1 INFO-VAX	Wed, 21 Nov 2001	Volume 2001 : Issue 647       Contents:/ RE: 3rd-party memory for AlphaServer DS systems < RE: analyze/disk crashes with message %system-f-accvio, acceH Re: analyze/disk crashes with message %system-f-accvio, access violationH Re: analyze/disk crashes with message %system-f-accvio, access violationH Re: analyze/disk crashes with message %system-f-accvio, access violation  Re: C-Kermit 8.0 Beta.04 for VMS  Re: C-Kermit 8.0 Beta.04 for VMS  Re: C-Kermit 8.0 Beta.04 for VMS/ DCL minute of the other day: WASD WEBserver map  DEC Supplied user account? Re: DEC Supplied user account? Re: F$GETQUI wildcard bug??  Re: F$GETQUI wildcard bug??  Re: F$GETQUI wildcard bug??  Re: F$GETQUI wildcard bug?? ) Intel Itanium versus AMD Hammer (& Alpha) - Re: Intel Itanium versus AMD Hammer (& Alpha) - Re: Intel Itanium versus AMD Hammer (& Alpha) 6 Re: IPF Calling Standard (was: ISV's and VMS on Intel)' 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? ' 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? ' Re: Is it a DEC C problem, or is it me? ' Re: Is it a DEC C problem, or is it me?  Re: Life After Alpha Re: Life After Alpha Re: Life After Alpha Re: Life After Alpha Re: Life After Alpha Re: Life After Alpha Re: Life After Alpha Re: Life After Alpha Mount remote filesystem?  NFS?" Re: Mount remote filesystem?  NFS?" Opposition to the HP/Compaq merger Re: RWAST/LAT port problem Re: Scripting / batch files?. Re: Server vs. Workstation:  You Make The Call. Re: Server vs. Workstation:  You Make The Call5 RE: Software to emulate someone sitting at a terminal 9 Re: Solaris: ready for prime time?  Keep your VMS system. 9 Re: Solaris: ready for prime time?  Keep your VMS system. 9 Re: Solaris: ready for prime time?  Keep your VMS system. & Re: Some merger issues (long - sorry!)& Re: Some merger issues (long - sorry!)@ TCP/IP services settings disappear after restarting the softwareD Re: TCP/IP services settings disappear after restarting the software% telnet very slow from pc/linux to vms ) Re: telnet very slow from pc/linux to vms ) Re: telnet very slow from pc/linux to vms ) Re: telnet very slow from pc/linux to vms + The DCL minute of the day: show known links   RE: TZ887 installation questions; RE: VMS on IBM power chip would make IBM No. 1 in high end!  Re: Xetra, on which OS?  You are invited to this  Re: You are invited to this  Re: You are invited to this ( Re: [TCPIP V5.1 ECO 3] Problems with DNS( Re: [TCPIP V5.1 ECO 3] Problems with DNS? Re: [VMS Motif] How to avoid starting the DECW$SERVER process ?   F ----------------------------------------------------------------------  % Date: Wed, 21 Nov 2001 15:26:43 +1100 6 From: "Arena, Steve" <Steve.Arena@pacificpower.com.au>8 Subject: RE: 3rd-party memory for AlphaServer DS systemsQ Message-ID: <D750FFBD4936D111842000805F15EFA4043A87DD@meppb1.pacificpower.com.au>    > "Arena, Steve" wrote:  > >   > > > -----Original Message----- > > > From: Arena, Steve, > > > Sent: Thursday, 15 November 2001 14:20 > > > To:   'Nic Clews' C > > > Subject:      RE: 3rd-party memory for AlphaServer DS systems  >    > Did the email fail? Curious. 	[-->]  Yeah  @ > > > I recently attended a MS Windows XP event which featured a > presentationE > > > from a well known CPU/motherboard manufacturer - lets call this 	 > company I > > > "I". During the break, I was talking with one of the technical( not  > sales)H > > > guys from "I". On the desk in front of us was a motherboard - just	 > sitting K > > > on the desk, no static precautions at all - the discussion got around  > toE > > > static. This guy from "I" assured me that static is no longer a 	 > problem J > > > despite all the warnings your read - "these warnings are just a hang > over > > > from the past".  > 8 > This event was tinged with more than a little fantasy.6 	[-->]  It was a sales event - Windows XP is for sale.  I > I would imagine you also believed all the "wonderful things they had to  > say about XP".C 	[-->]  No, but Windows XP is for sale and I need a new career. VMS D skills aren't in big demand in this part of the world. Windows XP isJ probably MS's best OS offering to date --- a bit like being tallest of theK seven dwarfs :) --- you could even almost call it multi-user multi-tasking.  	 E > It is possible that protection is built into such devices, but that C > would be the only way of saying this. Just remind yourself of the J > emphasis on the chip process, and the tiny, tiny sizes of the substratesH > they are pushing these days. Ask yourself why fusewire works. And thisI > is a nice formula to show the relationship between Voltage and current, G > Watts = Amps x Volts. With high voltage (static), you need next to no G > amps. I'm sure you're aware that static involves voltages starting at G > 10's of 1000's into several millions. The math(s) speaks for itself I  > think. > E > Humans can withstand quite high voltages, but it is current, in the I > order of milliamps, that kills, a couple of hundred across the heart is D > all it takes. Higher voltage is only needed to overcome the body's > natural resistance. B 	[-->]  Compaq have a very good technical description on their web siteE 	that includes typical voltages generated ( by moving you foot on the  floor for instance )    J > > > I still do the right thing when I install memory myself - but I know > that> > > > the memory has already been compromised before I get it. > C > Oh, kinda explains why I was having so much trouble with a memory J > upgrade I was applying to my PC recently. "The bad batch, handled by the > guy in nylon underwear".< 	[-->]  When you see how memory and CPU's are handled in theK distribution chain I'm surprised any of it works. I think it is true to say ) that internal protection has improved .    >   J > I suspect the guy you spoke to has as much of a grip on technical issues@ > than he does of real operating systems. This gives me no faith< > whatsoever. No wonder 'they' need an influx of good blood. > 4 	[-->]   You can only hope they make good use of it. > --  * > Regards, Nic Clews CSC Computer Sciences > nclews at csc dot com    ------------------------------  % Date: Tue, 20 Nov 2001 16:13:13 -0500 * From: WILLIAM WEBB <WWEBB1@email.usps.gov>E Subject: RE: analyze/disk crashes with message %system-f-accvio, acce - Message-ID: <0033000042258777000002L072*@MHS>   3 =0AEr, did you just add these disks to this system?   A I seem to recall an 8.something GB disk size limitation in VMS 5.   , This is from the hip so I could be mistaken.   WWWebb   -----Original Message-----/ From: Info-VAX-Request@Mvb.Saic.Com at INTERNET ( Sent: Tuesday, November 20, 2001 4:02 PMB To: Webb, William W Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNETE Subject: RE: analyze/disk crashes with message %system-f-accvio, acce     A What happens if you run analyze/disk without the /list qualifier?   
 Hans Vlems  H Dirk Van Bouchaute <Dirk.Van.Bouchaute@kender-thijssen.be> wrote in mes= sage7 news:67059dac.0111200859.490ec6d1@posting.google.com... 	 > Hello !  > @ > Has anybody seen those kind of problems when analyzing disks ?H > We get this command to crash every time we start it. On the other han= d, the disk" > seems to work without problems.. > B > The version OpenVms 5.5_2H4 and the disk is a RZ40 type, 9 Gbyte > : > Defragmenter DFO 2.5 is installed and might be involved. >  > The entire listing is: >  > $ ana /dis dka200: /list > Listing of index file on > 20-NOV-2001 12:23:54.71  > 0 > %ANALDISK-I-OPENQUOTA, error opening QUOTA.SYS$ > -SYSTEM-W-NOSUCHFILE, no such file$ > (00000001,00001,001)  INDEXF.SYS;1 > [1,4] $ > (00000002,00002,001)  BITMAP.SYS;1 > [1,4] $ > (00000003,00003,001)  BADBLK.SYS;1 > [1,4] $ > (00000004,00004,001)  000000.DIR;1 > [1,4] $ > (00000005,00005,001)  CORIMG.SYS;1 > [1,4] $ > (00000006,00006,001)  VOLSET.SYS;1 > [1,4] $ > (00000007,00007,001)  CONTIN.SYS;1 > [1,4] $ > (00000008,00008,001)  BACKUP.SYS;1 > [1,4] $ > (00000009,00009,001)  BADLOG.SYS;1 > [1,4] $ > (00000010,00001,001)  DZ_ARC.DIR;1 > [1,4] % > (00000011,00002,001)  ALM_BCK.DIR;1 
 > [130,77]% > (00000012,00002,001)  DCF_BCK.DIR;1 
 > [130,77]% > (00000013,00002,001)  HIS_BCK.DIR;1 
 > [130,77]% > (00000014,00002,001)  TLX_BCK.DIR;1 
 > [130,77]: > (00000015,00001,001)  ALM_LGF_00_10_11_TO_00_10_12.DAT;1
 > [130,77]: > (00000016,00001,001)  ALM_LGF_00_10_13_TO_00_10_15.DAT;1 > ...............  > ............... 
 > [130,77]: > (00000256,00001,001)  ALM_LGF_01_10_28_TO_01_10_28.DAT;1
 > [130,77]: > (00000257,00001,001)  ALM_LGF_01_10_28_TO_01_10_28.DAT;2
 > [130,77]: > (00000258,00001,001)  ALM_LGF_01_10_30_TO_01_10_30.DAT;1
 > [130,77]? > %SYSTEM-F-ACCVIO, access violation, reason mask=3D01, virtual 3 > address=3D57900358, PC=3D80000010, PSL=3D03C00004  > 4 >   Improperly handled condition, image exit forced. > 6 >         Signal arguments              Stack contents > 5 >         Number =3D 00000005                803CB440 5 >         Name   =3D 0000000C                00000002 3 >                  00000001                7FEA8F64 3 >                  57900358                7FEA8F4C 3 >                  80000010                00000004 3 >                  03C00004                7FEA91A0 3 >                                          00000003 3 >                                          00000007 3 >                                          0000000E 3 >                                          09000001  >  >         Register dump  > H >         R0 =3D 03C00000  R1 =3D 57900358  R2 =3D 00002000  R3 =3D 000= 00180 H >         R4 =3D 00000000  R5 =3D 00000000  R6 =3D 0073C540  R7 =3D 007= 3C5AA H >         R8 =3D 00000180  R9 =3D 0073C740  R10=3D 00741E4C  R11=3D 000= 00000 H >         AP =3D 7FEA8F00  FP =3D 7FEA8EC0  SP =3D 7FEA8F3C  PC =3D 800= 00010  >         PSL=3D 03C00004  >  >  > All advice is welcome !  >  > Many thanks in advance >  > Dirk Van Bouchaute > Software Support Engineer  > Kender Thijssen  > ' > Dirk.Van.Bouchaute@Kender-Thijssen.be  >  > phone : +32-52-33.01.60  >  > int ref: disbru.53488=   ------------------------------  % Date: Tue, 20 Nov 2001 21:19:00 +0100 , From: Didier Morandi <Didier.Morandi@gmx.ch>Q Subject: Re: analyze/disk crashes with message %system-f-accvio, access violation & Message-ID: <3BFABAB3.261BA8D0@gmx.ch>   Dirk Van Bouchaute wrote:  > 	 > Hello !  > @ > Has anybody seen those kind of problems when analyzing disks ?R > We get this command to crash every time we start it. On the other hand, the disk" > seems to work without problems.. > B > The version OpenVms 5.5_2H4 and the disk is a RZ40 type, 9 Gbyte > : > Defragmenter DFO 2.5 is installed and might be involved. >  > The entire listing is: >  > $ ana /dis dka200: /list   What happens if you  $ ana/disk/rep/con dka200: ?    D. --  G   --------------------------------------------------------------------- E MORANDI Consulting.  WEB: http://Didier.Morandi.Free.fr/index_us.html E Pflanzschulstrasse 53, 8004 Zurich, Switzerland. GSM: +41 79 705 4670 / 19, chemin de la Butte, 31400 Toulouse, France.   I Disaster Recovery Plans, Computer Security Audits, DEC OpenVMS Expertise H On parle franais, Man spricht Deutsch, Habla Castellano, English spoken   ------------------------------  # Date: Tue, 20 Nov 2001 20:10:38 GMT " From: "Hans Vlems" <hvlems@iae.nl>Q Subject: Re: analyze/disk crashes with message %system-f-accvio, access violation / Message-ID: <2NyK7.411$Dp.1283@typhoon.bart.nl>   A What happens if you run analyze/disk without the /list qualifier?   
 Hans Vlems  K Dirk Van Bouchaute <Dirk.Van.Bouchaute@kender-thijssen.be> wrote in message 7 news:67059dac.0111200859.490ec6d1@posting.google.com... 	 > Hello !  > @ > Has anybody seen those kind of problems when analyzing disks ?I > We get this command to crash every time we start it. On the other hand,  the disk" > seems to work without problems.. > B > The version OpenVms 5.5_2H4 and the disk is a RZ40 type, 9 Gbyte > : > Defragmenter DFO 2.5 is installed and might be involved. >  > The entire listing is: >  > $ ana /dis dka200: /list > Listing of index file on > 20-NOV-2001 12:23:54.71  > 0 > %ANALDISK-I-OPENQUOTA, error opening QUOTA.SYS$ > -SYSTEM-W-NOSUCHFILE, no such file$ > (00000001,00001,001)  INDEXF.SYS;1 > [1,4] $ > (00000002,00002,001)  BITMAP.SYS;1 > [1,4] $ > (00000003,00003,001)  BADBLK.SYS;1 > [1,4] $ > (00000004,00004,001)  000000.DIR;1 > [1,4] $ > (00000005,00005,001)  CORIMG.SYS;1 > [1,4] $ > (00000006,00006,001)  VOLSET.SYS;1 > [1,4] $ > (00000007,00007,001)  CONTIN.SYS;1 > [1,4] $ > (00000008,00008,001)  BACKUP.SYS;1 > [1,4] $ > (00000009,00009,001)  BADLOG.SYS;1 > [1,4] $ > (00000010,00001,001)  DZ_ARC.DIR;1 > [1,4] % > (00000011,00002,001)  ALM_BCK.DIR;1 
 > [130,77]% > (00000012,00002,001)  DCF_BCK.DIR;1 
 > [130,77]% > (00000013,00002,001)  HIS_BCK.DIR;1 
 > [130,77]% > (00000014,00002,001)  TLX_BCK.DIR;1 
 > [130,77]: > (00000015,00001,001)  ALM_LGF_00_10_11_TO_00_10_12.DAT;1
 > [130,77]: > (00000016,00001,001)  ALM_LGF_00_10_13_TO_00_10_15.DAT;1 > ...............  > ............... 
 > [130,77]: > (00000256,00001,001)  ALM_LGF_01_10_28_TO_01_10_28.DAT;1
 > [130,77]: > (00000257,00001,001)  ALM_LGF_01_10_28_TO_01_10_28.DAT;2
 > [130,77]: > (00000258,00001,001)  ALM_LGF_01_10_30_TO_01_10_30.DAT;1
 > [130,77]= > %SYSTEM-F-ACCVIO, access violation, reason mask=01, virtual - > address=57900358, PC=80000010, PSL=03C00004  > 4 >   Improperly handled condition, image exit forced. > 6 >         Signal arguments              Stack contents > 3 >         Number = 00000005                803CB440 3 >         Name   = 0000000C                00000002 3 >                  00000001                7FEA8F64 3 >                  57900358                7FEA8F4C 3 >                  80000010                00000004 3 >                  03C00004                7FEA91A0 3 >                                          00000003 3 >                                          00000007 3 >                                          0000000E 3 >                                          09000001  >  >         Register dump  > D >         R0 = 03C00000  R1 = 57900358  R2 = 00002000  R3 = 00000180D >         R4 = 00000000  R5 = 00000000  R6 = 0073C540  R7 = 0073C5AAD >         R8 = 00000180  R9 = 0073C740  R10= 00741E4C  R11= 00000000D >         AP = 7FEA8F00  FP = 7FEA8EC0  SP = 7FEA8F3C  PC = 80000010 >         PSL= 03C00004  >  >  > All advice is welcome !  >  > Many thanks in advance >  > Dirk Van Bouchaute > Software Support Engineer  > Kender Thijssen  > ' > Dirk.Van.Bouchaute@Kender-Thijssen.be  >  > phone : +32-52-33.01.60  >  > int ref: disbru.53488    ------------------------------  % Date: Tue, 20 Nov 2001 21:35:31 +0100 , From: Didier Morandi <Didier.Morandi@gmx.ch>Q Subject: Re: analyze/disk crashes with message %system-f-accvio, access violation % Message-ID: <3BFABE93.BC7CF8A@gmx.ch>    Hans Vlems wrote:  > C > What happens if you run analyze/disk without the /list qualifier?    you beat me by 9 minutes... :-(    D. --  G   --------------------------------------------------------------------- E MORANDI Consulting.  WEB: http://Didier.Morandi.Free.fr/index_us.html E Pflanzschulstrasse 53, 8004 Zurich, Switzerland. GSM: +41 79 705 4670 / 19, chemin de la Butte, 31400 Toulouse, France.   I Disaster Recovery Plans, Computer Security Audits, DEC OpenVMS Expertise H On parle franais, Man spricht Deutsch, Habla Castellano, English spoken   ------------------------------  % Date: Tue, 20 Nov 2001 21:20:10 +0100 , From: Didier Morandi <Didier.Morandi@gmx.ch>) Subject: Re: C-Kermit 8.0 Beta.04 for VMS & Message-ID: <3BFABAF9.59127F1E@gmx.ch>   Frank da Cruz wrote: > + > C-Kermit 8.0 is almost ready for release.    Frank,  C I would love to be here when you celebrate your 50th year of KERMIT  programming :-)    D. --  G   ---------------------------------------------------------------------OE MORANDI Consulting.  WEB: http://Didier.Morandi.Free.fr/index_us.html-E Pflanzschulstrasse 53, 8004 Zurich, Switzerland. GSM: +41 79 705 4670 / 19, chemin de la Butte, 31400 Toulouse, France.c  I Disaster Recovery Plans, Computer Security Audits, DEC OpenVMS ExpertisehH On parle franais, Man spricht Deutsch, Habla Castellano, English spoken   ------------------------------  % Date: Tue, 20 Nov 2001 20:54:42 +00004% From: Alan Greig <a.greig@virgin.net>V) Subject: Re: C-Kermit 8.0 Beta.04 for VMSi* Message-ID: <3BFAC311.9881194E@virgin.net>   Didier Morandi wrote:?   > Frank da Cruz wrote: > >?- > > C-Kermit 8.0 is almost ready for release.l >r > Frank, > E > I would love to be here when you celebrate your 50th year of KERMITe > programming :-)A >a  H What amazes me is that Frank has just provided a new release of the veryD first Kermit. Kermit-20 for TOPS-20. Updated to add support for longK packets to help transferring files in and out of TOPS-20 under an emulator.,   Now that's support!i   >E > D. > --I >   --------------------------------------------------------------------- G > MORANDI Consulting.  WEB: http://Didier.Morandi.Free.fr/index_us.htmliG > Pflanzschulstrasse 53, 8004 Zurich, Switzerland. GSM: +41 79 705 4670 1 > 19, chemin de la Butte, 31400 Toulouse, France.i > J > Disaster Recovery Plans, Computer Security Audits, DEC OpenVMS Expertise   --
 Alan Greig   ------------------------------  # Date: Wed, 21 Nov 2001 00:02:37 GMTA4 From: "Phillip D. Williams" <phdevaxvms@hotmail.com>) Subject: Re: C-Kermit 8.0 Beta.04 for VMSa= Message-ID: <xaCK7.162726$ez.20977127@news1.rdc1.nj.home.com>n  " How about a new manual for Kermit? PhillipP2 "Alan Greig" <a.greig@virgin.net> wrote in message$ news:3BFAC311.9881194E@virgin.net... >s >i > Didier Morandi wrote:e >e > > Frank da Cruz wrote: > > >h/ > > > C-Kermit 8.0 is almost ready for release.t > >r
 > > Frank, > >iG > > I would love to be here when you celebrate your 50th year of KERMITp > > programming :-)  > >y >mJ > What amazes me is that Frank has just provided a new release of the veryF > first Kermit. Kermit-20 for TOPS-20. Updated to add support for longC > packets to help transferring files in and out of TOPS-20 under an/	 emulator.  >t > Now that's support!s >t > >h > > D. > > --K > >   --------------------------------------------------------------------- I > > MORANDI Consulting.  WEB: http://Didier.Morandi.Free.fr/index_us.htmllI > > Pflanzschulstrasse 53, 8004 Zurich, Switzerland. GSM: +41 79 705 4670l3 > > 19, chemin de la Butte, 31400 Toulouse, France.e > >NL > > Disaster Recovery Plans, Computer Security Audits, DEC OpenVMS Expertise >r > -- > Alan Greig >  >    ------------------------------  % Date: Tue, 20 Nov 2001 21:53:40 +0100V, From: Didier Morandi <Didier.Morandi@gmx.ch>8 Subject: DCL minute of the other day: WASD WEBserver map& Message-ID: <3BFAC2D4.A75DBE45@gmx.ch>   $!+- $! WEBDOWEGO.COM5 $! A tool to build a map of an OpenVMS/WASD WEB site. . $! (WASD is here: http://wasd.vsm.com.au/wasd) $!
 $! <input>M $! The full filename specification of the HTML home page of the site to scan.AK $! If a particular HTML page specification is provided, all downwards links 8 $! will be processed, without any attempt to go upwards. $! $! <output>yH $! As of version 1, only text display, which can be redirected that way:? $! $ @WEBDOWEGO/output=result.txt WEB$GENERAL:[000000]INDEX.HTMm $! $! <side effects>- $! None. $! $! Revision historye $!( $! Version Date        Author     actionO $! ------- ----------- ---------- --------------------------------------------- * $! V1.0-0  27-sep-2001 D. Morandi creation $!-t& $ set noon						!disable error handler $!+e $! Initialization  $!-O $ say    = "write sys$output" & $ delete = "delete"					!genuine value& $ search = "search"					!genuine valueP $ dotted = "===================================================================" $!+t $! Ask for input. 4 $! Example: @WEBDOWEGO WEB$GENERAL:[000000]INDEX.HTM $!-sH $ if p1 .eqs. "" then inq p1 "HTML source file to scan (RETURN to exit)" $ if p1 .eqs. "" then exit $!+t& $! If file not found, say so and exit. $!-WD $ if f$search(p1) .eqs. "" then exit "%X00018292"	!exit with DCL msg8 $ device  = f$parse(p1,,,"device")			!where is the file?< $ dirspec = f$parse(p1,,,"directory")			!in which directory?7 $ dev_dir = device + dirspec				!full spec without namen2 $ my_pid = f$getjpi(0,"pid")				!to avoid mixing..@ $ temp = "sys$login:search_''my_pid'.temp"		!  ..temporary filesF $ if f$search(temp) .nes. "" then delete 'temp';*	!some cleaning first $!+tI $! Now we scan the HTML file for HREF entries with the SEARCH utility andyL $! store the result in a temporary file which will be automatically created.J $! The /NUMBER qualifier will provide the line number where the occurrence6 $! has been found for (further) documentation purpose. $!-o5 $ search/number 'p1' "href=","href =" /output='temp'	t $ say ""						!one blank linet, $ say p1," local hypertext links"				!header< $ say f$extract(0,f$len(p1),dotted),"======================", $ close/nolog ch					!make sure channel free. $ open/read ch 'temp'					!open resulting file $LOOP:% $ read/end=EOF ch line					!read datai9 $ line = f$edit(line,"trim,upcase,compress")	!format datap/ $ gosub GET_LINE_NUMBER					!obvious, isn't it?s1 $ gosub EXTRACT_DATA					!select only HREF stringy $!+ @ $! Here we skip non interesting records from the resulting file. $!-t, $ if real_href_data .eqs. ""  then goto LOOP, $ if real_href_data .eqs. " " then goto LOOP, $ if real_href_data .eqs. "/" then goto LOOPN $ if f$edit(f$extract(0,10,real_href_data),"upcase") .eqs. "JAVASCRIPT" then -      goto LOOPN $ if f$edit(f$extract(0,7,real_href_data),"upcase")  .eqs. "MAILTO:"    then -      goto LOOPN $ if f$edit(f$extract(0,7,real_href_data),"upcase")  .eqs. "HTTP://"    then -      goto LOOPN $ if f$edit(f$extract(0,7,real_href_data),"upcase")  .eqs. "/HTTPD/"    then -      goto LOOP $ say ""6 $ say "line ",nr,": ",real_href_data			!display result $!+iJ $! Now let's see what we have here. Is it another .HTML reference? Is it aB $! DCL script name to be executed? What is its physical path? etc. $!L $! 1. Do we have a directory specification? (a "/" somewhere in the string).- $!    (see comments in each subroutine code).  $!-t? $ if f$locate("/",real_href_data) .ne. f$length(real_href_data)e $ then						!yes, we do.; $    gosub CHECK_MAP_ENTRY			!is there an HTTPD MAP clause? < $    gosub CHECK_EXEC_ENTRY		!is there an HTTPD EXEC clause?5 $    gosub SUBSTITUTE				!now convert to VMS syntax..%9 $    gosub CHECK_INDEX			! ..and check for an INDEX.*HTM*L $ else. $    gosub CHECK_DOCUMENT			!is it a document? $ endif  $ goto LOOP					!done. $! $EOF: 
 $ close ch $ exitO $!-----------------------------------------------------------------------------O $GET_LINE_NUMBER:- $!+mH $! The SEARCH/NUMBER command gives (as an example) the following output: $!5 $! $ search/number web$general:[000000]index.htm href.E $! 74 href="/dsl_browse/dsl_browse/" onMouseOut="MM_swapImgRestore()"y $!M $! so it is easy to extract the line number using the space as the delimiter.aK $! If there is an error trying to get the line number, the value "error" issP $! sent back in local symbol "nr" to the program, otherwise we have the 4 digitsO $! formatted line number in it (we assume there will be no HTML files with mores $! than 9999 lines :-) $!-c6 $ line_nr = 'f$element(0," ",line)'			!get line number) $ nr = "error"							!should never appear G $ if line_nr .lt. 1000 then nr =   "0''line_nr'"	!pad number with "0".. I $ if line_nr .lt. 100  then nr =  "00''line_nr'"	!  ..for better format..eG $ if line_nr .lt. 10   then nr = "000''line_nr'"	!  ..without F$FAO :-)  $ return						!done.O $!-----------------------------------------------------------------------------E $EXTRACT_DATA: $!+05 $! In the previous example, the resulting line reads:0 $!E $! 74 href="/dsl_browse/dsl_browse/" onMouseOut="MM_swapImgRestore()"1 $!M $! We are only interested in the hyperlink, which is the string enclosed withGM $! double cotes just after the HREF= so let's extract this piece of code now.0( $! Local symbols used are the following: $!E $! 74 HREF="/DSL_BROWSE/DSL_BROWSE/" onMouseOut="MM_swapImgRestore()"1$ $!    ^    ^^___ real_href_data __^^0 $!    !    !____ cote1_pos         !__ cote2_pos $!    !_________ href_pos0 $!-18 $ href_pos  = f$locate("HREF",line)			!starting positionG $ href_data = f$extract(href_pos+4,f$length(line),line)	!hypertext data > $ cote1_pos = f$locate("""",href_data)			!double cote positionK $ new_href_data = f$extract(cote1_pos+1,f$length(line),href_data) !new dataTB $ cote2_pos = f$locate("""",new_href_data)		!closing cote positionJ $ real_href_data = f$extract(0,cote2_pos,new_href_data)	!extract only data $ returnO $!-----------------------------------------------------------------------------1 $CHECK_MAP_ENTRY:T $!+ F $! For the record, the MAP entry in the HTTPD$MAP.CONF file allows theL $! server to know that any location supplied on the HTTP:// command line mayF $! actually map a physical location in the system, either by mapping aD $! physical device/directory path, or a logical name pointing to it. $!D $! First we extract the first substring from the "/---/---/" string.; $! Example: line 0084: /DSL_UPLOAD/ means (if there is such F $! "MAP /DSL_UPLOAD/* /SOMEWHERE/*" entry in the file) that the stringJ $! DSL_UPLOAD has a correspondence in the HTTPD$MAP.CONF file, which valueI $! is the second parameter of the "MAP /DSL_UPLOAD/* /SOMEWHERE/*". So we F $! search for this occurrence and extract the second parameter. In ourB $! example, the line reads "MAP /DSL_UPLOAD/* /WEB$DSL_UPLOAD/*". L $! Then we have to translate the "WEB$DSL_UPLOAD" string to check if it is a' $! (generally concealed) logical name.   $!- A $ if f$locate("/",real_href_data) .eq. 0		!it is a root directory  $ then $    start_pos = 1 $ else $    start_pos = 0 $ endif4B $ entry = f$edit(f$element(start_pos,"/",real_href_data),"upcase")F $ map_temp = "sys$login:map_search_''my_pid'.temp"	!  ..temporary fileN $ if f$search(map_temp) .nes. "" then delete 'map_temp';* !some cleaning first6 $ def/user sys$output 'map_temp'			!get result in here4 $ def/user sys$error   nl:				!disable error displayK $ search ht_root:[local]httpd$map.conf "map /''entry'/*"! (see WASD doc :-)-' $ if $status .eq. 1					!found an entry0 $ then $    say ""rG $    say "WEBDOWEGO-I-MAP, ",entry," has a MAP entry in HTTPD$MAP.CONF"m $    close/nolog map_cha  $    open/read map_ch 'map_temp' $    read map_ch map_linei $    close map_cho3 $    value = f$element(2," ",map_line) - "/*" - "/"n# $    value = f$edit(value,"upcase") 9 $    say "                 ",entry," is mapped to ",valueh  $    if f$trnlnm(value) .nes. ""
 $    then : $       say "                 ",value," is a logical name"C $       say "                 ",value," value is ",f$trnlnm(value)  A $       filespec = "''value':[000000]INDEX.*HTM*"		!default valued $	found = f$search(filespec) $       if found .nes. ""d $       then" $	   say "Default page is  ",found $       else5 $          say "Any of files ",filespec," is MISSING"W6 $          dire_/prot/siz=all/dat 'value':[000000]*.*;
 $       endif,
 $    endif $ endifl0 $ gosub CHECK_SUBDIR				!is there a level below? $ returnO $!-----------------------------------------------------------------------------S $CHECK_EXEC_ENTRY: $!+aG $! For the record, the EXEC entry in the HTTPD$MAP.CONF file allows the- $! server to know that:a $!J $! 1. it should EXECute a script in a [.CGI_SCRIPT] subdirectory somewhereE $! 2. the location of the [.CGI_SCRIPT] is defined via a logical nameoI $! 3. there is a procedure to execute, which name is after the second "/"  $!D $! First we extract the first substring from the "/---/---/" string.E $! Example: line 0074: /DSL_BROWSE/DSL_BROWSE/ means (if there is an sF $! "EXEC /DSL_BROWSE/* /SOMEWHERE/*" entry in the file) that the firstJ $! DSL_BROWSE has a correspondence in the HTTPD$MAP.CONF file, which valueJ $! is the second parameter of the "EXEC /DSL_BROWSE/* /SOMEWHERE/*". So weF $! search for this occurrence and extract the second parameter. In ourC $! example, the line reads "EXEC /DSL_BROWSE/* /WEB$DSL_BROWSE/*". lL $! Then we have to translate the "WEB$DSL_BROWSE" string to check if it is a' $! (generally concealed) logical name.   $! $! Example::% $! line 0074: /dsl_browse/dsl_browse/rJ $!             ^          ^_________ script name (actually DSL_BROWSE.COM)? $!             !____________________ subdirectory specificationSE $! (I know, this name is stupid but I did not set this WEB server up)0 $!-0 $ search="search"1: $ entry = f$edit(f$element(1,"/",real_href_data),"upcase"). $ if f$search("a.a") .nes. "" then dele_ a.a;* $ def/user sys$output a.aT $ def/user sys$error nl:9 $ search ht_root:[local]httpd$map.conf "exec /''entry'/*", $ if $status .eq. 1  $ then $    say ""0J $    say "WEBDOWEGO-I-EXEC, ",entry," has an EXEC entry in HTTPD$MAP.CONF" $    close/nolog exec_ch $    open/read exec_ch a.a $    read exec_ch exec_line) $    close exec_ch4 $    value = f$element(2," ",exec_line) - "/*" - "/"# $    value = f$edit(value,"upcase")G: $    say "                  ",entry," is mapped to ",value  $    if f$trnlnm(value) .nes. ""
 $    then ; $       say "                  ",value," is a logical name"0D $       say "                  ",value," value is ",f$trnlnm(value) D $       procedure = f$edit(f$element(2,"/",real_href_data),"upcase")H $       filespec = "''value':[000000]''procedure'.COM"		!default to .COM $	found = f$search(filespec) $       if found .nes. ""r $       then# $	   say "Default script is ",foundm $       else/ $          say "Script ",filespec," is MISSING"i6 $          dire_/prot/siz=all/dat 'value':[000000]*.*;
 $       endif0
 $    endif $ endif30 $ gosub CHECK_SUBDIR				!is there a level below? $ returnO $!-----------------------------------------------------------------------------5 $CHECK_SUBDIR:? $ if f$locate("/",real_href_data) .ne. f$length(real_href_data)  $ then- $    subdir = f$element(0,"/",real_href_data) - $    file   = f$element(1,"/",real_href_data)02 $    fullpath = dev_dir - "]" + "." + subdir + "]" $    filespec = fullpath + file ) $    filespec = f$edit(filespec,"upcase") % $    subdir = f$edit(subdir,"upcase") # $    if f$search(filespec) .nes. "" 	 $    then0 $       say ""D $       say "WEBDOWEGO-I-SUB, ",subdir," is an OpenVMS subdirectory"2 $       say "                 located in ",dev_dir( $       say "Default page is  ",filespec
 $    endif $ endifP $ returnP $!------------------------------------------------------------------------------ $SUBSTITUTE:B $ if f$extract(0,7,real_href_data)  .eqs. "http://"    then returnB $ if f$extract(0,7,real_href_data)  .eqs. "mailto:"    then return$ $ pos = f$locate("/",real_href_data)9 $ if pos .eq. f$length(real_href_data) then return		!done-, $ if pos .eq. (f$length(real_href_data) - 1) $ then7 $!   there is a trailing "/" that we replace with a "]"c# $    real_href_data['pos',1] := "]"  $ else $    if pos .eq. 0	 $    theni? $!      the first character is a "/" that we replace with a "["o& $       real_href_data['pos',1] := "["	 $    elsetJ $!      there is a "/" which means a subdirectory, so we put a "." instead& $       real_href_data['pos',1] := "."
 $    endif $ endif-2 $ real_href_data = f$edit(real_href_data,"upcase") $ goto SUBSTITUTEeO $!----------------------------------------------------------------------------- 
 $CHECK_INDEX: 7 $ filespec = "WEB$GENERAL:''real_href_data'INDEX.*HTM*"l $ found = f$search(filespec) $ if found .nes. ""e $ then) $    rootdir = real_href_data - "[" - "]"a $    say ""l? $    say "WEBDOWEGO-I-PHY, ",rootdir," is an OpenVMS directory" ? $    say "                 located in ",f$trnlnm("web$general")>" $    say "Default page is  ",found $ endif  $ returnO $!-----------------------------------------------------------------------------t $CHECK_DOCUMENT:1 $ found = f$search("''dev_dir'''real_href_data'")l $ if found .nes. ""  $ then5 $    real_href_data = f$edit(real_href_data,"upcase")- $    say ""-< $    say "WEBDOWEGO-I-DOC, ",real_href_data," is a document"0 $    say "                 located in  ",dev_dir. $    say "                 fullspec is ",found $    eof = f$file(found,"EOF") $    alq = f$file(found,"ALQ") $    cdt = f$file(found,"CDT") $    rfm = f$file(found,"RFM"), $    say "                 created on  ",cdt, $    say "                 record fmt  ",rfm> $    say "                 file size   ",eof,"/",alq," blocks" $ endif0 $ return  E Works only with the WEB pages created for an OpenVMS WASD WEB Server.  Enjoy.   D. -- 3G   --------------------------------------------------------------------- E MORANDI Consulting.  WEB: http://Didier.Morandi.Free.fr/index_us.htmlnE Pflanzschulstrasse 53, 8004 Zurich, Switzerland. GSM: +41 79 705 4670K/ 19, chemin de la Butte, 31400 Toulouse, France.i  I Disaster Recovery Plans, Computer Security Audits, DEC OpenVMS ExpertisetH On parle franais, Man spricht Deutsch, Habla Castellano, English spoken   ------------------------------  % Date: Wed, 21 Nov 2001 12:20:04 +0800i5 From: Netsurfer <netsurfer@sentosa.singaporemail.com>-# Subject: DEC Supplied user account? 8 Message-ID: <6mamvt05oh1fsi041ej7ci027dd4vqisjl@4ax.com>  ? Will there be any negative impact if I remove account FIELD and.' DEFAULT from the OpenVMS user database?a        Regards,  	 Netsurferi        ====R For any personal email replies, please remove " sentosa. " from my E-mail address.   ------------------------------  % Date: Tue, 20 Nov 2001 23:42:41 -0500 2 From: rdeininger@mindspring.com (Robert Deininger)' Subject: Re: DEC Supplied user account?ML Message-ID: <rdeininger-2011012342410001@user-2iveakl.dialup.mindspring.com>  8 In article <6mamvt05oh1fsi041ej7ci027dd4vqisjl@4ax.com>,* netsurfer@sentosa.singaporemail.com wrote:  A > Will there be any negative impact if I remove account FIELD andn) > DEFAULT from the OpenVMS user database?0  I Leave DEFAULT alone.  It has something to do with filling in the defaultseI for new accounts that you create, IIRC.  I'm not sure AUTHORIZE will even  let you remove it.  G You can remove FIELD if you want, but I think it is more common to justmH set the DISUSER flag so the account can't be used.  If you ever need the  account, you can reset the flag.   -- - Robert Deininger rdeininger@mindspring.comp   ------------------------------    Date: 20 Nov 2001 13:28:05 -0800. From: SPAMSINK2001@YAHOO.COM (Alan E. Feldman)$ Subject: Re: F$GETQUI wildcard bug??= Message-ID: <343f30ae.0111201328.523046ff@posting.google.com>   p macarthur@encompasserve.org (Malcolm MacArthur) wrote in message news:<rnx4lOflI+TI@eisner.encompasserve.org>...V > In article <oQaknjI9qB50@eisner.encompasserve.org>, briggs@encompasserve.org writes:o > > In article <WZEoWmSE4Ci9@eisner.encompasserve.org>, macarthur@encompasserve.org (Malcolm MacArthur) writes:rO > >> I decided it might be a nice idea to differentiate between remote printersg [sinp]H > >> write a generic COM file using f$getqui to find queues with a givenQ > >> characteristic and execute a given command against them. But, issuing a SHOW!L > >> QUEUE command whilst in an F$GETQUI wildcard operation resets the queue? > >> context, putting me back to the first queue on the system.v > >> lO > >> Is this a bug in f$GETQUI? I've already figured out a way around it, but I  > >> shouldn't have to.r > >  [snip] > > D > > SHOW QUEUE uses SYS$GETQUI.  And destroys the context created by > > your F$GETQUI.   [snip]  TM > BTW, the work-around is to stuff all the queue names into a comma-separatedD6 > string, then work on them outwith the F$GETQUI call.  , You could also spawn the SHOW QUEUE command:   $ <commands using F$GETQUI>D $ SPAWN SHOW QUEUE  $ <more commands using F$GETQUI>  1 This works because it creates a separate process.=   -- Disclaimer: JMHO Alan E. Feldman	 afeldman&gfigroup.come   ------------------------------  % Date: Tue, 20 Nov 2001 22:52:56 -0000=3 From: "Malcolm" <malcolm@neverness.freeserve.co.uk>k$ Subject: Re: F$GETQUI wildcard bug??. Message-ID: <9temo6$k9r$1@news7.svr.pol.co.uk>  ; "Alan E. Feldman" <SPAMSINK2001@YAHOO.COM> wrote in message 7 news:343f30ae.0111201328.523046ff@posting.google.com... B > macarthur@encompasserve.org (Malcolm MacArthur) wrote in message/ news:<rnx4lOflI+TI@eisner.encompasserve.org>...e7 > > In article <oQaknjI9qB50@eisner.encompasserve.org>,i  briggs@encompasserve.org writes:9 > > > In article <WZEoWmSE4Ci9@eisner.encompasserve.org>, 7 macarthur@encompasserve.org (Malcolm MacArthur) writes:=H > > >> I decided it might be a nice idea to differentiate between remote printers > [sinp]J > > >> write a generic COM file using f$getqui to find queues with a givenL > > >> characteristic and execute a given command against them. But, issuing a SHOWH > > >> QUEUE command whilst in an F$GETQUI wildcard operation resets the queuenA > > >> context, putting me back to the first queue on the system.t > > >>K > > >> Is this a bug in f$GETQUI? I've already figured out a way around it," but Io > > >> shouldn't have to.l > > >p > [snip] > > >!F > > > SHOW QUEUE uses SYS$GETQUI.  And destroys the context created by > > > your F$GETQUI. >s > [snip] >r? > > BTW, the work-around is to stuff all the queue names into ad comma-separated	8 > > string, then work on them outwith the F$GETQUI call. >!. > You could also spawn the SHOW QUEUE command: >  > $ <commands using F$GETQUI>A > $ SPAWN SHOW QUEUE" > $ <more commands using F$GETQUI> >e3 > This works because it creates a separate process.r >_H Yeah, but I'm just interested in getting the queue name through F$GETQUIE and nothing else. Spawning would involve the overhead of creating ther processr  E What I am trying to do is to group print and batch queues using queuefG characteristics. I investigated multiple queue managers, but there is a J limit of five (why five?), which is too few for what I want to do. My nextI idea was to set each queue with a characteristic according to which groupa they were in. I couldTF then write a wrapper command file called WITH_CHAR, which would take aH characteristic name and a command and perform that command on each queue withI that characteristic. So I could do for instance, WITH_CHAR SITE1_PRINTERSp STOP/QUEUE/NEXT, and so on.   K A nice neat solution, and very easy to do. But I discovered that SHOW QUEUEs trips it up ;-)c  % Anyone think of a better alternative?(   -Malcolm MacArthur   > -- > Disclaimer: JMHO > Alan E. Feldman  > afeldman&gfigroup.comM   ------------------------------  % Date: Wed, 21 Nov 2001 11:03:36 +1030s% From: Jeremy Begg <jeremy@vsm.com.au>b$ Subject: Re: F$GETQUI wildcard bug??* Message-ID: <3BFAF660.3CC199F1@vsm.com.au>   Hi Malcolm,.  G > What I am trying to do is to group print and batch queues using queue I > characteristics. I investigated multiple queue managers, but there is a-L > limit of five (why five?), which is too few for what I want to do. My nextK > idea was to set each queue with a characteristic according to which groupu > they were in. I couldrH > then write a wrapper command file called WITH_CHAR, which would take aJ > characteristic name and a command and perform that command on each queue > withK > that characteristic. So I could do for instance, WITH_CHAR SITE1_PRINTERS  > STOP/QUEUE/NEXT, and so on.i > M > A nice neat solution, and very easy to do. But I discovered that SHOW QUEUEs > trips it up ;-)  > ' > Anyone think of a better alternative?   F The usual "alternative" is to incorporate the site name (or some otherI identifying information) in the queue name.  Of course, this is something C which is somewhat more difficult to do when your queues are already 7 established and your users know which queues are where!o   Regards,           Jeremy Beggt  =   +---------------------------------------------------------+n=   |            VSM Software Services Pty. Ltd.              |-=   |                 http://www.vsm.com.au/                  | =   |       "OpenVMS Systems Management & Programming"        |d=   |---------------------------------------------------------|!=   | P.O.Box 402, Walkerville, |  E-Mail:  jeremy@vsm.com.au |n=   | South Australia 5081      |   Phone:  +61 8 83592155    |r=   |---------------------------|  Mobile:  0414 422 947      |l=   |  A.C.N. 068 409 156       |     FAX:  +61 8 82231777    |_=   +---------------------------------------------------------+    ------------------------------   Date: 20 Nov 2001 23:02 CST_' From: carl@gerg.tamu.edu (Carl Perkins),$ Subject: Re: F$GETQUI wildcard bug??- Message-ID: <20NOV200123025959@gerg.tamu.edu>h  7 "Malcolm" <malcolm@neverness.freeserve.co.uk> writes...o< }"Alan E. Feldman" <SPAMSINK2001@YAHOO.COM> wrote in message8 }news:343f30ae.0111201328.523046ff@posting.google.com...C }> macarthur@encompasserve.org (Malcolm MacArthur) wrote in messaget0 }news:<rnx4lOflI+TI@eisner.encompasserve.org>...8 }> > In article <oQaknjI9qB50@eisner.encompasserve.org>,! }briggs@encompasserve.org writes: : }> > > In article <WZEoWmSE4Ci9@eisner.encompasserve.org>,8 }macarthur@encompasserve.org (Malcolm MacArthur) writes:I }> > >> I decided it might be a nice idea to differentiate between remote 	 }printersc	 }> [sinp]eK }> > >> write a generic COM file using f$getqui to find queues with a given M }> > >> characteristic and execute a given command against them. But, issuing- }a SHOWrI }> > >> QUEUE command whilst in an F$GETQUI wildcard operation resets the  }queueB }> > >> context, putting me back to the first queue on the system. }> > >>oL }> > >> Is this a bug in f$GETQUI? I've already figured out a way around it, }but I }> > >> shouldn't have to. }> > >	 }> [snip]s }> > >G }> > > SHOW QUEUE uses SYS$GETQUI.  And destroys the context created byh }> > > your F$GETQUI.U }>	 }> [snip]L }>@ }> > BTW, the work-around is to stuff all the queue names into a }comma-separated9 }> > string, then work on them outwith the F$GETQUI call.l }>/ }> You could also spawn the SHOW QUEUE command:t }> }> $ <commands using F$GETQUI> }> $ SPAWN SHOW QUEUE0# }> $ <more commands using F$GETQUI>t }>4 }> This works because it creates a separate process. }>I }Yeah, but I'm just interested in getting the queue name through F$GETQUI F }and nothing else. Spawning would involve the overhead of creating the }process } F }What I am trying to do is to group print and batch queues using queueH }characteristics. I investigated multiple queue managers, but there is aK }limit of five (why five?), which is too few for what I want to do. My next J }idea was to set each queue with a characteristic according to which group }they were in. I couldG }then write a wrapper command file called WITH_CHAR, which would take apI }characteristic name and a command and perform that command on each queuee }withuJ }that characteristic. So I could do for instance, WITH_CHAR SITE1_PRINTERS }STOP/QUEUE/NEXT, and so on. } L }A nice neat solution, and very easy to do. But I discovered that SHOW QUEUE }trips it up ;-) } & }Anyone think of a better alternative? }  }-Malcolm MacArthuru }  }> --I }> Disclaimer: JMHO  }> Alan E. Feldman }> afeldman&gfigroup.com  I A better alternative? How about not using SHOW QUEUE? Given what you said F you were trying to do, I don't see where the SHOW QUEUE comes into it.  F Anyhow, loop through the queues with the wildcarded F$GETQUI and buildJ a list. Then use the list of queues that match the criteria to do whateverH you want after you have looped through all the queues. F$ELEMENT is your friend.E  E Another alternative is to define logical names that list the printers D in each group. That saves you the time of scanning all the queues toE get the ones on the list since you already have the list defined. YounJ could even use cluster logicals for this, if you are on at least VMS V7.2.  . $ define/sys site1_queues queue3,queue4,queue6. $ define/sys site2_queues queue1,queue2,queue5  B And loop using F$TRNLNM to get each translation from the list. YouG can process each value as you pull it off, including doing a SHOW QUEUEyI before and/or after doing whatever it is you are doing to it, since therePG is no context to mess up other than the simple index pointer value thatL
 you maintain.W   --- Carl   ------------------------------  % Date: Tue, 20 Nov 2001 23:26:03 +0100  From: Dirk Munk <munk@home.nl>2 Subject: Intel Itanium versus AMD Hammer (& Alpha)' Message-ID: <3BFAD87B.F27BE65C@home.nl>g  E In the German magazine CT I've read some very interesting stuff abouta AMD's new Hammer CPU.d  G As you will know it is a 64 bit successor to the present Athlon series,cE and it has a excellent 32 bit x86 emulation, contrary to the Itanium.r  0 But look at this, expected SPECint 2000 figures:   Itanium @ 0.8 GHz -> 380 Alpha EV7 @ 1.2 GHz -> 800 AMD Hammer @ 2 GHz -> 1400 (!!)t  = It has a integrated memory controller with a bandwith of 5.33h Gigabyte/sec..E The best thing however is that it seems it will have 3 HyperTransport[E channels for interprocessor communication. One HyperTransport channel @ has a bandwith of 6.4 GigaByte/sec. It all reminds me of the EV7
 design....  ) Maybe we should ask for VMS on Hammer ???e   ------------------------------  % Date: Tue, 20 Nov 2001 17:40:59 -0500v- From: JF Mezei <jfmezei.spamnot@videotron.ca> 6 Subject: Re: Intel Itanium versus AMD Hammer (& Alpha), Message-ID: <3BFADBF2.814BEE83@videotron.ca>   Dirk Munk wrote:+ > Maybe we should ask for VMS on Hammer ???   8 No, VMS on Power is far better. It has a serious backer.  7 Besides, I could boot VMS on a MAC :-) :-) :-) :-) :-) c  L Ironic that Apple should have gone Alpha instead of PowerPC, but now, it may9 turn out that VMS may join Apple's decision to go Power !     R I still think that Digital and Apple should have merged, creating a Digital Apple.I VMS for servers, and MACs for desktops. It woudl have ruled the world andiE killed Microsoft before MS would have had a chance to get off of DOS.0   ------------------------------  # Date: Wed, 21 Nov 2001 01:19:25 GMT * From: "Bill Todd" <billtodd@metrocast.net>6 Subject: Re: Intel Itanium versus AMD Hammer (& Alpha)@ Message-ID: <xiDK7.59493$dk.4349524@bin1.nnrp.aus1.giganews.com>  K Dirk Munk <munk@home.nl> wrote in message news:3BFAD87B.F27BE65C@home.nl...$   ...(  + > Maybe we should ask for VMS on Hammer ???   I If Hammer lives up to just most of its potential it could become the most I important processor of this decade (and sink Itanic in the process).  The K handicap of the legacy x86 cruft appears to be far less burdensome than theEI EPIC excursion is for Itanic, and (unlike EPIC) it has a major up-side in F allowing Hammer to be used in current high-volume boxes (using currentA software) as well as in more specialized higher-end environments.-  H Also unlike Itanic, Hammer indeed *does* have a great deal of Alpha-likeH technology inside, and could easily (again, unlike Itanic) evolve into aJ very reasonable Alpha replacement (not that anyone was asking for one, butG if they have to have one it would be nice if it weren't such a definite. downgrade).(  E Of course, VMS on Hammer doesn't mesh well with Compaq's current "oneeI smoking brick of death to rule them all" rationale, so I wouldn't hold my.J breath unless current management gets the heave-ho (which brings to mind aI delightful picture of Qurly and his Qohorts soaring majestically over theaF walls of Qastle Qapellas - anyone know where we can find a Qatapult?).   - bill   ------------------------------  % Date: Tue, 20 Nov 2001 18:31:25 -0500a* From: John Reagan <john.reagan@compaq.com>? Subject: Re: IPF Calling Standard (was: ISV's and VMS on Intel)-) Message-ID: <3BFAE7CD.3020508@compaq.com>E  $ Brian Schenkenberger, VAXman- wrote:   > L > I see no register usage caveats in the calling standard description of the@ > Null Procedure.  Can you give a chapter and section reference? >     D Since Null Frame Procedures have no context and they can't allocate H stack space or even save/restore registers (unless you restrict unwinds E as noted in section 3.4.6).  In combination with table 3-1 that says -F registers R2-R15 are saved/restored if used, you can conclude that no A legal Null Frame procedure can conform to table 3-1 and modify a rI register at the same time.  Table 3-1 says if you want to modify R2-R15, rF you have to save/restore them. However, since a conforming Null Frame D cannot do that without a context, then no conforming Null Frame can D modify them.  As I mentioned there may indeed be tons of Null Frame H routines that don't conform to Chapter 3.  That is OK.  They can exist. I   They just aren't guaranteed to be usable in the general mixed-language  H programming environment.  I'll admit that the description of Null Frame F procedures is a little handwavying in places including the absolutely H vague sentence in 3.4.6 that reads "The circumstances ... are not fully F specified by this standard."  Kinda sounds like they are hinting that 6 most Null Frame procedures don't conform to Chapter 3.     -- M John Reagan4' Compaq Pascal/{A|I}MACRO Project Leaderu   ------------------------------  % Date: Tue, 20 Nov 2001 13:35:38 -0600V( From: David Harrold <DHarrold@wi.rr.com>0 Subject: Re: Is it a DEC C problem, or is it me?8 Message-ID: <7ublvt8gk64l1gcq6rh8t5bi065tj7fj94@4ax.com>  K On Tue, 20 Nov 2001 18:13:24 -0000, wspencer@ap.nospam.org (Warren Spencer): wrote:  G >>Compaq about it NOW, through offical channels. Calmly explain to themoI >>how this will impact your work.  If you don't have an official channel,AJ >>then emailing the right folks at Compaq will probably do almost as well.I >>Bleating in comp.os.vms is likely NOT an effective way to get Compaq toE	 >>notice.s > K >Thank you, and in my case, I'm aware of that.  I also fill out the online rM >questionaires as they come up.  Would you happen have the email address for r? >the person at Compaq in charge of caring about customer needs?. >l  J Several years ago when the "How am I driving?  Call 1-800-xxx-xxxx" labelsL started appearing on trucks, I saw someone had created a bumper sticker thatK said, "How is my driving? Call 1-800-EAT-SHIT".  I think you could probablye try that number... :-)   Dave Harrold   ------------------------------  % Date: Tue, 20 Nov 2001 20:22:21 +0000 % From: Alan Greig <a.greig@virgin.net> 0 Subject: Re: Is it a DEC C problem, or is it me?* Message-ID: <3BFABB7D.3A5879C3@virgin.net>   ed.vogel@compaq.com wrote:  D > "    Larry's statement is almost correct.  We are porting Compaq CD >     to IA64 for OpenVMS.  We expect this product to support almostI >     all of the current language.  There will be some differences in theI@ >     area of asm's and probably things such as #pragma linkage. >g  N Could you triple check that and, assuming still correct, make sure that futureA presentations are corrected.  The information we were given as anIO "authoritative" list of compilers was quite clear that there would be no Compaq.P C. I made a joke that on the list of languages one appeared to have been removedN as there was an extra comma in the slide. Hope Compaq C did not slide down the& plughole and I really was just joking.     --
 Alan Greig   ------------------------------  % Date: Tue, 20 Nov 2001 15:46:06 -0500q From: <ed.vogel@compaq.com>a0 Subject: Re: Is it a DEC C problem, or is it me?3 Message-ID: <jizK7.1776$RL6.58641@news.cpqcorp.net>a  2 "Alan Greig" <a.greig@virgin.net> wrote in message$ news:3BFABB7D.3A5879C3@virgin.net... |e |  | ed.vogel@compaq.com wrote: |sF | > "    Larry's statement is almost correct.  We are porting Compaq CF | >     to IA64 for OpenVMS.  We expect this product to support almostK | >     all of the current language.  There will be some differences in the B | >     area of asm's and probably things such as #pragma linkage. | >e | I | Could you triple check that and, assuming still correct, make sure thata futureC | presentations are corrected.  The information we were given as anEJ | "authoritative" list of compilers was quite clear that there would be no CompaqJ | C. I made a joke that on the list of languages one appeared to have been removed-L | as there was an extra comma in the slide. Hope Compaq C did not slide down theo( | plughole and I really was just joking. |F |T | -- | Alan Greig |s |-	     Alan,$  G     With your help I've tracked down the person who made this statementoC     and have asked the C/C++ product manager to contact this person.!     and get the message straight.M  5                                                    Eds   ------------------------------  % Date: Tue, 20 Nov 2001 20:50:44 +0000v% From: Alan Greig <a.greig@virgin.net>e0 Subject: Re: Is it a DEC C problem, or is it me?* Message-ID: <3BFAC224.9529E0FE@virgin.net>   Alan Greig wrote:t   > ed.vogel@compaq.com wrote: >aF > > "    Larry's statement is almost correct.  We are porting Compaq CF > >     to IA64 for OpenVMS.  We expect this product to support almostK > >     all of the current language.  There will be some differences in theHB > >     area of asm's and probably things such as #pragma linkage. > >t > P > Could you triple check that and, assuming still correct, make sure that futureC > presentations are corrected.  The information we were given as ano >   T Should have added that Doug Williams was the Compaq guy presenting. Don't recall his, exact title but believe he is fairly senior.     > -- > Alan Greig   --
 Alan Greig   ------------------------------  % Date: Tue, 20 Nov 2001 16:05:01 -0500-- From: JF Mezei <jfmezei.spamnot@videotron.ca>t0 Subject: Re: Is it a DEC C problem, or is it me?, Message-ID: <3BFAC57B.E8812F14@videotron.ca>   Warren Spencer wrote:HF > >You guys (and anyone else facing similar difficulties) need to tell > K > I appreciate the advice, and the sentiment Robert, however I have no suchqJ > need.  Instead I believe it is Compaq that *needs* to either communicateI > clearly (stop fibbing) and poll their customers for their requirements.t  N Compaq has set a long term course for VMS.  You may refuse to read it, but theN writing is clearly on the wall. When a company ignores an operating system andM constantly reminds us that it will be focusing on industry standard stuff andtM rationalize product lines etc, it doesn't take einstein to draw a conclusion.d  M I suspect that as part of this long term strategy, Compaq knew that a certaintL proportion of customers would stop upgrading at one point. The lack of VAX-CK support will just freeze a certain proportion of customers a few years fromiJ now. The longer in the future this happens, the greater the odds that theyN will migrate to NT on Compaq.  (remember that Compaq is convinced that NT will> eventually be able to take over all of the tasks done by VMS).  H Compaq is fully aware that lack of VAX-C will result in the loss of someN customers. It is perfectly aware that the lack of marketing of VMS will resultM in the further erosion of the installed base and of course, no new customers.oG It is perfectly aware that the murder of Alpha will dry up sales of itsvL profitable products. It has calculated that those variables would cause lessN harm that  than the benefits that will accrue from being an ever more obedient Microsoft pet.   ------------------------------  % Date: Tue, 20 Nov 2001 20:57:40 +0000w% From: Alan Greig <a.greig@virgin.net>h0 Subject: Re: Is it a DEC C problem, or is it me?* Message-ID: <3BFAC3C3.9F84FD96@virgin.net>   ed.vogel@compaq.com wrote:   >     Alan,f >iI >     With your help I've tracked down the person who made this statement E >     and have asked the C/C++ product manager to contact this personU# >     and get the message straight.t >k  K Ok, thanks for the prompt action. I thought it was a very strange decision.e   > 7 >                                                    Ede   --
 Alan Greig   ------------------------------    Date: 20 Nov 2001 15:48:17 -0600- From: koehler@encompasserve.org (Bob Koehler) 0 Subject: Re: Is it a DEC C problem, or is it me?3 Message-ID: <vPF0aBFAqJbX@eisner.encompasserve.org>-  T In article <3BFA6C1F.26CA0D12@127.0.0.1>, Nic Clews <sendspamhere@127.0.0.1> writes:  H > Using a _common_ compiler on multiple platforms (including VMS) allowsF > code from the lesser worlds to be ported much more easily to the VMS > world.  D    True, but it breaks VMS' long standing excellent record in upward*    compatability.  Makes one long for DEC.    1   ------------------------------    Date: 20 Nov 2001 15:51:07 -0600- From: koehler@encompasserve.org (Bob Koehler)60 Subject: Re: Is it a DEC C problem, or is it me?3 Message-ID: <orcO3+L$rArS@eisner.encompasserve.org>-   In article <rdeininger-2011011156020001@user-2iveans.dialup.mindspring.com>, rdeininger@mindspring.com (Robert Deininger) writes:G >  > Replying to to posts...D > L > You guys (and anyone else facing similar difficulties) need to tell Compaq) > about it NOW, through offical channels.K  H    I have very cleary told appropriate persons in Compaq that I want DECF    heritage in the compilers, not Intel heritage.  Now I wait and see.   ------------------------------    Date: 20 Nov 2001 15:52:19 -0600- From: koehler@encompasserve.org (Bob Koehler)g0 Subject: Re: Is it a DEC C problem, or is it me?3 Message-ID: <ABhUVotZQuS1@eisner.encompasserve.org>c   In article <rdeininger-2011011156020001@user-2iveans.dialup.mindspring.com>, rdeininger@mindspring.com (Robert Deininger) writes:i > L > You guys (and anyone else facing similar difficulties) need to tell Compaq) > about it NOW, through offical channels.e  G    OBTW, I also told them not to consider HP heritage in the compilers.    ------------------------------    Date: 20 Nov 2001 15:53:57 -0600- From: koehler@encompasserve.org (Bob Koehler)u0 Subject: Re: Is it a DEC C problem, or is it me?3 Message-ID: <Wq0u4bU+6e47@eisner.encompasserve.org>i  \ In article <3BFAC57B.E8812F14@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:  J > Compaq is fully aware that lack of VAX-C will result in the loss of some > customers.  B    Just wait 'till Compaq announces furhter support for the ANSI C    standard:  no $ in names.    o   ------------------------------  # Date: Wed, 21 Nov 2001 00:41:24 GMTQ* From: "Bill Todd" <billtodd@metrocast.net>0 Subject: Re: Is it a DEC C problem, or is it me?: Message-ID: <UKCK7.2373$YD.176317@news2.aus1.giganews.com>  8 Warren Spencer <wspencer@ap.nospam.org> wrote in message1 news:915F8EEE2warrenspencer1977@207.126.101.97...e7 > rdeininger@mindspring.com (Robert Deininger) wrote in C > <rdeininger-2011011156020001@user-2iveans.dialup.mindspring.com>:u   ...g  F > >You guys (and anyone else facing similar difficulties) need to tell >sK > I appreciate the advice, and the sentiment Robert, however I have no suchxJ > need.  Instead I believe it is Compaq that *needs* to either communicateI > clearly (stop fibbing) and poll their customers for their requirements.C  L Ah - you mean like they polled their customers about the Alphacide and found+ the response to be 'unbelievably positive'?i  @ I guess most of the people here weren't included in that poll...   - bill   ------------------------------   Date: 20 Nov 2001 21:29 CST ' From: carl@gerg.tamu.edu (Carl Perkins);0 Subject: Re: Is it a DEC C problem, or is it me?- Message-ID: <20NOV200121291240@gerg.tamu.edu>e  1 JF Mezei <jfmezei.spamnot@videotron.ca> writes...m }Alan Greig wrote:F }> One of the things from Friday's DECUS in London was that DEC C is aI }> dead end product. and will not be ported to IA64. Instead Intel C willt' }> become the  standard C on Intel VMS.E } H }But isn't that the same since Compaq donated all its compiler folks and }software to Intel ? }   }> Unlike DEC C which can handleB }> VAXC syntax, Intel C on VMS will not. At least not targeted for+ }> initial release and possibly not at all.n } M }So Intel will remove the code in DEC C that was for support of VAX C syntax.  } M }There goes the promise that customers would just have to recompile user-modec; }applications without any changes when/if VMS gets to IA64.s } L }What about the other languages such as Cobol, ADA etc etc ? (Does APL still	 }exist ?)p  G Intel C already exists (it is, reputedly as I havn't used it, not bad).hE It would not be a case of removing anything from DEC C, it would be a L case of not adding in all the stuff that DEC C already does that it doesn't.  F As far as I know, Intel doesn't already make compilers for Cobol, ADA,H etc. so there would be no Intel version to substitute for them. (If they do, I've never heard of them.)   --- Carl   ------------------------------  % Date: Tue, 20 Nov 2001 20:36:45 +0000.% From: Alan Greig <a.greig@virgin.net>a Subject: Re: Life After Alphae* Message-ID: <3BFABEDD.C1FA64A1@virgin.net>   "Terry C. Shannon" wrote:6  K Compaq *does* have an alternate plan, sort of. EV78 is in production today,8  I > early EV78 silicon powers the 32-way Marvel boxes CPQ is fooling around2 >   N There was a wonderful photo shown at DECUS in London last week that might justN imply something with far more EV7x processors than 32 has already been booted.M But let's keep Sun guessing for the moment. I  couldn't decide if Compaq wererO entirely serious with that slide anyway. At this rate expect to see the GS81920  by next Christmas ;-)    --
 Alan Greig   ------------------------------  # Date: Tue, 20 Nov 2001 21:37:51 GMTm  From: cjt <cheljuba@prodigy.net> Subject: Re: Life After Alpha0+ Message-ID: <3BFACD2A.290E73EA@prodigy.net>p   Alan Greig wrote:  > E > On Tue, 20 Nov 2001 15:25:21 GMT, cjt <cheljuba@prodigy.net> wrote:F > G > >> Compaq's analysis said that IA64 would blow away everyone else andnJ > >> IBM, AMD, Sun will all eventually have to jump on the IA64 bandwagon.K > >> The reason being that none of them could afford to keep up with Intel.  > >dO > >That sure sounds like a bogus argument to me.  Why would it suddenly be trueAK > >now, but not 10 years ago?  Why doesn't everybody have a "True Blue" IBMl2 > >PC on their desk?  How can VMS hope to survive? > C > I put this question directly and the answer was "the industry nowMD > wants fewer operating systems and fewer processors - you don't get  > anywhere by being different".   G Interesting.  Personally, I doubt "the industry" speaks with one voice.>  ) >I pointed out that DEC became the secondrH > largest computer company in the world by being different. No reply wasD > offered. I also asked why the statement about fewer processors andF > fewer OSs did not logically imply a Compaq/HP eventual target of one? > OS (Windows) on one bit of hardware (Intel) and did not get at > satisfactory reply.c > C > In fairness I am not sure the speaker really believed some of the B > Compaq party line but that was what he had to run with. He quiteG > clearly wished I would shutup at times. Like it or not decisions have E > been taken and, like the Compaq engineers, he said we would have to  > "get over it"  > -- > Alan   ------------------------------  % Date: Tue, 20 Nov 2001 23:28:24 +0100e' From: Carsten Gester <pentyern@gmx.net>o Subject: Re: Life After Alphaa- Message-ID: <9tele8$9ba$1@tyfon.itea.ntnu.no>a   Terry C. Shannon wrote:t   > 6 > "Carsten Gester" <pentyern@gmx.net> wrote in message) > news:9td9jv$c7v$1@tyfon.itea.ntnu.no...t >> >> >> Jerry Leslie wrote: >>) >> > John Plemons (john@mavin.com) wrote:rH >> > : Folks I think the tolling of the death bell for Alpha is a littleK >> > : early... If you check the Samsung website you will see its alive andsE >> > : well, Samsung is making a series of boards and has a worldwidec* >> > : network of VAR's to market them....K >> > : Check out http://www.samsungelectronics.com do a search on the 21264-K >> > : there is a whole section on the Alpha, with links to their worldwidee' >> > : distributors for the product....g >> > :8 >> > Interesting. The following URLs would support that: >> >K >> >   http://samsungelectronics.com/semiconductors/Alpha_CPU/Alpha_CPU.htme >> >& >> >   http://www.alpha-processor.com/ >> >, >> >   http://www.api-networks.com/products/ >> >I >> > but someone from API posted recently that they were out of the ALPHAn >> > business. >>J >> Yes, I've asked about Samsungs involvement in the future of Alpha and I > gottI >> the answer from Terry C. Shannon that they focused more on the netwprk 5 >> and transport stuff than on the Alpha chip itself.t >> > K > That's the word from API NetWorks and apparently from Samsung itself. One E > of Samsung's strongest supporters, Nebosja Novakovic, proprietor ofeF > www.novaglobal.com.sg, recently got out of the Alpha business and isK > flogging the NEW  fastest processor on the planet, IBM's Power4. Nova wassL > an Alpha advocate right from the get-go, but apparently dealing with Alpha@ > marketeers, Samsung, and Compaq itself finally got to the guy. > I > Samsung was approached back in July about continuing Alpha development,mC > but apparently opted not to do so. It's an expensive undertaking.  >  >>F >> > Hewlett Packard has licensed its OpenMail product to Samsung SDS: >> >? >> >    http://www.openmail.com/cyc/om/00/showfile.cgi?100-1757nI >> >    HP and Samsung SDS Sign Licensing Agreement for OpenMail Softwares >> >? >> >    http://www.openmail.com/cyc/om/00/showfile.cgi?100-1758I >> >    OpenMail Q and A >> >F >> > OpenMail on a supported unix variant is a drop-in replacement for >> > Microsoft Exchange. >> > >> >9 >> > --Jerry Leslie     (my opinions are strictly my own)D >>= >> I think the problem is that everybody cries Alpha is dead.0 > L > On Usenet, this is pretty much the case. But how many customers base their( > buying decisions on Usenet commentary?  C Not many, but the Usenet commentaries reflect the thoughts of many e potential customers.   > If I would beh > G >> someone who has to buy a new high performance system and I only hearoH >> about Alpha that this system is dead, I wouldn't buy an Alpha system. > L > Agreed. This a job for Compaq marketing. If they aren't doing the job, ask > them why.h  L I don't think it is only a job for Compaq marketing, because in my opinion, F if I had to spend a large amount of money on a cluster for example, I G wouldn't give to much on  the announces of a marketing department of a  I company. I would rather inform myself by asking other people, which have yK know-how, with in this case clustercomputing. And the problem here is that .H in this circle often is to here Alpha is dead. After all we don't speak D about the newest game console, that you can sell just making a good J advertisment. I don't want to say Compaq's marketing department has to do F nothing, but if everybody else say Alpha is dead it will probably die. Regardsc CarstenW   ------------------------------  # Date: Tue, 20 Nov 2001 22:42:08 GMTt4 From: "Terry C. Shannon" <terryshannon@mediaone.net> Subject: Re: Life After Alphas; Message-ID: <4%AK7.1662$bR6.329760@typhoon.ne.mediaone.net>a  4 "Carsten Gester" <pentyern@gmx.net> wrote in message' news:9tele8$9ba$1@tyfon.itea.ntnu.no...e > Terry C. Shannon wrote:e >  <snip> > >oH > > On Usenet, this is pretty much the case. But how many customers base theire* > > buying decisions on Usenet commentary? >nD > Not many, but the Usenet commentaries reflect the thoughts of many > potential customers.  I How true! The thing is, I don't know if Compaq is cognizant of this fact.kK Aside from a few dedicated folks in VMSland who monitor relevant newsgroupsoL on their own time, Compaq doesn't seem to pay much attention to goings-on in cyberspace.r   >t > > If I would bej > >1I > >> someone who has to buy a new high performance system and I only hear J > >> about Alpha that this system is dead, I wouldn't buy an Alpha system. > >rJ > > Agreed. This a job for Compaq marketing. If they aren't doing the job, ask 
 > > them why.  > D > I don't think it is only a job for Compaq marketing, because in my opinion,G > if I had to spend a large amount of money on a cluster for example, InH > wouldn't give to much on  the announces of a marketing department of aJ > company. I would rather inform myself by asking other people, which haveL > know-how, with in this case clustercomputing. And the problem here is thatI > in this circle often is to here Alpha is dead. After all we don't speak E > about the newest game console, that you can sell just making a good-K > advertisment. I don't want to say Compaq's marketing department has to donH > nothing, but if everybody else say Alpha is dead it will probably die.  H Again, quite true. Perhaps I should have said "marketing and messaging."I Press releases aren't very influential, but there are other ways to get atH message across. Proactively engaging the press comes to mind. After all,I without a heads-up, it's unlikely that the press will give much regard toa1 things like the Terascale Computing System, etc..e   ------------------------------    Date: 20 Nov 2001 14:59:40 -0800( From: bob@instantwhip.com (Bob Ceculski) Subject: Re: Life After Alpha = Message-ID: <d7791aa1.0111201459.7d8b8286@posting.google.com>   x "Terry C. Shannon" <terryshannon@mediaone.net> wrote in message news:<U3vK7.6894$eh7.3942482@typhoon.ne.mediaone.net>.../ > "cjt" <cheljuba@prodigy.net> wrote in messagef' > news:3BFA75DA.59557AE2@prodigy.net...n > > Alan Greig wrote:a > > >m: > > > On Mon, 19 Nov 2001 03:48:25 GMT, "Terry C. Shannon"( > > > <terryshannon@mediaone.net> wrote: > > >s	 > > > ...r > > > >> > > > >nK > > > >Given that Power4 is a pretty decent piece of work, do you think IBMs >  would > > >rL > > > But according to Compaq's analysis Power4  is a dead end architecture.K > > > Can't remember the name of the (senior) Compaq Alpha guy who gave theeI > > > presentation in London on Friday. Should have been Dave Fenwick butoK > > > there was a last minute change. Anyone want to jump in with the name?7 > > >FH > > > Compaq's analysis said that IA64 would blow away everyone else andK > > > IBM, AMD, Sun will all eventually have to jump on the IA64 bandwagon.tL > > > The reason being that none of them could afford to keep up with Intel. > >aK > > That sure sounds like a bogus argument to me.  Why would it suddenly be  >  true L > > now, but not 10 years ago?  Why doesn't everybody have a "True Blue" IBM3 > > PC on their desk?  How can VMS hope to survive?h > >l > N > It may be a bogus argument, but the questions you pose don't really shed anyK > light on the relative "bogusity" of the claim. I presume there's a lot ofiH > stuff that's true today that wasn't true ten years ago. As for the IBMM > peecee, which came out twenty years ago IIRC, that franchise got whacked by J > an Attaq of the Killer Clones. One of which is (for the moment) based inH > Houston and sports a name which represents a contraction of COMPAtible
 > Quality. > N > And as for VMS, it'll survive on its merits, not the merits of the processorK > it runs upon. VMS actually is pretty oblivious to the Alpha architecture, L > that's why PALcode was created. All the PALcode goodies get moved into theG > OS itself with the IPF port, hence VMS should be more portable in theo	 > future.  > H > That said, marketing, or lack thereof, will play a pivotal role in the( > continued success--or failure--of VMS.  H I hope you are right ... I get sick at the thoughts of someday having toF port to windoze, unix, linux ... gag!  excuse me, I need to run to the bathroom as I am getting sick!   ------------------------------  # Date: Wed, 21 Nov 2001 00:29:59 GMT-* From: "Bill Todd" <billtodd@metrocast.net> Subject: Re: Life After Alpha@@ Message-ID: <bACK7.36814$8q.5476935@bin2.nnrp.aus1.giganews.com>  = Terry C. Shannon <terryshannon@mediaone.net> wrote in messager5 news:4%AK7.1662$bR6.329760@typhoon.ne.mediaone.net...c > 6 > "Carsten Gester" <pentyern@gmx.net> wrote in message) > news:9tele8$9ba$1@tyfon.itea.ntnu.no...    ...r  F > > I don't think it is only a job for Compaq marketing, because in my
 > opinion,I > > if I had to spend a large amount of money on a cluster for example, I J > > wouldn't give to much on  the announces of a marketing department of aL > > company. I would rather inform myself by asking other people, which haveI > > know-how, with in this case clustercomputing. And the problem here isv thatK > > in this circle often is to here Alpha is dead. After all we don't speak G > > about the newest game console, that you can sell just making a good J > > advertisment. I don't want to say Compaq's marketing department has to doJ > > nothing, but if everybody else say Alpha is dead it will probably die. >oJ > Again, quite true. Perhaps I should have said "marketing and messaging."K > Press releases aren't very influential, but there are other ways to get a-J > message across. Proactively engaging the press comes to mind. After all,K > without a heads-up, it's unlikely that the press will give much regard tog3 > things like the Terascale Computing System, etc..v  C DEC never had much in the way of competent marketing (though it wasrK sometimes better than what exists now) - but 'way back in its heyday it hadaJ a *visible commitment* that was 100% credible:  it devoted major resourcesK to hardware and software development and made it clear that survival of thea1 platform was critical to survival of the company.e  G Lacking similar commitment by Compaq, Alpha is indeed dead, or at least L dying with the end well in sight (though the date may be uncertain).  And noG amount of marketing drivel or press schmoozing is going to convince the@, people you're talking about above otherwise.   - bill   ------------------------------  # Date: Wed, 21 Nov 2001 03:17:48 GMTc  From: cjt <cheljuba@prodigy.net> Subject: Re: Life After Alphaa+ Message-ID: <3BFB1CD5.A9830F6F@prodigy.net>q   Bob Ceculski wrote:g > z > "Terry C. Shannon" <terryshannon@mediaone.net> wrote in message news:<U3vK7.6894$eh7.3942482@typhoon.ne.mediaone.net>...1 > > "cjt" <cheljuba@prodigy.net> wrote in messagee) > > news:3BFA75DA.59557AE2@prodigy.net...: > > > Alan Greig wrote:m > > > >t< > > > > On Mon, 19 Nov 2001 03:48:25 GMT, "Terry C. Shannon"* > > > > <terryshannon@mediaone.net> wrote: > > > >  > > > > ...n
 > > > > >>	 > > > > >:M > > > > >Given that Power4 is a pretty decent piece of work, do you think IBM 
 > >  would > > > >sN > > > > But according to Compaq's analysis Power4  is a dead end architecture.M > > > > Can't remember the name of the (senior) Compaq Alpha guy who gave the:K > > > > presentation in London on Friday. Should have been Dave Fenwick butDM > > > > there was a last minute change. Anyone want to jump in with the name?w > > > > J > > > > Compaq's analysis said that IA64 would blow away everyone else andM > > > > IBM, AMD, Sun will all eventually have to jump on the IA64 bandwagon.)N > > > > The reason being that none of them could afford to keep up with Intel. > > >gM > > > That sure sounds like a bogus argument to me.  Why would it suddenly be 	 > >  trueaN > > > now, but not 10 years ago?  Why doesn't everybody have a "True Blue" IBM5 > > > PC on their desk?  How can VMS hope to survive?i > > >  > >lP > > It may be a bogus argument, but the questions you pose don't really shed any3 > > light on the relative "bogusity" of the claim. e  P For some reason, my news server hasn't got Terry's post, so let me respond here.  M The argument to which I was replying was (I paraphrase) that the player with -K the most resources would necessarily prevail.  Intel had massive resources  M 10 years ago.  IBM had massive resources when it produced the first PC.  VMS 0M has relatively meager resources backing it.  Thus if the argument were true, rM Intel and IBM would already completely dominate, and VMS would be long gone. s  L Those with the massive resources sometimes trip, stumble and fall (remember N that song?).  Sometimes a good idea can prevail even if it comes from a source! that isn't initially the richest.a  9 You present examples of this phenomenon in the following:e   >I presume there's a lot ofSJ > > stuff that's true today that wasn't true ten years ago. As for the IBMO > > peecee, which came out twenty years ago IIRC, that franchise got whacked by L > > an Attaq of the Killer Clones. One of which is (for the moment) based inJ > > Houston and sports a name which represents a contraction of COMPAtible > > Quality. > >hP > > And as for VMS, it'll survive on its merits, not the merits of the processorM > > it runs upon. VMS actually is pretty oblivious to the Alpha architecture,nN > > that's why PALcode was created. All the PALcode goodies get moved into theI > > OS itself with the IPF port, hence VMS should be more portable in then > > future.s > >lJ > > That said, marketing, or lack thereof, will play a pivotal role in the* > > continued success--or failure--of VMS. > J > I hope you are right ... I get sick at the thoughts of someday having toH > port to windoze, unix, linux ... gag!  excuse me, I need to run to the  > bathroom as I am getting sick!   ------------------------------  % Date: Tue, 20 Nov 2001 22:55:13 -0500p- From: JF Mezei <jfmezei.spamnot@videotron.ca>  Subject: Re: Life After Alphaa, Message-ID: <3BFB257E.655668E7@videotron.ca>   Bill Todd wrote:E > DEC never had much in the way of competent marketing (though it wase) > sometimes better than what exists now) r  G Was browsing through some old stuff I had and found an early 1980s CIPSgA membership directory, and low and behond, Digital had a full pageyH advertisement on the back cover for, you'll never believe this: ALL-IN-1J !!!!!!!!!!!  I nearly fell down when i saw this. (back then, I didn't know' about itso I wouldn't have noticed it).r  $ (CIPS was a canadian IT association)  % >- but 'way back in its heyday it hadCL > a *visible commitment* that was 100% credible:  it devoted major resourcesM > to hardware and software development and made it clear that survival of the 3 > platform was critical to survival of the company.M  G And it didn't have a problem admitting that it was dependent on VMS foreN profits. It woudld also expand a lot of resources to bring applications to the
 VMS platform.   I > Lacking similar commitment by Compaq, Alpha is indeed dead, or at least F > dying with the end well in sight (though the date may be uncertain).  A Yo, there is no uncertainty. Compaq announced the death of Alpha.i   ------------------------------  % Date: Tue, 20 Nov 2001 15:52:59 -0800m$ From: "Mike Scott" <mscott@axys.com>' Subject: Mount remote filesystem?  NFS?e Message-ID: <3bfaf1cd@nubby2.>  H I have 5 OpenVMS 7.1 workstations on a Fast-Ethernet LAN.  Often, all ofH them need access to the same data.  Currently, we just clone the data toG each system with the <host::device:[dir]*> DECnet mechanism.  I wish toaI setup a separate computer to act as a 'fileserver' to all these stations, L and make it as transparent as possible (i.e. a local directory that "really"L points to a remote disk: my$disk:[remotedata]).  I only see limited options:G Unix or OpenVMS server with NFS.  Is there something better/faster/moree	 reliable?  -with thanks   ------------------------------  % Date: Wed, 21 Nov 2001 11:12:15 +1030e% From: Jeremy Begg <jeremy@vsm.com.au>e+ Subject: Re: Mount remote filesystem?  NFS?-* Message-ID: <3BFAF867.2060127A@vsm.com.au>   Hello Mike,   J > I have 5 OpenVMS 7.1 workstations on a Fast-Ethernet LAN.  Often, all ofJ > them need access to the same data.  Currently, we just clone the data toI > each system with the <host::device:[dir]*> DECnet mechanism.  I wish todK > setup a separate computer to act as a 'fileserver' to all these stations,hN > and make it as transparent as possible (i.e. a local directory that "really"N > points to a remote disk: my$disk:[remotedata]).  I only see limited options:I > Unix or OpenVMS server with NFS.  Is there something better/faster/mored > reliable?o  K Is a VMScluster out of the question?  It's certainly better & more reliablew+ (and probably faster) than NFS for OpenVMS.1           Jeremy Begg   =   +---------------------------------------------------------+:=   |            VSM Software Services Pty. Ltd.              |h=   |                 http://www.vsm.com.au/                  |n=   |       "OpenVMS Systems Management & Programming"        | =   |---------------------------------------------------------|.=   | P.O.Box 402, Walkerville, |  E-Mail:  jeremy@vsm.com.au |t=   | South Australia 5081      |   Phone:  +61 8 83592155    |>=   |---------------------------|  Mobile:  0414 422 947      |f=   |  A.C.N. 068 409 156       |     FAX:  +61 8 82231777    |t=   +---------------------------------------------------------+n   ------------------------------    Date: 20 Nov 2001 12:09:27 -0800" From: amalloy@techtarget.com (Amy)+ Subject: Opposition to the HP/Compaq mergerb= Message-ID: <dfbaf67b.0111201209.799d1a50@posting.google.com>l  F I'm the site editor for searchHP. We posted a story today based on theF results of our last poll and comments from both Compaq and HP users. I/ thought some of you might want to check it out. N http://searchhp.techtarget.com/originalContent/0,289142,sid6_gci782615,00.html  C Also, we have a new poll up about the merger. So, feel free to also ? share your opinion on that. http://searchhp.techtarget.com/poll    Hope you find this interesting.n
 Amy Malloy   ------------------------------  % Date: Wed, 21 Nov 2001 13:18:17 +110001 From: "Jeremy and Jan Speet" <speetj@bigpond.com>S# Subject: Re: RWAST/LAT port probleml6 Message-ID: <37EK7.721$Lf2.9519@newsfeeds.bigpond.com>  ; I presume these dedicated ports point to a terminal server.>  I If so, check that there are no sessions running on the target port on theo terminal server.  F It might also be handy to check the terminal server software revision.   Regards,
 Jeremy Speet.h  / "LSchiller" <lxschi@ispat.com> wrote in message,7 news:8af707a9.0111200851.354678ce@posting.google.com...e > Hello! >iB > I am hoping someone can help me with a problem that begain on myC > system a few days ago.  I have a program that reads messages on aIG > queue and uses sys$brkthru to send these messages to terminals.  ThisnH > program was set up as a detached process a few years ago, and has beenH > working great until last week.  Now it will go into an RWAST state and$ > does not clear up unless I reboot. >  > Additional background:F > I have a different program that also runs as a detached process.  ItD > is the operator menu program that is assigned to run on a specificE > terminal (using a LAT port).  I actually have 3 of these processes,rD > executing the identical program, on 3 different LAT ports, but I'm> > only having problems with 1.  When the system starts up, theC > message-sending process is started first.  Then the operator menupC > process are started up.  Last week, I recompiled and relinked thegG > operator menu programs, and when I restarted them, the third operator H > menu process went into a COM state when it was started up, practicallyD > continuously (with an occasional RWMPB state popping up).  FurtherB > investigation seemed to point to the message-sending program notH > releasing the LAT port.  However, when I showed that device, the ownerG > process was non-existant, and therefore I could not kill it.  I'm not E > sure if I reached the correct conclusions.  I switched the order ofSD > start-up so the operator menus were first, and at first everythingD > looked okay, but after a little while, the message-sending programG > would go into an RWAST state.  I rebooted, and darn if the same thingh > didn't happen again. >tG > Now, the 2 programs do share a common user-library that was modified.z: > Only the operator-menu program was relinked though.  The7 > message-sending program hasn't been touched in years.C >aG > Any ideas?  If you need more iformation, please let me know.  Thanks.  >  > LSchiller (lxschi@ispat.como   ------------------------------  % Date: Tue, 20 Nov 2001 21:23:06 +0100t, From: Didier Morandi <Didier.Morandi@gmx.ch>% Subject: Re: Scripting / batch files?h& Message-ID: <3BFABBAA.7270076A@gmx.ch>   Syltrem wrote: > ? > Scripting language on VMS is very powerful and is called DCL.m >  > Lookup the DCL dictionary.   Yeah.fB Look up too the DCL Programming book (second edition) available ati http://www.amazon.com/exec/obidos/ASIN/1555581919/qid=1006287616/sr=1-1/ref=sr_1_12_1/107-4349714-6205312m   (mind the CUT/PASTE)  4 Hoff, I make how much each time I promote your book? :-)u   D. -- hG   ---------------------------------------------------------------------AE MORANDI Consulting.  WEB: http://Didier.Morandi.Free.fr/index_us.html E Pflanzschulstrasse 53, 8004 Zurich, Switzerland. GSM: +41 79 705 4670m/ 19, chemin de la Butte, 31400 Toulouse, France.   I Disaster Recovery Plans, Computer Security Audits, DEC OpenVMS ExpertisePH On parle franais, Man spricht Deutsch, Habla Castellano, English spoken   ------------------------------  % Date: Wed, 21 Nov 2001 13:25:59 +1100a1 From: "Jeremy and Jan Speet" <speetj@bigpond.com>,7 Subject: Re: Server vs. Workstation:  You Make The Calli7 Message-ID: <geEK7.724$Lf2.10111@newsfeeds.bigpond.com>y  2 A uVAX 3300 was a server - that's a long time ago!  ? "Robert Deininger" <rdeininger@mindspring.com> wrote in messagekF news:rdeininger-2011011009050001@user-2ive7v8.dialup.mindspring.com...6 > In article <3BFA696A.83C13347@UIowa.EDU>, Rick Dyson > <Rick-Dyson@UIowa.EDU> wrote:s > H > >     For the purposes of licensing the Compaq CSLG/ESL, they define a4 > > "system" as a server *OR* five (5) workstations. > >pE > >     Can anyone help me "officially" clarify whether the following0C > > Digital/Compaq hardware are either a 'server' or 'workstation'?  > >: > >     DEC 3000/400F > Sold in both server and workstation configurations.  The hardware isL > identical, as far as I can tell.  There's a "SERVER" environment variable,G > but it doesn't seem to matter. The S3 switch chooses the console. The I > server configuration is sometimes named a 3000/400S, and probably has an. > slightly different model number on the back. >  > >     Alpha 3000/500K > If you mean the DEC 3000/500, it's "mainly" a workstation, since it has ayF > built-in graphics adapter.  But it will work as a server if desired.* > (I've never heard of an "Alpha 3000/500" >b > >     XP1000 > Intended as a workstation. >h > >     ES45 > >     ES40 68/833e? > AFAIK, all the "modern" alpha systems will work as servers orwK > workstations.  Supposedly you can add a graphics card to a GS320 and makev > a really big workstation.M >  > >     uVAX 3300  > Dunno. >h >   > Of course, I'm not "official". >  > -- > Robert Deininger > rdeininger@mindspring.com    ------------------------------   Date: 20 Nov 2001 23:18 CSTu' From: carl@gerg.tamu.edu (Carl Perkins) 7 Subject: Re: Server vs. Workstation:  You Make The Calle- Message-ID: <20NOV200123183157@gerg.tamu.edu>t  U In article <3BFA696A.83C13347@UIowa.EDU>, Rick Dyson <Rick-Dyson@UIowa.EDU> writes...eE }    For the purposes of licensing the Compaq CSLG/ESL, they define an1 }"system" as a server *OR* five (5) workstations.t } B }    Can anyone help me "officially" clarify whether the following@ }Digital/Compaq hardware are either a 'server' or 'workstation'? }  }    DEC 3000/400s }    Alpha 3000/500c }    XP1000e	 }    ES45o }    ES40 68/833 }  }    uVAX 3300 }  }Thanks in Advance!o } 	 }Regards,  }Ricke  D For this purpose, I think you'll find the first three are considered) workstations and the last three servers. o   --- Carl   ------------------------------  % Date: Tue, 20 Nov 2001 22:40:42 -0000t- From: Martin Walker <Martin.Walker@csf.co.uk> > Subject: RE: Software to emulate someone sitting at a terminalN Message-ID: <0262A6086BFBD411959500508B69C5EA02A5192C@ThisAddressDoesNotExist>  J There was a post on c.o.v dated 20-Oct-01 (sybject VTEST on Alpha) statingG that Cyrano had gone into receivership.  No further info that I've seena though.a   -----Original Message-----> From: tessier-ashpool@usa.net [mailto:tessier-ashpool@usa.net] Sent: 20 November 2001 16:38 To: Info-VAX@Mvb.Saic.Com9> Subject: Re: Software to emulate someone sitting at a terminal     >     NEXUS Software Servicesn >       Gainsborough House >        109 Portland St >       Manchester   M1 6DNu >        Tel: 061 237 3126 > H > (I'm going to try calling that number tonight, but I've been told that< > previous attempts to contact the business failed totally.)  E 1. UK phone numbers changed about 4 years back, with an extra digit 1 D after the 0 in the area code. Try this format: +44 161 237 3126 (you5 drop the zero when dialling into UK internationally).l  > Stop Press - don't bother trying. The BT automated response toC dialling that number: "sorry, this number has not been recognised",b ie: it's a dead phone number.i  E 2. Don't Computer Ass-ociates own MANMAN nowadays? Can they help you?a. http://interbiz.com/Products/SC/ERP/manman.asp  C 3. Another 'emulate humans typing stuff in' product is Cyrano Test.sD See http://www.cyrano.com/products/test.html NB: used to be known as) Performance Software V-TEST a while back.e   Not much, but HTH. Cheers mate.   
 Chris Bardelle Pommie-landt    A This e-mail including any attachments is confidential and may be iF legally privileged. If you have received it in error please advise the@ sender immediately by return email and then delete it from your  system. B The unauthorised use, distribution, copying or alteration of this F email is strictly forbidden. If you need assistance please contact the  help desk on (+44)(0)870 8704820   ------------------------------    Date: 20 Nov 2001 13:11:07 -0600+ From: young_r@encompasserve.org (Rob Young)uB Subject: Re: Solaris: ready for prime time?  Keep your VMS system.3 Message-ID: <ruQRaVkKBWDA@eisner.encompasserve.org>   d In article <ILtK7.110$YD.8935@news2.aus1.giganews.com>, "Bill Todd" <billtodd@metrocast.net> writes: >    > ! >   But one gotcha in a "failoverLC >> cluster" I know if is that once box A fails over to box B a full D >> integrity check must be run, which takes quite a while to trundle >> through all those blocks. > M > No, it takes a few seconds if the file system is protected by a log - e.g.,eK > as the log-enhanced UFS is on Solaris, ext3fs is on Linux, VxFS is on alloH > the platforms Veritas supports, AdvFS is on Tru64, JFS is on AIX, etc. >   A 	Ah... but you are at the wrong level (OS).  I'm talking about an > 	integrity check (not fsck) because the *application* requiresD 	it.  Why doesn't it require it in a VMS cluster?  Different methods2 	of course.  Remember too, that I said this above:  7 		But one gotcha in a "failover cluster" I know [of] ise  8 	that wasn't/isn't a guess.. so when you quickly jump in? 	and say "No" , you don't know what I know about the situation.h   >>@ >> They aren't equivalent paradigms.  One is far superior to the> >> other and I'll give you a few seconds to take a stab at it. >> >> >K >> > As long as VMS bigots keep calling anything that isn't a VMS cluster ab > poorH >> > imitation, they'll miss the need for VMS to match in other respects > those H >> > other forms of clusters that *are* competing successfully with VMS,2 >> > regardless of what people like Rob may think. >> > >>C >> I'm a VMS bigot because I understand both sides of the argument,H' >> (remember Bill?) not in spite of it.l > H > From your comments above, you certainly don't appear to understand the > non-VMS side.n >   = 	No.  You don't understand what I'm talking about because I'ms; 	not describing it fully.  Guess you might be interested in * 	the application.  Sorry.  Won't go there.   				Robt   ------------------------------  % Date: Tue, 20 Nov 2001 22:56:05 +0100d& From: Michael Joosten <joost@c-lab.de>B Subject: Re: Solaris: ready for prime time?  Keep your VMS system.# Message-ID: <3BFAD175.FF6@c-lab.de>    Bob Koehler wrote: >    > >r. > > That seems appropriate -- not 'wiped out.' > J >    The users have no files after the crash, but the file system is full.G >    That's what I mean by wiped out.  After all I was listening to the C >    beeps that came with the kernel messages and neither I nor theeF >    actuall users were particularly worried.  After all, Solaris is a >    modern UNIX, right? >   F Sounds a little like the case I reported. 'fsck' ran with errors? (NotH sure what happens in such a case when UFS with logging is used). Did theG filesystem mount, 'df' reported few empty blocks, but still no files totF see? Some catastrophic failures lead to fsck putting all the files andH directories in lost+found, but these are then often SCSI-related faults.6 All in all, this rather reminds me to a hefty crash...   -- M* Michael Joosten, SBS C-LAB, joost@c-lab.de* Fuerstenallee 11, 33094 Paderborn, Germany, Phone: +49 5251 606127, Fax: +49 5251 6060658 C-LAB is a cooperation of University Paderborn & SIEMENS   ------------------------------  # Date: Wed, 21 Nov 2001 00:57:09 GMTl* From: "Bill Todd" <billtodd@metrocast.net>B Subject: Re: Solaris: ready for prime time?  Keep your VMS system.A Message-ID: <FZCK7.36420$qx2.2190861@bin5.nnrp.aus1.giganews.com>   6 Rob Young <young_r@encompasserve.org> wrote in message- news:ruQRaVkKBWDA@eisner.encompasserve.org...-E > In article <ILtK7.110$YD.8935@news2.aus1.giganews.com>, "Bill Todd"m  <billtodd@metrocast.net> writes: > >a >  > >c# > >   But one gotcha in a "failover E > >> cluster" I know if is that once box A fails over to box B a fulluF > >> integrity check must be run, which takes quite a while to trundle > >> through all those blocks. > >iI > > No, it takes a few seconds if the file system is protected by a log -d e.g.,yI > > as the log-enhanced UFS is on Solaris, ext3fs is on Linux, VxFS is ono alloJ > > the platforms Veritas supports, AdvFS is on Tru64, JFS is on AIX, etc. > >w >iB > Ah... but you are at the wrong level (OS).  I'm talking about an? > integrity check (not fsck) because the *application* requires E > it.  Why doesn't it require it in a VMS cluster?  Different methodst3 > of course.  Remember too, that I said this above:d  K I specifically described in the previous post the fact that applications onsI *either* kind of cluster have the option to work in *either* manner:  howwL the application fails over (or otherwise continues) is orthogonal to how the system does.  I A single-instance application can run on VMS or one of the other kinds of J clusters and fail over, in which case it needs to perform exactly the sameG kinds of integrity checks it performs after an unclean shutdown (checks I which can be made to run as quickly as the application developer chooses, L with the usual trade-offs between run-time overhead and re-start speed).  AnJ application with multiple cooperating instances can also run on VMS or oneK of the other kinds of clusters, in which case it has internal run-time (vs.hD start-up) code that must be able to handle the failure of one of theL instances (and with the same kind of trade-off between run-time overhead andD the speed with which the remaining instances can resume processing).  C In sum, there is nothing special about a VMS cluster that allows an J application on it to continue after a failure with less internal integrityG checking than it needs to perform on other platforms - because there iseI nothing in a VMS cluster that leaves the application's data in any better1L shape after an instance (node) failure than it is left by an instance (node)C failure on other platforms (as long as those platforms use either a L log-protected file system or careful replacement as ODS-2 does to ensure its
 consistency).    - bill   ------------------------------  % Date: Tue, 20 Nov 2001 23:08:06 +0100s From: Dirk Munk <munk@home.nl>/ Subject: Re: Some merger issues (long - sorry!)a' Message-ID: <3BFAD445.3466C99F@home.nl>l  N I've read somewhere that (informal) talks between Ms. Fiorina and Mr. Capellas started somewhere in 2000.....   JF Mezei wrote:k   > John McLean wrote:H > > between business unit representatives of both companies, Ms. FiorinaG > > contacted Mr. Capellas in June 2001 to discuss Compaq's interest ino > > licensing HP-UX. > P > That is interesting. Start to THINK about a merger in end of June, a couple ofP > dys before the Alpha murder announcement and then in early september, announceP > a done deal that has been in the works for months to a point where most of the= > assimilation was planned before the bankers were called in.M >gM > I have this strange feeling that the talks between HP and Compaq began much-M > before that. I wonder if Compaq had targetted HP as a buyer long before and O > offered itself to HP. I find it rather unnerving that in the 2 months between-L > the end of june and early september, HP and Compaq could have wrapped up aM > merger deal and gotten the accountants to see if it was a mistake for HP to-L > buy Compaq, and gotten the bankers involved to structure the deal etc etc. >8Q > Or is HP buying Compaq in the dark not really knowing what it is getting into ?b >e4 > > After several days of deliberation, Mr. CapellasG > > contacted Ms. Fiorina to suggest that the synergies between the two L > > companies were broader than HP-UX and that HP consider whether a broader5 > > strategic relationship might be a viable option."p >sM > read: after a few drinks, Curly though he could merge with Carly, but Carly N > wasn't drunk enough and interpreted his proposal as a company merger instead( > of the body merger ... :-) :-) ;-) :-) >lP > Note that if the initial meeting took place on June 22, "after several days ofM > deliberation" would have put this after the Alpha murder announcement. ThiseM > would make it look like the Alpha deal was negotiated well before Curly metLJ > with Carly since such deals are generally not negotiated in a few hours. >rH > If Curly negotiated the Alpha murder in only a few hours, it is eitherL > absolutely stupid to kill off your main profit generator, or it was a wellH > planned strategic move where the impacts of murdering Alpha were fullyP > considered and thought to be less than the benefits of giving the dead body to > Intel for dissection.  > I > > A: Not at all. That decision was made long before any discussion of am > > merger.g > O > Curly to Carly months before, during a cocktail party: "Tell me something, iftO > I wanted to make Compaq attractive as a takeover target, what should I do ?".fO > Carly responds with "Get rid of that Alpha thing and get in bed with Intel...  >RP > (read: Curly's equivalent to "if I wanted you to go out on a date with me, howP > much weight should I lose to be attractive to you ?". Remember that Palmer hadK > asked the same question to Pfeiffer during the 3 years that preceeded then1 > official announcement of Compaq buying Digital.s > G > Another thing to consider: the announcement that life support will benO > widthdrawn from Tru64. Surely Compaq had considered this decision well beforec
 > June 22. >fL > Again, I get the distinct feeling that on June 22, Curly came prepared notM > with flowers, but with a well detailed proposal and engagement ring, with aeM > list of all the things that Compaq would sacrifice to show its love for HP.p( > (Alpha, Tru64 and go knows what else). >lK > > and operating synergies".   This supposes that the merger will begin tosI > > realize the benefits of these synergies in the fourth quarter of HP's K > > fiscal year 2002, but will not fully realize these synergies until HP's  > > fiscal year 2004.u >iP > Ha ! What a joke. Anyone who believes that "synergies" thing is naive. MergersN > of large corporations do not create any synergies, they only make one largerM > and more bloated company. And lets not talk about the turf wars on which ofbL > the HP or Compaq products will survive etc etc. They will be mired in that$ > merger stuff for more than a year. >aL > Meanwhile, customer won't know whether to buy the HP product or the CompaqM > product, not knowing which product line will survive in the longer term. SoyH > Dell or IBM will look like more solid offerings for wintel crap stuff. >hK > > ** WOW !  That's a nice bit of crystal-ball gazing.  Maybe they will be3L > > 100% realised and maybe it will be 11.3 billion or maybe it will be more > > 100% higher.  >sO > But they don't even consider the possibioity that synergies won't materialise O > for a long time and that the merger will aleniate many customers. If you wereh: > a Tru64 customer right now, where would you want to go ? >aL > > - Both pay the other $675 million (within 2 days) if the merger does notI > > go ahead because a third-party has made a counter-proposal to acquireeI > > the company.  (This applies to mergers that start within 12 months of-J > > any breakdown between HP and Compaq or that such a merger is concluded > > within 2 years). >aO > IS that legal ? If Carly were to have a fight with Curly and call the wedding O > off next week, and get all the paper signed to cancel the wedding, why shouldcM > that preclude Curly from seeking another mate the day after the wedding wasi > called off officially ?l >aL > On the one hand the deal specifies alimony of 675 million if a third partyO > gets involved before the deal is completed. That I can understand. But once acK > deal is officially off, I don't see why HP should still have some alimony2: > patyments should Compaq find another partner afterwards. >aL > Can HP really prevent Compaq from merging with someone else for 2 years ifM > Carly decides that the wedding is off ? That would be tantamount to a death.N > sentence for Compaq. For Compaq to agree to those terms, they must have been > very desperate.  > " > > The merger may be cancelled if9 > > - the BoD accepts an alternative acquisition proposalRH > > - the BoD starts negotiations with a third party regards acquisition >f: > Doesn't that contradict the 675 million penalty clause ?   ------------------------------  # Date: Wed, 21 Nov 2001 01:03:24 GMT * From: "Bill Todd" <billtodd@metrocast.net>/ Subject: Re: Some merger issues (long - sorry!)l@ Message-ID: <w3DK7.39024$uB.6104112@bin3.nnrp.aus1.giganews.com>  K Dirk Munk <munk@home.nl> wrote in message news:3BFAD445.3466C99F@home.nl... G > I've read somewhere that (informal) talks between Ms. Fiorina and Mr.d Capellas  > started somewhere in 2000.....  G Such details appear to vary with the explanation they're pushing on anyi particular day.-   - bill   ------------------------------  % Date: Tue, 20 Nov 2001 22:39:13 +0100f/ From: "Beate Schmitz" <beate.schmitz@hetnet.nl>dI Subject: TCP/IP services settings disappear after restarting the softwaret/ Message-ID: <OhSd1vgcBHA.321@net037s.hetnet.nl>w  	 Hi there,o  < If I define a alias in TCP/IP sercvices 5.1 with the command  TCPIP> ifconfig alias eth0 xxxxxJ then this works correctly, but after restarting TCP/IP services this alias is gone.   My question on you folks is:J How can i define a alias permanently so that after restarting the software the alias is also created.5 I can not find anything in the TCPIP> HELP SET CONFIG: Please can you help me ??g  ! Some spec's that might be needed:u Alpha AS800 with OpenVMS7.3C TCP/IP services 5.1 eco 2n   kind regards Salvi Schrijen salvi.schrijen@libertel.nl   ------------------------------  # Date: Tue, 20 Nov 2001 23:49:00 GMT*4 From: "Matt Muggeridge" <Matt.Muggeridge@compaq.com>M Subject: Re: TCP/IP services settings disappear after restarting the software A Message-ID: <MZBK7.355215$bY5.1512585@news-server.bigpond.net.au>0   You have a couple of options...   = If you prefer to use 'ifconfig', then place these commands in K TCPIP$SYSTARTUP.COM, so each time TCPIP is restarted, the ifconfig commandsr
 are executed.o   Alternately, take a look at:       TCPIP HELP SET INTERFACE  H To save the data in the permanent database, use SET CONFIG INTERFACE ...  K e.g. if you have interfaces WE0, WE1, WE2, and you want to define an aliases for WE1        $ tcpip set interf WEB10A /addr=10.10.10.10/broad=10.10.10.255/net=255.255.255.0  ! dynamic "     $ tcpip set config interf WEB1B /addr=10.10.10.10/broad=10.10.10.255/net=255.255.255.0 ! permanent  . NOTE the syntax I used for the interface name:  J "B" designates the second controller (WE1).  For another alias on WE1, you would use WEB2, WEB3, ...v< Similarly, for aliases on WE0, you would use WEA1, WEA2, ...G And to complete the picture, for aliases on WE2, you would use... (thisS& section to be filled in by the reader)   Matt.9  : "Beate Schmitz" <beate.schmitz@hetnet.nl> wrote in message) news:OhSd1vgcBHA.321@net037s.hetnet.nl...h > Hi there,i >,> > If I define a alias in TCP/IP sercvices 5.1 with the command" > TCPIP> ifconfig alias eth0 xxxxxL > then this works correctly, but after restarting TCP/IP services this alias
 > is gone. >r > My question on you folks is:L > How can i define a alias permanently so that after restarting the software > the alias is also created.7 > I can not find anything in the TCPIP> HELP SET CONFIGm > Please can you help me ??b >e# > Some spec's that might be needed:n > Alpha AS800 with OpenVMS7.3M > TCP/IP services 5.1 eco 2u >  > kind regards > Salvi Schrijen > salvi.schrijen@libertel.nl >t >u >s >p   ------------------------------   Date: 21 Nov 2001 05:49:50 GMT- From: forkosh@panix2.panix.com (John Forkosh)r. Subject: telnet very slow from pc/linux to vms) Message-ID: <9tff9u$6bi$1@news.panix.com>   2 I have a thinwire lan using a single piece of coax7 backbone, off which are hung several pc's running linuxu4 and FreeBSD, and several VAXstations running vms 7.2 (from Montagar cd).e8      They're all interconnected via tcp/ip.  If I telnet8 from vms-->vms, or vms-->pc (either to linux or to bsd), then everything is fine.4      But telnetting from pc-->vms chews up _lots_ of9 cpu cycles on the vms side.  It functions fine (as far as 6 I can tell), but way too slow.  Mon sys shows 100% cpu7 (on a vs4000/vlc), but no process seems to be using it.r7      Two extra weird things: (1)telnetting from FreeBSDD9 is _much_ worse than from linux -- the vlc hangs entirely37 until I Ctrl-[ and drop the connection.  (2)if I telnets4 vms-->pc and then back from _that_ session pc-->vms,( everything is fine, just like vms-->vms.5      I've tried playing around with parameters at them6 telnet> prompt, but to no avail.  I can't speed up the6 pc-->vax connection (and can't slow down the vms-->vms7 connection, which I'd hoped might provide useful info).t@      I know next to nothing about telnet internals/protocol/etc,= so was just playing around and hoping to stumble on solution.i0 But that hasn't happened.  Thanks for any ideas, John (forkosh@panix.com)   ------------------------------  % Date: Tue, 20 Nov 2001 23:00:02 -0700s+ From: "Barry Treahy, Jr." <Treahy@mmaz.com>u2 Subject: Re: telnet very slow from pc/linux to vms' Message-ID: <3BFB42E2.9070803@mmaz.com>c   John Forkosh wrote:e  4 >They're all interconnected via tcp/ip.  If I telnet9 >from vms-->vms, or vms-->pc (either to linux or to bsd),  >then everything is fine.h5 >     But telnetting from pc-->vms chews up _lots_ ofr: >cpu cycles on the vms side.  It functions fine (as far as7 >I can tell), but way too slow.  Mon sys shows 100% cpus8 >(on a vs4000/vlc), but no process seems to be using it. >>I There is nothing strange about what you are doing, in fact I telnet into hD a VLC system from a PC all the time, as I also do to and from Linux G systems from PC's and other VMS systems.  What you neglected to stated >G is which TCP/IP stack are you running on VMS as well as on the PC's (I eE presume that they are Win9X or newer?) or what the Telnet program is  ! that you are using on the PC...  d  I Something to look at on the VAX side, are you logging a lot of errors on uG your ethernet port?  Thin-net isn't the greatest to use because damage iG to the dielectric in the cable can cause all sorts of problems between -= two systems, and not between any other system on the segment.r   Barrye   -- e  @ Barry Treahy, Jr  *  Midwest Microwave  *  Vice President & CIO   A E-mail: Treahy@mmaz.com * Phone: 480/314-1320 * FAX: 480/661-7028e   ------------------------------   Date: 21 Nov 2001 06:48:15 GMT- From: forkosh@panix2.panix.com (John Forkosh)a2 Subject: Re: telnet very slow from pc/linux to vms) Message-ID: <9tfinf$7a0$1@news.panix.com>i  * Barry Treahy, Jr. (Treahy@mmaz.com) wrote: : John Forkosh wrote: 6 : >They're all interconnected via tcp/ip.  If I telnet; : >from vms-->vms, or vms-->pc (either to linux or to bsd),o : >then everything is fine. 7 : >     But telnetting from pc-->vms chews up _lots_ of < : >cpu cycles on the vms side.  It functions fine (as far as9 : >I can tell), but way too slow.  Mon sys shows 100% cpul: : >(on a vs4000/vlc), but no process seems to be using it.  K : There is nothing strange about what you are doing, in fact I telnet into lF : a VLC system from a PC all the time, as I also do to and from Linux I : systems from PC's and other VMS systems.  What you neglected to stated aI : is which TCP/IP stack are you running on VMS as well as on the PC's (I cG : presume that they are Win9X or newer?) or what the Telnet program is -# : that you are using on the PC...  0>      Unfortunately, I don't know how to precisely answer.  VMS> was installed straight from Montagar 7.2 hobbyist cd (and from= its [.kits] directory), and configured using tcpip$config.como as vanilla as possible.s>      PC side has no Windows on it at all.  Linux was installed? from a Slackware 4.0 distribution cd, and FreeBSD 3.2 from thato8 cd (both are a year or two old).  I guess that isn't too+ informative.  I'll try to get more details.   K : Something to look at on the VAX side, are you logging a lot of errors on iI : your ethernet port?  Thin-net isn't the greatest to use because damage uI : to the dielectric in the cable can cause all sorts of problems between e? : two systems, and not between any other system on the segment.s      No errors at all. Thanks,V John (forkosh@panix.com)   ------------------------------  + Date: Wed, 21 Nov 2001 01:52:04 -0500 (EST)h% From: John Forkosh <john@forkosh.com>.2 Subject: Re: telnet very slow from pc/linux to vmsF Message-ID: <Pine.NEB.4.40.0111210151500.2122-100000@panix2.panix.com>  * Barry Treahy, Jr. (Treahy@mmaz.com) wrote: : John Forkosh wrote:t  6 : >They're all interconnected via tcp/ip.  If I telnet; : >from vms-->vms, or vms-->pc (either to linux or to bsd),c : >then everything is fine.k7 : >     But telnetting from pc-->vms chews up _lots_ ofr< : >cpu cycles on the vms side.  It functions fine (as far as9 : >I can tell), but way too slow.  Mon sys shows 100% cpue: : >(on a vs4000/vlc), but no process seems to be using it.  J : There is nothing strange about what you are doing, in fact I telnet intoE : a VLC system from a PC all the time, as I also do to and from LinuxeH : systems from PC's and other VMS systems.  What you neglected to statedH : is which TCP/IP stack are you running on VMS as well as on the PC's (IF : presume that they are Win9X or newer?) or what the Telnet program is! : that you are using on the PC...i>      Unfortunately, I don't know how to precisely answer.  VMS> was installed straight from Montagar 7.2 hobbyist cd (and from= its [.kits] directory), and configured using tcpip$config.comg as vanilla as possible.c>      PC side has no Windows on it at all.  Linux was installed? from a Slackware 4.0 distribution cd, and FreeBSD 3.2 from thati8 cd (both are a year or two old).  I guess that isn't too+ informative.  I'll try to get more details.o  J : Something to look at on the VAX side, are you logging a lot of errors onH : your ethernet port?  Thin-net isn't the greatest to use because damageH : to the dielectric in the cable can cause all sorts of problems between? : two systems, and not between any other system on the segment.u      No errors at all. Thanks,  John (forkosh@panix.com)   ------------------------------  % Date: Tue, 20 Nov 2001 21:29:30 +0100e, From: Didier Morandi <Didier.Morandi@gmx.ch>4 Subject: The DCL minute of the day: show known links& Message-ID: <3BFABD29.FCA551E9@gmx.ch>   $!+e $! SHOW_KNOWN_LINKS.COMe $!N $! A DECnet-Plus equivalent of the DECnet Phase IV "NCP SHOW KNOWN LINKS" cmd.2 $! (see also SHOW_LINKS.COM in the FREEWARE CD v4) $!
 $! <input> $! P1: I  input links only $!     O  output links onlym5 $!     B  both links (default value if p1 is missing)  $!7 $! P2: any character(s) will trigger LOOP display mode.c& $! Example: $ @SHOW_KNOWN_LINKS B LOOP $! Loop exits on CTRL C. $! $! <output>n $! Display example: O $! Known Link Volatile Summary as of 20-NOV-2001 13:00:43 on FPT306 DECnet-Plusn $!O $! PID      Local User   Remote User  Access   from/to Node    type/Port numberhP $! -------- ------------ ------------ -------- --------------- -----------------P $! 00000175 (FAL)        PIDM4X       Incoming (unknown)       NSP$PORT_00002054P $! 00000091 (REMACP)     SYSTEM       Incoming 260.259.233.250 OSI$PORT_0_006B  P $! 00000091 (REMACP)     PIDM4X       Incoming (unknown)       NSP$PORT_00002056P $! 0000019F (PHONE)      SYSTEM       Incoming LOCAL:.FPT288   NSP$PORT_00002058P $! 0000015B SYSTEM       (unknown)    Outgoing 260.259.233.250 OSI$PORT_0_006A  P $! 0000018D SYSTEM       (unknown)    Outgoing LOCAL:.FPT288   NSP$PORT_00002057 $!M $! Results have been saved in files INCOMING_LINKS.TXT and OUTGOING_LINKS.TXTh) $! (you may wish to purge these files :-)U $! $! <side effects>w $! noneb $! $! <disclaimer> D $! This procedure is an eMailWare production from MORANDI & Friends.. $! http://Didier.Morandi.Free.fr/index_us.html $  set noonn8 $  mail/subj="SKL used" NL: smtp%"Didier.Morandi@gmx.ch"1 $! (you may wish to remove the two previous line)) $! $! Revision historyt $!( $! Version Date        Author     actionO $! ------- ----------- ---------- ---------------------------------------------n% $! V1.0-0  19-nov-2001 DMo	  creationl $!-h $ on control then exit
 $ set noon $ if .not. f$priv("bypass")s $ then> $    write sys$output "You need BYPASS to run this procedure."	 $    exita $ endifl $ goto START $!
 $ASK_USER:0 $ inq p1 "Incoming, Outgoing or Both? (I/O/[B])" $START:r4 $ choice = f$edit(p1,"trim,upcase,collapse")			!ptz2 $ if choice .eqs. "" then choice = "B"				!defaultE $ if f$locate(choice,"IOB") .eq. 3 then goto ASK_USER		!invalid inpute $!+d? $! Display header (to have the user wait kindly during we work)i $!-s) $ this_node = f$trnlnm("sys$node") - "::"  $ gosub GET_DECNET_VERSION# $ type/page nl:							!clear screena	 $RESTART:m $ gosub INIT $ gosub TEMP_FILES_CLEANUP  $ gosub GET_SESSION_CONTROL_DATA! $ time = f$extract(0,20,f$time())p $ say "E[H"							!home only $ say -aJ   "Known Link Volatile Summary as of ",time, " on ",this_node," DECnet",dv $ type sys$input $DECK  E[1mO PID      Local User   Remote User  Access   from/to Node    type/Port numberE[mmM -------- ------------ ------------ -------- --------------- -----------------  $EOD $ if there_are_no_linksf $ then $    say ""l% $    say "No information in database"e $    say "" - $    if p2 .nes. "" then goto RESTART			!loopr	 $    exit  $ endifd $ close in_res $ close out_resfG $ if choice .eqs. "I" .or. choice .eqs. "B" then type 'in_results_file',H $ if choice .eqs. "O" .or. choice .eqs. "B" then type 'out_results_file' $ say "E[J"e* $ if p2 .nes. "" then goto RESTART			!loop  $ if choice .eqs. "I" then say -4     "Result has been saved in file ",in_results_file  $ if choice .eqs. "O" then say -5     "Result has been saved in file ",out_results_file   $ if choice .eqs. "B" then say -P     "Results have been saved in files ",in_results_file," and ",out_results_file $ exit $! $GET_SESSION_CONTROL_DATA: $!+cK $! This is the main loop which will gather a minimum of information on all tN $! incoming and outgoing links via DECnet Phase IV and "Phase V" (OSI), nativeO $! and "over IP". A further version of the tool will also process TCP/IP links.u $!N $! To ease development, we gather all links, then for each link we gather OSI N $! and NSP (Phase IV) data for both directions. The display will be different M $! according to the user choice (incoming, outgoing or both) at display time.w $!-p# $ define/user sys$output 'outfile1'r% $ ncl show session control port * alle $ define sys$output nl:  $ define sys$error  nl: G $ search/window=(6,7)/out='tmp_search_file' 'outfile1' "transport port"3 $ status = $status $ deassign sys$output. $ deassign sys$error5 $ if status .eqs. "%X08D78053"				!no strings matchede $ then $    there_are_no_links = 1e $    returna $ endif  $ close/nolog ch_scl& $ open/read   ch_scl 'tmp_search_file' $!
 $LOOP_NCL: $ read/end=EOF_SCL ch_scl line3 $ if line .eqs. "" then goto LOOP_NCL			!empty linet $!+hJ $! The SEARCH gives star separated data. We find a separator, so we recordL $! the previous result we computed before we read this line, then we go back* $! to read the next set of SEARCHed lines. $!-o$ $ if f$extract(0,3,line) .eqs. "***" $ then $    gosub RECORD_RESULT $    goto LOOP_NCL $ endif  $!+  $! scan line $! ========= $!-h% $ line = f$edit(line,"trim,compress")d $!+ ( $! get SCL (Session ControL) entry name. $!-o% $ if f$extract(0,4,line) .eqs. "Name"t $ then, $    scl_entry = f$element(1,"=",line) - " " $    goto LOOP_NCL $ endifo $!+f $! get local user name $! =================== $!-a0 $ if f$extract(0,14,line) .eqs. "Local End User" $ then6 $    loc_user = f$element(2,"=",line) - " [0,0]" - " "6 $    if loc_user .eqs. "42" then loc_user = "(REMACP)"3 $    if loc_user .eqs. "17" then loc_user = "(FAL)"a5 $    if loc_user .eqs. "29" then loc_user = "(PHONE)" , $!   add your favourite network objects here $    goto LOOP_NCL $ endife $!+t $! get port number $! =============== $!-t0 $ if f$extract(0,14,line) .eqs. "Transport Port" $ then$ $    port_nr = f$element(1,"=",line)6 $    if f$locate("OSI",port_nr) .ne. f$length(port_nr)	 $    then * $       port_nr = f$element(4," ",port_nr) $       gosub GET_OSI_DATA	 $    else * $       port_nr = f$element(3," ",port_nr) $       gosub GET_NSP_DATA
 $    endif $    goto LOOP_NCL $ endif- $!+-. $! get access direction (incoming or outgoing) $! ====================d $!-l* $ if f$extract(0,9,line) .eqs. "Direction" $ then, $    direction = f$element(1,"=",line) - " " $    goto LOOP_NCL $ endifn $!+o $! get remote user namee $! ====================s $!-D1 $ if f$extract(0,15,line) .eqs. "Remote End User"t $ then% $    rem_user = f$element(2,"=",line) ) $    rem_user = f$element(1,"]",rem_user) - $    if rem_user .eqs. "]" then rem_user = ""i6 $    if rem_user .eqs. "42" then rem_user = "(REMACP)"3 $    if rem_user .eqs. "17" then rem_user = "(FAL)" 5 $    if rem_user .eqs. "29" then rem_user = "(PHONE)"i $    goto LOOP_NCL $ endifI $!+v $! get remote node namef $! ====================r $!-e0 $ if f$extract(0,14,line) .eqs. "Node Name Sent" $ thenO $    if rem_node .nes. "" then old_rem_node = rem_node	!save previous (?) valuei' $    node = f$element(1,"=",line) - " "c9 $    if node .nes. "<Default value>" then rem_node = noden9 $!    if "''rem_node'" .nes. "''old_rem_node'" then say	-s:  !       "GET_SCL_DATA: Previous value """,old_rem_node,	-,  !       """ superceded by """,rem_node,"""" $    goto LOOP_NCL $ endif  $!+e $! get local process IDo $! ====================e $!-u4 $ if f$extract(0,18,line) .eqs. "Process Identifier" $ then4 $    pid = f$element(1,"=",line) - " " - """" - """" $    goto LOOP_NCL $ endif- $!+-: $! Nothing (very) interesting, we loop to the next record. $!-  $ goto LOOP_NCLc $!	 $EOF_SCL:j $ close ch_scl $ gosub RECORD_RESULTp $ returnO $!-----------------------------------------------------------------------------n $GET_OSI_DATA: $ gosub TEMP_FILES_CLEANUP# $ define/user sys$output 'outfile2',2 $ ncl show osi transport port 'port_nr' all statusP $ search/wind=(0,0)/out='tmp_search_file' 'outfile2' "Remote RFC1006 IP Address" $ close/nolog ch_osi$ $ open/read ch_osi 'tmp_search_file' $ read ch_osi line_osi- $ line_osi = f$edit(line_osi,"trim,compress") L $ if rem_node .nes. "" then old_rem_node = rem_node	!save previous (?) value( $ node = f$element(1,"=",line_osi) - " "6 $ if node .nes. "<Default value>" then rem_node = node6 $! if "''rem_node'" .nes. "''old_rem_node'" then say	-:  !       "GET_OSI_DATA: Previous value """,old_rem_node,	-,  !       """ superceded by """,rem_node,"""" $ close ch_osi $ returnP $!------------------------------------------------------------------------------ $GET_NSP_DATA: $ gosub TEMP_FILES_CLEANUP# $ define/user sys$output 'outfile3' ( $ ncl show nsp port 'port_nr' all status/ $ if f$search("a.a") .nes. "" then delete a.a;*a4 $ search/window=(0,2)/out=A.A 'outfile3' 'scl_entry' $ close/nolog ch_nsp $ open/read ch_nsp A.A $ read ch_nsp line_nsp $ read ch_nsp line_nsp $ read ch_nsp line_nsp- $ line_nsp = f$edit(line_nsp,"trim,compress")i2 $ rem_node = f$element(4," ",line_nsp) - "(" - ")"J $ if rem_node .eqs. " " then rem_node = ""	    !missing in decnet_register $ close ch_nsp $ returnO $!-----------------------------------------------------------------------------  $RECORD_RESULT: 2 $ if loc_user .eqs. "" then loc_user = "(unknown)"2 $ if rem_user .eqs. "" then rem_user = "(unknown)"2 $ if rem_node .eqs. "" then rem_node = "(unknown)"B $! if rem_node .eqs. "<Default value>" then rem_node = "(unknown)" $ if direction .eqs. "Incoming"l $ then $    write in_res	--      pid," ",						-5      f$extract(0,12,"''loc_user'            ")," ",	-p5      f$extract(0,12,"''rem_user'            ")," ",	-m      direction," ",					-D5      f$extract(0,15,"''rem_node'            ")," ",	-a.      f$extract(0,17,"''port_nr'             ") $ else $    write out_res	-      pid," ",						-5      f$extract(0,12,"''loc_user'            ")," ",	-g5      f$extract(0,12,"''rem_user'            ")," ",	-I      direction," ",					- 5      f$extract(0,15,"''rem_node'            ")," ",	-n.      f$extract(0,17,"''port_nr'             ") $ endifo $ loc_user = ""s $ rem_user = ""  $ old_rem_node = ""s $ rem_node = ""e $ port_nr = "" $ direction = "" $ scl_entry = "" $ returnO $!-----------------------------------------------------------------------------t $INIT: $ delete = "delete"  $ ncl = "mc ncl" $ say = "write sys$output" $ search ="search"+ $ scl_entry = ""				!key to access SCL data-3 $ loc_user = ""					!local username initiating linkr4 $ rem_user = ""					!remote username initiating link6 $ old_rem_node = ""				!original value for remote node4 $ rem_node = ""					!displayed value for remote node2 $ port_nr = ""					!port number (NSP$xx or OSI$xx)) $ direction = ""				!Incoming or Outgoing)2 $ there_are_no_links = 0			!flag for fancy display- $ outfile1 = "show_session_control_port.temp"d+ $ outfile2 = "show_osi_transport_port.temp"s! $ outfile3 = "show_nsp_port.temp",1 $ tmp_search_file  = "temporary_search_file.temp"-) $ in_results_file  = "INCOMING_LINKS.TXT" ) $ out_results_file = "OUTGOING_LINKS.TXT"m $ close/nolog in_res $ close/nolog out_resp& $ open/write in_res  'in_results_file'' $ open/write out_res 'out_results_file'  $ returnO $!-----------------------------------------------------------------------------r $TEMP_FILES_CLEANUP:9 $ if f$search(outfile1) .nes. "" then delete 'outfile1';*-9 $ if f$search(outfile2) .nes. "" then delete 'outfile2';*f9 $ if f$search(outfile3) .nes. "" then delete 'outfile3';*MG $ if f$search(tmp_search_file) .nes. "" then delete 'tmp_search_file';*i $ returnO $!-----------------------------------------------------------------------------L $GET_DECNET_VERSION: $ dv = " Phase IV"G $ if f$search("sys$system:net$local_name_database.dat") .nes. "" then -u      dv = "-Plus"d $ return    M Warning: change the "E[" to "<ESC>[" with your (favourite) editor before run.e   Enjoy.   D. -- )G   ---------------------------------------------------------------------oE MORANDI Consulting.  WEB: http://Didier.Morandi.Free.fr/index_us.htmluE Pflanzschulstrasse 53, 8004 Zurich, Switzerland. GSM: +41 79 705 4670e/ 19, chemin de la Butte, 31400 Toulouse, France.s  I Disaster Recovery Plans, Computer Security Audits, DEC OpenVMS ExpertisetH On parle franais, Man spricht Deutsch, Habla Castellano, English spoken   ------------------------------  % Date: Tue, 20 Nov 2001 22:40:45 -0000 - From: Martin Walker <Martin.Walker@csf.co.uk>-) Subject: RE: TZ887 installation questionsmN Message-ID: <0262A6086BFBD411959500508B69C5EA02A5192D@ThisAddressDoesNotExist>  K Engineers would recommend shutdown before plugging SCSI cables.  Of course, ' you might well get away with it - YMMV!w   -----Original Message-----. From: Rick Dyson [mailto:Rick-Dyson@UIowa.EDU] Sent: 20 November 2001 17:15 To: Info-VAX@Mvb.Saic.Comv) Subject: Re: TZ887 installation questionsf     Linda Luik wrote:lJ > I am hooking a TZ887 to a Alpha 2000 running Alpha O/S 6.2-1h3 with full > patches (as far as I know).MI > There are three disks installed (dka100, 200, 600) and an internal  4mmiH > tape drive (mka500). This leaves scsi ports 3 and 4 open for business.C > The system has a storageworks box containing more disks, but thisa3 > shouldn't be important to the tape drive install.t >  > Do I:u( > 1) set the TZ887 dip switch to 3 or 4? > <<< install the tape drive by plugging it in , etc... >>> 2) runautoconfigure or will a reboot be needed first?	I would just choose one.  Itdoesn't matter in the long run.  Themore important concerns are total SCSIcabling length and proper termination.I will assume you already are up tospeed with those issues.	No need to reboot, unless you want to...$ MCR Sysman		SYSMAN> IO Autoconfig /Log	I might suggest youlook into the MRU (Media Robot Utility, I think),with it's ROBOT and XROBOTinterfaces for controll ing the robotic unit onthe drive from a remoteterminal.  Very handy unless you like driving intowork to move a tape fromone slot to another at 3am. :)Regards,Rick    A This e-mail including any attachments is confidential and may be eF legally privileged. If you have received it in error please advise the@ sender immediately by return email and then delete it from your  system. B The unauthorised use, distribution, copying or alteration of this F email is strictly forbidden. If you need assistance please contact the  help desk on (+44)(0)870 8704820   ------------------------------   Date: 20 Nov 2001 22:27 CSTp' From: carl@gerg.tamu.edu (Carl Perkins)0D Subject: RE: VMS on IBM power chip would make IBM No. 1 in high end!- Message-ID: <20NOV200122275644@gerg.tamu.edu>   ' "Tom Linden" <tom@kednos.com> writes...I/ }> From: Alan Greig [mailto:a.greig@virgin.net]r+ }> Sent: Tuesday, November 20, 2001 3:21 AM  }> To: Info-VAX@Mvb.Saic.ComG }> Subject: Re: VMS on IBM power chip would make IBM No. 1 in high end!n }> }>I }> On Fri, 16 Nov 2001 22:44:11 -0500, David Froble <davef@tsoft-inc.com> 	 }> wrote:4 }> }> >John Eisenschmidt wrote: }> >K }> > > Alpha is already gone. IBM won't be doing anything with that - whiche+ }> > >  is fine, Power is a great platform., }>F }> But part of the justification Compaq used to decide that IA64 wouldI }> become industry standard was that nobody else would be able to keep upVF }> with Intel investment. A power point slide at Friday#s London DECUSE }> actually said that "IBM will not be able to afford to keep up withUA }> Intel". This was challenged from the audience - one of the fewlI }> challenges that didn't come from me...The question was put as "How can)E }> you say that the biggest computer company in the world will not be B }> able to keep up with Intel".The answer was just "that's what weI }> predict". I also took issue with the use of "Industry Standard" in alla* }> the powerpoint slides to describe IA64. } K }For someone to claim that IBM couldn't keep with Intel in spending strikeseI }me as foolhardy and irresponsible.  The fact that the speaker offered no 	 }analysistH }backing his prediction, suggests it has no value.  It is BS and drivel.  E It strikes me as being just plain wrong, and basically stupid. In thenH last twelve months, IBM has had over 3 times the sales as Intel and alsoH about 2.7 times the profit (Sales: IBM = $88.7B, Intel = $28.3B; Profit:H IBM = $8.04B, Intel = $2.98B; the biggest bright point in the financialsE for Intel is that they currently have a few billion in cash more than 0 IBM does, and a slightly higher profit margin).   K If Intel were to increase their spending by about $3 billion per year, theySJ would just break even. If IBM then matched that increase, they would stillJ make $5 billion in profit - or about $2 billion more profit than Intel nowI actually makes. In terms of profit margin, that would put Intel at 0% andcH leave IBM at over 5.5%. Alternatively, if they were to both reduce theirI profit margins to 5% by spending more, Intel would get to spend some $1.6OG billion/year more than in the last year but IBM would get to spend sometG $3.6 billion more. All assuming that the future incomes and profits for-1 both are similar to the recent levels, of course.-  G Remember: just about any moron can make a PowerPoint slide, and some ofP them do.   --- Carl   ------------------------------  # Date: Tue, 20 Nov 2001 20:05:36 GMT " From: "Hans Vlems" <hvlems@iae.nl>  Subject: Re: Xetra, on which OS?/ Message-ID: <kIyK7.410$Dp.1281@typhoon.bart.nl>   H The Amsterdam stock exchange also runs on VMS. That did not keep it fromK going off-line over embarrassing long periods of time, even if the hardwaren( configuration called for zero down time.G The rumours I've heard, again just rumours, were that the clusters wered	 seriouslyIG underpowered and that the application software was not exactly bug frees either.nI VMS is a good operating system, equaled only by OS/390 (MVS), but it does % require some resources to run on ....r  
 Hans Vlems  1 Rudolf Wingert <win@fom.fgan.de> wrote in messageL0 news:200111200659.HAA01972@sinet1.fom.fgan.de... > Hello, >tC > today I red in a German newspaper, that here in Germany the stock D > system Xetra was down for two hours. On which OS runs Xetra? Or isE > Xetra the OS? Was the underlaying OS OpenVMS? If yes, why did it bel > down  for two hours? >x  > TIA and regards Rudolf Wingert >p   ------------------------------  % Date: Tue, 20 Nov 2001 14:09:19 -0500e2 From: "Sue Skonetski" <susan.skonetski@compaq.com>  Subject: You are invited to this3 Message-ID: <rSxK7.1770$RL6.58594@news.cpqcorp.net>r  J Attunity and Compaq are pleased to invite you to participate in a live webI seminar, e-Business Integration for OpenVMS, scheduled for December 13 at  2:00pm.D  H More information and registration see http://www.attunity.com/mail.html.   ------------------------------  % Date: Tue, 20 Nov 2001 14:35:59 -0500 2 From: "Sue Skonetski" <susan.skonetski@compaq.com>$ Subject: Re: You are invited to this3 Message-ID: <qfyK7.1772$RL6.58525@news.cpqcorp.net>e   that's EST (GMT - 5:00)m    = "Sue Skonetski" <susan.skonetski@compaq.com> wrote in message - news:rSxK7.1770$RL6.58594@news.cpqcorp.net...C >  >mL > Attunity and Compaq are pleased to invite you to participate in a live webK > seminar, e-Business Integration for OpenVMS, scheduled for December 13 at-	 > 2:00pm.D > J > More information and registration see http://www.attunity.com/mail.html. >a >e   ------------------------------  # Date: Tue, 20 Nov 2001 23:36:24 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) $ Subject: Re: You are invited to this0 Message-ID: <00A0557A.B8CFFFE8@SendSpamHere.ORG>  h In article <rSxK7.1770$RL6.58594@news.cpqcorp.net>, "Sue Skonetski" <susan.skonetski@compaq.com> writes: >  >sK >Attunity and Compaq are pleased to invite you to participate in a live webcJ >seminar, e-Business Integration for OpenVMS, scheduled for December 13 at >2:00pm. >cI >More information and registration see http://www.attunity.com/mail.html.l >f >,  A ... and, of course, it will work with Netscape 3.03 on OpenVMS?  c --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMc            gJ   "And of course, I'm a genius, so people are naturally drawn to my fiery I   intellect.  Their admiration overwhelms their envy!" -- Calvin & Hobbes    ------------------------------  % Date: Tue, 20 Nov 2001 14:06:17 -0500!0 From: Jason Fountain <Jason.Fountain@compaq.com>1 Subject: Re: [TCPIP V5.1 ECO 3] Problems with DNSr* Message-ID: <3BFAA9A9.3261CD54@compaq.com>  N > The TCPIP V5.1 (at least with ECO 3) DNS server does have real problems withM > completing zonetransfers. Most of the time, they do complete, but sometimesoM > (and much much more often than tolerable) they simply hang just before they L > even begin (TCP connection is established and nothing happens until client- > times out). No logfile entries seen so far.n  O Are you talking about zone transfers from the zones changing or from doing a SHm
 HOST command? R Do you see temp files in the BIND directory?  Are all of your zones loaded without, errors (check the beginning of the logfile)?   -jar   ------------------------------  % Date: Tue, 20 Nov 2001 21:04:17 +0100p, From: "Nico van der Boom" <njvdboom@caiw.nl>1 Subject: Re: [TCPIP V5.1 ECO 3] Problems with DNS 4 Message-ID: <3bfab62b$0$222$4d4ebb8e@news.nl.uu.net>  7 "Peter LANGSTOEGER" <eplan@kapsch.net> wrote in messagea$ news:3bf99b1f$1@news.kapsch.co.at...G > Today I changed my DNS server from TCPIP V5.1 ECO 3 to TCPware V5.5-3m again. > H > Reason is, I host over 150 domains on my DNS server and I was tired ofJ > checking which domain is now not current on the secondary DNS server (my ISP).= >=I > The TCPIP V5.1 (at least with ECO 3) DNS server does have real problemsn withC > completing zonetransfers. Most of the time, they do complete, but.	 sometimeshH > (and much much more often than tolerable) they simply hang just before theyL > even begin (TCP connection is established and nothing happens until client- > times out). No logfile entries seen so far.n >e- > Does anyone have seen this behaviour, too ?  > Does anyone have a clue ?e8 > Is there an ECO 4 to fix it and when is it available ?  * I have the same problem on TCPIP 5.1 ECO 23 After starting the bind server everything works ok. K after a while (could be hours, could be days) the zone transfer (AXFR) fromt a secondary stops working. - no log entry is made._) - nothing wrong with bind server process.i) - normal query's (A, PTR, NS) still work.   L I am running in a cluster. AXFR stops working on the cluster ip adress, AXFR on a member still works.  I I have contacted compaq support about this problem (10 days ago, they areg researching this)e   In the meantime..=5 i run a dcl in detached which checks AXFR every hour.  when it stops working: - ndc statsm( - ndc dump, and save dump in unique file - ndc restart (restart bind)  and everything is working again.  I I thought about installing ECO3, your message tells me it problably isn't. the answer.nE (a few months back we ran BIND on VAX, worked for 6 years without anyNL problem, after migrating to alpha-cluster with tcpip, problem after problem)   CU,= Nico van der Boomf   ------------------------------  # Date: Tue, 20 Nov 2001 20:00:01 GMT." From: "Hans Vlems" <hvlems@iae.nl>H Subject: Re: [VMS Motif] How to avoid starting the DECW$SERVER process ?. Message-ID: <5DyK7.401$Dp.981@typhoon.bart.nl>  @ Interesting, my VS3100/GPX has DECW$IGNORE_DECWINDOWS defined in SYLOGICALS.COMJ and get exactly what you want: I can open X sessions on other stations but the graphics consoleD remains in ASCII mode (LA36 IIRC). SYLOGICALS.COM is basically this:. $       define/sys/exec decw$server_biolm 1200- $       define/sys/exec decw$server_diolm 90013 $       define/sys/exec decw$install_xterminal TRUE 4 $       define/sys/exec DECW$IGNORE_WORKSTATION TRUE/ $       define/sys/exec sys$decdtm_inhibit TRUEE  J I'm not quite sure why that did not work for you. You did install DW-MOTIF didn't you?-  
 Hans Vlems    5 Peter LANGSTOEGER <eplan@kapsch.net> wrote in message $ news:3bf98fcc$1@news.kapsch.co.at...J > I just want to start a system (a workstation) with MOTIF but without theH > DECW$SERVER process (I don't use the graphics console and want to save mem).l >/% > What is the best/suggested method ?e >hF > Defining DECW$IGNORE_DECWINDOWS is not ok, because MOTIF is then not started % > at all (images not installed, ...).= >mC > SYSGEN WINDOW_SYSTEM=0 does normally only lead to endless AUTOGEN"
 questions. >n > Anyone, please ? >h > --> > Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651= > Network and OpenVMS system manager  Fax.    +43 1 81111-888-> > <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netJ > A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"   ------------------------------   End of INFO-VAX 2001.647 ************************