------  # Date: Thu, 15 Nov 2001 09:28:37 GMTe* From: "Bill Todd" <billtodd@metrocast.net>" Subject: Re: IBM to bid for CPQ ??@ Message-ID: <9VLI7.29421$jp.2181262@bin2.nnrp.aus1.giganews.com>  2 Pio Baettig <baettig@hotmail.com> wrote in message% news:3bf37015@siufuxsun02.unifr.ch...gB > Could it be possible, that at some point, IF both architectures,? > Hammer & IA64 prove equally succesful (in sold quantity) thatoA > Features would be cross licensed between Intel and AMD, or that B > one firm would be forced to make its processor compatible to theA > other one (as AMD didn't enhance 3Dnow but took ISSE into theird > "XP" processors)?o  E No.  A main virtue of Hammer is that it doesn't carry all the largelymL useless baggage that IA64 does, hence is cheaper, runs cooler, and is easierI to enhance.  It would have to give much of that up to support IA64 code -SK and while it would be a real hoot (though not a huge surprise) if IA64 code,J on a multi-core, OoO RISCy platform ran faster than it would on its nativeI platform, it's unlikely that AMD would have much incentive to put that to-J the test given that the only code likely to run on IA64 in the foreseeableL future (and perhaps ever) that won't have an IA32 version that will run justH fine on Hammer (at least if Microsoft supports Hammer) will be high-end,G lower-volume applications where doing a recompile to get them to run on H Hammer won't be the problem that store-shelf shrink-wrap software faces.  D > I don't know if software vendors want to support their apps on two< > mainstream architectures and put two boxes on the shelves.  I That's a significant problem with IA64 all by itself:  since it won't rundJ IA32 code at anything like the speed people are accustomed to, if a vendorK wants to support both the IA32 and the IA64 platform they'll often need twodI boxes (or two versions in the same box) to do so, whereas since IA32 will L run at full native speed on Hammer there's no such problem on that platform.   - bill   ------------------------------  % Date: Thu, 15 Nov 2001 05:29:23 -0500p- From: JF Mezei <jfmezei.spamnot@videotron.ca>l" Subject: Re: IBM to bid for CPQ ??, Message-ID: <3BF39902.4D768C77@videotron.ca>   Pio Baettig wrote: > B > Could it be possible, that at some point, IF both architectures,? > Hammer & IA64 prove equally succesful (in sold quantity) thatr9 > Features would be cross licensed between Intel and AMD,a  R If both architectures are as succesful technically, then the cheaper one will win.  J The one caveat in this: If IA64 being a new chip (although laden with someK legacy support) has a longer projected lifetime whereas Hammer has an image A that it will hit its ceiling sooner, then perhaps IA64 might win.h   ------------------------------  % Date: Thu, 15 Nov 2001 05:32:59 -05000- From: JF Mezei <jfmezei.spamnot@videotron.ca>)" Subject: Re: IBM to bid for CPQ ??, Message-ID: <3BF399DA.83FA0BB3@videotron.ca>  I With both Compaq and HP having announced their unconditional surrender tonK Intel, will there be any serious "enterprise" vendors that will sell Hammer L based systems, or will Hammer be relegated to the home computing market (eg:8 high end 3d graphics with live rendering etc for games).  C Will IBM try to regain NT onto Power, or will it become a Microsoft 7 competitor, pushing Linux on Power against NT on IA64 ?o   ------------------------------  # Date: Thu, 15 Nov 2001 15:43:01 GMTt. From: "Duane Sand" <duane.sand@mindspring.com>" Subject: Re: IBM to bid for CPQ ??> Message-ID: <9oRI7.15689$RG1.8303836@news1.rdc1.sfba.home.com>   "Pio Baettig" wrotemB > Could it be possible, that at some point, IF both architectures,? > Hammer & IA64 prove equally succesful (in sold quantity) that 9 > features would be cross licensed between Intel and AMD,C8 > or that one firm would be forced to make its processor: > compatible to the other one (as AMD didn't enhance 3Dnow, > but took ISSE into their "XP" processors)?@ > I don't know if software vendors want to support their apps on@ > two mainstream architectures and put two boxes on the shelves.  ? The Itanium instruction set architecture is chock full of Intelm: patents and HP patents (handed over to Intel).  One of the< main goals Intel gave itself in designing this complex thing8 is that it be impossible for anyone to legally produce a; compatible chip without violating Intel-controlled patents.g= There is no way that Intel will ever license this instructionu? set to others.  They will never repeat the mistake of licensingt4 AMD, as they did for the 8086 back when IBM required8 a relatively smaller Intel to license a second source as factory insurance.  ; HP probably retains the rights to implement its own version 5 of IA64 chips, but I wonder if they have the right tou8 manufacture them in non-Intel fabs, or the right to have3 3rd companies like IBM cook them.  Probably not; HPi4 was very much wanting to get out of the fab-building9 business itself, and Intel would have been crazy to allowa9 the IA64 design to possibly slip away to IBM and Koreans.S   ------------------------------  # Date: Thu, 15 Nov 2001 15:55:31 GMT % From: Milton <mbhewitt@optonline.net>a" Subject: Re: IBM to bid for CPQ ??8 Message-ID: <gmo7vt4nttdlkfgnkievsa7fv2eq0moveh@4ax.com>  , On Thu, 15 Nov 2001 05:32:59 -0500, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:   J >With both Compaq and HP having announced their unconditional surrender toL >Intel, will there be any serious "enterprise" vendors that will sell HammerM >based systems, or will Hammer be relegated to the home computing market (eg:.9 >high end 3d graphics with live rendering etc for games).s >DD >Will IBM try to regain NT onto Power, or will it become a Microsoft8 >competitor, pushing Linux on Power against NT on IA64 ?  6 Why would IBM make the same mistake that Digital made,8 and have to support a binary incompatable version of NT?E Plus, I doubt that M$ would allow IBM accesss to the source code that < it would have to have, to make it work, in the first place.   . It will be Linux on Power4 against NT on IA64.  0 Peformance and reliability vs. you know what. :)   Cheers,  Milton        ------------------------------   Date: 15 Nov 2001 16:21:58 GMT) From: leslie@clio.rice.edu (Jerry Leslie)p" Subject: Re: IBM to bid for CPQ ??' Message-ID: <9t0q36$936$1@joe.rice.edu>y  & Milton (mbhewitt@optonline.net) wrote: :m8 : Why would IBM make the same mistake that Digital made,: : and have to support a binary incompatable version of NT?   You mean work again.               ------  G : Plus, I doubt that M$ would allow IBM accesss to the source code that > : it would have to have, to make it work, in the first place.  :aA The PowerPC and the MIPS chips were supported by NT at one time. D  E Their support was discontinued when the chip manufacturers got tired /B of paying Microsoft, who expects them to pay for their processor's support.  4 --Jerry Leslie     (my opinions are strictly my own)   ------------------------------    Date: 15 Nov 2001 09:47:07 -0800( From: bob@instantwhip.com (Bob Ceculski)" Subject: Re: IBM to bid for CPQ ??= Message-ID: <d7791aa1.0111150947.1af18950@posting.google.com>:  p "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote in message news:<kcyI7.1564$RL6.50540@news.cpqcorp.net>...L > Ignoring all the crap.  My opinion is that Power4 is the baddest thing outJ > there short of an EV7, and IBM is a gorilla with staying power and cash.M > What Power is missing is Windows64.  Power and Itanium will be the dominanteM > 64-bit architectures 5 years from now.  I'll be happy to be on either.  I'mf > planning on being on Itanium.r >  >  > ? > JF Mezei wrote in message <3BF1A6F3.8EAEC33D@videotron.ca>...o > >Alan Greig wrote:+ > >> Better get that PowerPC port underway., > >lL > >I was thinking about why would IBM really want Compaq for. But with AlphaJ > >dead, porting tru64 and VMS to power pc might put volumes just over the >  top,fJ > >allowing IBM to make Power PC a truly competitive chip and kill Intel's >  IA64. > >9K > >With HP and Compaq having stated that they wanted to be like IBM, if IBM. >  weregM > >to remove the stuff inside of Compaq that could make Compaq like IBM ( VMSn >  andM > >Tandem and Alpha), then IBM would ensure it remans well ahead of the pack.  > >>L > >By buying Compaq, IBM could also gain control and access to all of the IP >  ofi3 > >Digital, some of which might be of value to IBM.  > >tH > >The big question is whether that would be worth the purchase price of	 > Compaq.t  ; unless if IBM buys compaq and vms, then you'll be on power!H   ------------------------------   Date: 15 Nov 2001 17:35:26 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon) " Subject: Re: IBM to bid for CPQ ??, Message-ID: <9t0ucu$1i9t$1@info.cs.uofs.edu>  + In article <3BF1678E.93D12DD9@prodigy.net>,0#  cjt <cheljuba@prodigy.net> writes:rO |> I assume the CPQ folks would then again change gears and we'd start to hear P, |> about how great NonStop Power-4s will be? |> n |> John McLean wrote:. |> >  3 |> > Not being one to let a good rumour go past ...f |> > c |> > IBM to bid for CPQ? |> > hJ |> > Article at http://www.theinquirer.net/13110115.htm  suggests that IBMH |> > might mount a hostile bid for Compaq "and cut off HP at the gulch". |> > e% |> > Now that would be interesting ..   > Wasn't IBM one of the companies licensed to make Alpha chips??   bill   -- oJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   g   ------------------------------   Date: 15 Nov 2001 17:37:07 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon) " Subject: Re: IBM to bid for CPQ ??, Message-ID: <9t0ug3$1i9t$2@info.cs.uofs.edu>  ) In article <3BF1818F.D683C6AB@rdrop.com>,s(  Dean Woodward <deanw@rdrop.com> writes:
 |> cjt wrote:u |> > tP |> > I assume the CPQ folks would then again change gears and we'd start to hear. |> > about how great NonStop Power-4s will be? |> oI |> At least IBM has, and appreciates, it's own R&D division.  Too bad the J |> EV8 team went to Intel.  (Where do I go to start a rumor that IBM wants |> to hire 'em away? ;-) |>  * |> Ah, well.  I'm just wishin' out loud...  E Going back to one of my previous topics, if Compaq has a "no re-hire" B contract with Intel, would that preclude IBM from hiring the Aplha engineers away from Intel??    bill     --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   o   ------------------------------  % Date: Thu, 15 Nov 2001 08:58:45 -0500o- From: "John Eisenschmidt" <jeisensc@aaas.org> , Subject: Is it a DEC C problem, or is it me?+ Message-ID: <sbf383d9.063@AAASMTA.aaas.org>t   I'm going to take a flyer -=20  F We have a 3rd party application we use here that has a TCP/IP socket =L listener written in C. It's currently running on our 7.1 system, but we're =L having problems trying to migrate it to 7.2-1. The vendor has supplied new =I header files since 7.1 uses UCX and 7.2-1 uses TCPIP. Anyway we compile =mG the code, link it, all goes well, but when the program gets called it =-	 explodes.a  I The vendor can't seem to find their problem (Solaris is their strength, =uJ not VMS), and I wanted to make absolutely sure it wasn't something on my = end.   $ TCPIP SHOW VERSION  A   DIGITAL TCP/IP Services for OpenVMS Alpha Version V5.0A - ECO 3e=   on a COMPAQ AlphaServer DS10 617 MHz running OpenVMS V7.2-1u  
 $ CC /VERSION & DEC C V5.7-004 on OpenVMS Alpha V7.2-1  : Are there any issues with this version of DEC C and 7.2-1?  ? The program is compiled and linked with the following switches:a7 $ CC /STANDARD=3DVAXC/NOMEMBER_ALIGN/ASSUME=3DNOALIGNEDs  ' When called, the program abends thusly:o  L %SYSTEM-F-ACCVIO, access violation, reason mask=3D00, virtual address=3D000=3 0000000000000, PC=3DFFFFFFFF8095C8E8, PS=3D0000001Be/ %TRACE-F-TRACEBACK, symbolic stack dump followstJ   image    module    routine             line      rel PC           abs PCL                                             0 0000000000000000 FFFFFFFF8095= C8E8L  ADMC_GUI                                   0 00000000000228E8 000000000003= 28E8L  ADMC_GUI                                   0 0000000000021E24 000000000003= 1E24L  ADMC_GUI                                   0 0000000000021868 000000000003= 1868L  ADMC_GUI                                   0 00000000000203D4 000000000003= 03D4L  ADMC_GUI                                   0 0000000000020070 000000000003= 0070L  PTHREAD$RTL                                0 00000000000312FC 000000007BBA= B2FCL  PTHREAD$RTL                                0 0000000000012B48 000000007BB8= CB48L                                             0 FFFFFFFF854ED3F4 FFFFFFFF854E= D3F4   Thanks,1 John   ------------------------------  % Date: Thu, 15 Nov 2001 17:29:28 +0300h4 From: "Ruslan R. Laishev" <laishev@smtp.deltatel.ru>0 Subject: Re: Is it a DEC C problem, or is it me?0 Message-ID: <3BF3D148.8D8769A4@smtp.deltatel.ru>   Hi John,   	a piece of code for example ?   John Eisenschmidt wrote: >  > I'm going to take a flyer -t > t> We have a 3rd party application we use here that has a TCP/IP socket listener written in C. It's currently running on our 7.1 system, but we're having problems trying to migrate it to 7.2-1. The vendor has supplied new header files since 7.1 uses UCX and 7.2-1 uses TCPIP. Anyway we compile the code, link it, all goes well, but when the program gets called it explodes. >  > The vendor can't seem to find their problem (Solaris is their strength, not VMS), and I wanted to make absolutely sure it wasn't something on my end.  >  > $ TCPIP SHOW VERSION > C >   DIGITAL TCP/IP Services for OpenVMS Alpha Version V5.0A - ECO 3o? >   on a COMPAQ AlphaServer DS10 617 MHz running OpenVMS V7.2-1  >  > $ CC /VERSIONd( > DEC C V5.7-004 on OpenVMS Alpha V7.2-1 > < > Are there any issues with this version of DEC C and 7.2-1? > A > The program is compiled and linked with the following switches:*5 > $ CC /STANDARD=VAXC/NOMEMBER_ALIGN/ASSUME=NOALIGNED  > ) > When called, the program abends thusly:e > x > %SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=0000000000000000, PC=FFFFFFFF8095C8E8, PS=0000001B1 > %TRACE-F-TRACEBACK, symbolic stack dump followseL >   image    module    routine             line      rel PC           abs PCQ >                                             0 0000000000000000 FFFFFFFF8095C8E8rQ >  ADMC_GUI                                   0 00000000000228E8 00000000000328E8hQ >  ADMC_GUI                                   0 0000000000021E24 0000000000031E246Q >  ADMC_GUI                                   0 0000000000021868 0000000000031868eQ >  ADMC_GUI                                   0 00000000000203D4 00000000000303D4tQ >  ADMC_GUI                                   0 0000000000020070 0000000000030070(Q >  PTHREAD$RTL                                0 00000000000312FC 000000007BBAB2FCtQ >  PTHREAD$RTL                                0 0000000000012B48 000000007BB8CB48sQ >                                             0 FFFFFFFF854ED3F4 FFFFFFFF854ED3F4n > 	 > Thanks,  > John   -- u Cheers, Ruslan.sD +---------------------pure personal opinion------------------------+;       RADIUS Server for OpenVMS project - www.radiusvms.come8         vms-isps@dls.net - Forum for ISP running OpenVMS*                  Mobile: +7 (901) 971-3222A    TKD (WTF) in Russia, St.-Petersburg - www.TaeKwonDo-WTF.SPb.RUr   ------------------------------  % Date: Thu, 15 Nov 2001 15:49:03 +0100 $ From: "Jakob Erber" <erberj@post.ch>0 Subject: Re: Is it a DEC C problem, or is it me? Message-ID: <3bf3d5e0$1@hcwe67>   , Any reason, why you are using VAXC standard?  > "John Eisenschmidt" <jeisensc@aaas.org> schrieb im Newsbeitrag% news:sbf383d9.063@AAASMTA.aaas.org...  I'm going to take a flyer -   D We have a 3rd party application we use here that has a TCP/IP socketJ listener written in C. It's currently running on our 7.1 system, but we'reJ having problems trying to migrate it to 7.2-1. The vendor has supplied newK header files since 7.1 uses UCX and 7.2-1 uses TCPIP. Anyway we compile thesK code, link it, all goes well, but when the program gets called it explodes.b  K The vendor can't seem to find their problem (Solaris is their strength, not I VMS), and I wanted to make absolutely sure it wasn't something on my end.&   $ TCPIP SHOW VERSION  A   DIGITAL TCP/IP Services for OpenVMS Alpha Version V5.0A - ECO 3t=   on a COMPAQ AlphaServer DS10 617 MHz running OpenVMS V7.2-1e  
 $ CC /VERSIONm& DEC C V5.7-004 on OpenVMS Alpha V7.2-1  : Are there any issues with this version of DEC C and 7.2-1?  ? The program is compiled and linked with the following switches: 3 $ CC /STANDARD=VAXC/NOMEMBER_ALIGN/ASSUME=NOALIGNEDe  ' When called, the program abends thusly:e  ; %SYSTEM-F-ACCVIO, access violation, reason mask=00, virtualS: address=0000000000000000, PC=FFFFFFFF8095C8E8, PS=0000001B/ %TRACE-F-TRACEBACK, symbolic stack dump followsaJ   image    module    routine             line      rel PC           abs PC>                                             0 0000000000000000 FFFFFFFF8095C8E8>  ADMC_GUI                                   0 00000000000228E8 00000000000328E8>  ADMC_GUI                                   0 0000000000021E24 0000000000031E24>  ADMC_GUI                                   0 0000000000021868 0000000000031868>  ADMC_GUI                                   0 00000000000203D4 00000000000303D4>  ADMC_GUI                                   0 0000000000020070 0000000000030070>  PTHREAD$RTL                                0 00000000000312FC 000000007BBAB2FC>  PTHREAD$RTL                                0 0000000000012B48 000000007BB8CB48>                                             0 FFFFFFFF854ED3F4 FFFFFFFF854ED3F4   Thanks,I John   ------------------------------  % Date: Thu, 15 Nov 2001 10:10:19 -0500d- From: JF Mezei <jfmezei.spamnot@videotron.ca>i0 Subject: Re: Is it a DEC C problem, or is it me?+ Message-ID: <3BF3DAD7.D09C77F@videotron.ca>-   Jakob Erber wrote: > . > Any reason, why you are using VAXC standard?  ( Yeah, I was going to ask the same thing.  M Also, I noticed that in the traceback stack, there were the PTHREAD routines.tB Have you installed the various pthread patches applicable to 7.2 ?   ------------------------------  % Date: Thu, 15 Nov 2001 11:01:47 -0500 - From: "John Eisenschmidt" <jeisensc@aaas.org>a0 Subject: Re: Is it a DEC C problem, or is it me?+ Message-ID: <sbf3a0a1.079@AAASMTA.aaas.org>   K That's them - to be honest I couldn't tell you. If we ever get it working =eH on 7.2-1 I was going to try recompiling without that to see if it works.  < >>> "Jakob Erber" <erberj@post.ch> 11/15/2001 9:49:03 AM >>>, Any reason, why you are using VAXC standard?  > "John Eisenschmidt" <jeisensc@aaas.org> schrieb im Newsbeitrag% news:sbf383d9.063@AAASMTA.aaas.org...i I'm going to take a flyer -F  D We have a 3rd party application we use here that has a TCP/IP socketJ listener written in C. It's currently running on our 7.1 system, but we'reJ having problems trying to migrate it to 7.2-1. The vendor has supplied newI header files since 7.1 uses UCX and 7.2-1 uses TCPIP. Anyway we compile =e theeC code, link it, all goes well, but when the program gets called it =c	 explodes.s  I The vendor can't seem to find their problem (Solaris is their strength, =- not-I VMS), and I wanted to make absolutely sure it wasn't something on my end.u   $ TCPIP SHOW VERSION  A   DIGITAL TCP/IP Services for OpenVMS Alpha Version V5.0A - ECO 30=   on a COMPAQ AlphaServer DS10 617 MHz running OpenVMS V7.2-1   
 $ CC /VERSIONh& DEC C V5.7-004 on OpenVMS Alpha V7.2-1  : Are there any issues with this version of DEC C and 7.2-1?  ? The program is compiled and linked with the following switches: 7 $ CC /STANDARD=3DVAXC/NOMEMBER_ALIGN/ASSUME=3DNOALIGNED   ' When called, the program abends thusly:e  = %SYSTEM-F-ACCVIO, access violation, reason mask=3D00, virtualo@ address=3D0000000000000000, PC=3DFFFFFFFF8095C8E8, PS=3D0000001B/ %TRACE-F-TRACEBACK, symbolic stack dump follows J   image    module    routine             line      rel PC           abs PC>                                             0 0000000000000000 FFFFFFFF8095C8E8>  ADMC_GUI                                   0 00000000000228E8 00000000000328E8>  ADMC_GUI                                   0 0000000000021E24 0000000000031E24>  ADMC_GUI                                   0 0000000000021868 0000000000031868>  ADMC_GUI                                   0 00000000000203D4 00000000000303D4>  ADMC_GUI                                   0 0000000000020070 0000000000030070>  PTHREAD$RTL                                0 00000000000312FC 000000007BBAB2FC>  PTHREAD$RTL                                0 0000000000012B48 000000007BB8CB48>                                             0 FFFFFFFF854ED3F4 FFFFFFFF854ED3F4   Thanks,c John   ------------------------------  % Date: Thu, 15 Nov 2001 11:29:36 -0500e( From: David Froble <davef@tsoft-inc.com>0 Subject: Re: Is it a DEC C problem, or is it me?, Message-ID: <3BF3ED70.8000506@tsoft-inc.com>   John Eisenschmidt wrote:    > I'm going to take a flyer -  >G  > We have a 3rd party application we use here that has a TCP/IP socket@D  >  listener written in C. It's currently running on our 7.1 system,F  > but we're having problems trying to migrate it to 7.2-1. The vendorC  >  has supplied new header files since 7.1 uses UCX and 7.2-1 useseF  > TCPIP. Anyway we compile the code, link it, all goes well, but when(  >  the program gets called it explodes.  >@  > The vendor can't seem to find their problem (Solaris is theirE  > strength, not VMS), and I wanted to make absolutely sure it wasn't   > something on my end.x  >  > $ TCPIP SHOW VERSION?  >G  > DIGITAL TCP/IP Services for OpenVMS Alpha Version V5.0A - ECO 3 on ar:  >  COMPAQ AlphaServer DS10 617 MHz running OpenVMS V7.2-1  >7  > $ CC /VERSION DEC C V5.7-004 on OpenVMS Alpha V7.2-1h  >=  > Are there any issues with this version of DEC C and 7.2-1?n  >G  > The program is compiled and linked with the following switches: $ CCh2  >  /STANDARD=VAXC/NOMEMBER_ALIGN/ASSUME=NOALIGNED  >*  > When called, the program abends thusly:  >>  > %SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual>  > address=0000000000000000, PC=FFFFFFFF8095C8E8, PS=0000001B  %TRACE-F-TRACEBACK,n:  >  symbolic stack dump follows image    module    routine8  >  line      rel PC           abs PC 0 0000000000000000@  > FFFFFFFF8095C8E8 ADMC_GUI                                   0-  > 00000000000228E8 00000000000328E8 ADMC_GUIo9  >           0 0000000000021E24 0000000000031E24 ADMC_GUI E  >                       0 0000000000021868 0000000000031868 ADMC_GUIp7  >                                   0 00000000000203D4u@  > 00000000000303D4 ADMC_GUI                                   00  > 0000000000020070 0000000000030070 PTHREAD$RTL<  >           0 00000000000312FC 000000007BBAB2FC PTHREAD$RTL>  >                       0 0000000000012B48 000000007BB8CB48 0$  > FFFFFFFF854ED3F4 FFFFFFFF854ED3F4  >  > Thanks, John:  >  >  C Not directly related, since I used the system service interface foreH socket communications.  The one time I had such I had mis-defined one ofD the structures used to pass data to a routine.  Don't remember whichJ one, possibly one that defines the address and port and such for a socket.  C The new header files is where I'd start looking.  Specifically the iH structures used for specifying socket information.  From looking at the F TCP/IP V5.? documentation and include files, at least for Basic, they D were pretty much the same, and included variables with both the UCX I prefix and the TCPIP prefix, to allow older code to run without changing   a bunch of constant names.   Dave   -- I4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com6 T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486   ------------------------------  # Date: Thu, 15 Nov 2001 11:42:39 GMTt% From: Milton <mbhewitt@optonline.net> - Subject: Itanium flunking Compaq server testsd8 Message-ID: <p7a7vtof48grehva4rk4n561qhna8jn4qt@4ax.com>  W http://dailynews.yahoo.com/h/cn/20011114/tc/itanium_flunking_compaq_server_tests_1.htmlc  E Intel's Itanium processor is failing to pass Compaq Computer's stresso@ tests, according to a Compaq representative, thus holding up the$ release of Compaq's Itanium servers.     = A Compaq representative said that the company has experienced B "sightings" with Itanium, Intel's 64-bit processor for servers, inF Compaq's internal testing of its ProLiant DL590/64. The representativeE would not go so far as to call the issue a flaw, but said the problemsA appeared to be caused by the processor. The problem crops up withh? servers running both the 733MHz and 800MHz version of the chip.n  7 >[1]See footnote for Intel's definition of a *sighting*g  E Sightings are a broad computer industry term for equipment that fails>C to perform as expected in testing, but with no immediately apparent  defects.  B Compaq has tested servers from other small manufacturers and foundD they too did not pass Compaq's rigorous tests--a normal prerequisite for commercial release.>  B "In our opinion, it is an Intel Itanium issue," the representativeE said. Compaq and Intel are working to determine the root cause of the  problem.  C An Intel representative confirmed the sightings but added that most F such problems are not chip-related. Most sightings also do not rise toB the level of a flaw. He also noted that Itanium servers from other manufacturers continue to ship.-  > The problem is the latest stumble for Itanium, one of the mostE ambitious product launches in Intel's history. The chip, co-developed D by Intel and Hewlett-Packard, is designed to power servers that willA compete against machines from Sun Microsystems. Itanium, however, C contains an entirely new architecture and instruction set that took A years to complete and debug. Few chip companies have been able to A successfully switch from one chip design to another easily. Most, = including Intel, typically enhance the existing core product.r  > Originally due in the mid- to late 90s, the chip only came outA commercially in May 2001. Partly because of the delays and partlyoF because of performance, most server manufacturers have barely marketedF servers containing the first version of the chip and are concentratingD on the next version, code-named McKinley, which comes out next year.  A Some computer executives have become increasingly critical of the F chip. Because the chip is based on an entirely new design, very littleF software exists today that runs on the server--a situation exacerbated by the harsh economic climate.  @ Customer interest for Itanium servers is "effectively zero," JoeD Marengi, general manager of Dell Computer's enterprise business unitD in North America, said in an interview Tuesday at the Comdex (news -@ web sites) Fall 2001 trade show. "The investment involved in the transition in huge.r  F "I don't see the speed and benefit to what the processor brings to the equation," he added.  E Intel is also facing a potential competitive threat from AMD's Hammers; processor for servers, coming out late next year. Hammer islC effectively based on Intel's familiar X86 architecture but can alson run 64-bit applications.  D While no major U.S. manufacturer has ever used an AMD processor in aB server, some analysts have said that the potential performance, asF well as comparative familiarity of the architecture, could make Hammer interesting to them. e  < > I found this part of the article particularly interesting:  A "Customer interest for Itanium servers is "effectively zero," JoerD Marengi, general manager of Dell Computer's enterprise business unitD in North America, said in an interview Tuesday at the Comdex (news -@ web sites) Fall 2001 trade show. "The investment involved in the transition in huge."  7 >Not very reassuring to Compaq's enterprise customers. 0  R >Especially those that are to be forced marched onto this new processor architect.   [1] F What is a sighting? "A sighting is any issue affecting the performanceF or reliability of an enterprise server system that is witnessed at the= system, software and/or platform levels. Sightings are common0F occurrences and most are root-caused to something other than a siliconD issue. If an Intel processor or product's behaviour deviates outside= of normal specifications, a sighting may become an erratum. "M' http://www.theinquirer.net/14110105.htm    Cheers,2 Milton   ------------------------------  % Date: Thu, 15 Nov 2001 08:46:57 -05002- From: JF Mezei <jfmezei.spamnot@videotron.ca> 1 Subject: Re: Itanium flunking Compaq server testsB, Message-ID: <3BF3C750.CA1481D5@videotron.ca>  
 Milton wrote:0 > Y > http://dailynews.yahoo.com/h/cn/20011114/tc/itanium_flunking_compaq_server_tests_1.htmlF > G > Intel's Itanium processor is failing to pass Compaq Computer's stress B > tests, according to a Compaq representative, thus holding up the& > release of Compaq's Itanium servers.  K Does that surprise anyone ? Besides, Itanium won't power any serious Compaq @ systems until 2004-2005 so current bugs/problems are irrelevant.   ------------------------------  # Date: Thu, 15 Nov 2001 14:12:26 GMTl% From: Milton <mbhewitt@optonline.net>o1 Subject: Re: Itanium flunking Compaq server testso8 Message-ID: <rbi7vt43lorbmshcsdeafngjm6ucqgvknt@4ax.com>  , On Thu, 15 Nov 2001 08:46:57 -0500, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:u   >Milton wrote: >> mZ >> http://dailynews.yahoo.com/h/cn/20011114/tc/itanium_flunking_compaq_server_tests_1.html >> eH >> Intel's Itanium processor is failing to pass Compaq Computer's stressC >> tests, according to a Compaq representative, thus holding up the ' >> release of Compaq's Itanium servers.I >  >Does that surprise anyone ? h  
 Not me. :)  0 >Besides, Itanium won't power any serious CompaqA >systems until 2004-2005 so current bugs/problems are irrelevant.   E That's if Compag is still around by then, in it's current incarnation - or it's new improved proposed merger version.   F It just seems like an inauspicious start for a processor that has beenD in developement for 11 years. From '89 at HP or if you wish from '94D HP/Intel co-develpment. That's still 7 years developement. One wouldC think, that would have been plenty of time to have gotten the minorE bugs squashed.  D Remember that this is the processor that both HP and Compaq have bet their future on.  " Hope they made the right decision.   Cheers,  Milton   ------------------------------  % Date: Thu, 15 Nov 2001 10:08:38 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca>M1 Subject: Re: Itanium flunking Compaq server tests0, Message-ID: <3BF3DA72.908139CE@videotron.ca>  
 Milton wrote: F > HP/Intel co-develpment. That's still 7 years developement. One wouldE > think, that would have been plenty of time to have gotten the minor  > bugs squashed.    N When you try to be all to everyone as the IA64 has tried, you get stuck with a* bloated chip that is too complex to debug.   long live alpha.  F > Remember that this is the processor that both HP and Compaq have bet > their future on.  - Yep. That is why I won't bet on Compaq or HP.F   ------------------------------  # Date: Thu, 15 Nov 2001 15:45:07 GMT-% From: Milton <mbhewitt@optonline.net>-1 Subject: Re: Itanium flunking Compaq server testsr8 Message-ID: <64o7vt0v8vj5m387ddbk0oj3tecmu7nl9f@4ax.com>  , On Thu, 15 Nov 2001 10:08:38 -0500, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:   O >When you try to be all to everyone as the IA64 has tried, you get stuck with a-+ >bloated chip that is too complex to debug.r  E If you think it's bloated now, what do you suppose it will look like,r% when "the essence of Alpha" is added?e  ( http://www.tru64.org/skc/IPFSpecial.html   Cheers,  Milton   ------------------------------  # Date: Thu, 15 Nov 2001 16:01:40 GMT@% From: Milton <mbhewitt@optonline.net> 1 Subject: Re: Itanium flunking Compaq server testse8 Message-ID: <c9p7vt0cpm5ioot14f1q1dsnirijk6e5t1@4ax.com>  , On Thu, 15 Nov 2001 08:46:57 -0500, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:t   >Milton wrote: >>  Z >> http://dailynews.yahoo.com/h/cn/20011114/tc/itanium_flunking_compaq_server_tests_1.html >>  H >> Intel's Itanium processor is failing to pass Compaq Computer's stressC >> tests, according to a Compaq representative, thus holding up theh' >> release of Compaq's Itanium servers.t >  >Does that surprise anyone ? e  E But shouldn't these flaws have been exposed thru the rigorous testingh# already done over the past 2 years?A  E "Compaqs ISSG engineers began putting prototype Blazer boxes through ? the Meatgrinder test suite nearly two years ago and testing hash< continued unabated during an extensive IPF shakedown cruise.( http://www.tru64.org/skc/IPFSpecial.html   ------------------------------  # Date: Thu, 15 Nov 2001 17:55:52 GMTr3 From: Tim Llewellyn <tim.llewellyn@cableinet.co.uk>i1 Subject: Re: Itanium flunking Compaq server tests0/ Message-ID: <3BF400B5.41B92CF8@cableinet.co.uk>A   JF Mezei wrote:a >  > Milton wrote:m > > [ > > http://dailynews.yahoo.com/h/cn/20011114/tc/itanium_flunking_compaq_server_tests_1.html  > > I > > Intel's Itanium processor is failing to pass Compaq Computer's stress D > > tests, according to a Compaq representative, thus holding up the( > > release of Compaq's Itanium servers. > M > Does that surprise anyone ? Besides, Itanium won't power any serious Compaq8B > systems until 2004-2005 so current bugs/problems are irrelevant.  G the surprising thing (:-)) to me is that Compaq could come out smelling  better than the28 vendors who are marketting before they test adequately.    --   Tim.Llewellyn@cableinet.co.uk  A  C Standard disclaimer applies. My views in no way represent those of 0! my employers or service provider.    ------------------------------  # Date: Thu, 15 Nov 2001 18:01:36 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)11 Subject: Re: Itanium flunking Compaq server testsi0 Message-ID: <00A0515E.1F33C4C3@SendSpamHere.ORG>  ` In article <64o7vt0v8vj5m387ddbk0oj3tecmu7nl9f@4ax.com>, Milton <mbhewitt@optonline.net> writes:- >On Thu, 15 Nov 2001 10:08:38 -0500, JF Mezei & ><jfmezei.spamnot@videotron.ca> wrote: >iP >>When you try to be all to everyone as the IA64 has tried, you get stuck with a, >>bloated chip that is too complex to debug. >2F >If you think it's bloated now, what do you suppose it will look like,& >when "the essence of Alpha" is added? >@) >http://www.tru64.org/skc/IPFSpecial.htmll >  >Cheers, >Miltong  H There was great wisdom in the Alpha design.  Too bad the intel-lect that+ now has the Alpha brain trust can't see it!n --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COME            @J   "And of course, I'm a genius, so people are naturally drawn to my fiery I   intellect.  Their admiration overwhelms their envy!" -- Calvin & Hobbess   ------------------------------    Date: 15 Nov 2001 08:21:42 -0800 From: ch257r@yahoo.com (Chris) Subject: Re: Items for sale!< Message-ID: <f42b2d56.0111150821.480dd2f@posting.google.com>  d mjenkins@jcn.net (MikeWJ) wrote in message news:<42d0e8ed.0111130649.7539aba5@posting.google.com>...E > I have these items to sell.  If interested, please call me, Mike ateH > 314-552-8666 or Dave at 314-552-8506.  Serious inquiries only, please. >  > Looking to sell: > ) >  24 VT terminals (VT220, VT320, VT420),i" >  31 VT keyboards (LK201, LK401), >   3 Remote Console Switches, >   3 Decserver 200/MC,d >   1 Decserver 250, >   9 LA75 printers, >   1 LA424 Printer, >   1 LP29 printer,  >   1 LG02 printer,A >   1 LG31 printer,7 >   1 LPS20 printer, >   1 LPS32 printer, >   1 LN06 (Declaser 2200),  >   2 VXT2000 Desktop units, >   1 VXT1000 Desktop unit,E$ >   1 StorageWorks 800 media cabinetE >     (HSD50 controller, 4 BA350 shelves, (28) 2-GB HD's, (2) TL7L's E! >      & (1) TLZ9 4mm tape units)h% >   2 VAX 4505a's (each has 2 RF72's) ' >   1 Microvax 3300 w/HD extension box.s) >      Other miscellaneous VAX equipment.0 > H > May have another 3300 in the parts bin.  If so, will add to this list. > 	 > Thanks.  >  > MikeWJ         Mike,e  E When you are posting stuff for sale, it is infinately more helpful if A you also post where you are. Some poeple might be concerned about A shipping/freight expenses. I am interested in the MicroVax(s) and8 related vaxen stuff.     Chris.   ------------------------------  % Date: Thu, 15 Nov 2001 18:23:41 +1100 / From: "Phil Howell" <phowell@snowyhydro.com.au>3J Subject: Re: Linking shareable library  using another shareable library !?1 Message-ID: <75KI7.1288$2Z5.61207@ozemail.com.au>   5 "Lars Funck Jensen" <reslfj@ofir.dk> wrote in message 7 news:49807d45.0111140747.5ac097db@posting.google.com...BG > I have the following macro32 program that i try to link as a sharable0G > image. The program is a "jacket" procedure that put parameters passedEH > in registers (ap and r2) onto the stack and "calls" a c program. The cH > program calls MQSeries to put and get messages. The c program needs to > access the mqm.exe library.a >CJ > The problem I have is linking the module and getting the dynamic linking toL > work. As my problem can be reproduced without calling the c program I have > commented the calls out. > B > If I link it with the option file, install it and try to call it  > my calling program just hangs. >hE > If i remove the line "SYS$SHARE:MQM/SHAREABLE" from the option filei< > it returns with the expected return code (3,4,... or 101). >  > Any help ??????h >o5 (Disclaimer - I am not sure what you are trying to doe, so if I am talking rubbish please excuse me) > /lars funck jensen >l >(8 > The security of cma_msqe.exe and sys$share:mqm.exe are1 > CMA_MQSE.EXE;7       [SYSTEM] (RWED,RWED,RE,RE)e1 > MQM.EXE;1            MQM      (RWED,RWED,RE,RE)T >I >e > $ macro/mig rnatmqse, so this produces an object file rnatmqse.obj9 should you either include the object in your options fileR$ or insert it into an object library?( > $ LINK/SHARE/MAP/FULL CMA_MQSE.OPT/OPT; link/map/full is your friend - what does the link map show?lC > $ install replace/open/header/share SAG$ROOT:[CMA.MQ]CMA_MQSE.EXEN= > $ define/system/exec rnatmqse SAG$ROOT:[CMA.MQ]CMA_MQSE.EXE  > ...t > $ type cma_mqse.optF > GSMATCH=LEQUAL,1,1# > SAG$ROOT:<CMA.MQ>CMA_MQSE/LIBRARYe# you usually use /library for a .olb8' use /share to link to a shareable image>$ > symbol_vector=(RNATMQSE=procedure) > !l > SYS$SHARE:MQM/SHAREABLE  > $C6 where is the object code of rnatmqse? in cma_mqse.olb? > .... > $ type rnatmqse.mar < >                 .psect RNATMQSE_code,exe,nowrt,rel,pic,shr >>. >                 .call_entry max_args=31,   -. >                             home_args=TRUE,-. >                             input=<R2>,    -J >                             preserve=<R2,R3,R4,R5,R6,R7,R8,R9,R10,R11>,-, >                             label=RNATMQSE > ;7J >         movzbl  (ap),r0                ; place no. of NATURAL parameters in R0.= >         cmpl    r0,#2                  ; must be at least 3BK >         bgtr    10$                    ; if no. parameters > 2  - process  them.iL >         brw     err                    ; if less than 2 parameters - done. > 10$:3 >         pushl   r2                     ; *parmdech@ >         addl3   #4,ap,r1               ; skip # param longword4 >         pushl   r1                     ; **parmptrB >         pushl   r0                     ; # of NATURAL parameters > ;. >pA > ;       calls   #3,rcxxmqse            ; call HGP's C   program.I > ;                                      ; Return code in r0 is passes tos NATURALi > ;eD >         ret                            ; Return to NATURAL program > ;rI > err:    movl    #101,r0                ; Macro interface errcode:"Not 3  parm's"c
 >         ret  > ;e > ;- >                 .end   ------------------------------    Date: 15 Nov 2001 06:00:36 -0800( From: reslfj@ofir.dk (Lars Funck Jensen)J Subject: Re: Linking shareable library  using another shareable library !?= Message-ID: <49807d45.0111150600.2cdc3720@posting.google.com>h  g "Phil Howell" <phowell@snowyhydro.com.au> wrote in message news:<75KI7.1288$2Z5.61207@ozemail.com.au>..e7 > (Disclaimer - I am not sure what you are trying to do@. > so if I am talking rubbish please excuse me)  0 Any help is much needed!  Thank you for posting.  @ I have a language (Natural v2.17 from Software AG) that can link"  to shared libraries dynamically. 5 In Natural I write:   CALL 'RNATMQSE' PARM1 PARM2 ...'  : I want to access MQSeries API available in the C language.4 The jacket procedure in macro32 is there to push the5 parameters passed in R2, AP+4 AND (AP) onto the stackr so the  C procedure:  4 long RCXXMQSE(word nparm, byte **argp, finfo *argdp)
 can be calledi  8 This works for other C procedures calling VMS system or < even IP sockets using the OpenVMS linkers default libraries.  = The C procedure rcxxmqse calls the MQ API that is implementede in sys$share:mqm.exe.   5 In my first posting I have made a comment of the liner? ;       calls   #3,rcxxmqse            ; call HGP's C   programn7 to get the linker statements to work, without actually h+ extracting/linking  any C or MQ API modulese  . The MQSerise manual has the following example:# $ cc/include=mqs_include testmqse.c # $ cc/include=mqs_include rcxxmqse.co2 $ link testmqse.obj,rcxxmqse.obj,sys$input/options sys$share:mqm/shareablea ctrl/z $t $ run testmqse.exe  5 (Where I have substituted the name of my testprogram r and my c procedure).( This works OK as a standalone program.    h "Phil Howell" <phowell@snowyhydro.com.au> wrote in message news:<75KI7.1288$2Z5.61207@ozemail.com.au>...7 > "Lars Funck Jensen" <reslfj@ofir.dk> wrote in messages9 > news:49807d45.0111140747.5ac097db@posting.google.com... I > > I have the following macro32 program that i try to link as a sharableHI > > image. The program is a "jacket" procedure that put parameters passedhJ > > in registers (ap and r2) onto the stack and "calls" a c program. The cJ > > program calls MQSeries to put and get messages. The c program needs to > > access the mqm.exe library.  > >sL > > The problem I have is linking the module and getting the dynamic linking >  tolN > > work. As my problem can be reproduced without calling the c program I have > > commented the calls out. > >lD > > If I link it with the option file, install it and try to call it" > > my calling program just hangs. > >eG > > If i remove the line "SYS$SHARE:MQM/SHAREABLE" from the option filee> > > it returns with the expected return code (3,4,... or 101). > >  > > Any help ??????e > >u7 > (Disclaimer - I am not sure what you are trying to do . > so if I am talking rubbish please excuse me) > > /lars funck jensen > >a > >i: > > The security of cma_msqe.exe and sys$share:mqm.exe are3 > > CMA_MQSE.EXE;7       [SYSTEM] (RWED,RWED,RE,RE)o3 > > MQM.EXE;1            MQM      (RWED,RWED,RE,RE), > >r > >n > > $ macro/mig rnatmqse. > so this produces an object file rnatmqse.obj; > should you either include the object in your options file & > or insert it into an object library?  + I forgot to include ( in my first posting) t6 $ LIBRARY/OBJECT/REPLACE/LOG CMA_MQSE.OLB rnatmqse.OBJ  * > > $ LINK/SHARE/MAP/FULL CMA_MQSE.OPT/OPT> >  link/map/full is your friend - what does the link map show?E > > $ install replace/open/header/share SAG$ROOT:[CMA.MQ]CMA_MQSE.EXEI? > > $ define/system/exec rnatmqse SAG$ROOT:[CMA.MQ]CMA_MQSE.EXEa > > ...  > > $ type cma_mqse.opta > > GSMATCH=LEQUAL,1,1% > > SAG$ROOT:<CMA.MQ>CMA_MQSE/LIBRARYt% > you usually use /library for a .olbn) > use /share to link to a shareable imaget  > My modules (rnatmqse.obj and rcxxmqse.obj) are in cma_mqse.olb  & > > symbol_vector=(RNATMQSE=procedure) > > !e > > SYS$SHARE:MQM/SHAREABLEf > > $i9 >  where is the object code of rnatmqse? in cma_mqse.olb?n > > .... > > $ type rnatmqse.marb> > >                 .psect RNATMQSE_code,exe,nowrt,rel,pic,shr > >o0 > >                 .call_entry max_args=31,   -0 > >                             home_args=TRUE,-0 > >                             input=<R2>,    -L > >                             preserve=<R2,R3,R4,R5,R6,R7,R8,R9,R10,R11>,-. > >                             label=RNATMQSE > > ;tL > >         movzbl  (ap),r0                ; place no. of NATURAL parameters	 >  in R0. ? > >         cmpl    r0,#2                  ; must be at least 3.M > >         bgtr    10$                    ; if no. parameters > 2  - process  >  them.N > >         brw     err                    ; if less than 2 parameters - done. > > 10$:5 > >         pushl   r2                     ; *parmdec/B > >         addl3   #4,ap,r1               ; skip # param longword6 > >         pushl   r1                     ; **parmptrD > >         pushl   r0                     ; # of NATURAL parameters > > ;f > >nC > > ;       calls   #3,rcxxmqse            ; call HGP's C   program K > > ;                                      ; Return code in r0 is passes tol
 >  NATURAL > > ;rF > >         ret                            ; Return to NATURAL program > > ;tK > > err:    movl    #101,r0                ; Macro interface errcode:"Not 3t
 >  parm's" > >         retn > > ;v > > ;  > >                 .end   ------------------------------   Date: 15 Nov 2001 15:30:35 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon) 3 Subject: Re: Need help, was: Rob's British Champion0, Message-ID: <9t0n2r$1f0l$1@info.cs.uofs.edu>  : In article <OF791CC05C.DDE25F33-ON00256B03.0045DCDC@btyp>,#  Steve.Spires@yellgroup.com writes:g |> qK |> Or, for more information try www.192.com which will search all electoralnL |> registers for the name you input, and give you back address, phone number |> etc.0 |> /F |> Paul Sture <paul.sture@bluewin.ch> on 11/09/2001 07:05:29 PM wrote: |> eC |> If you have his location, try Directory Enquiries at www.bt.com.  |> C  C Both of these seem like good ideas, until I provide the last bit ofiC data.  My friends name is "Smith".  "David S, Smith" to be precise.sC How many of them do you thnk there might be in England today??  :-)5   Thanks for the help, though.   bill   -- hJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   f   ------------------------------  # Date: Thu, 15 Nov 2001 14:05:57 GMTm1 From: "Mark D. Jilson" <jilly@clarityconnect.com>o: Subject: Re: OpenVMS o/s cd-rom insufficient quotas/limits2 Message-ID: <3BF3CB9D.9CA52CDE@clarityconnect.com>  C As I said, what you attempted will not solve the issue as a spawnedeH subprocess will not get a new pagefile quota, it uses the pagefile quotaF pool of the parent.  Remember that these CDs must be able to be bootedB and used on all systems listed in the SPD at the minimum supportedE memory size.  Historically this has lead to these conservative SYSGENsF values and booting conversationally and using SYSBOOT is the method toE use set parameters for systems larger than the minimum configuration.t   Jim Strehlow wrote:e > 4 > I set the sysgen parameters, I did a WRITE ACTIVE,% > spawned a new subprocess, and stilll& > I got the insufficient memory error.( > So, I still do not have a work-around. > > > Compaq should ship a better product "out of the box" so thatB > $ ANALYZE/DISK/REPAIR should work at the DCL prompt immediately. >  > Thank you for your help. > F > I will later log a call with Compaq support and vent my frustration. > / > Jim Strehlow, Data911 OpenVMS Systems Manager 
 > Alameda, CAv > > > "Mark D. Jilson" <jilly@clarityconnect.com> wrote in message- > news:3BF28363.BDA4895@clarityconnect.com...eI > > What you show below has absolutely no effect on the parameters of theoI > > current process.  Now if you SPAWN a subprocess then you will get the D > > larger for the non-pooled quotas but not for the pooled ones andG > > pagefile quota which I'm almost positive is the low quota here is a- > > pooled quota.o > >1 > > Jim Strehlow wrote:t > > >e1 > > > When I boot from the OpenVMS v7.2-1 cd-rom, A > > > I modified the following dynamic parameters without successt > > >o > > > $$$ MCR SYSGEN > > > SET PQL_DFILLM       1000m > > > SET PQL_MFILLM       1000t > > > SET PQL_DBIOLM       1000: > > > SET PQL_MBIOLM       10001 > > > SET PQL_DDIOLM       1000e > > > SET PQL_MDIOLM       1000I > > >p > > > SET PQL_DASTLM       3000q > > > SET PQL_MASTLM       3000o > > > SET PQL_DTQELM       3000  > > > SET PQL_MTQELM       3000o > > > SET PQL_DENQLM       4000  > > > SET PQL_MENQLM       4000  > > >u > > > SET PQL_DBYTLM     500000e > > > SET PQL_MBYTLM     500000e > > > SET PQL_DJTQUOTA     8192  > > > SET PQL_MJTQUOTA     8192p > > >p > > > SET PQL_DWSDEFAULT   4000h > > > SET PQL_MWSDEFAULT   4000w > > > SET PQL_DWSQUOTA     6000t > > > SET PQL_MWSQUOTA     6000e > > > SET PQL_DWSEXTENT   32768w > > > SET PQL_MWSEXTENT   32768  > > > SET PQL_DPGFLQUOTA 250000  > > > SET PQL_MPGFLQUOTA 250000M > > > WRITE ACTIVE > > >m@ > > > CREATE  SYS$SYSTEM:PAGEFILE.SYS /SIZE= 2105300 /CONTIGUOUS/ > > > INSTALL SYS$SYSTEM:PAGEFILE.SYS /PAGEFILEb@ > > > CREATE  SYS$SYSTEM:SWAPFILE.SYS /SIZE=   67300 /CONTIGUOUS/ > > > INSTALL SYS$SYSTEM:SWAPFILE.SYS /SWAPFILEn > > >n > > > When I0 > > > $$$ ANALYZE /DISK_STRUCTURE /REPAIR  dunn: > > >s- > > > I still get insufficient virtual memory ; > > > %ANALDISK-F-ALLOCMEM, error allocating virtual memoryt3 > > > -LIB-F-INSVIRMEM, insufficient virtual memoryo > > > J > > > When I alter SYSTEM to the following when I boot from a system disk,	 > analyze4 > > > works.I > > > MODIFY SYSTEM /FILLM= 1000 /SHRFILLM= 0 /BIOLM= 1000 /DIOLM= 1000 -t/ > > >  /ASTLM= 3000 /TQELM= 3000 /ENQLM= 4000 -21 > > >  /BYTLM= 500000 /PBYTLM= 0 /JTQUOTA= 8192 -aJ > > >  /WSDEFAULT= 2048 /WSQUOTA= 4096 /WSEXTENT= 32768 /PGFLQUOTA= 250000 > > > M > > > Might my problem be related to WSMAX which is not dynamic at 24000 from- > the-
 > > > cd-rom?- > > >a > > > Jim Strehlow, Data911  > > >rB > > > "Mark D. Jilson" <jilly@clarityconnect.com> wrote in message4 > > > news:<3BF037FE.454146CB@clarityconnect.com>...B > > > > You can also boot conversational and then do the following2 > > > > SYSBOOT> SET PQL_MPGFLQUOTA {large number}% > > > > SYSBOOT> SET WRITESYSPARAMS 00 > > > > SYSBOOT> CONTe > > ><L > > > > Jilly - Working from Home in the Chemung River Valley - Lockwood, NY5 > > > > - jilly@clarityconnect.com - Brett Bodine fan 3 > > > > - Mark.Jilson@Compaq.com - since 1975 or so 3 > > > > - http://www.jilly.baka.com               -k   -- eD Jilly	- Working from Home in the Chemung River Valley - Lockwood, NY0 	- jilly@clarityconnect.com			- Brett Bodine fan. 	- Mark.Jilson@Compaq.com			- since 1975 or so, 	- http://www.jilly.baka.com               -   ------------------------------    Date: 15 Nov 2001 09:03:57 -0600- From: koehler@encompasserve.org (Bob Koehler)  Subject: P/OS anyone?e3 Message-ID: <AasdSvhBQain@eisner.encompasserve.org>n  I    My recently aquired Pro 350 won't boot.  It comes to a graphics screen-C    showing a highlighted floppy with a question mark under it, and  M    highlights the first two CTi boards (hard drive and floppy controllers).  1H    The numeric diagnostic codes seem to match the hard drive controller.  F    I asume this means there's no OS on the hard drive (P/OS 3.2 or so F    could boot from the hard drive IIRC).  Anybody know where I can get8    a copy of P/OS?  I'd rather mess with that than UNIX.  G    I found a note on Eisner that P/OS 3.2 was in the DECUS library, but29    friends tell me the library is not workable right now.o  F    I tried reseating everything and swapping in a different (but still    empty) floppy drive.n  @    Kudos to the fellow who did the mechanical design of the CTI.   ------------------------------  # Date: Thu, 15 Nov 2001 17:59:06 GMT)= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)  Subject: Re: P/OS anyone?a0 Message-ID: <00A0515D.C5BAAA05@SendSpamHere.ORG>  c In article <AasdSvhBQain@eisner.encompasserve.org>, koehler@encompasserve.org (Bob Koehler) writes:s > J >   My recently aquired Pro 350 won't boot.  It comes to a graphics screenD >   showing a highlighted floppy with a question mark under it, and N >   highlights the first two CTi boards (hard drive and floppy controllers).  I >   The numeric diagnostic codes seem to match the hard drive controller.r >oG >   I asume this means there's no OS on the hard drive (P/OS 3.2 or so IG >   could boot from the hard drive IIRC).  Anybody know where I can get 9 >   a copy of P/OS?  I'd rather mess with that than UNIX.1 >3H >   I found a note on Eisner that P/OS 3.2 was in the DECUS library, but: >   friends tell me the library is not workable right now. > G >   I tried reseating everything and swapping in a different (but still  >   empty) floppy drive. >sA >   Kudos to the fellow who did the mechanical design of the CTI.  >e  K P/OS was on its drive.  Open the case and make sure everything is connectedp7 and clean.  Something might have come loose in transit.E --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM             sJ   "And of course, I'm a genius, so people are naturally drawn to my fiery I   intellect.  Their admiration overwhelms their envy!" -- Calvin & Hobbesc   ------------------------------  % Date: Thu, 15 Nov 2001 10:02:24 +0000 % From: Alan Greig <a.greig@virgin.net>u Subject: Re: Pathworks8 Message-ID: <0547vtkbo6gidi2203h9hq78226von1lo7@4ax.com>  B On Thu, 15 Nov 2001 05:29:15 GMT, "Aaron Cardon" <aaron@vwbus.org> wrote:  2 > Thank you for taking the time to answer my post.3 > It sounds like Pathworks is the optimal solution.? > K >I plan to write a COBOL application (using Microfocus NetExpress 3.1) thatgI >will run on a PC and be used to update COBOL indexed files located on an I >Open VMS server.  I want to use Pathworks to map a drive to the Open VMSEM >volume containing the COBOL data, thus giving my COBOL application access to L >that data.  Do you know if there will be any data incompatibilities between0 >the PC and Open VMS data?  Is this plan viable?  B Ah you want to update indexed files. That's a bit different as youF will be able to open the file on the PC via Pathworks but Windows will? not recognize the file format. So you would need to implement aN. minimal subset of RMS knowledge in the PC app.  = However the supported way to do this is to run something likeLF "Attunity Connect" (formerly ISG Navigator) on the VMS box and the PC.E The PC app can then use standard ODBC type drivers to read/write dataAF to the VMS RMS file just as if it was in a database such as MS-ACCESS, SQL Server, Oracle, RDB etc.   See www.attunity.com  F There are other products but Attunity is the one I can speak for as we8 use it to provide access to RDB, DBMS and RMS databases.  B For an extremely cheap and cheerful solution you could knock up  aB TCPIP socket server app on the VMS side and have the Cobol programF send its data to this socket from the PC. A simple such solution could> even be written in DCL on the VMS side running under UCX/TCPIPA auxiliary server (inetd). It would just read/write to sys$net and , read/wiite to the RMS file after processing.     >  >Aaron >  >  >;; >"MikeG" <michael.greenNOSP@Mvideotron.ca> wrote in message + >news:3bf3422d.6211872@news.videotron.ca...e	 >> Aaron,.G >> you are on the right track.  There are multiples way however to push I >> and pull files from a pc to an OpenVMS file system, but Pathworks will1G >> allow you to do so with the most transparency.   Allows you to treat F >> your shared VMS directories as shared drive letters on your WindowsF >> desktop.   You only need procure and install the Pathworks/AdvancedH >> Server software and purchase required client access license (CAL) forF >> number of clients to connect and you've got a file and print server	 >> going.  >>H >> You could look at NFS server or FTP if your needs are less demanding,H >> plus at lower cost.  In the case of NFS you need an both a server andD >> client component.  The NFS client you can get from Hummingbird orH >> Microsoft's Unix Services CD, and the NFS server component you merelyE >> enable via TCP/IP or UCX configuration menus.   It's a bit muckieriF >> then Pathworks, and I believe is single access only.  FTP client isH >> available out of box with Windows variants and can again be turned on) >> via TCP/IP or UCX configuration menus.p >>
 >> Cheers,	 >> Mike Gp >> >h   -- Alan   ------------------------------  % Date: Thu, 15 Nov 2001 11:53:10 -0000t+ From: "Tim Jackson" <tim.jackson@amsjv.com>y# Subject: Re: Proxy Removal Problem.Q& Message-ID: <3bf3a9ee$1@pull.gecm.com>  B When I had this problem (on V7.3) I ended up stopping the securityF server, deleting the NET*PROXY.DAT files, starting the security serverE again and then using the AUTHORIZE CREATE/PROXY command to create newgE empty databases.  To ensure the new files were created in the clusteraE common system directory I had to make sure the NETPROXY and NET$PROXY C logical names were defined and pointing to SYS$COMMON:[SYSEXE] (viam2 SYLOGICALS.COM) else the files would be created in SYS$SPECIFIC:[SYSEXE].   HTHlD ------------------ Purely Personal Opinion -------------------------D Tim Jackson                                    tim.jackson@amsjv.com Air Systems GroupM Alenia Marconi Systems Ltd.e  6 "Rob Buxton" <rob.buxton@wcc.govt.nz> wrote in message+ news:3bf2e5bb.501246524@news.wcc.govt.nz...m
 > Hi Folks >t$ > Running VMS 7.2 & DECNet OSI 7.2-10 > Ive got some old proxies I'm trying to delete. >t= > There's a couple of stubborn ones that I can't get trashed.m9 > All troublesome Proxies are set to the DEFAULT Account.g >. > So >i > UAF> show /prox *::userx >o' >  Default proxies are flagged with (D)Z >1 > NODE1::USERX
 >     DEFAULT  >c > NODE2::USERX
 >     DEFAULTe >  > But, >  > UAF> remove /prox *::userx; > %UAF-E-NAFREMERR, error removing proxy from *::userx to *pC > -SECSRV-E-NOSUCHPROXY, no proxy record matches your specification  >c > or >e  > UAF> remove /prox NODE1::USERXF > %UAF-E-NAFREMERR, error removing proxy from LOCAL:.NODE1::USERX to *C > -SECSRV-E-NOSUCHPROXY, no proxy record matches your specification  >dH > I know that when a User is removed the associated Proxies get removed.H > But I don't want to (and according to the help in Authorize you can't) > remove the DEFAULT Account.u? > The USERX Account has already been removed from this machine.l > F > I've also tried adding a new proxy of the same spec, it adds it withD > the LOCAL. Prefix. When I then delete this it just deletes the one > with the LOCAL. Prefix.  >i > Any help appreciated.w >  > TIA. >d > Rob.   ------------------------------    Date: 16 Nov 2001 01:55:00 +0800, From: Paul Repacholi <prep@prep.synonet.com># Subject: Re: Proxy Removal Problem.x- Message-ID: <878zd76hsb.fsf@prep.synonet.com>   + rob.buxton@wcc.govt.nz (Rob Buxton) writes:m   > UAF> show /prox *::userx > ' >  Default proxies are flagged with (D); >  > NODE1::USERX
 >     DEFAULT, >  > NODE2::USERX
 >     DEFAULTn >  > But, >  > UAF> remove /prox *::userx; > %UAF-E-NAFREMERR, error removing proxy from *::userx to *oC > -SECSRV-E-NOSUCHPROXY, no proxy record matches your specificationd  3 Don't you need to  explicitly specify the /default?G   UAF> remove /prox *::userx/def     >  > or >   > UAF> remove /prox NODE1::USERXF > %UAF-E-NAFREMERR, error removing proxy from LOCAL:.NODE1::USERX to *C > -SECSRV-E-NOSUCHPROXY, no proxy record matches your specificatione   -- r< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.Q@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Thu, 15 Nov 2001 07:29:18 -0800w* From: James Gessling <jgessling@yahoo.com>0 Subject: Re: Sources of class information needed) Message-ID: <3BF3DF4E.7A7CA013@yahoo.com>n   There info here:  ) http://www.compaq.com/training/cd-os.html3   Jim    Lyndon Bartels wrote:   > > I have an opportunity to take some VMS management classes... > : > THough I don't know of too many places that teach them.. >,F > If y'all can email me phone numbers/addresses, (or post them here as! > well) I'd really appreciate it.  > E > I'm interested in Performance tuning, Advanced Administration, etc.  ># > Thanks in advance, >  > Lyndon >. > --I > My opinions are mine and mine alone. They seldom align with those of my  > employer.  > J > The only good thing about putting the cart before the horse is you don't > have > to look at the horse's butt.   ------------------------------  % Date: Thu, 15 Nov 2001 09:24:30 +0000a% From: Alan Greig <a.greig@virgin.net> N Subject: Re: Special IPF-Inside Issue of Shannon Knows Compaq at www.tru64.org8 Message-ID: <2o17vtsjct7675jfc4akk5vpnaog6uak49@4ax.com>  5 On Wed, 14 Nov 2001 15:36:22 -0500, "Fred Kleinsorge" $ <kleinsorge@star.zko.dec.com> wrote:   >eE >Paul Repacholi wrote in message <87herzbu9i.fsf@prep.synonet.com>...  > H >>If you believe their numbers at all. BTW, does anyone know the name of> >>'the engineer' who said that Alpha would lose it performance >>advantage yet? >> >  > M >Yes.  And he is a very, very credible source - and someone quite invested in0 >Alpha technology.  D There's a psychological technique to bring people on side. SometimesE known as the "foot in the face" technique ("foot in the door" being adA more common method). This involves giving someone really bad newsSF which upsets them then later providing another option which gives some@ path forward. The target then will jump enthusiastically on thatC option even though they might have rejected it out of hand earlier.   = Compaq wanted out of the Alpha business and very senior Alpha9@ management have known this with near certainty (or possibly even= certainty) for a long time. Faced with the prospect of losings? everything (including VMS) or supporting (and even appearing toiC initiate) the new policy most will choose the latter. They may event4 convince themselves it was the right path all along.  2 Sort of like the Stockholm syndrome with hostages.  D Had Compaq really wanted to stick with Alpha it would have gone backD to this "senior engineer" and said "Hmm we have problems with Alpha.B How much extra funding do you need to fix this and here it is." By@ massaging business unit profitability figures given to employeesE Compaq made it appear that the additional money option could not even ! be considered. Ergo cancel Alpha.u   QEDt   >a >i   -- Alan   ------------------------------  % Date: Thu, 15 Nov 2001 10:59:30 -0500e* From: Trent Worthington <fuckoff@none.nul>N Subject: Re: Special IPF-Inside Issue of Shannon Knows Compaq at www.tru64.org8 Message-ID: <uap7vtk10qm3scq8fodo2m46659o86h0j4@4ax.com>  F On Wed, 14 Nov 2001 19:48:09 GMT, "Bill Todd" <billtodd@metrocast.net> wrote:   > 6 >Trent Worthington <fuckoff@none.nul> wrote in message3 >news:4m85vt8d3pitflamqp1ikr3ptqoo48dsti@4ax.com...O2 >> On Wed, 14 Nov 2001 03:55:04 GMT, Tim Llewellyn >e >... >eC >> >you mean until they could scrounge up some reasonable evidence?g& >>                            ^^^^^^^^' >> ITYM fabricate. He's guilty alright:  >>J >> http://www.cnn.com/2001/WORLD/europe/11/14/inv.britain.proof/index.html > J >Ah - so now expressing approval of something, and saying it's the kind ofK >action you've encouraged, constitutes an admission of guilt.  That kind ofrC >dilution of legal standards is certainly consistent with the other L >diminutions of freedoms that we're seeing domestically, but reminds me moreI >of you-know-who's Germany than of what America is supposed to stand for.   E Did you even read the article, you hapless, retarded, bleeding-heart,  Taliban apologist?:o  " (transcript excerpts of the video)  C "The document quotes bin Laden as saying: "It is what we instigatedgF for a while, in self-defence. If avenging the killing of our people is= terrorism, let history be a witness that we are terrorists."    ? Pay special attention now, so I don't have to repeat this post:h  F "The battle has been moved inside America, and we shall continue until= we win this battle, or die in the cause and meet our maker." G  ? Let me know if there are any words you don't understand, M'kay?0   ------------------------------  % Date: Thu, 15 Nov 2001 12:47:50 -0500 5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> N Subject: Re: Special IPF-Inside Issue of Shannon Knows Compaq at www.tru64.org3 Message-ID: <efTI7.1628$RL6.52895@news.cpqcorp.net>T  K Alan Greig wrote in message <2o17vtsjct7675jfc4akk5vpnaog6uak49@4ax.com>...46 >On Wed, 14 Nov 2001 15:36:22 -0500, "Fred Kleinsorge"% ><kleinsorge@star.zko.dec.com> wrote:_ >Q >>F >>Paul Repacholi wrote in message <87herzbu9i.fsf@prep.synonet.com>... >>I >>>If you believe their numbers at all. BTW, does anyone know the name of0? >>>'the engineer' who said that Alpha would lose it performance2 >>>advantage yet?_ >>>T >> >>K >>Yes.  And he is a very, very credible source - and someone quite invested  in >>Alpha technology.  > E >There's a psychological technique to bring people on side. SometimesLF >known as the "foot in the face" technique ("foot in the door" being aB >more common method). This involves giving someone really bad newsG >which upsets them then later providing another option which gives somePA >path forward. The target then will jump enthusiastically on that D >option even though they might have rejected it out of hand earlier. >F> >Compaq wanted out of the Alpha business and very senior AlphaA >management have known this with near certainty (or possibly evenl> >certainty) for a long time. Faced with the prospect of losing@ >everything (including VMS) or supporting (and even appearing toD >initiate) the new policy most will choose the latter. They may even5 >convince themselves it was the right path all along.P >Q3 >Sort of like the Stockholm syndrome with hostages.e >oE >Had Compaq really wanted to stick with Alpha it would have gone back E >to this "senior engineer" and said "Hmm we have problems with Alpha.@C >How much extra funding do you need to fix this and here it is." BylA >massaging business unit profitability figures given to employeestF >Compaq made it appear that the additional money option could not even" >be considered. Ergo cancel Alpha. >0 >QED    J It does not match the story that I heard directly from him.  This wasn't aL fishing expedition to kill Alpha.  As far as I know, it started by trying toJ figure out how to get our Fab issues sorted out (find a silicon "partner")D to contain costs and make sure we could deliver on our Alpha roadmap
 schedules.  K Think back to how much whining there was about EV6 schedule, or getting then! faster versions in new processes.c  F But conspiracy theorist can come up with alternatives.  In the end, itJ doesn't matter (at least to me).  A decision was made for good or ill, andG the consequences are playing out.  My personal feeling on the result ofQH being ported to Itanium is that while I would have preferred to spend myH time on other things, the end result will be a reliable, fast, competive- platform for VMS for the next decade or more.h   ------------------------------  % Date: Thu, 15 Nov 2001 19:04:16 +0100r& From: John McLean <mcleanj@dplanet.ch>N Subject: Re: Special IPF-Inside Issue of Shannon Knows Compaq at www.tru64.org* Message-ID: <3BF403A0.8D59E0C8@dplanet.ch>   Trent Worthington wrote: > H > On Wed, 14 Nov 2001 19:48:09 GMT, "Bill Todd" <billtodd@metrocast.net> > wrote: >  > >E8 > >Trent Worthington <fuckoff@none.nul> wrote in message5 > >news:4m85vt8d3pitflamqp1ikr3ptqoo48dsti@4ax.com... 4 > >> On Wed, 14 Nov 2001 03:55:04 GMT, Tim Llewellyn > >  > >... > > E > >> >you mean until they could scrounge up some reasonable evidence? ( > >>                            ^^^^^^^^) > >> ITYM fabricate. He's guilty alright:0 > >>L > >> http://www.cnn.com/2001/WORLD/europe/11/14/inv.britain.proof/index.html > > L > >Ah - so now expressing approval of something, and saying it's the kind ofM > >action you've encouraged, constitutes an admission of guilt.  That kind oftE > >dilution of legal standards is certainly consistent with the otheraN > >diminutions of freedoms that we're seeing domestically, but reminds me moreK > >of you-know-who's Germany than of what America is supposed to stand for.s > G > Did you even read the article, you hapless, retarded, bleeding-heart,r > Taliban apologist?:s > $ > (transcript excerpts of the video) > E > "The document quotes bin Laden as saying: "It is what we instigatedrH > for a while, in self-defence. If avenging the killing of our people is> > terrorism, let history be a witness that we are terrorists." > A > Pay special attention now, so I don't have to repeat this post:. > H > "The battle has been moved inside America, and we shall continue until> > we win this battle, or die in the cause and meet our maker." > A > Let me know if there are any words you don't understand, M'kay?t     Trent,   The facts are   F (a) there is still no evidence bin Laden that would stand up in court;E there are assertions, there are circumstantial and vague connections;T there is NO public evidence.  G (b) The statement that you quote is inconclusive regardless of how much @ you might wish/hope it was otherwise.  If you were to say, as anD American, "We beat Russia in the basketball." it would be impossibleC from this context to determine if you were actually in the team andPC played an active part, or if you were speaking generally and really H meaning that America beat Russia.  The big difference is the size of the. collective group encompassed by the word "we".  D In bin Laden's case he could easily have been speaking in terms of aF broad group (or groups) of terrorists, none of which directly involves  him in any kind of active role.     E The reaction of the USA is a big concern to many people.  Rather thanPG become embroiled in recycled arguments of a few months back, let me putt  an analogy in computing terms...  F If you managed a computer system and one day found that a certain user@ could breach the security whenever they wanted to, would you ...E (a)  disable the user's account and perhaps punish them in additional  waysG (b)  discover how they did it and take steps to close the security holeo ?z   Your answer is ???    F With regard to terrorists, the USA appears to be mainly concerned withG (a) and are paying lip-service to (b).  Part (b) is very difficult on atG national basis.  The USA is losing the war against drugs and if this isCH an example of the best levels of effort, then it is almost inconceivableE that the country can be truly secured against all forms of terrorism.v  G In this case (as with drugs) one must work to remove the motivation andtH incentive for these illegal activities.  Remove the demand and you don'tC just stop the current incident but you lessen the chances of othersu) undertaking similar activities in future.r  C This removal of motivation is where many of us feel that the USA isi< failing and consequently we fear of other acts of terrorism.     John McLeanh   ------------------------------  % Date: Thu, 15 Nov 2001 13:19:55 -0500 * From: Trent Worthington <fuckoff@none.nul>N Subject: Re: Special IPF-Inside Issue of Shannon Knows Compaq at www.tru64.org8 Message-ID: <4618vtke2j4e71mfo48je6m3vomgpfd7rg@4ax.com>  D On Thu, 15 Nov 2001 19:04:16 +0100, John McLean <mcleanj@dplanet.ch> wrote:   >t >e >Trent Worthington wrote:a >> lI >> On Wed, 14 Nov 2001 19:48:09 GMT, "Bill Todd" <billtodd@metrocast.net>y	 >> wrote:  >> t >> >9 >> >Trent Worthington <fuckoff@none.nul> wrote in messaged6 >> >news:4m85vt8d3pitflamqp1ikr3ptqoo48dsti@4ax.com...5 >> >> On Wed, 14 Nov 2001 03:55:04 GMT, Tim Llewellynk >> > >> >...S >> >F >> >> >you mean until they could scrounge up some reasonable evidence?) >> >>                            ^^^^^^^^l* >> >> ITYM fabricate. He's guilty alright: >> >>oM >> >> http://www.cnn.com/2001/WORLD/europe/11/14/inv.britain.proof/index.html. >> >M >> >Ah - so now expressing approval of something, and saying it's the kind ofeN >> >action you've encouraged, constitutes an admission of guilt.  That kind ofF >> >dilution of legal standards is certainly consistent with the otherO >> >diminutions of freedoms that we're seeing domestically, but reminds me morenL >> >of you-know-who's Germany than of what America is supposed to stand for. >> tH >> Did you even read the article, you hapless, retarded, bleeding-heart, >> Taliban apologist?: >>  % >> (transcript excerpts of the video)  >> 1F >> "The document quotes bin Laden as saying: "It is what we instigatedI >> for a while, in self-defence. If avenging the killing of our people iso? >> terrorism, let history be a witness that we are terrorists."e >> yB >> Pay special attention now, so I don't have to repeat this post: >>  I >> "The battle has been moved inside America, and we shall continue untilt? >> we win this battle, or die in the cause and meet our maker."a >> mB >> Let me know if there are any words you don't understand, M'kay? >i >  >Trent,e >f >The facts are   >SG >(a) there is still no evidence bin Laden that would stand up in court; F >there are assertions, there are circumstantial and vague connections; >there is NO public evidence.-  " OK, let me try this another way...     Hmmm....  
 Ummmm........m              , HE ADMITTED IT YOU FUCKING IDIOT!!!!!!!!!!!1  H >(b) The statement that you quote is inconclusive regardless of how muchA >you might wish/hope it was otherwise.  If you were to say, as an7E >American, "We beat Russia in the basketball." it would be impossible D >from this context to determine if you were actually in the team andD >played an active part, or if you were speaking generally and reallyI >meaning that America beat Russia.  The big difference is the size of theE/ >collective group encompassed by the word "we".A  - And what part of "we" did you not understand?x  E >In bin Laden's case he could easily have been speaking in terms of a-G >broad group (or groups) of terrorists, none of which directly involves ! >him in any kind of active role.     BWAHAHAHAHAHAHAHAHA!!!!!!!!!   (wipes tears from eyes...)  F >The reaction of the USA is a big concern to many people.  Rather thanH >become embroiled in recycled arguments of a few months back, let me put! >an analogy in computing terms...a  % Oh great, another ridiculous ANALogy.'  G >If you managed a computer system and one day found that a certain usermA >could breach the security whenever they wanted to, would you ...oF >(a)  disable the user's account and perhaps punish them in additional >waysiH >(b)  discover how they did it and take steps to close the security hole >? >  >Your answer is ???.   Both. First A, then B.  G >With regard to terrorists, the USA appears to be mainly concerned withe >(a)  , Excellent. This is what I call a good start.  $ > and are paying lip-service to (b).  A Bullshit. Just wait, Iraq will be next, and then perhaps Lebanon.   " >  Part (b) is very difficult on a >national basis.  / Of course it is. Are you suggesting we not try?E   *sheesh*  9 >  The USA is losing the war against drugs and if this is I >an example of the best levels of effort, then it is almost inconceivableUF >that the country can be truly secured against all forms of terrorism. >eH >In this case (as with drugs) one must work to remove the motivation and( >incentive for these illegal activities.  B Just tell all those drug addicts to stop, ehh? Yeah, that'll work.	 *snicker*C   <snip>   ------------------------------    Date: 15 Nov 2001 10:27:05 -0800( From: bob@instantwhip.com (Bob Ceculski)N Subject: Re: Special IPF-Inside Issue of Shannon Knows Compaq at www.tru64.org= Message-ID: <d7791aa1.0111151027.260d2149@posting.google.com>o  e Alan Greig <a.greig@virgin.net> wrote in message news:<2o17vtsjct7675jfc4akk5vpnaog6uak49@4ax.com>...l7 > On Wed, 14 Nov 2001 15:36:22 -0500, "Fred Kleinsorge"o& > <kleinsorge@star.zko.dec.com> wrote: >  > >sG > >Paul Repacholi wrote in message <87herzbu9i.fsf@prep.synonet.com>...g > > J > >>If you believe their numbers at all. BTW, does anyone know the name of@ > >>'the engineer' who said that Alpha would lose it performance > >>advantage yet? > >> > >S > >gO > >Yes.  And he is a very, very credible source - and someone quite invested in  > >Alpha technology. > F > There's a psychological technique to bring people on side. SometimesG > known as the "foot in the face" technique ("foot in the door" being asC > more common method). This involves giving someone really bad newsyH > which upsets them then later providing another option which gives someB > path forward. The target then will jump enthusiastically on thatE > option even though they might have rejected it out of hand earlier.- > ? > Compaq wanted out of the Alpha business and very senior AlpharB > management have known this with near certainty (or possibly even? > certainty) for a long time. Faced with the prospect of losing1A > everything (including VMS) or supporting (and even appearing to3E > initiate) the new policy most will choose the latter. They may even 6 > convince themselves it was the right path all along. > 4 > Sort of like the Stockholm syndrome with hostages. > F > Had Compaq really wanted to stick with Alpha it would have gone backF > to this "senior engineer" and said "Hmm we have problems with Alpha.D > How much extra funding do you need to fix this and here it is." ByB > massaging business unit profitability figures given to employeesG > Compaq made it appear that the additional money option could not evenh# > be considered. Ergo cancel Alpha.. >  > QEDv >   K I keep hearing that alpha was losing ground to Intel ... if you look at alloI the recent Itanic problems, what kind of time frame were they using to doc the analysis with, 20 years?   ------------------------------  % Date: Thu, 15 Nov 2001 08:48:09 +0100f= From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <noone@dummy.com>wL Subject: Re: Unix and Linux certs advisory - VMS remains only unhackable OS!) Message-ID: <3BF37339.40A4D8B8@dummy.com>(  > Isn't it the CDE image that recevies some data and then itself> writes it into it's own buffer space, including some (program)@ space after the buffer. The "user" at some other CPU don't realy> *do* anything on the server and he never obtain any privs. The= CDE image is using it's normal privs to run the code from theiC overwritten buffer. The CDE do not realy *know* what's going on... i   Jan-Erik Sderholm.h   Don Sykes wrote: >  > E > I don't pretend to know Unix well, but how can a user overwriting adG > memory buffer obtain the root privileges of the CDE image reading theeE > buffer? I can't do that in VMS, which I'm sure is your point, but I-3 > thought Unix had process/privilege isolation too.0 >    ------------------------------  % Date: Thu, 15 Nov 2001 09:08:04 -0000 * From: "Richard Brodie" <R.Brodie@rl.ac.uk>L Subject: Re: Unix and Linux certs advisory - VMS remains only unhackable OS!+ Message-ID: <9t00ll$kus@newton.cc.rl.ac.uk>r  Z "Don Sykes" <anonymous@pacbell.net> wrote in message news:3BF3050C.560B906A@pacbell.net...    E > I don't pretend to know Unix well, but how can a user overwriting aMG > memory buffer obtain the root privileges of the CDE image reading the 	 > buffer?a  H Roughly: you play around until you manage to overwrite a stack frame andI the PC ends up with a garbage value. Then tweak until the buffer contains. code and the PC points to it.A  J Unfortunately, while its easy to protect code pages, you can still execute data pages.c   > I can't do that in VMS Don't worry, others can.   ------------------------------    Date: 15 Nov 2001 08:31:00 -0600+ From: young_r@encompasserve.org (Rob Young)iL Subject: Re: Unix and Linux certs advisory - VMS remains only unhackable OS!3 Message-ID: <ZVLg5j8dgovV@eisner.encompasserve.org>a  U In article <3BF3050C.560B906A@pacbell.net>, Don Sykes <anonymous@pacbell.net> writes:p >   E >> There is a remotely exploitable buffer overflow vulnerability in awD >> shared library that is used by dtspcd. During client negotiation,D >> dtspcd accepts a length value and subsequent data from the clientI >> without performing adequate input validation. As a result, a malicious I >> client can manipulate data sent to dtspcd and cause a buffer overflow,u3 >> potentially executing code with root privileges.t > E > I don't pretend to know Unix well, but how can a user overwriting aMG > memory buffer obtain the root privileges of the CDE image reading the5E > buffer? I can't do that in VMS, which I'm sure is your point, but Ic3 > thought Unix had process/privilege isolation too.a >   ! 	Problem is things run as "root".2  ; 	We had clients that for security purposes wouldn't give us2? 	root access.  No problem.  Our application ran as "root".  And0> 	of course the developers were savvy enough to stick in a backA 	door.  A back door that no one would find or accidently trigger..= 	It required a specific keystroke combination.  But once thatt@ 	keystroke combo was pressed - "voila!"  You were shelled out... 	with "root" privileges.  : 	I'm sure there are many applications out there today like 	that.   	You gotta love Eunuchs!   				Robe   ------------------------------  % Date: Thu, 15 Nov 2001 15:46:45 +0100c$ From: "Jakob Erber" <erberj@post.ch>) Subject: usage of new products on vms axpn Message-ID: <3bf3d557@hcwe67>a   Hi All,m  E most of the entries in this group are about complaining about the badhJ destiny of VMS. Bad support from Compaq and less and less products. But ifJ companies provide attractive new products for VMS nobody seems to be using them. Examples?d  D Due to the Java support on VMS, Beas J2EE product WebLogic Server isE supported for VMS. Also Ionas powerfull ORB is available for VMS. Ando others.y  K This products would allow for VMS not to be treaded as legacy system but tohD be fully integraded in a new system architecture. But the todays VMSD community seems to consist "only" of System managers and C- or Marco programmers.  @ Please excuse if this sounds hard, I just want to provoke a bit.   best regards   JakobH   ------------------------------    Date: 15 Nov 2001 11:13:24 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)e- Subject: Re: usage of new products on vms axp 3 Message-ID: <GStID2jIiJZq@eisner.encompasserve.org>i  D In article <3bf3d557@hcwe67>, "Jakob Erber" <erberj@post.ch> writes:	 > Hi All,t > G > most of the entries in this group are about complaining about the badlL > destiny of VMS. Bad support from Compaq and less and less products. But ifL > companies provide attractive new products for VMS nobody seems to be using > them. Examples?i > F > Due to the Java support on VMS, Beas J2EE product WebLogic Server isG > supported for VMS. Also Ionas powerfull ORB is available for VMS. Andu	 > others.n > M > This products would allow for VMS not to be treaded as legacy system but torF > be fully integraded in a new system architecture. But the todays VMSF > community seems to consist "only" of System managers and C- or Marco > programmers.  E I didn't see any evidence for your statement that "nobody seems to be C using them".  Don't confuse the VMS customer base with the group ofeB folks hanging around comp.os.vms.  People who don't have a problemB and have lots of work to do are unlikely to spend their time here.   ------------------------------  % Date: Thu, 15 Nov 2001 00:48:49 -05000( From: David Froble <davef@tsoft-inc.com>H Subject: Re: Using INETACP_FUNC$C_GETHOSTBYNAME using fortran- examples?, Message-ID: <3BF35741.9040909@tsoft-inc.com>   Tim Llewellyn wrote:   >  > David Froble wrote:N >  >>John Gemignani, Jr. wrote: >>I >> > Why are you asking the ACP for this information?  You are better offe/ >> >  calling the DECC function to do the work.p >> >
 >> > -John >>. >>Only if you can get yourself to work in 'C'. >> > G > Hey Dave, this is VMS, you can call C routines from Fortran, I'm suree > that is what John C > meant. OK, you might have to mess with %VAL and %REF and friends.  >   I Yes, that's true.  But, using the C routines from other than C has to be iG harder than working in C.  I just didn't have the courage to go there. ,B I've done system programming from time to time, and using the RTL I routines and system services is something I'm comfortable with.  My real  I problem was learning TCP/IP and socket programming.  It was fun, and I'm  E now keeping a lookout for other places I can use it.  What I haven't tC done yet is any type of simultaneous processing of messages on the a server side.  G A bit of a question for those who have had to address such.  I've read aI about passing messages to other processes for handling.  I could imagine iG threads being used in some applications.  What have those who have had nC to do simultaneous processing of messages found to be good methods?    Dave   -- r4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com6 T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486   ------------------------------    Date: 15 Nov 2001 01:50:25 -08004 From: jamie.stallwood@wcigroup.com (Jamie Stallwood)H Subject: Re: Using INETACP_FUNC$C_GETHOSTBYNAME using fortran- examples?= Message-ID: <eaaf81e4.0111150150.54feeb32@posting.google.com>c  z Hiroyuki_Tanaka4@excite.co.jp (Hiroyuki Tanaka) wrote in message news:<68cfa44d.0111141236.613f605d@posting.google.com>... > PROGRAM DRIVER >       IMPLICIT NONE   A Thanks for that, I'll give it a jolly good go. And thanks for thet8 lievly discussion everyone, it always cheers me up... :)   ------------------------------   Date: 15 Nov 2001 09:42 CST ' From: carl@gerg.tamu.edu (Carl Perkins)bH Subject: Re: Using INETACP_FUNC$C_GETHOSTBYNAME using fortran- examples?- Message-ID: <15NOV200109421889@gerg.tamu.edu>a  , David Froble <davef@tsoft-inc.com> writes... }What I haven't D }done yet is any type of simultaneous processing of messages on the 
 }server side.i } H }A bit of a question for those who have had to address such.  I've read J }about passing messages to other processes for handling.  I could imagine H }threads being used in some applications.  What have those who have had D }to do simultaneous processing of messages found to be good methods? }  }Dave   E I think the easiest method may be to do it all in one process and oneaC thread using asynchronous QIOs and ASTs. When any QIO completes, ityC gathers the relevant data, puts it in a queue (one for each of readUG completion, write completion, or new connection, for example), and does E a SYS$WAKE. When the process wakes up, it pulls things off queues andiD processes them. When there isn't anthing left in any queue it does aG SYS$HIBER. You can handle a moderately large number of connections thatlH don't shift too much data, or do too much processing of it, this way all6 in one process (watching out for your process quotas).  A There are examples of parts of this in various places (such as insD SYS$SYSROOT:[SYSHLP.EXAMPLES.TCPIP...]) but they are pretty much allC in C. Perhaps the best example is SYS$EXAMPLES:DB_SERVER.C. I found H that translating it to Fortran wasn't very difficult (although I do knowD C, which probably makes it considerably easier). It is a server thatH processes IO over multiple connections via DECnet. There is some sort ofI example in Fortran called SYS$EXAMPLES:OSIT$ECHO.FOR which does somethingsF vaguely similar (I didn't look very closely, but it doesn't seem to beI as complete an example as DB_SERVER.C is) also over DECnet. All the TCPIPn5 examples seem to use only QIOW, which is not so nice.>  F Then again I've never actually used the multiple processes or multipleH threads methods so I can't say that if methods are, or are not, actually$ significantly more difficult to use.   --- Carl   ------------------------------  # Date: Thu, 15 Nov 2001 18:18:28 GMTy3 From: Tim Llewellyn <tim.llewellyn@cableinet.co.uk>cH Subject: Re: Using INETACP_FUNC$C_GETHOSTBYNAME using fortran- examples?/ Message-ID: <3BF40601.D55C2C0B@cableinet.co.uk>l   Carl Perkins wrote:t > . > David Froble <davef@tsoft-inc.com> writes... > }What I haven't E > }done yet is any type of simultaneous processing of messages on the  > }server side.> > }nI > }A bit of a question for those who have had to address such.  I've readiK > }about passing messages to other processes for handling.  I could imagineiI > }threads being used in some applications.  What have those who have hadfF > }to do simultaneous processing of messages found to be good methods? > }  > }Daveu > G > I think the easiest method may be to do it all in one process and oneoE > thread using asynchronous QIOs and ASTs. When any QIO completes, itoE > gathers the relevant data, puts it in a queue (one for each of readwI > completion, write completion, or new connection, for example), and doesiG > a SYS$WAKE. When the process wakes up, it pulls things off queues andiF > processes them. When there isn't anthing left in any queue it does aI > SYS$HIBER. You can handle a moderately large number of connections that J > don't shift too much data, or do too much processing of it, this way all8 > in one process (watching out for your process quotas).  , I too have found this approach to work well.   > C > There are examples of parts of this in various places (such as ineF > SYS$SYSROOT:[SYSHLP.EXAMPLES.TCPIP...]) but they are pretty much allE > in C. Perhaps the best example is SYS$EXAMPLES:DB_SERVER.C. I found.J > that translating it to Fortran wasn't very difficult (although I do know  	 You too! e  F > C, which probably makes it considerably easier). It is a server thatJ > processes IO over multiple connections via DECnet. There is some sort ofK > example in Fortran called SYS$EXAMPLES:OSIT$ECHO.FOR which does something H > vaguely similar (I didn't look very closely, but it doesn't seem to beK > as complete an example as DB_SERVER.C is) also over DECnet. All the TCPIP.7 > examples seem to use only QIOW, which is not so nice.u   > H > Then again I've never actually used the multiple processes or multipleJ > threads methods so I can't say that if methods are, or are not, actually& > significantly more difficult to use. >   D The multiple processes approach requires more context switches (save process state,C load new process state) but with modern hardware this may not be ani issue unlessF you are pushing the capabilities of current top-of-the-range hardware. RemovingB this overhead is, I suppose, one reason why threads libraries were developed, theE other being to reduce implementation errors using QIO's, AST's, event 	 flags anda> linked lists, which presumably is what the thread libraries do internally.   	 regards     
 > --- Carl   -- d Tim.Llewellyn@cableinet.co.uk  l  C Standard disclaimer applies. My views in no way represent those of c! my employers or service provider.    ------------------------------    Date: 16 Nov 2001 02:28:49 +0800, From: Paul Repacholi <prep@prep.synonet.com>H Subject: Re: Using INETACP_FUNC$C_GETHOSTBYNAME using fortran- examples?- Message-ID: <87zo5n51ni.fsf@prep.synonet.com>   ) carl@gerg.tamu.edu (Carl Perkins) writes:b  H > Then again I've never actually used the multiple processes or multipleJ > threads methods so I can't say that if methods are, or are not, actually& > significantly more difficult to use.  G The WASD web server will provide a nice example to study. And serve the  odd page or two as well :)   -- o< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.l@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------   Date: 15 Nov 2001 18:27:56 GMT% From: "Paul Dembry" <pade@trifox.com>o Subject: Using vfork and exit50 Message-ID: <9t11fc$i3e@dispatch.concentric.net>  J According to the DEC C RTL manual, exit() is supposed to return control toK wherever vfork() was called, similar to what a successful exec() call does.eI However it is not doing so for me, it just exits the parent process.  HasaG anyone used this combination successfully?  We are running OpenVMS 7.2.n Thanks,  Paul   ------------------------------  % Date: Thu, 15 Nov 2001 12:06:35 -0500u2 From: rdeininger@mindspring.com (Robert Deininger) Subject: Re: VMS CrashL Message-ID: <rdeininger-1511011206360001@user-2iveb9o.dialup.mindspring.com>  A In article <1011114143943.2914B-100000@Ives.egh.com>, John Santosn <JOHN@egh.com> wrote:w    B > In the 99% case, when I'm shutting down a cluster node, it is toG > load an ECO or for minor hardware reasons, such as re-routing a power F > cable or swapping out a tape drive, and the system will only be downD > a few minutes.  Since I have 3 voting nodes, I don't have to worryF > about cluster quorum.  I could stop all apps that are using a servedA > disk (on the other two nodes), remove any installed images, andaE > dismount the disks before rebooting, and then remount the disks andtA > restart the apps after, but it is *much* easier to just let thegD > served disks go into mount-verification if either of the other twoF > nodes hits them while the serving node is down.  MSCP-serving startsB > up very early in the boot process, so the disks come back pretty > fast.n  I This is what I consider the most "normal" kind of cluster transition.  IfoG a node is only going down for a few minutes, just do the reboot.  MountbG verification is your friend.  Even if you lose quorum without the node,rJ it's typically less than 2 or 3 minutes before the vote comes back.  UsingI REMOVE_NODE, or simulating its effects by hand, is more trouble than it'se worth in many cases.    B > Each system has a local system disk and at least one other localB > disk, and there is no shared storage.  (Do the turbochannel SCSI9 > interfaces in an Alpha 3000-300 support shared SCSI?)  l  E Definitely not supported.  Probably wouldn't even work since the SCSIo; interface and/or driver likely doesn't support target mode.h  J If you need shared SCSI on a turbochannel alpha system, the only supportedJ way is using a KZTSA module.  They were pretty expensive when current, dueF to their excellent performance, so they aren't really plentiful.  I'veJ snagged one; when I find another, I'll be able to tell you for certain how& shared SCSI works on DEC 3000 systems.   -- a Robert Deininger rdeininger@mindspring.com    ------------------------------    Date: 15 Nov 2001 09:05:11 -08000 From: pearsonb@windows-devel.com (Brian Pearson)A Subject: VMS debug image runs OK but nocheck, check version crash:= Message-ID: <7042227b.0111150905.199e056a@posting.google.com>e  > Here's my dilema: I have a VMS EXE that crashes with an access? violation when compiled with check and nocheck (in other words, C nodebug) versions. When I compile and link a debug version, it runsiE perfectly. This EXE dials out to remote location, hooks into a remoteoC display and presents the display back to us, at the central office.uB The EXE does lots of fun things like logical IO to change terminal= characteristics, etc. The problems started when we upgrade tok- OpenVMS/VAX Version 7.2. Here's my questions:w  A 1. Anyone experiences this problem after upgrading to OpenVMS/VAXx Version 7.2.  B 2. Anyone have experience debugging an nondebug version of an EXE.  > 3. Access violation usually implies memory read/write problemsC (example, off the end of an array or to address 0). Anybody know ofwB any other access violations associated with devices or logical IO.   ------------------------------  % Date: Thu, 15 Nov 2001 12:53:23 -0500e5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>tE Subject: Re: VMS debug image runs OK but nocheck, check version crash 3 Message-ID: <skTI7.1629$RL6.52774@news.cpqcorp.net>o  J Try building the non-debug version no optimize and see what happens.  AlsoL check for uninitialized variables (use a recent DEC C compiler and ask it to check for bad things).   Brian Pearson wrote in message4 <7042227b.0111150905.199e056a@posting.google.com>...? >Here's my dilema: I have a VMS EXE that crashes with an accessa@ >violation when compiled with check and nocheck (in other words,D >nodebug) versions. When I compile and link a debug version, it runsF >perfectly. This EXE dials out to remote location, hooks into a remoteD >display and presents the display back to us, at the central office.C >The EXE does lots of fun things like logical IO to change terminal > >characteristics, etc. The problems started when we upgrade to. >OpenVMS/VAX Version 7.2. Here's my questions: >hB >1. Anyone experiences this problem after upgrading to OpenVMS/VAX
 >Version 7.2.p >tC >2. Anyone have experience debugging an nondebug version of an EXE.  >o? >3. Access violation usually implies memory read/write problemstD >(example, off the end of an array or to address 0). Anybody know ofC >any other access violations associated with devices or logical IO.    ------------------------------   Date: 15 Nov 2001 11:48 CST ' From: carl@gerg.tamu.edu (Carl Perkins)iE Subject: Re: VMS debug image runs OK but nocheck, check version crashd- Message-ID: <15NOV200111484378@gerg.tamu.edu>o  4 pearsonb@windows-devel.com (Brian Pearson) writes...? }Here's my dilema: I have a VMS EXE that crashes with an accesso@ }violation when compiled with check and nocheck (in other words,D }nodebug) versions. When I compile and link a debug version, it runsF }perfectly. This EXE dials out to remote location, hooks into a remoteD }display and presents the display back to us, at the central office.C }The EXE does lots of fun things like logical IO to change terminal > }characteristics, etc. The problems started when we upgrade to. }OpenVMS/VAX Version 7.2. Here's my questions: } B }1. Anyone experiences this problem after upgrading to OpenVMS/VAX
 }Version 7.2.o } C }2. Anyone have experience debugging an nondebug version of an EXE.o } ? }3. Access violation usually implies memory read/write problemsyD }(example, off the end of an array or to address 0). Anybody know ofC }any other access violations associated with devices or logical IO.m  E You should post the exact error message. An ACCVIO has a reason mask. G Look at HELP/MESSAGE ACCVIO to see what the reason mask is telling you.vG Carefully examine what the code is doing at and shortly before it dies.j  A The stereotypical reason for something like this is that you have@A an uninitialized variable being used. The either random junk or 0E> is flinging you off into someplace that you should't be going.  H The other reason things tend to pop up on an upgrade is timing problems.C Something that used to take long enough to avoid the problem is noweB running quicker than before and causing problems because of it (orG the opposite - something is now slower, but that is a lot less common).cG This is often tied into the first thing - the data isn't there yet whena it is trying to use it.t  J Both of these types of problems can disappear in a debug version, althoughF it may be the /nooptimize that generally goes with the /debug that canH cause the problem to go away. You might check to see what happens if youE compile with /noopt but don't use /debug. (If this makes it work, youeF could just always build it that way - but I wouldn't, I would find out what is wrong and fix it.)  F Another possibile reason is differing quotas. You can get an ACCVIO ifI you can't allocate more space on the stack. Make sure your process quotas-G are at least as big as they used to be. (The HELP/MESS ACCVIO tells you ( what may hints at this being the cause.)   --- Carl   ------------------------------  % Date: Thu, 15 Nov 2001 13:04:16 -050004 From: John Malmberg <Malmberg@dskwld.zko.dec.compaq>E Subject: Re: VMS debug image runs OK but nocheck, check version crashy4 Message-ID: <3BF403A0.5060306@dskwld.zko.dec.compaq>   Brian Pearson wrote:  @ > Here's my dilema: I have a VMS EXE that crashes with an accessA > violation when compiled with check and nocheck (in other words,,E > nodebug) versions. When I compile and link a debug version, it runs G > perfectly. This EXE dials out to remote location, hooks into a remote E > display and presents the display back to us, at the central office. D > The EXE does lots of fun things like logical IO to change terminal? > characteristics, etc. The problems started when we upgrade ton/ > OpenVMS/VAX Version 7.2. Here's my questions:h > C > 1. Anyone experiences this problem after upgrading to OpenVMS/VAXk > Version 7.2.    - No, but I have seen it when testing programs.t    lD > 2. Anyone have experience debugging an nondebug version of an EXE.    : You basically need to understand a little bit of Macro-32.A Having listings with the assembly language interspersed can help.u    @ > 3. Access violation usually implies memory read/write problemsE > (example, off the end of an array or to address 0). Anybody know ofdD > any other access violations associated with devices or logical IO.    ; The big clue is that it works in debug, and not in nodebug.2  E In my personal experience it indicates an uninitialized variable thatg% the program is assuming to contain 0.o  F Another issue that can cause similar problems is AST routines that areF improperly coded to access variables that are no longer in scope. This% can show up as an intermittant error.w  > Check your use of system services.  Make sure that you are notF inadvertantly calling the asynchronous version where you meant to callE the synchronous version.  An example would be using SYS$QIO where youe meant to be using SYS$QIOW.3  A You can also look in the OpenVMS FAQ or search the Ask The Wizard   7 answered questions.  See http://www.openvms.compaq.com/k     -John  malmberg@dskwld.zko.dec.compaq Personal Opinion Only)   ------------------------------  % Date: Thu, 15 Nov 2001 16:22:14 +0100a5 From: Rainer Brestan <rainer.brestan@sea.ericsson.se>E Subject: VMS emulator needed/ Message-ID: <3BF3DDA6.736A34F3@sea.ericsson.se>8  G Does anybody know emulators, who can run VMS binaries on other machines  than OpenVMS machines ?y   Thanks, Rainer   ------------------------------   Date: 15 Nov 2001 15:42:05 GMT) From: leslie@clio.rice.edu (Jerry Leslie)s  Subject: Re: VMS emulator needed' Message-ID: <9t0nod$6sg$1@joe.rice.edu>   6 Rainer Brestan (rainer.brestan@sea.ericsson.se) wrote:I : Does anybody know emulators, who can run VMS binaries on other machiness : than OpenVMS machines ?  :  : Thanks, Rainer  ) For OpenVMS VAX binaries, there's Charon:      http://www.charon-vax.com/   --Jerry Leslie   ------------------------------  % Date: Thu, 15 Nov 2001 17:00:08 +0100 7 From: "Dijk, Jeroen van" <Jeroen.vanDijk@Getronics.com>i  Subject: RE: VMS emulator neededO Message-ID: <2795B75EF003D311801A00A0C906B511011C6A12@cucexec.gbc.getronics.nl>n  2 There is one for the uVAX. CharonVAX it is called.   > -----Original Message-----> > From: Rainer Brestan [mailto:rainer.brestan@sea.ericsson.se]( > Sent: donderdag 15 november 2001 16:22 > To: Info-VAX@Mvb.Saic.Com  > Subject: VMS emulator needed >  > ; > Does anybody know emulators, who can run VMS binaries on o > other machines > than OpenVMS machines ?X >  > Thanks, Rainer >    ------------------------------  % Date: Thu, 15 Nov 2001 09:44:22 -0700V From: Kevin Handy <kth@srv.net>i  Subject: Re: VMS emulator needed& Message-ID: <3BF3F0E6.500C3B5@srv.net>   Rainer Brestan wrote:i > I > Does anybody know emulators, who can run VMS binaries on other machinese > than OpenVMS machines ?   B Besides Charon there are two others that I know of in development.; Bob Supnik has started one to include in his simh series ofm? emulators, and Tim Stark has started one to include in his TS10t@ series of emulators.  Neither is yet functional (mostly problems( figuring out how to emulate MSCP disks).   ------------------------------  % Date: Thu, 15 Nov 2001 07:51:33 +0100 2 From: Berger Harald <hotline@harryworld.dydns.org>= Subject: Re: VMS/RMS - destroyed sequential files - NEED HELP-8 Message-ID: <31p6vt4vkiq6qpt8cu6cc8lit1k7obl4bc@4ax.com>  5 the next "funny" thing that has happend on the files:i   .) a type /tail	is okB .) a type /tail=60 is ok .) a type /tail=80 is ok .) a type /tail = 100 is ok ' .) a type /tail = 150  following error:e> %TYPE-W-READERR, error reading DSA501:<ET_KATALOG>EXP_TX.TXT;18 -SYSTEM-E-UNSUPPORTED, unsupported operation or function  > and now the "funny" thing. when now making a  type /tail = 100C i got the same error. but 5 minutes before the type/tail was OK....h          , On Wed, 14 Nov 2001 08:18:37 -0500, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:    >re: problem with files: > 7 >have you tried backup/verify myfile.dat yourfile.dat ?e >dN >Also, try type/tail of the file. It *should* work on a variable length recordL >file , but if there is anything screwy with it, it will fail with a message" >such as unsupported organisation.   ------------------------------  % Date: Thu, 15 Nov 2001 07:58:29 +0100i2 From: Berger Harald <hotline@harryworld.dydns.org>= Subject: Re: VMS/RMS - destroyed sequential files - NEED HELP 8 Message-ID: <arp6vt0jrf5jammtfj4ou8mp821l0sus38@4ax.com>  0 $ backup /veri exp_art.txt exp_art_test.txt /log@ %BACKUP-S-CREATED, created DSA501:[ET_KATALOG]EXP_ART_TEST.TXT;11 %BACKUP-I-STARTVERIFY, starting verification passl7 %BACKUP-E-VERIFYERR, verification error for block 16 of.% DSA501:[ET_KATALOG]EXP_ART_TEST.TXT;1o7 %BACKUP-E-VERIFYERR, verification error for block 80 ofv% DSA501:[ET_KATALOG]EXP_ART_TEST.TXT;1O8 %BACKUP-E-VERIFYERR, verification error for block 192 of% DSA501:[ET_KATALOG]EXP_ART_TEST.TXT;1Y8 %BACKUP-E-VERIFYERR, verification error for block 208 of% DSA501:[ET_KATALOG]EXP_ART_TEST.TXT;1h8 %BACKUP-E-VERIFYERR, verification error for block 288 of% DSA501:[ET_KATALOG]EXP_ART_TEST.TXT;1tB %BACKUP-S-COMPARED, compared DSA501:[ET_KATALOG]EXP_ART_TEST.TXT;1      , On Wed, 14 Nov 2001 08:18:37 -0500, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:a   >re: problem with files: >r7 >have you tried backup/verify myfile.dat yourfile.dat ?a >eN >Also, try type/tail of the file. It *should* work on a variable length recordL >file , but if there is anything screwy with it, it will fail with a message" >such as unsupported organisation.   ------------------------------  % Date: Thu, 15 Nov 2001 07:56:11 +0100 2 From: Berger Harald <hotline@harryworld.dydns.org>= Subject: Re: VMS/RMS - destroyed sequential files - NEED HELPr8 Message-ID: <0hp6vt8f1hu67p4p6l6jh7jjqqkhkkfhbj@4ax.com>   sorry, but what in NPP, ??    E On 15 Nov 2001 05:54:17 +0800, Paul Repacholi <prep@prep.synonet.com>w wrote:  5 >Berger Harald <hotline@harryworld.dydns.org> writes:d >i@ >> I hope someone can help me by the following (stupid) problem. >rA >> The Error: When creating the files local on Node (A) -> all OKsB >> 	   When creating the files on node (B), single Disk, -> all OKA >> 	   When creating the files on node (B), shadow set (2 Disks),dA >> 	    -> following error will occour when i make a type xxx.txto > - >> %TYPE-W-READERR, error reading xxxxx.TXT;1-E >> -RMS-F-IRC, illegal record encountered; VBN or record number = 222p > 9 >> 	    -> on 3 other files i got the following message:	N > 0 >> %TYPE-F-WRITEERR, error writing SYS$OUTPUT:.;0 >> -RMS-F-SYS, QIO system service request failed, >> -SYSTEM-F-EXQUOTA, process quota exceeded, >            ^^^^^^^  ^^^^^^^^^^^^^^^^^^^^^^ > E >Have you looked at the quotas on both systems and checked them?  ThelF >SYSGEN quotas in particular. I suspect that the larger NPP use in the >DS driver is biting.o   ------------------------------  % Date: Thu, 15 Nov 2001 02:43:51 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca>6= Subject: Re: VMS/RMS - destroyed sequential files - NEED HELP , Message-ID: <3BF37237.B316D873@videotron.ca>  7 If backup generates errors, I suspect something smelly.   L If you have good/full control over the system, I would consider removing oneN drive from the showset and then testing the other drive (still in shadow set).J You could then mount the second drive independanly and test that one too. 0 Perhaps one drive is OK ad the other is corrupt.  L If you find that one drive is faulty and the other is good, you could modifyN that drive (to invalidate its membership to the shadow set), and then recreateK the shadow set with only the good drive and later on add the bad drive (yous might want to reformat it).O  J before you do that, you might want to do a DUMP/BLOCK of the file multipleA times and see if that generates a error at a consistent location.    ------------------------------  + Date: Thu, 15 Nov 2001 12:15:49 +0000 (UTC)- From: david20@alpha1.mdx.ac.uk= Subject: Re: VMS/RMS - destroyed sequential files - NEED HELPn+ Message-ID: <9t0blk$818$1@aquila.mdx.ac.uk>%  m In article <31p6vt4vkiq6qpt8cu6cc8lit1k7obl4bc@4ax.com>, Berger Harald <hotline@harryworld.dydns.org> writes:o6 >the next "funny" thing that has happend on the files: >h >..) a type /tail	is okn >..) a type /tail=60 is ok >..) a type /tail=80 is ok >..) a type /tail = 100 is ok ) >..) a type /tail = 150  following error:n? >%TYPE-W-READERR, error reading DSA501:<ET_KATALOG>EXP_TX.TXT;1 9 >-SYSTEM-E-UNSUPPORTED, unsupported operation or functione >e? >and now the "funny" thing. when now making a  type /tail = 100eD >i got the same error. but 5 minutes before the type/tail was OK.... >e >f  K Type/tail doesn't work particularly well. I much prefer the old tail public E domain program which seems to handle just about all sequential files.t     From the helpfile :h   "c  % !  TAIL,  Monitor last part of a file K !      This .HLP file written by:  Andy Harper, King's College London, 1991b !. 1 TAILH Prints the last screen of a file to the terminal. Also monitor a file to1 watch it as new lines are added and display them.e   Format:   "      $ TAIL [-|+size[rf]] filespec
 2 Examples      $ TAIL FRED.DAT6           Display the last page of the file "FRED.DAT"        $ TAIL -10 *.DAT =           Display the last 10 lines of each file that matcheso           "*.DAT".        $ TAIL -20f FRED.DAT C           Display the last 20 lines of the file "FRED.DAT" and thendA           go into a loop waiting 2 seconds and displaying any newi-           lines added to the end of the file.y
 2 FilespecA A wildcarded filename. The last page of each file that matches isu
 displayed.	 2 Optionsh. The options must be preceded by one of + or -.  @ The size value gives the number of lines. If preceded by `-', itH determines the number of lines from the end of the file to be displayed.@ If preceded by a `+', it determines the number of lines from theC beginning of the file to start displaying. The default is 15 lines.e  H The `r' option specified that lines from the file are to be displayed in reverse order.  H The `f' option specifies that, following display of the last part of theE file, a loop is entered to monitor the file and display any new linesa> that are added.  The file is monitored once every two seconds.   "e  7 If you can't find it elsewhere you can pick it up from    $ ftp://www.axp.mdx.ac.uk/vms/tail.zip    
 David Webb VMS and Unix team leader CCSS Middlesex University       >p >  > - >On Wed, 14 Nov 2001 08:18:37 -0500, JF Mezei & ><jfmezei.spamnot@videotron.ca> wrote: >i >>re: problem with files:m >>8 >>have you tried backup/verify myfile.dat yourfile.dat ? >>O >>Also, try type/tail of the file. It *should* work on a variable length recordpM >>file , but if there is anything screwy with it, it will fail with a messagec# >>such as unsupported organisation.  >.   ------------------------------  + Date: Thu, 15 Nov 2001 12:37:05 +0000 (UTC)  From: david20@alpha1.mdx.ac.uk= Subject: Re: VMS/RMS - destroyed sequential files - NEED HELPw+ Message-ID: <9t0cth$895$1@aquila.mdx.ac.uk>-  L In article <9t0blk$818$1@aquila.mdx.ac.uk>, david20@alpha1.mdx.ac.uk writes:n >In article <31p6vt4vkiq6qpt8cu6cc8lit1k7obl4bc@4ax.com>, Berger Harald <hotline@harryworld.dydns.org> writes:7 >>the next "funny" thing that has happend on the files:l >> >>..) a type /tail	is ok >>..) a type /tail=60 is ok. >>..) a type /tail=80 is oke >>..) a type /tail = 100 is ok* >>..) a type /tail = 150  following error:@ >>%TYPE-W-READERR, error reading DSA501:<ET_KATALOG>EXP_TX.TXT;1: >>-SYSTEM-E-UNSUPPORTED, unsupported operation or function >>@ >>and now the "funny" thing. when now making a  type /tail = 100E >>i got the same error. but 5 minutes before the type/tail was OK....  >> >> >sL >Type/tail doesn't work particularly well. I much prefer the old tail publicF >domain program which seems to handle just about all sequential files. >.< Just reread this and saw that I hadn't made clear that this 8 -SYSTEM-E-UNSUPPORTED, unsupported operation or functionH occurs fairly frequently with type/tail on files which are perfectly ok.4 One cause appears to be if the file is fairly large.P Hence I can quite believe that a type/tail = 100 was OK and then failed a little later on a growing file.  
 David Webb VMS and Unix team leader CCSS Middlesex University       >e >From the helpfile : >d >" >.& >!  TAIL,  Monitor last part of a fileL >!      This .HLP file written by:  Andy Harper, King's College London, 1991 >! >1 TAIL5I >Prints the last screen of a file to the terminal. Also monitor a file to:2 >watch it as new lines are added and display them. >s >Format: >D# >     $ TAIL [-|+size[rf]] filespec  >2 Examplesn >     $ TAIL FRED.DATe7 >          Display the last page of the file "FRED.DAT"t >a >     $ TAIL -10 *.DAT> >          Display the last 10 lines of each file that matches >          "*.DAT".h >I >     $ TAIL -20f FRED.DATD >          Display the last 20 lines of the file "FRED.DAT" and thenB >          go into a loop waiting 2 seconds and displaying any new. >          lines added to the end of the file. >2 FilespeciB >A wildcarded filename. The last page of each file that matches is >displayed.p
 >2 Options/ >The options must be preceded by one of + or -.  > A >The size value gives the number of lines. If preceded by `-', itrI >determines the number of lines from the end of the file to be displayed. A >If preceded by a `+', it determines the number of lines from theeD >beginning of the file to start displaying. The default is 15 lines. >kI >The `r' option specified that lines from the file are to be displayed inm >reverse order.s >hI >The `f' option specifies that, following display of the last part of theAF >file, a loop is entered to monitor the file and display any new lines? >that are added.  The file is monitored once every two seconds.o >i >" >s8 >If you can't find it elsewhere you can pick it up from  >d% >ftp://www.axp.mdx.ac.uk/vms/tail.zip  >k >  >David Webbn >VMS and Unix team leadera >CCSSh >Middlesex University  >e >e >n >> >> >>. >>On Wed, 14 Nov 2001 08:18:37 -0500, JF Mezei' >><jfmezei.spamnot@videotron.ca> wrote:a >> >>>re: problem with files: >>>c9 >>>have you tried backup/verify myfile.dat yourfile.dat ?  >>>iP >>>Also, try type/tail of the file. It *should* work on a variable length recordN >>>file , but if there is anything screwy with it, it will fail with a message$ >>>such as unsupported organisation. >>   ------------------------------  % Date: Thu, 15 Nov 2001 14:03:05 +0100l2 From: Berger Harald <hotline@harryworld.dydns.org>= Subject: Re: VMS/RMS - destroyed sequential files - NEED HELP 8 Message-ID: <bue7vt0nk3d8na60e0fhi4qmjlni6s0skl@4ax.com>  B On Thu, 15 Nov 2001 12:37:05 +0000 (UTC), david20@alpha1.mdx.ac.uk wrote:  M >In article <9t0blk$818$1@aquila.mdx.ac.uk>, david20@alpha1.mdx.ac.uk writes:ao >>In article <31p6vt4vkiq6qpt8cu6cc8lit1k7obl4bc@4ax.com>, Berger Harald <hotline@harryworld.dydns.org> writes:i8 >>>the next "funny" thing that has happend on the files: >>>y >>>..) a type /tail	is oku >>>..) a type /tail=60 is ok >>>..) a type /tail=80 is ok >>>..) a type /tail = 100 is ok + >>>..) a type /tail = 150  following error:cA >>>%TYPE-W-READERR, error reading DSA501:<ET_KATALOG>EXP_TX.TXT;1s; >>>-SYSTEM-E-UNSUPPORTED, unsupported operation or functiona >>> A >>>and now the "funny" thing. when now making a  type /tail = 100rF >>>i got the same error. but 5 minutes before the type/tail was OK.... >>>h >>>r >>M >>Type/tail doesn't work particularly well. I much prefer the old tail publicfG >>domain program which seems to handle just about all sequential files.p >>= >Just reread this and saw that I hadn't made clear that this  9 >-SYSTEM-E-UNSUPPORTED, unsupported operation or functioneI >occurs fairly frequently with type/tail on files which are perfectly ok. 5 >One cause appears to be if the file is fairly large..Q >Hence I can quite believe that a type/tail = 100 was OK and then failed a little  >later on a growing file.e >e >David Webbs >VMS and Unix team leadere >CCSSc >Middlesex University  >s >  >a >> >>From the helpfile :e >> >>"e >>' >>!  TAIL,  Monitor last part of a fileiM >>!      This .HLP file written by:  Andy Harper, King's College London, 1991o >>!a >>1 TAILJ >>Prints the last screen of a file to the terminal. Also monitor a file to3 >>watch it as new lines are added and display them.n >>	 >>Format:t >>$ >>     $ TAIL [-|+size[rf]] filespec >>2 Examples >>     $ TAIL FRED.DAT8 >>          Display the last page of the file "FRED.DAT" >> >>     $ TAIL -10 *.DATh? >>          Display the last 10 lines of each file that matchesu >>          "*.DAT". >> >>     $ TAIL -20f FRED.DATyE >>          Display the last 20 lines of the file "FRED.DAT" and thenuC >>          go into a loop waiting 2 seconds and displaying any newa/ >>          lines added to the end of the file.h >>2 FilespecC >>A wildcarded filename. The last page of each file that matches ise >>displayed. >>2 Optionsr0 >>The options must be preceded by one of + or -. >>B >>The size value gives the number of lines. If preceded by `-', itJ >>determines the number of lines from the end of the file to be displayed.B >>If preceded by a `+', it determines the number of lines from theE >>beginning of the file to start displaying. The default is 15 lines.n >>J >>The `r' option specified that lines from the file are to be displayed in >>reverse order. >>J >>The `f' option specifies that, following display of the last part of theG >>file, a loop is entered to monitor the file and display any new lines-@ >>that are added.  The file is monitored once every two seconds. >> >>"e >>9 >>If you can't find it elsewhere you can pick it up from 0 >>& >>ftp://www.axp.mdx.ac.uk/vms/tail.zip >> >> >>David Webb >>VMS and Unix team leader >>CCSS >>Middlesex University >> >> >> >>>o >>>m >>>e/ >>>On Wed, 14 Nov 2001 08:18:37 -0500, JF Mezein( >>><jfmezei.spamnot@videotron.ca> wrote: >>>  >>>>re: problem with files:o >>>>: >>>>have you tried backup/verify myfile.dat yourfile.dat ? >>>>Q >>>>Also, try type/tail of the file. It *should* work on a variable length record O >>>>file , but if there is anything screwy with it, it will fail with a message1% >>>>such as unsupported organisation.r >>>o  F first of all .. thanks for spending your time.... im really a VAX/VMSC fan.. and nothing is better than a Vax, and ive done the test witht9 the /tail=100 once more..... but the result was the same..  4 im not joking... i only tell you what ive seen...    ------------------------------  % Date: Thu, 15 Nov 2001 14:46:54 +0100n2 From: Berger Harald <hotline@harryworld.dydns.org>= Subject: Re: VMS/RMS - destroyed sequential files - NEED HELP 8 Message-ID: <6ah7vtsmnvu1mcbv9kq9kjton3sms61kn1@4ax.com>   Heres the 1 part of the test:  B ive removed (dismounted) one Drive and startet my little Programm* once more on the Single Member Shadow-Set.  4 a type of the file(s) (ex. type exp_art.txt) is ok..  * a type *.txt  produces the QIO-error !!!??  . (ive checked several times that this is a Vax# and not a Windows-based System ;-))d  6 Tomorrow (2nd part of the test) i will check this with the other disc...t  < but i think that theres a problem with a sysgen parameter... and the hardware is ok..                  , On Thu, 15 Nov 2001 02:43:51 -0500, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:o  8 >If backup generates errors, I suspect something smelly. >eM >If you have good/full control over the system, I would consider removing one O >drive from the showset and then testing the other drive (still in shadow set).aK >You could then mount the second drive independanly and test that one too.  1 >Perhaps one drive is OK ad the other is corrupt.E >RM >If you find that one drive is faulty and the other is good, you could modifyoO >that drive (to invalidate its membership to the shadow set), and then recreate/L >the shadow set with only the good drive and later on add the bad drive (you >might want to reformat it). > K >before you do that, you might want to do a DUMP/BLOCK of the file multiplelB >times and see if that generates a error at a consistent location.   ------------------------------  % Date: Thu, 15 Nov 2001 09:10:57 -0500b- From: JF Mezei <jfmezei.spamnot@videotron.ca>i= Subject: Re: VMS/RMS - destroyed sequential files - NEED HELPo, Message-ID: <3BF3CCEF.4C16110F@videotron.ca>   ok another twist:1  N have you tried edit/tpu and etit/edt (then "change", then scroll down) for the file ?  O TPU reads the whole file into memory as you start. EDT reads it as you need it.f  6 Out of curiosity, have you done ANA/RMS for the file ?   ------------------------------  % Date: Thu, 15 Nov 2001 09:40:30 -0600_+ From: Christopher Smith <csmith@amdocs.com>e# Subject: RE: Wanted: Small MicroVAX,L Message-ID: <3B55D7F383B0D31197D9009027541CBF1170DE58@cmiexch1.cmi.itds.com>   > -----Original Message-----: > From: Robert DiRosario [mailto:rdirosario@starpower.net]  @ > A VS2000 might be a poor choice.  It is slow, can only take a  > maximum ofD > 14 Megs of RAM, using a DEC memory board, and can only hold one RDF > (MFM) disk drive.  One additional RD drive can be connected using anH > external expansion box.  The largest drive it will support is an RD54,@ > about 160 Megs, and quite slow compared to newer SCSI drives.   J Actually, they will accept two half-space drives internally.  Having takenF several apart, I see no reason that -- given the proper cable -- these/ couldn't both be half-height "MFM" type drives.y  B Of course, the RD54 being the largest "MFM" drive made, AFAIR, twoG half-height drives would likely not solve the capacity problem.  Also IaL believe the 160M number is unformatted, and the useable space is, of course, less.,  A On the other hand, see the comment about the SCSI driver below....   >  The onlyjH > load/backup device it will take is a TK50 tape drive.  There is NO VMSF > support for any type of SCSI devices.  You would need to find VMS onH > TK50 tapes or find someone to copy standalone backup  and the VMS save > sets onto tape.e  0 I just happen to have most of VMS 5.5 on tk50 ;)  F .. and to add to what you say, Wolfgang Moeller (I hope I spelled thatK right) has written a SCSI driver for the machine, which should allow you to I attach disks, tapes, and perhaps even cd-roms.  I have meant to try it ongJ one of mine for a while.  Now that I have one with a floppy I can probably test it out relatively easily.  J The basic deal with the SCSI driver, IIRC, is that it's in two parts.  YouK need a firmware "patch," which can either be burned into the 2000's rom, oreL booted from a floppy or the like.  You also need a VMS driver, which must be6 added to the configuration of VMS that you're using.    I I suppose that getting VMS to actually boot from a SCSI device would be a H two step process, then, of bringing the machine up with a supported bootK device, and copying the driver into the VMS setup (or perhaps using anotherrE VMS machine...), and then booting properly with the altered firmware.-   Regards,   Chris   ! Christopher Smith, Perl Developerz Amdocs - Champaign, IL   /usr/bin/perl -e '? print((~"\x95\xc4\xe3"^"Just Another Perl Hacker.")."\x08!\n");  'F  3   ------------------------------    Date: 15 Nov 2001 06:49:48 -0800# From: pmartyn@csc.com (Phil Martyn)l9 Subject: Re: Web Connecting from an Alpha (using TCPWARE)r= Message-ID: <23088ca1.0111150649.7922a65f@posting.google.com>    OK, thanks for the assist.   ------------------------------   Date: 15 Nov 2001 09:56:20 GMT) From: leslie@clio.rice.edu (Jerry Leslie) * Subject: Welcome To Microsoft's IT Academy' Message-ID: <9t03g4$ime$1@joe.rice.edu>i) Keywords: microsoft,understands,marketingd  3 Microsoft understands how to breed loyal minions...e  4    http://news.cnet.com/news/0-1003-200-7875365.htmlD    Microsoft expands tech training programs -  Tech News -  CNET.com  +    Microsoft expands tech training programsf    By Joe Wilcox    Staff Writer, CNET News.com#    November 14, 2001, 12:50 p.m. PT   I   "Microsoft on Wednesday added yet another way to recruit Windows users: F    A program to train students in high schools, vocational schools andG    two- and four-year colleges to work with the software maker's latest8    technology.  H    Dubbed IT Academy, the program is billed as a way to prepare studentsB    for careers in high tech. While the program is expected to helpI    schools train students for technology jobs, Microsoft also benefits byiF    adding yet another tool in its recruitment arsenal. As it does withB    software developer programs, which offer tools and training forI    creating applications that run on Windows operating systems, Microsoftg:    will be able to generate more interest in its products.  D    Microsoft's ambitious training program underscores the increasingH    importance of leveraging school programs to maintain market share andE    public awareness, analysts say. For years, Apple Computer used itsmH    near lock on system and software sales to elementary and high schools)    as way of keeping up interest in Macs.g    .    .    .H    But Microsoft does not provide training in competing or non-Microsoft    technologies.  @    Yes, they are just trained in Microsoft products, Carew said.I    "However, the schools do have the freedom to develop their curriculum, A    so there are soft skills in the program, so that it's not just     product-focused."  I    Microsoft is offering the program in two levels, costing either $1,500$C    or $5,000 annually. The program runs from Jan. 1 to Dec. 31. All G    accredited U.S. or Canadian high schools or colleges are eligible tolE    participate in the first program, but the higher-level one is onlyp     available to high schools..."    4 --Jerry Leslie     (my opinions are strictly my own)   ------------------------------   End of INFO-VAX 2001.636 ************************