1 INFO-VAX	Wed, 15 May 2002	Volume 2002 : Issue 268       Contents:: Re: $CHECK_ACCESS fails if there is no read access to file: Re: $CHECK_ACCESS fails if there is no read access to file: Re: $CHECK_ACCESS fails if there is no read access to fileG Re: A (free) VMS environment in a Linux/Unix system ?, was: Re: freeVMS G Re: A (free) VMS environment in a Linux/Unix system ?, was: Re: freeVMS G Re: A (free) VMS environment in a Linux/Unix system ?, was: Re: freeVMS G Re: A (free) VMS environment in a Linux/Unix system ?, was: Re: freeVMS B Acrobat (Was: Re: Scott Stallard, forget your VMS to HP UX dream!)F Re: Acrobat (Was: Re: Scott Stallard, forget your VMS to HP UX dream!)F Re: Acrobat (Was: Re: Scott Stallard, forget your VMS to HP UX dream!)F Re: Acrobat (Was: Re: Scott Stallard, forget your VMS to HP UX dream!)F Re: Acrobat (Was: Re: Scott Stallard, forget your VMS to HP UX dream!)$ Re: Bob Palmer and the demise of DEC3 Re: Capellas: Linux, Windows Will 'Eviscerate' Unix 3 Re: Capellas: Linux, Windows Will 'Eviscerate' Unix 3 Re: Capellas: Linux, Windows Will 'Eviscerate' Unix 3 Re: Capellas: Linux, Windows Will 'Eviscerate' Unix 3 Re: Capellas: Linux, Windows Will 'Eviscerate' Unix 3 Re: Capellas: Linux, Windows Will 'Eviscerate' Unix 3 Re: Capellas: Linux, Windows Will 'Eviscerate' Unix 3 Re: Capellas: Linux, Windows Will 'Eviscerate' Unix 3 Re: Capellas: Linux, Windows Will 'Eviscerate' Unix 3 Re: Capellas: Linux, Windows Will 'Eviscerate' Unix $ DCPS and Poscript Printing..........( Re: DCPS and Poscript Printing..........( Re: DCPS and Poscript Printing..........( Re: DCPS and Poscript Printing..........( Re: DCPS and Poscript Printing..........( Re: DCPS and Poscript Printing..........- Re: Forced migration to HPHUX - Storm Warning - Re: Forced migration to HPHUX - Storm Warning - Re: Forced migration to HPHUX - Storm Warning - Re: Forced migration to HPHUX - Storm Warning - Re: Forced migration to HPHUX - Storm Warning - Re: Forced migration to HPHUX - Storm Warning - Re: Forced migration to HPHUX - Storm Warning - Re: Forced migration to HPHUX - Storm Warning - Re: Forced migration to HPHUX - Storm Warning - Re: Forced migration to HPHUX - Storm Warning  FTP mget on latest files ? Re: FTP mget on latest files ? Re: FTP mget on latest files ? Re: FTP mget on latest files ?/ Irish Railways [was Re: Stallards smoking gun!] P Re: Let go of VMS for the good of your career? (was "Re: System manager  availabP Re: Let go of VMS for the good of your career? (was "Re: System manager  availabP Let go of VMS for the good of your career? (was "Re: System manager available")  Newbie using Openvms and C... ! Re: Newbie using Openvms and C... ! Re: Newbie using Openvms and C... ! Re: Newbie using Openvms and C... ! Re: Newbie using Openvms and C...  RE: No new Alpha sales RE: No new Alpha sales Re: No new Alpha sales Re: No new Alpha sales Re: No new Alpha sales Re: No new Alpha sales Re: No new Alpha sales Re: No new Alpha sales Re: No new Alpha sales Re: OpenVMS Hobbyist License Re: OpenVMS SNMP TRAP , Problem with the internal clock of an XP10000 Re: Problem with the internal clock of an XP10000 problems setting socket nonblocking with fcntl() Proxy question Re: Proxy questionP Re: Reads vs. Writes (was Re: What is good model for disk i/o w        /shadowin Re: System manager available Re: System manager available Re: System manager available Re: System manager available TCP/IP for VMS...HELP! RE: TCP/IP for VMS...HELP! Re: TCP/IP for VMS...HELP! Re: TCP/IP for VMS...HELP! Re: TCP/IP for VMS...HELP! Re: TCP/IP for VMS...HELP! Re: TCP/IP for VMS...HELP! Re: TCP/IP for VMS...HELP!3 Telecom fix-line billing systems running on OpenVMS 7 Re: Telecom fix-line billing systems running on OpenVMS  UNIX Security a good bet UNIX Security a good bet Version of the KZPBA Re: Version of the KZPBA Re: Version of the KZPBA Re: Who cares about marketing! Re: Who cares about marketing! Re: Who cares about marketing! Re: Who cares about marketing! Re: Who cares about marketing! Re: Who cares about marketing! Re: Who cares about marketing!' Re: WIS??? Did it die with the HP deal? $ Re: www.compaq.com is link to hp.com  F ----------------------------------------------------------------------    Date: 15 May 2002 02:04:03 -0700 From: cbdeja@my-deja.com (-)C Subject: Re: $CHECK_ACCESS fails if there is no read access to file = Message-ID: <611952a3.0205150104.70f9acb6@posting.google.com>   5 rob.buxton@wcc.govt.nz (Rob Buxton) wrote in message  G > Just guessing, but suspect the difference here is VAX running 5.5 and & > Alpha running something a bit later. > B > There was a big change to how Security worked between VMS 5 & 6.  > This may be part of the issue?  E It may be something along those lines. An additional quirk on VMS 5.5 F is that the access() C runtime library function will always return theB correct response regardless of whether you have read access to the% file, while sys$check_access() fails.   C I presume that access() is using a different system call to get the F information, but as VMS isn't open source I can't find out what it is.C Anyway, access() is documented as not checking ACLs (just the basic C file permissions) and I need to check ACLs - so the source code for * access() probably wouldn't help me anyway!  D I'll enquire with our IT department to see whether this problem withD read access and sys$check_access has had a patch issued at some time in the past.  @ But if anyone can think of anything else, please do let me know.   Colin    ------------------------------    Date: 15 May 2002 06:35:54 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) C Subject: Re: $CHECK_ACCESS fails if there is no read access to file 3 Message-ID: <XvLB6PSpeP1O@eisner.encompasserve.org>   \ In article <611952a3.0205150104.70f9acb6@posting.google.com>, cbdeja@my-deja.com (-) writes:7 > rob.buxton@wcc.govt.nz (Rob Buxton) wrote in message  H >> Just guessing, but suspect the difference here is VAX running 5.5 and' >> Alpha running something a bit later.  >>  C >> There was a big change to how Security worked between VMS 5 & 6. ! >> This may be part of the issue?  > G > It may be something along those lines. An additional quirk on VMS 5.5 H > is that the access() C runtime library function will always return theD > correct response regardless of whether you have read access to the' > file, while sys$check_access() fails.   E Certainly my comments presumed you were running an up-to-date copy of  VMS (7.2-1 or later).    ------------------------------  % Date: Wed, 15 May 2002 10:17:04 -0400 1 From: "Mark D. Jilson" <jilly@clarityconnect.com> C Subject: Re: $CHECK_ACCESS fails if there is no read access to file 2 Message-ID: <3CE26DE0.B409D051@clarityconnect.com>  G The only change in this area that I can (quickly) find is documented in G OpenVMS VAX Version 6.0 Release Notes, (AA-PV6ZA-TE), MAY 1993, Section C 2.2.14 Security Changes and deals with SYS$CHECK_ACCESS and CONTROL C access.  See also the DSNlink article (may be available in askvms)  < [OpenVMS] $CHECK_ACCESS Returns SS$_NOPRIV On Control Access   - wrote: > 6 > rob.buxton@wcc.govt.nz (Rob Buxton) wrote in messageI > > Just guessing, but suspect the difference here is VAX running 5.5 and ( > > Alpha running something a bit later. > > D > > There was a big change to how Security worked between VMS 5 & 6." > > This may be part of the issue? > G > It may be something along those lines. An additional quirk on VMS 5.5 H > is that the access() C runtime library function will always return theD > correct response regardless of whether you have read access to the' > file, while sys$check_access() fails.  > E > I presume that access() is using a different system call to get the H > information, but as VMS isn't open source I can't find out what it is.E > Anyway, access() is documented as not checking ACLs (just the basic E > file permissions) and I need to check ACLs - so the source code for , > access() probably wouldn't help me anyway! > F > I'll enquire with our IT department to see whether this problem withF > read access and sys$check_access has had a patch issued at some time > in the past. > B > But if anyone can think of anything else, please do let me know. >  > Colin    --  C Jilly	- Working from Home in the Chemung River Valley - Waverly, NY 0 	- jilly@clarityconnect.com			- Brett Bodine fan+ 	- Mark.Jilson@HP.com				- since 1975 or so  	- http://www.jilly.baka.com   ------------------------------  % Date: Wed, 15 May 2002 08:27:49 -0700 ' From: David Mathog <mathog@caltech.edu> P Subject: Re: A (free) VMS environment in a Linux/Unix system ?, was: Re: freeVMS+ Message-ID: <3CE27E75.735ED934@caltech.edu>    Simon Clubley wrote: > k > In article <abr2he$5ia$1@tyfon.itea.ntnu.no>, Roar =?iso-8859-1?Q?Thron=E6s?= <roart@nvg.ntnu.no> writes: G > > Simon Clubley <clubley@remove_me.eisner.decus.org-earth.ufp> wrote:  > > O > > : I really like what they are doing and hope that they succeed, but I can't M > > : help wondering if the best short term approach might be to provide some @ > > : kind of a VMS environment on Linux and other Unix systems. > >  > > : What do others think ? > > N > > : If you agree, then what do you think the most important parts of the VMSA > > : environment are to get running on these platforms and why ?  > > < > > You want something like http://www.accelr8.com/dcl.html? > >  > J > I am aware of this product as well as the limited version free download.K > Although DCL would be of interest, I would only use it to do those things = > that are especially cumbersome in Unix and easy in VMS, ie:  > $ >         $ search [...]*.txt string  I I estimate that a search utility that would accept appropriate/equivalent O expressions could be written in Perl or C in under a day.  I'm not volunteering O mind you, just stating that if you want this the effort to achieve it is fairly C minimal.  It is, however, substantially greater than just using the D longer, cumbersome (agreed) find command.  If you use my extract and execinput programs from   =   ftp://saf.bio.caltech.edu/pub/software/linux_or_unix_tools/   : you can use this longer but slightly more flexible method:  J   % find . | grep \.txt$ | extract -mt -cols 'echo searching in [1] ; grep string [1]' | execinput   G Of course this will screw up on file names with spaces, semicolons, and C the line in them.  Hmm, this points out a feature this needs that I ? had not thought of up until now - another format option for the D extracted string to pad out the string with  unix escape characters. I'llE go put that in now, so that -cols '[ue:1]' will handle these sorts of  pathological strings.    Regards,   David Mathog mathog@caltech.edu   ------------------------------  + Date: Wed, 15 May 2002 15:48:47 +0000 (UTC) % From: John Forkosh <john@invalid.com> P Subject: Re: A (free) VMS environment in a Linux/Unix system ?, was: Re: freeVMS, Message-ID: <abu00v$q3d$1@reader1.panix.com>  C Simon Clubley <clubley@remove_me.eisner.decus.org-Earth.UFP> wrote:  <<snip>>8 : The kinds of things that I would really like to see on : Linux/Unix are:  : I : 2) RMS. The purpose here is to have a decent free ISAM facility (I know J :    that C-ISAM is commercially available) in it's own right, and RMS has@ :    proven to be very reliable and feature rich over the years.  9 Try Berkeley db.  That's actually the current "standard". D The current software is freely available (for noncommercial/hobbyist? purposes) from  www.sleepycat.com, and works real well for most  isam requirements. --  > John Forkosh  ( mailto:  j@f.com  where j=john and f=forkosh )   ------------------------------  + Date: Wed, 15 May 2002 16:02:05 +0000 (UTC) % From: John Forkosh <john@invalid.com> P Subject: Re: A (free) VMS environment in a Linux/Unix system ?, was: Re: freeVMS, Message-ID: <abu0pt$q3d$2@reader1.panix.com>  & John Forkosh <john@invalid.com> wrote:E : Simon Clubley <clubley@remove_me.eisner.decus.org-Earth.UFP> wrote: 
 : <<snip>>: : : The kinds of things that I would really like to see on : : Linux/Unix are:  : : K : : 2) RMS. The purpose here is to have a decent free ISAM facility (I know L : :    that C-ISAM is commercially available) in it's own right, and RMS hasB : :    proven to be very reliable and feature rich over the years.  ; : Try Berkeley db.  That's actually the current "standard". F : The current software is freely available (for noncommercial/hobbyistA : purposes) from  www.sleepycat.com, and works real well for most  : isam requirements.  7 I forgot to mention that I think (haven't tried it yet) 5 Berkeley db should build okay under vms, too.  So you 8 could first migrate apps away from rms and onto the more6 portable Berkeley db, while continuing to use whatever5 other vms facilities the apps require.  Then, if/when 8 the time comes, migrating the apps entirely over to unix8 would hopefully be a little less painful.  It's not that: I'm advocating doing that, but just recognizing the likely5 inevitability.  So a minimally painful migration plan  is a good thing. --  > John Forkosh  ( mailto:  j@f.com  where j=john and f=forkosh )   ------------------------------  % Date: Wed, 15 May 2002 09:53:10 -0700 ' From: David Mathog <mathog@caltech.edu> P Subject: Re: A (free) VMS environment in a Linux/Unix system ?, was: Re: freeVMS+ Message-ID: <3CE29276.C69829A9@caltech.edu>    David Mathog wrote:  >  > Simon Clubley wrote: > > m > > In article <abr2he$5ia$1@tyfon.itea.ntnu.no>, Roar =?iso-8859-1?Q?Thron=E6s?= <roart@nvg.ntnu.no> writes: I > > > Simon Clubley <clubley@remove_me.eisner.decus.org-earth.ufp> wrote:  > > > Q > > > : I really like what they are doing and hope that they succeed, but I can't O > > > : help wondering if the best short term approach might be to provide some B > > > : kind of a VMS environment on Linux and other Unix systems. > > >  > > > : What do others think ? > > > P > > > : If you agree, then what do you think the most important parts of the VMSC > > > : environment are to get running on these platforms and why ?  > > > > > > > You want something like http://www.accelr8.com/dcl.html? > > >  > > L > > I am aware of this product as well as the limited version free download.M > > Although DCL would be of interest, I would only use it to do those things ? > > that are especially cumbersome in Unix and easy in VMS, ie:  > > & > >         $ search [...]*.txt string >  >  If you use my extract and > execinput programs from  > ? >   ftp://saf.bio.caltech.edu/pub/software/linux_or_unix_tools/  > < > you can use this longer but slightly more flexible method: > L >   % find . | grep \.txt$ | extract -mt -cols 'echo searching in [1] ; grep > string [1]' | execinput  > I > Of course this will screw up on file names with spaces, semicolons, and  > the line in them.   J Not any more.  The new command line switch is -bs (hauntingly appropriate) since:  ' % echo "windows document" | extract -bs  windows\ document   N and the modified find line, which should work with any file name no matter how hideous is:   J  % find . | grep \.txt$ | extract -bs -cols 'echo searching in [1,] ; grep string [1,]' | execinput   Regards,   David Mathog mathog@caltech.edu   ------------------------------  # Date: Thu, 09 May 2002 21:50:29 GMT 3 From: sy18889@rabbit.fmr.com (Bradford J. Hamilton) K Subject: Acrobat (Was: Re: Scott Stallard, forget your VMS to HP UX dream!) . Message-ID: <FaCC8.20$O3.46@news-srv1.fmr.com>  $ Actually, you may both be wrong		:-)  N Documents produced with Acrobat allow the creator to select various "security"J features, including disallowing the ability to select text for cutting andF pasting.  We use this feature at work to publish PDF's on our website,O appropriately "locked down" so that others cannot "steal" our work and claim it I as their own (this actually happened to us, which is why we purchased the 
 software!)  N I'll refrain from passing judgement here, except to note that personal attacks& are frowned upon in Usenet postings...  m In article <7nAC8.73653$M7.6945774@bin7.nnrp.aus1.giganews.com>, "Bill Todd" <billtodd@metrocast.net> writes:  > 6 >"Bob Ceculski" <bob@instantwhip.com> wrote in message8 >news:d7791aa1.0205091030.5857582b@posting.google.com... >u >....  >s? >> Sorry, I couldn't get stupid windoze to cut from the pdf ...r > K >And you could have done this on VMS?  I suspect it wasn't Windows that wase >stupid. >S9 >Hint:  the 'text select' tool in Acrobat is your friend.S >v >- billi >  >u >n   Bradford J. Hamilton& braMdhamAilPtoSn@aMtAtPbi.cSom		(home)& sMy1A88P89S@rabMbit.fAmPr.coSm		(work)  ; "All opinions that I express are my own, not my employer's"e "Lose the MAPS"    ------------------------------  % Date: Wed, 15 May 2002 16:16:05 +0000 2 From: John Eisenschmidt <jweisen@eisenschmidt.org>O Subject: Re: Acrobat (Was: Re: Scott Stallard, forget your VMS to HP UX dream!)e4 Message-ID: <20020515161605.B32643@eisenschmidt.org>   --oTHb8nViIGeoXxdp* Content-Type: text/plain; charset=us-ascii Content-Disposition: inlineo+ Content-Transfer-Encoding: quoted-printable    And then some.  L We bought a product last year called Asura (which is owned by a bank in Ger=L many and refused to sell their product to Adobe). What a gem that is! It re=L renders a PDF as all its composite objects and lets you edit them on the fl=L y - reflow text, replace images, edit images, etc etc. Things that Acrobat =) CANNOT do (which is why Adobe bought it).i  L There was a problem a couple years ago about the CIA releasing FOIA documen=L ts that were redacted inside Acrobat - if you paused the application as it =L was paiting the screen you could see the information under the blackline. T=L he only way most of those "security" products for PDF work is to rerender t=G he document as a flat object, but some of the ones I have seen do work.   L As a medium of information exchange, as someone who has worked in print pub=L lishing for several years, PDF is a godsend. As far as the tools Adobe has =? to work with them (editing, securing, etc), pretty pathetic.=20   L Unless the Voices are Mistaken, Bradford J. Hamilton (sy18889@rabbit.fmr.co=	 m) Wrote:n& > Actually, you may both be wrong		:-) >=20L > Documents produced with Acrobat allow the creator to select various "secu= rity" L > features, including disallowing the ability to select text for cutting andH > pasting.  We use this feature at work to publish PDF's on our website,L > appropriately "locked down" so that others cannot "steal" our work and cl= aim itK > as their own (this actually happened to us, which is why we purchased thed > software!) >=20L > I'll refrain from passing judgement here, except to note that personal at= tacksR( > are frowned upon in Usenet postings... >=20L > In article <7nAC8.73653$M7.6945774@bin7.nnrp.aus1.giganews.com>, "Bill To=$ dd" <billtodd@metrocast.net> writes: > >d8 > >"Bob Ceculski" <bob@instantwhip.com> wrote in message: > >news:d7791aa1.0205091030.5857582b@posting.google.com... > >n > >....s > > A > >> Sorry, I couldn't get stupid windoze to cut from the pdf ...l > >eL > >And you could have done this on VMS?  I suspect it wasn't Windows that w= as
 > >stupid. > >o; > >Hint:  the 'text select' tool in Acrobat is your friend.N > >A	 > >- bill  > >t > >t > >  >=20 > Bradford J. Hamilton( > braMdhamAilPtoSn@aMtAtPbi.cSom		(home)( > sMy1A88P89S@rabMbit.fAmPr.coSm		(work) >=20= > "All opinions that I express are my own, not my employer's"e > "Lose the MAPS"l   --=20 / John W. Eisenschmidt <jweisen@eisenschmidt.org>b6  Homepage URL    | http://www.eisenschmidt.org/jweisenL  PGP Public Key  | http://www.eisenschmidt.org/jweisen/misc/jeisenschmidt.a= scD  PGP Fingerprint | 5F9B F916 5AD1 3295 CF99 BC1E 1F97 E6A3 37E3 BEF2  L FOO MANE PADME HUM: "Our first obligation is to keep the FOO counters turni= ng."   --oTHb8nViIGeoXxdp' Content-Type: application/pgp-signature  Content-Disposition: inlinee   -----BEGIN PGP SIGNATURE-----M Version: GnuPG v1.0.6 (OpenBSD)X* Comment: For info see http://www.gnupg.org  @ iD8DBQE84onFH5fmozfjvvIRApbPAKCcjPoEtNABPa22DKpAYl/TS40JBwCggPSi S/scux5JHu0T2qn+zLe7JJg= =ZCJSu -----END PGP SIGNATURE-----b   --oTHb8nViIGeoXxdp--   ------------------------------  % Date: Wed, 15 May 2002 12:39:49 -0400 2 From: Atlant Schmidt <atlantnospam@mindspring.com>O Subject: Re: Acrobat (Was: Re: Scott Stallard, forget your VMS to HP UX dream!)-- Message-ID: <3CE28F55.19F9454@mindspring.com>-   "Bradford J. Hamilton" wrote:0  - > Actually, you may both be wrong         :-)u >cP > Documents produced with Acrobat allow the creator to select various "security"L > features, including disallowing the ability to select text for cutting andH > pasting.  We use this feature at work to publish PDF's on our website,Q > appropriately "locked down" so that others cannot "steal" our work and claim it K > as their own (this actually happened to us, which is why we purchased theh > software!)  B In which case you just use Acrobat's "print to a file" capability,< then cut the text out of the resulting unprotected .ps file.   Note:   =   Unless you read Forth, you may need a ps-to-text converter.e     Atlant   ------------------------------  % Date: Wed, 15 May 2002 16:41:38 +0000t2 From: John Eisenschmidt <jweisen@eisenschmidt.org>O Subject: Re: Acrobat (Was: Re: Scott Stallard, forget your VMS to HP UX dream!)o4 Message-ID: <20020515164138.A26079@eisenschmidt.org>   --tKW2IUtsqtDRztdT* Content-Type: text/plain; charset=us-ascii Content-Disposition: inline + Content-Transfer-Encoding: quoted-printablee   Print Inhibit Security.d  	 Next =3D)y  L Unless the Voices are Mistaken, Atlant Schmidt (atlantnospam@mindspring.com= ) Wrote: > "Bradford J. Hamilton" wrote:  >=20/ > > Actually, you may both be wrong         :-)  > > L > > Documents produced with Acrobat allow the creator to select various "se= curity":L > > features, including disallowing the ability to select text for cutting = and@J > > pasting.  We use this feature at work to publish PDF's on our website,L > > appropriately "locked down" so that others cannot "steal" our work and = claim itL > > as their own (this actually happened to us, which is why we purchased t= he > > software!) >=20D > In which case you just use Acrobat's "print to a file" capability,> > then cut the text out of the resulting unprotected .ps file. >=20 > Note:t >=20? >   Unless you read Forth, you may need a ps-to-text converter.  >=20 >=20 > Atlant >=20   --=20y/ John W. Eisenschmidt <jweisen@eisenschmidt.org>n6  Homepage URL    | http://www.eisenschmidt.org/jweisenL  PGP Public Key  | http://www.eisenschmidt.org/jweisen/misc/jeisenschmidt.a= scD  PGP Fingerprint | 5F9B F916 5AD1 3295 CF99 BC1E 1F97 E6A3 37E3 BEF2  L FOO MANE PADME HUM: "Our first obligation is to keep the FOO counters turni= ng."   --tKW2IUtsqtDRztdT' Content-Type: application/pgp-signature  Content-Disposition: inlinel   -----BEGIN PGP SIGNATURE-----1 Version: GnuPG v1.0.6 (OpenBSD)c* Comment: For info see http://www.gnupg.org  @ iD8DBQE84o/BH5fmozfjvvIRAuRwAJwNOfPHfrU4sO0JeO6cC5MsX4HIAgCgwMcZ QKCOd6cYzBLbr1H4Uh5cjEs= =xav4S -----END PGP SIGNATURE-----h   --tKW2IUtsqtDRztdT--   ------------------------------  % Date: Wed, 15 May 2002 13:08:37 -0400 2 From: Atlant Schmidt <atlantnospam@mindspring.com>O Subject: Re: Acrobat (Was: Re: Scott Stallard, forget your VMS to HP UX dream!)e. Message-ID: <3CE29615.935C12A6@mindspring.com>   John Eisenschmidt wrote:   > Print Inhibit Security.  >i	 > Next =)    Fine.h     1. Screen-shot, fed to OCR.i  ;   2. Human, 3-cents/hour employee in a third-world country,?       transcribing it.  8 Face it, if it's displayed to a human, it can be copied.     Atlant   ------------------------------  + Date: Wed, 15 May 2002 09:05:42 +0000 (UTC)r From: david20@alpha2.mdx.ac.uk- Subject: Re: Bob Palmer and the demise of DECf+ Message-ID: <abt8d6$lu5$1@aquila.mdx.ac.uk>g  Z In article <ue2jh22vbodm22@corp.supernews.com>, "Mike Foley" <mike-spam@yelof.com> writes: >t? >    Prior to the NT "Affinity" program, there was quite a pushL9 >    within VMS to get those VMS customers who were going  >    to Unix to go to DEC Unix.  >e> >    Unfortunately, I'd say a fair amount ported to Unix using= >    the tools available, only to end up ready to run on Sun.s >n; >    If as much effort went into keeping customers happy ont; >    VMS, we'd ALLbe alot happier. (ie: pushing a marketinga7 >    message that staying with VMS is a smart choice in  >    business, etc..)g >vE >                                                                mikef >    Lots of parallels there. n  E Users "persuaded" to move from VMS to DEC Unix (at that time Ultrix).-I Then along comes Alpha. Users faced with port to a new "DEC UNIX" runningcI on a different chip. Both Unixes owned by same company but port difficult?, enough that many users look at alternatives.  L And that was with porting to a chip and new OS which both looked at least asL good performance wise and in functionality as the chip and OS being replaced3 - a far cry from what Tru64 users are facing today.-      
 David Webb VMS and Unix team leader CCSS Middlesex University        5 >"Carl Perkins" <carl@gerg.tamu.edu> wrote in message ( >news:13MAY200218222753@gerg.tamu.edu...8 >> "Terry C. Shannon" <terryshannon@attbi.com> writes...0 >> }> > Andrew Harrison SUNUK Consultancy wrote:> >> }> > > If is makes you feel any better Sun has a OpenVMS to+ >> }> > > Solaris Affinity Program as well.r >> }G >> }Andrew, you have my email address. I'd be interested (seriously) ink	 >learningw( >> }more about the Sun affinity program. >> } >> }terry st >>1 >> The difference between Sun and Digital/Compaq:u >>A >> Sun's "affinity" program is to more people from a competitor's & >> OS (VMS) to their own OS (Solaris). >>D >> Digital/Compaq's "affinity" program was to move people from their2 >> own OS (VMS) to a competitor's OS (Windows NT). >>A >> HP may be, in some sense, taking a setp up - at least they mayy@ >> be trying to shift people from one of their OSes (VMS) (well,D >> two of them actually as you can add in Tru64) to another of theirH >> own OSes (HPUX). This may be a bit less stupid than what has happenedD >> before, from the persective of a company trying to make money forB >> themselves rather than for Microsoft. However, note that "a bit+ >> less stupid" is not the same as "smart".m >> >> --- Carld >| >e   ------------------------------  % Date: Wed, 15 May 2002 09:06:21 +0100x& From: Ken Green <Ken.Green@kgcc.co.uk>< Subject: Re: Capellas: Linux, Windows Will 'Eviscerate' Unix* Message-ID: <3CE216FD.65DA2AB1@kgcc.co.uk>   Bob Ceculski wrote:h   > Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> wrote in message news:<abrg09$1df$1@new-usenet.uk.sun.com>...  > > Bob Koehler wrote: > >a > > > In article <abgrtf$1ib$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> writes: > > >2 > > >>MPE has clusters > > >> > > >> > > >nB > > >    Yeah, right.  MPE has clusters like Solaris has clusters. > > >mL > > >    Back to reality:  HP has clusters.  Back when DEC still existed, HPJ > > >    said VMS clusters ARE the definitive implementation and that they- > > >    would like to have as good in HP-UX.e > > >  > >/ > >a> > > Nope OpenVMS clusters are not the definitive cluster, theyA > > represent on strand of cluster developement. There are othersn: > > IBM Sysplex, BeoWolf, Application Server Clusters etc. > > G > > Understanding this might help you understand why not everyone needsr@ > > or apsires to having an OpenVMS cluster in their datacenter. > >e > > Regardst
 > > Andrew >mI > Andrew, everyone knows that VMS clusters are still superior to anything J > out there ... many have tried to copy it, but no one has ... tests proveF > this year after year ... to speak otherwise is to make a fool out of > yourself!r  J Bob, what Andrew is saying is that the word cluster is heavily overloaded.N It means different things to different people, no one solution is likely to be perfect for all of them.  J For HP-UX  there has been at least three totally different non overlapping uses of the word cluster.i  J Diskless clusters (although that doesn't mean systems couldn't have disks)H This made a group of workstations act like a large single system, it wasM nothing like NFS diskless. All the systems had a single common root directorysJ and filesystem. 99.9% of admin was just the same as it would have been for a single stand alone system.H I suspect if you asked most workstation or PC users what they would like. a network to look like, it would be like this.   Computer clusters/snake farms.G A bunch of workstations (or any other box) with a distributed schedulere   HA clusters/ Service GuardC A group of systems that watch each others backs, and lend a helpings% hand if one system gets into trouble.b   These are all called clusters.H Yet they all provide very different functionality, they were designed toD meet the needs of different groups of users/apps/businesses. Not all9 of these would want the functionality the others provide.d  C I realise that DEC came out with a product called "Clusters" a longl> time before most Unix vendors had even thought linking systemsB together, but the word has taken on lots of different meanings all of which are equally valid.,   Cheers   Ken3   ------------------------------  % Date: Wed, 15 May 2002 09:43:03 +0100 & From: Ken Green <Ken.Green@kgcc.co.uk>< Subject: Re: Capellas: Linux, Windows Will 'Eviscerate' Unix* Message-ID: <3CE21F97.88DBE38C@kgcc.co.uk>   Rob Young wrote:   > >tA > > BTW HP-UX had shared disk cluster from HP-UX 6.0 through to 96 > > before dropping them.t > >  > * >         Couldn't get it right, it seems. >. >-  ; Not at all, they got it right first time, but times change.   E The Diskless Ux environment gave full SVID compliant system behaviour G for a smallish group of workstaions. It included full coherency for the G buffer caches, in most case this worked fine and quickly, but we tended F to find some programs (mostly Fortran but some C) which tended to openN files read/write unnessessarily, and so force the double caching to be dropped  in favour of disk owner caching.  F However in order to make this work at a reasonable speed when it first@ came out the decision was made to use raw IEEE 802.3 LAN packetsH rather than a nice high level protocol like TCP/IP or DECNET, since most> systems LAN performance was CPU power limited rather than wire speed limited.  H At the time this came out I was working for DECs largest UK distributor,I and will never forget the face on one of the salesman faces when half way/F through a demo to some customers of LAVC he pointed to the LAN trafficJ monitor behind him (without looking) and said something along the lines ofL "And as you can see, with all these VAXs in the building clustered we rarelyK see the network above 4% utilisation, so you can see Ethernet has plenty ofhG bandwidth", to which the customer said "Hmmmm...... well it look prettyrE well saturated to me", no one had warned me, and I was was doing somed> stress testing forcing a workstation to page over the network.  H The downside to this approach was that it was a single LAN based design.G It couldn't be routed, all the clients needed to be on the same networkeO segment, it didn't even like a lot of bridges, since at the time a lot of thesea. couldn't really keep up with a full 10Mb link.  I But eventually it was killed because it wasn't standard, one big customeraI (The US Navy I think) wanted NFS diskless clusters for security (not suresF how you can ever equate NFS with security) and so DUX was killed. OnlyJ to be replaced with NFS based diskless. However as soon as people realisedM that this wasn't like a DUX cluster they bought disks for all there machines.e9 (Including the customer who had made sure DUX was knifed)r  D Diskless clusters were invented to tackle the problem that disk wereE expensive. It so happens the solution to this problem also solved thea) a part of the Admin is expensive problem.a  B When the disks are expensive problem went away, and someone killed9 the easy admin solution. Diskless clusters went away too.6   As I say... times change   Cheers   Kend   ------------------------------  % Date: Wed, 15 May 2002 10:30:04 +0100c& From: Ken Green <Ken.Green@kgcc.co.uk>< Subject: Re: Capellas: Linux, Windows Will 'Eviscerate' Unix* Message-ID: <3CE22A9C.D4CC9952@kgcc.co.uk>   Bill Todd wrote:  5 > "Ken Green" <Ken.Green@kgcc.co.uk> wrote in messaget& > news:3CE18838.45DC9379@kgcc.co.uk... > > Bill Todd wrote: > > 9 > > > "Ken Green" <Ken.Green@kgcc.co.uk> wrote in messageo* > > > news:3CE16112.EC41AF98@kgcc.co.uk... > > >m	 > > > ...P > > >M7 > > > > Not everyone wants shared disk, sharing disk ise > SSSSLLLLOOOOOWWWWW!!!!!e > > > : > > > Not at all, at least if you know how (and VMS does). > > >0 > > > - bill > >  > > OK I'll bite,m > >jN > > so how do you get a cohertent buffer cache with multiple systems accessing5 > > the same disk without using some sort of network?. >6M > So how do multiple systems access any shared data at all (whether on shareds2 > disk or not) without using some sort of network? >E  G Sorry this was what I was talking about, some apps need shared and someaG don't, non shared is quicker than shared. Even if you can eliminate anylJ overhead in having the shred capability in place when it is not being used: you can still find it is being used when it is not needed.   >wM > With non-shared disks, if you want to share data among systems they need towM > cooperate, just as they need to using shared disks.  The primary differencetG > due to the sharing of the disk is that when desired data isn't cached N > somewhere the system that wants it can, after obtaining any necessary locks,M > get it directly from (or send it directly to) the disk rather than ask some ( > other system to get it and send it on.  C The problem comes when someone else is caching the data. How do youmF cope. Do you manage cache coherency on a per IO basis or on a per file basis.  O If you try to do it on a per IO basis, then every read or write needs to check,t  L so you loose the advantage of local caching. You could minimise this to onlyI impacting writes I guess but that would require all systems to maintain a & cache directory for the whole cluster.  J If you do it on a per file basis, it's much easier. At open time you checkM whether you are the only node accessing the file, or that all nodes accessing K are only open for reading, in this case there is no problem with caching atpF both ends. However once a node opens a shared file for writes then theD caching needs to be done in just one place, or as above checking forF conflicts needs to be performed on a per IO basis, or write only basis with a local cache directory.e  G Sorry I don't know how VMS tackles these issues, it has been more yearstE than I care to remember since I used a VMS cluster, and back then hadd? no idea how an OS worked. I'm sure VMS clusters have moved on ai long way since then anyway.a   >  >wM > In large systems, this means that  1) there's no need to partition the data>L > among the various systems that 'serve' it (because all systems have directJ > access to all data),  2) there's no problem after partitioning said dataM > with a server getting overloaded due to hot spots (because, again, all data H > is directly accessible to all who need it), and  3) the storage can beJ > scaled independently of the processing capacity (and without the need to > repartition).n >h  L Not sure about 2), if they are sharing data unnessessarily then the data canL hot spot. If local copies of the data were being maintained then it would beK quicker. This is of course dependant on whether you think you need to sharenG the data. My whole point was that some times you do, and some times youC don't.   >3M > And eliminating the need for 'server'-style operation does not increase therL > network traffic (thus does not increase access latency).  Obtaining a lockJ > from whoever is in a position to grant it (which is usually known by theK > requestor - but I'm not going to reprise the VMS distributed lock manager8K > algorithms here) takes one exchange, after which the requestor can obtain7M > the data from disk, which is the same number of operations that would occur-N > if instead the requestor asked a 'server' to obtain the data and forward it.J > (Actually, it can be less, because the VMS DLM algorithms make it likelyI > that most lock requests are local, hence the disk access can take placed > directly.) >B  M If the data is not clean in the cache of the system that is "in a position toc	 grant it"tM then either the data needs to be exchanged across the network, or it needs tor beN written back to the disk, before the requestor can ask for the upto date copy.K A different app design where the data didn't need sharing could be quicker.l     >oN > With a good distributed cache (such as the one Oracle recently developed forL > its 'cache fusion' architecture) you can do even better due to the abilityN > to combine cache-management messages with lock-management messages (and onceJ > you have such a distributed cache, shared storage becomes a definite win > over served storage).-  L I guess my main issue is how you get a good distributed cache without addingM significantly to the latency of a cache access. I'll need to looking into theu Oracle cache fusion architecture.  6 >  VMS unfortunately dropped the ball on this for manyH > years, but is finally getting there with XFC (when they get it workingJ > right).  Meanwhile, there is the legitimate knock that a node that wantsJ > access to data that another node has cached must still go to disk for it  F HP-UX used to use double caching here, but only where there was either! no sharing, or read only sharing.o   >oF > (whereas if it was cached at a 'server' then no disk access would beM > required) - but this is an anomaly of the VMS implementation rather than any0 > inherent problem with concurrent disk-sharing. >o > - bill  K There are other problems with disk sharing, such as sharing the IO rate andF connection bandwidth too.r    K Please don't think I am against all sharing and clustering, that is not the  caseE as I have written in another posting, the origenal issue was that VMSe
 ClusteringM is not the only type of clustering, the term is horribly overloaded and it is,	 doubtfullt@ that anyone solution will be right for all the uses of the term.  G It just like some application work best on a large SMP or ccNUMA server I such as a SuperDome or GS (and I think these are very different from whathM I can see) and some applications work best on a rack full of small systems or   blades. It's horses for courses.  M BTW, as I said it is many years since I last used a VMS cluster, back then weCK had our 11/785s linked by a series of cables a good couple of inches thick,pM for the CI bus. Since the clustering technology IS going to be added to HP-UX/O I want to learn all I can about it, can you recommend any good places to start,e  N I know I have a lot to learn, in my own case it's going to need to be from theL high level overview down to VMS distributed lock manager algorithms, if they7 are going to show up in the HP-UX kernel in the future.U     Cheers   Ken-   ------------------------------   Date: 15 May 2002 12:17:18 GMT& From: peter@abbnm.com (Peter da Silva)< Subject: Re: Capellas: Linux, Windows Will 'Eviscerate' Unix- Message-ID: <abtjke$pfv@web.eng.baileynm.com>3  3 In article <4Y3fAhA+Hm4y@eisner.encompasserve.org>,r. Bob Koehler <koehler@encompasserve.org> wrote:X > In article <abgoh6$1lj@web.eng.baileynm.com>, peter@abbnm.com (Peter da Silva) writes:N > > How does "see where the configuration information gets converted to actualO > > changes in system state" translate to anything even vaguely resembling "alle8 > > configuration files are written by startup scripts"?  : >    Oh, so you expect me to disassemble to HPPARC binary?  H You can if you want, but I don't think that will provide any useful infoI on your son's Linux box. Which is, after all, what we were talking about.r  " >    I've got better things to do.  = Keeping track of the discussion is obviously not one of them.w   -- nO I've seen things you people can't imagine. Chimneysweeps on fire over the roofsnO of London. I've watched kite-strings glitter in the sun at Hyde Park Gate.  AllSL these things will be lost in time, like chalk-paintings in the rain.   `-_-'K Time for your nap.  | Peter da Silva | Har du kramat din varg, idag?    'U`y   ------------------------------    Date: 15 May 2002 08:46:07 -0600- From: koehler@encompasserve.org (Bob Koehler) < Subject: Re: Capellas: Linux, Windows Will 'Eviscerate' Unix3 Message-ID: <Tt69DzL8K18j@eisner.encompasserve.org>U   In article <abrg09$1df$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> writes:  < > Nope OpenVMS clusters are not the definitive cluster, they/ > represent on strand of cluster developement.    :    So you disagree with HP, which now have what they want.    Not suprizing.h   ------------------------------    Date: 15 May 2002 09:32:58 -0600+ From: young_r@encompasserve.org (Rob Young)t< Subject: Re: Capellas: Linux, Windows Will 'Eviscerate' Unix3 Message-ID: <MrQWAycLmtYL@eisner.encompasserve.org>2  S In article <3CE21F97.88DBE38C@kgcc.co.uk>, Ken Green <Ken.Green@kgcc.co.uk> writes:  > Rob Young wrote: >  >> >B >> > BTW HP-UX had shared disk cluster from HP-UX 6.0 through to 9 >> > before dropping them. >> > >>+ >>         Couldn't get it right, it seems.  >> >> > = > Not at all, they got it right first time, but times change.o > G > The Diskless Ux environment gave full SVID compliant system behaviour I > for a smallish group of workstaions. It included full coherency for theeI > buffer caches, in most case this worked fine and quickly, but we tended>H > to find some programs (mostly Fortran but some C) which tended to openP > files read/write unnessessarily, and so force the double caching to be dropped" > in favour of disk owner caching. > H > However in order to make this work at a reasonable speed when it firstB > came out the decision was made to use raw IEEE 802.3 LAN packetsJ > rather than a nice high level protocol like TCP/IP or DECNET, since most@ > systems LAN performance was CPU power limited rather than wire > speed limited. > J > At the time this came out I was working for DECs largest UK distributor,K > and will never forget the face on one of the salesman faces when half wayeH > through a demo to some customers of LAVC he pointed to the LAN trafficL > monitor behind him (without looking) and said something along the lines ofN > "And as you can see, with all these VAXs in the building clustered we rarelyM > see the network above 4% utilisation, so you can see Ethernet has plenty ofrI > bandwidth", to which the customer said "Hmmmm...... well it look prettynG > well saturated to me", no one had warned me, and I was was doing someu@ > stress testing forcing a workstation to page over the network. > J > The downside to this approach was that it was a single LAN based design.I > It couldn't be routed, all the clients needed to be on the same networkHQ > segment, it didn't even like a lot of bridges, since at the time a lot of thesee0 > couldn't really keep up with a full 10Mb link. > K > But eventually it was killed because it wasn't standard, one big customertK > (The US Navy I think) wanted NFS diskless clusters for security (not suredH > how you can ever equate NFS with security) and so DUX was killed. OnlyL > to be replaced with NFS based diskless. However as soon as people realisedO > that this wasn't like a DUX cluster they bought disks for all there machines.e; > (Including the customer who had made sure DUX was knifed)o > F > Diskless clusters were invented to tackle the problem that disk wereG > expensive. It so happens the solution to this problem also solved thee+ > a part of the Admin is expensive problem.u > D > When the disks are expensive problem went away, and someone killed; > the easy admin solution. Diskless clusters went away too.t >   ? 	I've read this and your other reply.  You are abusing the termo? 	cluster but this isn't anything new.  It led Pfister to createlC 	a nice tome as in his travels and talks he came to realize dealingt= 	with technical people he was astounded at their same varyingf4 	definitions and so penned:  "In Search of Clusters"  C 	A VAX/VMS cluster of 17 years ago had solved the "problem" you aree@ 	describing in the last two paragraphs.  It is called a CI basedD 	cluster where all drives are seen by all members of the cluster andE 	typically share a common single system disk.  One nearby has 3 largei> 	VAXes.  None of the VAXes has a disk inboard.  The CI acts asC 	a private network to the storage with lock management handling thee. 	disk traffic (keep from stomping each other).  D 	Now maybe you want a "spread cluster" in 1990.  Okay, you are usingC 	a LAN and have 10 MBit to contend with and certainly aren't doing p@ 	large amounts of IO.  The CI is 140 MBit.  So yes... scientificA 	computing is a nice target for the "cluster" you are describing.t? 	Certainly not a business performing a lot of transactions evens 	15-20 years ago.   A 	I am sure part of the impetus for creating a "true" cluster (allr? 	disk resources shared) was the fact that storage was not cheapk 	20 years ago.   					Rob   ------------------------------  % Date: Wed, 15 May 2002 16:27:28 +0100-& From: Ken Green <Ken.Green@kgcc.co.uk>< Subject: Re: Capellas: Linux, Windows Will 'Eviscerate' Unix* Message-ID: <3CE27E60.853D1927@kgcc.co.uk>   Rob Young wrote:  U > In article <3CE21F97.88DBE38C@kgcc.co.uk>, Ken Green <Ken.Green@kgcc.co.uk> writes:d > > Rob Young wrote: > >0 > >> >D > >> > BTW HP-UX had shared disk cluster from HP-UX 6.0 through to 9 > >> > before dropping them. > >> > > >>- > >>         Couldn't get it right, it seems.  > >> > >> > > ? > > Not at all, they got it right first time, but times change.l > >gI > > The Diskless Ux environment gave full SVID compliant system behaviour K > > for a smallish group of workstaions. It included full coherency for the K > > buffer caches, in most case this worked fine and quickly, but we tendedrJ > > to find some programs (mostly Fortran but some C) which tended to openR > > files read/write unnessessarily, and so force the double caching to be dropped$ > > in favour of disk owner caching. > >EJ > > However in order to make this work at a reasonable speed when it firstD > > came out the decision was made to use raw IEEE 802.3 LAN packetsL > > rather than a nice high level protocol like TCP/IP or DECNET, since mostB > > systems LAN performance was CPU power limited rather than wire > > speed limited. > >hL > > At the time this came out I was working for DECs largest UK distributor,M > > and will never forget the face on one of the salesman faces when half way J > > through a demo to some customers of LAVC he pointed to the LAN trafficN > > monitor behind him (without looking) and said something along the lines ofP > > "And as you can see, with all these VAXs in the building clustered we rarelyO > > see the network above 4% utilisation, so you can see Ethernet has plenty ofrK > > bandwidth", to which the customer said "Hmmmm...... well it look pretty I > > well saturated to me", no one had warned me, and I was was doing somecB > > stress testing forcing a workstation to page over the network. > >lL > > The downside to this approach was that it was a single LAN based design.K > > It couldn't be routed, all the clients needed to be on the same network,S > > segment, it didn't even like a lot of bridges, since at the time a lot of thesen2 > > couldn't really keep up with a full 10Mb link. > >sM > > But eventually it was killed because it wasn't standard, one big customer M > > (The US Navy I think) wanted NFS diskless clusters for security (not surelJ > > how you can ever equate NFS with security) and so DUX was killed. OnlyN > > to be replaced with NFS based diskless. However as soon as people realisedQ > > that this wasn't like a DUX cluster they bought disks for all there machines.d= > > (Including the customer who had made sure DUX was knifed)l > >nH > > Diskless clusters were invented to tackle the problem that disk wereI > > expensive. It so happens the solution to this problem also solved the-- > > a part of the Admin is expensive problem.0 > >0F > > When the disks are expensive problem went away, and someone killed= > > the easy admin solution. Diskless clusters went away too.9 > >c >oH >         I've read this and your other reply.  You are abusing the term. >         cluster but this isn't anything new.  F As I said, the term cluster has become horribly overloaded. This isn'tD good. I wish it hadn't, but we have to live in a world where lots of1 people use the term for lots of different things.   J This is what I have largely been trying to say, Your definition of clusterG is not the only one. It might be a good one, but it's not the only one.   ) I guess if pushed I'd define a cluster as   ) "A group of systems cooperating together"a     > It led Pfister to createL >         a nice tome as in his travels and talks he came to realize dealingF >         with technical people he was astounded at their same varying= >         definitions and so penned:  "In Search of Clusters"u >s   I'll look it out.a   >mL >         A VAX/VMS cluster of 17 years ago had solved the "problem" you areI >         describing in the last two paragraphs.  It is called a CI based M >         cluster where all drives are seen by all members of the cluster andaN >         typically share a common single system disk.  One nearby has 3 largeG >         VAXes.  None of the VAXes has a disk inboard.  The CI acts asoL >         a private network to the storage with lock management handling the7 >         disk traffic (keep from stomping each other).r >oM >         Now maybe you want a "spread cluster" in 1990.  Okay, you are usingsK >         a LAN and have 10 MBit to contend with and certainly aren't doingoI >         large amounts of IO.  The CI is 140 MBit.  So yes... scientificcJ >         computing is a nice target for the "cluster" you are describing.  F The main user was CAD systems, if you sell workstations these guys are your bread and butter.  J At the time 10Mbits was a reasonable speed for disk. The HP systems at theC time used HP-IB (IEEE488) with a tweaked handshake. This could onlyeJ deliver 1MB/s so a 10Mb lan was as fast as the disk interface. The typicalI workstation disk (say a HP7957 or 58) probably only deliver about 350KB/scF The server disks (HP7936/37) would do the whole 1MB/s, I'm sure when IL compared these to the RA81s on VAX I used to run, the HP disks were quicker.' So the LAN link speed was not an issue.t  E These cluster were designed to be cheap, CI was not a cheap solution.aB Using operational systems as the disk controllers was cheaper thanJ having dedicated disk controllers like the CI bus clusters. (Sorry I can'tC remember the name of the box that connected the shared disks to thed cluster)  P As I said there is more than one definition of a cluster and different solutions# to different to different problems.s  E LAVC, was running DEC clusters over 10Mb LANs, and could get no whereu1 need link speed on the uVax II s we had at Rapid.    >tH >         Certainly not a business performing a lot of transactions even >         15-20 years ago. >r  H Your right, for HP-UX 8 some joker put the DUX stuff onto the 800 seriesL I never did find anyone who had even thought of using it. It was way to slowO to do any IO intensive stuff on. On the other hand almost all workstation sitesaK used it, many even when all systems had local disks. It was easy to manage.        >cJ >         I am sure part of the impetus for creating a "true" cluster (allH >         disk resources shared) was the fact that storage was not cheap >         20 years ago.t >s   And nor I'm sure was CIC   >'- >                                         Robr  H I'm sure the DEC/COMPAQ clustering is great at what it does. I'm glad toC see that it will be added into HP-UX over the next couple of years.s  G However the term cluster has become over used, and I'm sure that no one M solution is likely to be the perfect match for all the things the word is nowe
 used to mean.r     Cheers   Kenr   ------------------------------  # Date: Wed, 15 May 2002 16:29:14 GMTs* From: "Bill Todd" <billtodd@metrocast.net>< Subject: Re: Capellas: Linux, Windows Will 'Eviscerate' UnixA Message-ID: <u1wE8.26077$Ze4.2538864@bin5.nnrp.aus1.giganews.com>y  3 "Ken Green" <Ken.Green@kgcc.co.uk> wrote in messagec$ news:3CE216FD.65DA2AB1@kgcc.co.uk...   ...t  L > For HP-UX  there has been at least three totally different non overlapping > uses of the word cluster.n >sL > Diskless clusters (although that doesn't mean systems couldn't have disks)J > This made a group of workstations act like a large single system, it wasE > nothing like NFS diskless. All the systems had a single common root 	 directory L > and filesystem. 99.9% of admin was just the same as it would have been for > a single stand alone system.J > I suspect if you asked most workstation or PC users what they would like0 > a network to look like, it would be like this. >   > Computer clusters/snake farms.I > A bunch of workstations (or any other box) with a distributed scheduler  >h > HA clusters/ Service GuardE > A group of systems that watch each others backs, and lend a helping'' > hand if one system gets into trouble.h >u  > These are all called clusters.  I I suspect that the point you are missing is that VMS clustering (from thehK very beginning) completely covered the first and third kinds of 'different'gK clusters you describe above, and to at least some degree the second as wellcL (though the distributed scheduler would have to have been an add-on).  SinceL the second may be the easiest to create outside the base OS (see Beowulf), I' don't consider that a significant lack.n  D It really is the case that no one has significantly extended the VMSI clustering mechanisms since they were created 18 years ago, and that mosthC have only implemented subsets of them.  VMS even provides the basicuI mechanisms (via the lock manager) for multiple distributed instances of atB single application to coordinate its own clustered cooperation andK operational redundancy, though this is an area where other systems *may* (IlJ just don't happen to know) have come up with more application hand-holdingG facilities than VMS supports.  And AFAIK *no one* except IBM's ParalleluJ Sysplex has a fully-distributed clustered file system equivalent to VMS's:F other file systems (both vendor and third-party) depend upon fail-overH (rather than concurrent) shared storage access for at least metadata andK usually for data as well, with the resulting melange of cascading fail-overdL scripts to determine who gets what portion of the file system after multipleL failures occur (plus the potential for hot spots due to the served nature ofA the metadata and, usually, data).  Plus, VMS also supports use of I host-private storage exported as shared (with full distributed redundancyaF mechanisms) for situations where that may be preferable to use of true concurrently-shared storage.   - bill   ------------------------------  # Date: Wed, 15 May 2002 17:15:04 GMTp# From: "John Smith" <a@nonymous.com> < Subject: Re: Capellas: Linux, Windows Will 'Eviscerate' UnixG Message-ID: <sIwE8.20211$t8_.7539@news01.bloor.is.net.cable.rogers.com>:  9 http://saveateaglestore.site.yahoo.net/corworofsuit2.html-  . Corel Wordperfect Office 2002 Professional OEM $14Y  L reads/writes Word formats and is arguably a better word processor than Word.    : "Warren Spencer" <wspencer@ap.nospam.org> wrote in message1 news:92098C765warrenspencer1977@209.249.90.100...   3 > >Presumably it is worth that to those who buy it.c >lH > For those that buy it, I'd be interested to know if they're new to theJ > product, and thus buy it based on its stand-alone value, or whether manyF > purchases are coerced via MS's relentless push on the upgrade wheel. SeemstJ > to me that some or perhaps many purchases are not based solely on stand- > alone value. >  > ws >e > -- snip -- >r >  > -- >  > Warren Spencer) > Senior Software Engineer (not a writer)s > The Associated Press >w> > ** Time flies like an arrow.  Fruit flies like a bananna. ** >o   ------------------------------  # Date: Wed, 15 May 2002 17:19:12 GMTh* From: "Bill Todd" <billtodd@metrocast.net>< Subject: Re: Capellas: Linux, Windows Will 'Eviscerate' UnixA Message-ID: <kMwE8.26809$fU2.2751161@bin8.nnrp.aus1.giganews.com>i  3 "Ken Green" <Ken.Green@kgcc.co.uk> wrote in message $ news:3CE22A9C.D4CC9952@kgcc.co.uk... > Bill Todd wrote:   ...K  H > > So how do multiple systems access any shared data at all (whether on shared4 > > disk or not) without using some sort of network? > >g >1I > Sorry this was what I was talking about, some apps need shared and somea+ > don't, non shared is quicker than shared.   A Not unless  1) the data you need happens to be local to where thenG application is running (in which case, there's no need for that data tocI reside on a distributed file system at all:  just use a local one for it)-K and  2) the OS is not smart enough to optimize out sharing overhead when nob  sharing in fact is taking place.    Even if you can eliminate anyL > overhead in having the shred capability in place when it is not being used< > you can still find it is being used when it is not needed.  I I suppose in some systems that could be true.  But it's not in VMS:  on aoJ per-file basis, if on file open it is noted that no one else is interestedE in the file, the opening host obtains complete control over it and no K distributed activity is required thereafter unless some other host developsfI a concurrent interest in it (and if that host is found to be accessing it.E more intensely than the original one, then the DLM moves coordinatings responsibility over there).    >n > > L > > With non-shared disks, if you want to share data among systems they need toD > > cooperate, just as they need to using shared disks.  The primary
 differenceI > > due to the sharing of the disk is that when desired data isn't cachedeI > > somewhere the system that wants it can, after obtaining any necessaryE locks,J > > get it directly from (or send it directly to) the disk rather than ask some* > > other system to get it and send it on. > E > The problem comes when someone else is caching the data. How do you H > cope. Do you manage cache coherency on a per IO basis or on a per file > basis.   Per-I/O basis, for VMS.e   >nJ > If you try to do it on a per IO basis, then every read or write needs to check,  L Only if inter-host sharing actually occurs (see above).  And, as I explainedK later, even when it is necessary this does not increase the network messagey- traffic over that required for 'served' data.u   >s. > so you loose the advantage of local caching.  L No.  Once you have the data, you can cache it until some other host wants toF modify it, and reuse it without additional interactions for that time.     You could minimise this to onlyK > impacting writes I guess but that would require all systems to maintain ad( > cache directory for the whole cluster.  L Nope.  I recommend the 1987 Digital Technical Journal explanation of how theE VMS DLM works (it used to be on line at Compaq, anyway).  Or the moreuE extensive explanations in the VMS Internals and Data Structures book.    >tL > If you do it on a per file basis, it's much easier. At open time you checkE > whether you are the only node accessing the file, or that all nodest	 accessingUJ > are only open for reading, in this case there is no problem with caching atH > both ends. However once a node opens a shared file for writes then theF > caching needs to be done in just one place, or as above checking forH > conflicts needs to be performed on a per IO basis, or write only basis > with a local cache directory.   K Nope.  You seem to be sufficiently interested in this that you would likely C find understanding the VMS DLM to be a worthwhile use of your time.n   >yI > Sorry I don't know how VMS tackles these issues, it has been more yearsyG > than I care to remember since I used a VMS cluster, and back then hadtA > no idea how an OS worked. I'm sure VMS clusters have moved on at > long way since then anyway.e  I The basic design/operation of the VMS DLM was there in the first clustersn: release in 1983/4.  Since then most improvements have been performance-related.   >V > >S > >tJ > > In large systems, this means that  1) there's no need to partition the dataG > > among the various systems that 'serve' it (because all systems haveb directL > > access to all data),  2) there's no problem after partitioning said dataJ > > with a server getting overloaded due to hot spots (because, again, all dataJ > > is directly accessible to all who need it), and  3) the storage can beL > > scaled independently of the processing capacity (and without the need to > > repartition).a > >u >eJ > Not sure about 2), if they are sharing data unnessessarily then the data canr > hot spot.   L Whether it's 'necessary' to share data is an application decision.  It's theF OS's job to make that sharing as efficient as possible, and VMS does -F including balancing access across multiple copies (when the VMS volumeI shadowing approach is used) to minimize the impact of any such hot spots.   C  If local copies of the data were being maintained then it would be-
 > quicker.  K VMS certainly allows you to do that as well when you choose to export localGK disks to the cluster rather than use true shared storage:  it biases accessxH to the local copy unless the local disk is far more loaded than a remote copy.<  C  This is of course dependant on whether you think you need to sharedI > the data. My whole point was that some times you do, and some times yout > don't. >  > >hK > > And eliminating the need for 'server'-style operation does not increase  thetI > > network traffic (thus does not increase access latency).  Obtaining as lockL > > from whoever is in a position to grant it (which is usually known by theE > > requestor - but I'm not going to reprise the VMS distributed lockn managerlF > > algorithms here) takes one exchange, after which the requestor can obtainI > > the data from disk, which is the same number of operations that wouldh occurdL > > if instead the requestor asked a 'server' to obtain the data and forward it. L > > (Actually, it can be less, because the VMS DLM algorithms make it likelyK > > that most lock requests are local, hence the disk access can take placeo > > directly.) > >  >EL > If the data is not clean in the cache of the system that is "in a position to > grant it"tL > then either the data needs to be exchanged across the network, or it needs to > beJ > written back to the disk, before the requestor can ask for the upto date copy.,D > A different app design where the data didn't need sharing could be quicker.  J And would be quicker on VMS as well:  as described previously, if the dataL doesn't need sharing, such overheads are avoided (and by definition no data,: dirty or not, will be sitting in some other host's cache).   >u >r > > L > > With a good distributed cache (such as the one Oracle recently developed foryF > > its 'cache fusion' architecture) you can do even better due to the ability K > > to combine cache-management messages with lock-management messages (and  onceL > > you have such a distributed cache, shared storage becomes a definite win > > over served storage).  >tG > I guess my main issue is how you get a good distributed cache withouti addingK > significantly to the latency of a cache access. I'll need to looking intod thes > Oracle > cache fusion architecture.   I think you'll appreciate it.   I First, a distributed cache (rather than a server-local cache that clients,K take advantage of) is *always* preferable as long as the (very low latency)rJ benefit you obtain from being able to cache your own hot data locally (andG access it without any distributed interactions) exceeds the overhead of I having to coordinate access to data across the cluster (which only reallyuL applies to data that at least someone is updating frequently and others needH frequently).  Note that even in many interchange-intensive applications,K while a 'served' copy of the data *may* be somewhat more efficient than the H distributed cache approach, it may well be even better to centralize theH operation entirely (i.e., ship the access *functions* to a central pointH rather than keep bouncing the data back and forth between the server and those who need it).a  J So even with the slow links of a decade ago distributed caching could make& sense for many common access patterns.  H Second, even when the data you need is not cached locally, it costs veryH little more to obtain it from some other node that *does* have it cachedK than from a central server:  a 3-cornered message (requestor to coordinatorsD to actual cacher, and response to requestor) rather than a 2-messageK request/response.  This is I believe what Oracle now does and what XFC willaL eventually do.  Since with optimized protocol stacks one-way message latencyH can be reduced to under 100 us. (and under 10 us. with hardware-enhancedK mechanisms such as VI), and since sending even just a 4 KB 'block' across agH 100 Mbit link takes about 400 us., the difference in that environment atJ worst nearly negligible (and with something like VI nearly negligible evenK using a gigabit link, though in that case the additional interrupt requiredt5 in the 3-corned case may start to become noticeable).v   >e8 > >  VMS unfortunately dropped the ball on this for manyJ > > years, but is finally getting there with XFC (when they get it workingL > > right).  Meanwhile, there is the legitimate knock that a node that wantsL > > access to data that another node has cached must still go to disk for it >fH > HP-UX used to use double caching here, but only where there was either# > no sharing, or read only sharing.r  G VMS allows double-caching as well, even for write-shared files (it justeL invalidates any cached copies at other nodes when a write operation overlaps them).   >  > > H > > (whereas if it was cached at a 'server' then no disk access would beL > > required) - but this is an anomaly of the VMS implementation rather than an2 > > inherent problem with concurrent disk-sharing. > >d
 > > - bill >tI > There are other problems with disk sharing, such as sharing the IO rateo and  > connection bandwidth too.a  G If you need to get to the data on a disk, the I/O rate will be the same7F whether a bunch of hosts access it directly or the same bunch of hostsI access it through a server (modulo the inefficiences introduced by servednK access).  The only exceptions being isochronous access where the server can K decide whom to screw if the total rate exceeds the disk's capacity to serveu it.a  I Similar observations apply to sharing connection bandwidth:  the disk candL supply only as much bandwidth as whatever pipe it's connected to can manage,H regardless, so while a switched back-end storage network used for sharedJ disks must be able to distribute that bandwidth to whatever combination ofI hosts require it this doesn't differ much from the need for the front-endoF network that served storage uses to distribute the load (again, unless  isochronous access is at issue).   >t >lI > Please don't think I am against all sharing and clustering, that is notr the. > caseG > as I have written in another posting, the origenal issue was that VMS  > ClusteringL > is not the only type of clustering, the term is horribly overloaded and it is > doubtfull.B > that anyone solution will be right for all the uses of the term.  I And as I wrote elsewhere, VMS comes damn close to supporting all of them,  and supporting them well.e   >oI > It just like some application work best on a large SMP or ccNUMA serverfK > such as a SuperDome or GS (and I think these are very different from whatuL > I can see) and some applications work best on a rack full of small systems or" > blades. It's horses for courses. >tL > BTW, as I said it is many years since I last used a VMS cluster, back then weF > had our 11/785s linked by a series of cables a good couple of inches thick,I > for the CI bus. Since the clustering technology IS going to be added toi HP-UX J > I want to learn all I can about it, can you recommend any good places to start,  E Greg Pfister's "In Search of Clusters" (2nd edition) has already beenjC mentioned, and is the best starting point I know of.  See above for D recommendations specific to VMS clusters.  The Oracle 'cache fusion'L approach is similar to an IBM research project for the 'Calypso' file systemL back in the mid-'90s (there used to be papers available on line, but findingJ them starting at ibm.com was a challenge - starting at almaden.ibm.com may be better).m   >yL > I know I have a lot to learn, in my own case it's going to need to be from thenI > high level overview down to VMS distributed lock manager algorithms, ifl they9 > are going to show up in the HP-UX kernel in the future.j  ' A pity HP isn't using the Tru64 kernel.    - bill   ------------------------------    Date: 15 May 2002 03:34:28 -07006 From: mgattaura@collins-stewart.com (Mandeep Gattaura)- Subject: DCPS and Poscript Printing.......... = Message-ID: <781290e2.0205150234.774c7626@posting.google.com>t  A This is a follow up post to a message "DCPS Help Required" on thes 30-April-2002,  E I am trying to print a file (which has been converted to postscript),oA using a form with a postscript setup module. The form needs to bep? overlaid on to all of the pages of the print file. I have triednD creating a form (using xpdf), and modifying for use in this scenarioF with no success. I have know taken a different route of trying to draw= the postscript from hand and wrapping it up in a function andn: re-defining showpage to print the "form" function as well.  , But when I print the file, using the command@ "PRINT/QUE=PS_QUEUE/FORM=PS_FORM PRINTFILE.PS" . But the printerE prints out data (with no form) on the first sheet, and then a load ofr% blank forms on the subsequent sheets.   & An extract of the print file is......." (        PROF & MRS I M CREWE  ) s (        LAKESIDE HOUSE ) s  (        WIVENHOE PARK ) s (        COLCHESTER ) si (        ESSEX ) s (        CO4 3SQ ) sD (        This transaction is subject to the rules and regulations of The ) sa" (        London Stock Exchange.) s showpage %%Page: 5 5p, sheetwidth 0 inch translate     % new origin2 90 rotate                       % landscape format# /x0 upperx 0 get bodymargin add defe4 /y0 uppery bodymargin bodyfontsize add 0 add sub def x0 y0 moveto bodyfont setfont ( ) s  ( ) si: (    CANCEL YOUR PURCHASE SCOT POWER ORD GBP0.50 \( ISIN : GB0006900707 \)) s ( ) sl  (    ERNEST ABRAHAM DARE ESQ ) s      6 and a condensed version of the postscript form is..... %! /contract_note { /inch {72 mul} def 0.5 setlinewidthB newpath 1.3 inch 0.18 inch moveto 1.3 inch 11.2 inch lineto strokeB newpath 1.5 inch 0.18 inch moveto 1.5 inch 11.2 inch lineto stroke= newpath 2 inch 0.18 inch moveto 2 inch 5.6 inch lineto stroketA newpath 3.9 inch 9.2 inch moveto 3.9 inch 11.2 inch lineto strokeaB newpath 4.5 inch 0.18 inch moveto 4.5 inch 11.2 inch lineto strokeC newpath 5.65 inch 0.18 inch moveto 5.65 inch 4.8 inch lineto strokemD newpath 7.55 inch 0.18 inch moveto 7.55 inch 11.2 inch lineto strokeC newpath 1.3 inch 0.18 inch moveto 7.55 inch 0.18 inch lineto strokeo> newpath 1.3 inch 1.8 inch moveto 2 inch 1.8 inch lineto stroke> newpath 1.3 inch 3.6 inch moveto 2 inch 3.6 inch lineto strokeA newpath 4.5 inch 4.8 inch moveto 7.55 inch 4.8 inch lineto strokee> newpath 1.3 inch 5.6 inch moveto 2 inch 5.6 inch lineto stroke@ newpath 1.3 inch 7.3 inch moveto 1.5 inch 7.3 inch lineto stroke@ newpath 1.3 inch 9.2 inch moveto 1.5 inch 9.2 inch lineto stroke@ newpath 3.9 inch 9.2 inch moveto 4.5 inch 9.2 inch lineto strokeC newpath 1.3 inch 11.2 inch moveto 7.55 inch 11.2 inch lineto stroken
 } bind def  /old_showpage /showpage load def, /showpage { old_showpage contract_note } def    : Can anyone give me some advice on where I am going wrong ?   TIA  Mandeep    ------------------------------  % Date: Wed, 15 May 2002 13:40:42 +0200r9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>h1 Subject: Re: DCPS and Poscript Printing..........o' Message-ID: <3CE2493A.E3632B6D@aaa.com>    Hi.-1 Well, I'v no answer to your specific problem, but:3 I'd take good care *not* to post data to c.o.v that>2 seams to include customer or client information !!   Jan-Erik Sderholm.    ------------------------------  % Date: Wed, 15 May 2002 13:45:50 +0100G From: Roy Omond <Roy@Omond.net> 1 Subject: Re: DCPS and Poscript Printing.......... ( Message-ID: <3CE2587E.610AE7F@Omond.net>   Mandeep Gattaura wrote:t  C > This is a follow up post to a message "DCPS Help Required" on theo > 30-April-2002, > G > I am trying to print a file (which has been converted to postscript),aC > using a form with a postscript setup module. The form needs to bedA > overlaid on to all of the pages of the print file. I have triediF > creating a form (using xpdf), and modifying for use in this scenarioH > with no success. I have know taken a different route of trying to draw? > the postscript from hand and wrapping it up in a function andr< > re-defining showpage to print the "form" function as well. > . > But when I print the file, using the commandB > "PRINT/QUE=PS_QUEUE/FORM=PS_FORM PRINTFILE.PS" . But the printerG > prints out data (with no form) on the first sheet, and then a load oft' > blank forms on the subsequent sheets.  >o( > An extract of the print file is.......   [... judicial snip ...]t  F Bloody hell, couldn't you have at least anonymised the name involved ?I I *know* the person concerned, and I'm sure he'd be pissed off seeing hisi  4 name and address being banded willy-nilly like that.  G *sheesh* ... please try to engage the brain before doing something likes that.n  	 Roy Omond  Blue Bubble Ltd.   ------------------------------    Date: 15 May 2002 14:02:34 -0000= From: Doc.Cypher <Use-Author-Supplied-Address-Header@[127.1]>d1 Subject: Re: DCPS and Poscript Printing..........a6 Message-ID: <20020515140234.12137.qmail@gacracker.org>  5 On Wed, 15 May 2002, Roy Omond <Roy@Omond.net> wrote:n >Mandeep Gattaura wrote:   <snip>   >[... judicial snip ...] >rG >Bloody hell, couldn't you have at least anonymised the name involved ?sJ >I *know* the person concerned, and I'm sure he'd be pissed off seeing his >d5 >name and address being banded willy-nilly like that.I >aH >*sheesh* ... please try to engage the brain before doing something like >that.  E It might be an idea for the original poster to go to the URL below iny1 Google and cancel that post out of their archive.   N http://services.google.com:8882/urlconsole/controller?cmd=reload&lastcmd=login  K Note, the original post has not yet showed up in the Google archive, but itu will.t     Doc. --  6 The bigger the humbug, the better people will like it.K ~ Phineas Taylor Barnum.                             https://vmsbox.cjb.neti   ------------------------------  + Date: Wed, 15 May 2002 14:06:54 +0000 (UTC)  From: david20@alpha2.mdx.ac.uk1 Subject: Re: DCPS and Poscript Printing.......... + Message-ID: <abtq1t$rdv$1@aquila.mdx.ac.uk>t  J In article <3CE2587E.610AE7F@Omond.net>, Roy Omond <Roy@Omond.net> writes: >Mandeep Gattaura wrote: >n >>) >> An extract of the print file is.......  >  >[... judicial snip ...] >uG >Bloody hell, couldn't you have at least anonymised the name involved ?tJ >I *know* the person concerned, and I'm sure he'd be pissed off seeing his >a5 >name and address being banded willy-nilly like that.i >hH >*sheesh* ... please try to engage the brain before doing something like >that. >a
 >Roy Omond >Blue Bubble Ltd.e >   N The address involved is in the UK. If the poster or his company is also in theK UK then I'm pretty sure the poster has just violated the UK Data protectionc act.  
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  % Date: Wed, 15 May 2002 13:56:07 -0400s- From: JF Mezei <jfmezei.spamnot@videotron.ca>w1 Subject: Re: DCPS and Poscript Printing.......... , Message-ID: <3CE2A136.42C8C47A@videotron.ca>   Mandeep Gattaura wrote: . > But when I print the file, using the commandB > "PRINT/QUE=PS_QUEUE/FORM=PS_FORM PRINTFILE.PS" . But the printerG > prints out data (with no form) on the first sheet, and then a load of1' > blank forms on the subsequent sheets.   / Do you have distiller ? (to turn PS into PDF) ?n  N If so, you can add commands such as "(here) print" and "(there) print" in yourI postscript, and as Distiller executes the postscript, you will see in the - distiller status log, those statements appear L (note you may wish to add "\n" at the end of each string since print doesn't do a newline).  J Also, have you tested the showpage redefinition ? perhaps it isn't exactly! right and showpage calls itself. e  K Have you tried /old_showpage  /showpage bind def ?     This would make sure H that old_showpage points to the code associated with showpage instead of> pointing to the name. (and makes it perform a tad faster too).  M Not sure if you have access to the application which generates the postscriptfM "data". If you do, it would perhaps be far simpler to have the application doc something such as :o  
 setup_page
 contract_forms#  (        PROF & MRS I M CREWE  ) sh  (        LAKESIDE HOUSE ) s  (        WIVENHOE PARK ) sn  (        COLCHESTER ) s  (        ESSEX ) st  (        CO4 3SQ ) st showpage    L and then define setup_page to have the various moveto, translates etc needed to start the page,+ and contract form to image the form itself.l  T Or you could even remove the "contract_form" and have setup_page call contract_form.  M The generated data is still generic enough from the application point of viewuK and you still have full control over the layout by modyfying the postscriptrA header, and no need to putz around with redefinition of commands.f    . > sheetwidth 0 inch translate     % new origin4 > 90 rotate                       % landscape format   have you tried:a  
 -90 rotate -792 0 translate    L If you do the "landspace" wrong, your text will all be rasterized outside of the paper area.e   ------------------------------  % Date: Wed, 15 May 2002 10:26:39 +0100cU From: Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com>k6 Subject: Re: Forced migration to HPHUX - Storm Warning0 Message-ID: <abt9kf$ggt$1@new-usenet.uk.sun.com>   Mark E. Levy wrote:   % > "Andrew Harrison SUNUK Consultancy"n@ > <andrew_nospam.harrison_remove_this@sun#.com> wrote in message, > news:abs7fj$7ad$1@new-usenet.uk.sun.com... >  >> >>Fred Kleinsorge wrote: >> >>H >>>Your results sound atypical.  It would be interesting to know exactly >>>s > what > H >>>your VMS system does, and a post mortem on your downtime.  Of course, >>>y > the  > 2 >>>same information for your UNIX systems as well. >>>  >>>u >>>. >> >>Sorry  >>? >>We have just been asked to replace an OpenVMS based system ats< >>a customers site (cluster 7.3 GS140's) guess why ????????? >>
 >>Downtime >> > ( > Oh, I get it. They want MORE downtime. >     @ No the upgrade to 7.3 went badly and they lost their environment0 completely for more than an hour more than once.  > This has not made them happy bunnies hence the desire to scrap OpenVMS.     Regardsg   Andrew Harrison    ------------------------------  % Date: Wed, 15 May 2002 10:56:49 +0100, From: Roy Omond <Roy@Omond.net>e6 Subject: Re: Forced migration to HPHUX - Storm Warning) Message-ID: <3CE230E1.680EE246@Omond.net>h  ( Andrew Harrison SUNUK Consultancy wrote:  B > No the upgrade to 7.3 went badly and they lost their environment2 > completely for more than an hour more than once. >t@ > This has not made them happy bunnies hence the desire to scrap
 > OpenVMS.  A I can think of only one scenario where this could have pertained,y@ namely gross incompetence, both at the technical level ("upgrade/ to VMS 7.3") and at the (PHB) managerial level.i  A I am a bit suspicious of your claim.  Sounds like you made it up.e  	 Roy Omondt Blue Bubble Ltd.   ------------------------------  % Date: Wed, 15 May 2002 12:59:01 +0100rU From: Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com>p6 Subject: Re: Forced migration to HPHUX - Storm Warning0 Message-ID: <abtii5$j7j$1@new-usenet.uk.sun.com>   Roy Omond wrote:  * > Andrew Harrison SUNUK Consultancy wrote: >  > B >>No the upgrade to 7.3 went badly and they lost their environment2 >>completely for more than an hour more than once. >>@ >>This has not made them happy bunnies hence the desire to scrap
 >>OpenVMS. >> > C > I can think of only one scenario where this could have pertained,dB > namely gross incompetence, both at the technical level ("upgrade1 > to VMS 7.3") and at the (PHB) managerial level.* > C > I am a bit suspicious of your claim.  Sounds like you made it up.r >     > I didn't. The upgrade to 7.3 was done so that it would make it/ easier to migrate off the GS140's onto GS320's.s  ' The failures were caused by bugs in 7.3C   Regardsn Andrew Harrison    ------------------------------  % Date: Wed, 15 May 2002 13:39:08 +0100n From: Roy Omond <Roy@Omond.net>h6 Subject: Re: Forced migration to HPHUX - Storm Warning) Message-ID: <3CE256EB.2A084B2C@Omond.net>0  ( Andrew Harrison SUNUK Consultancy wrote:   > Roy Omond wrote: >t, > > Andrew Harrison SUNUK Consultancy wrote: > >V > >rD > >>No the upgrade to 7.3 went badly and they lost their environment4 > >>completely for more than an hour more than once. > >>B > >>This has not made them happy bunnies hence the desire to scrap > >>OpenVMS. > >> > >eE > > I can think of only one scenario where this could have pertained, D > > namely gross incompetence, both at the technical level ("upgrade3 > > to VMS 7.3") and at the (PHB) managerial level.  > >sE > > I am a bit suspicious of your claim.  Sounds like you made it up.t > >  >r@ > I didn't. The upgrade to 7.3 was done so that it would make it1 > easier to migrate off the GS140's onto GS320's.f >s) > The failures were caused by bugs in 7.3t  8 Now I'm convinced you made it up.  I'll call your bluff.  A I'll bet we don't get to know which customer this is meant to be. = If it were true (and I'm convinced it's not), then I'd expectl? we'd get to hear of this in the meeja.  Why not drop Mike Mageed an anonymous hint ?   ! Ha, didn't think you would.  FUD.e  	 Roy Omonds Blue Bubble Ltd.   ------------------------------  % Date: Wed, 15 May 2002 07:56:48 -0500t( From: David Harrold <DHarrold@wi.rr.com>6 Subject: Re: Forced migration to HPHUX - Storm Warning2 Message-ID: <abtlug$llu$1@newsreader.mailgate.org>  & (I know I'm going to regret this.... )  K On Tue, 14 May 2002 15:26:27 -0400, JF Mezei <jfmezei.spamnot@videotron.ca>r wrote:   >"Stuart, Ed" wrote: >>  O >> The "Special Edition of the TRU64 UNIX and OpenVMS Times" has been modified.p* >> The OpenVMS/HP-UX migration Q&A is now: >aN >I find this to be unethical. A company should not be able to modify publishedM >documents. It might be able to add an addendum at the end to clarify, but itkH >should not be allowed to modify already published materials, especiallyM >something as big as a change in corporate policy. Did the version number forrO >that document change ? It should have, otherwise you have two nearly identicale@ >documents circulating with a major difference in policy inside.  L You have been asking, demanding and begging for Compaq and then HP to changeM the way they have been marketing, messaging and handling VMS for a long time.yK You have pointed to the various commitments made and then broken.  Now, you L point out another problem with HP's message about the handling and future of VMS now that they own it.a  J So, they listing to you (and others) and change the message to one that is# better.  Now they're unethical???  n  - No wonder they don't listen to the newsgroup.D   Dave Harrold  ( (Yes, I know, now *I'm* the idiot... :-(      N ..............................................................................N David Harrold                              E-Mail: David_Harrold at aurora.orgI Sr. Software Systems Engineer              Phone:          (414) 647-6204oI                                            Pager:          (414) 941-4634eG Aurora Health Care                         Fax:          (414) 647-4999  3031 W. Montana Street Milwaukee, WI 53215e   ------------------------------  + Date: Wed, 15 May 2002 12:58:36 +0000 (UTC)b From: david20@alpha1.mdx.ac.uk6 Subject: Re: Forced migration to HPHUX - Storm Warning+ Message-ID: <abtm1s$q83$1@aquila.mdx.ac.uk>c  m In article <JSeE8.11445$t8_.9140@news01.bloor.is.net.cable.rogers.com>, "John Smith" <a@nonymous.com> writes:hH >My remarks pretty much sum up what I have seen happen in the commercial
 >marketplace.l >eK >Academia may be different, but since I am not in that 'space' I will defer  >to you in that regard.i > M Not in my experience. I can only speak for Middlesex University but the adminoM systems are treated pretty much as they would be in a commercial environment. O The admin systems here aren't clustered hence there are times when they need to-@ be taken down - memory upgrade, reboot when applying patches etc. However uptimes of a year or more are typical.N Also sometimes the DBAs need to take the Oracle running on those systems down I to increase the size of the SGA etc. Which is obviously a loss of service . although the systems are still up and running.  M The Academic alphas which provide student mail, student web areas, Oracle fortO teaching etc for some 25000+ students is a cluster and again I try and maintaine it's availabilty 24x7x52.   N The Unix systems here tend to have much shorter uptimes not because of crashesA but simply because of the need to apply so many security patches.   
 David Webb VMS and Unix team leader CCSS Middlesex University       >h? >"Bill Gunshannon" <bill@triangle.cs.uofs.edu> wrote in messaget' >news:abrjmk$2n3t$1@info.cs.uofs.edu..._ >> How about a reality check?? >>H >> In article <pqaE8.4357$t8_.325@news01.bloor.is.net.cable.rogers.com>,) >>  "John Smith" <a@nonymous.com> writes: L >> |> "Dissatisfied with unix? Down-time costing you too much lost business? >>C >> The University's VMS system has had more downtime (scheduled and @ >> un-scheduled) than my Unix servers in every year since I have
 >> been here.a >> >> |>b >Tired, >> |> of paying top dollar for "unix guru's" >>D >> The bottom rung VMS guy here earns more than I do and I have over >> 20 years of experience. >>M >> |>                                        who change employers faster thane% >> |> humming birds beat their wings?a >>@ >> I have been here longer than any of the VMS staff, in fact, IB >> have watched the whole VMS staff turn over at least three times >> since I've been here. >>M >> |>                                 Choose the operating system that "let's. >you- >> |> sleep nights and stay calm" - OpenVMS."n >>E >> Considering that most people who are running Unix on their systems I >> are perfectly happy and none of what you had above compares to reality J >> how do you think statements like that make you look to management types >> making decisions??1 >> >> billo >> >> --eM >> Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesiG >> bill@cs.scranton.edu     |  and a sheep voting on what's for dinner.  >> University of Scranton   | A >> Scranton, Pennsylvania   |         #include <std.disclaimer.h>8 >c >i   ------------------------------  # Date: Wed, 15 May 2002 13:54:20 GMTo5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>n6 Subject: Re: Forced migration to HPHUX - Storm Warning9 Message-ID: <gMtE8.21$6U2.236632@cacnews.cac.cpqcorp.net>i  6 Andrew Harrison SUNUK Consultancy wrote in message ...     >i> >We have just been asked to replace an OpenVMS based system at; >a customers site (cluster 7.3 GS140's) guess why ?????????o >5	 >Downtimen >l  I What I think is that if the customer is replacing a business critical VMSrL system with a Sun system, they will be scratching their heads wondering whatH they were thinking 18 months from now - if they even manage to get it toI work.  There are many, many stories about how the Sun promise was just so1L much vapor - and left customers with no viable solution after a lot of work.   ------------------------------  % Date: Wed, 15 May 2002 17:38:25 +01003U From: Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com> 6 Subject: Re: Forced migration to HPHUX - Storm Warning0 Message-ID: <abu2u2$o5f$1@new-usenet.uk.sun.com>   Roy Omond wrote:  * > Andrew Harrison SUNUK Consultancy wrote: >  >  >>Roy Omond wrote: >> >>+ >>>Andrew Harrison SUNUK Consultancy wrote:D >>>n >>>c >>>PD >>>>No the upgrade to 7.3 went badly and they lost their environment4 >>>>completely for more than an hour more than once. >>>>B >>>>This has not made them happy bunnies hence the desire to scrap >>>>OpenVMS. >>>> >>>>D >>>I can think of only one scenario where this could have pertained,C >>>namely gross incompetence, both at the technical level ("upgradef2 >>>to VMS 7.3") and at the (PHB) managerial level. >>>xD >>>I am a bit suspicious of your claim.  Sounds like you made it up. >>>w >>>n@ >>I didn't. The upgrade to 7.3 was done so that it would make it1 >>easier to migrate off the GS140's onto GS320's.  >>) >>The failures were caused by bugs in 7.3  >> > : > Now I'm convinced you made it up.  I'll call your bluff. > C > I'll bet we don't get to know which customer this is meant to be.t? > If it were true (and I'm convinced it's not), then I'd expect A > we'd get to hear of this in the meeja.  Why not drop Mike Magee. > an anonymous hint ?n >     < Its a large investment bank. I really cannot say who or what= they are doing with the Alphas. I am under non disclosure andb< while it would be nice to drop an anonymous hint it would be
 unethical.  = You can choose to believe what you want but as you know thereg= is a longish snagging list with 7.3 and my customer fell foul- of that.   Regards- Andrew HarrisonM   ------------------------------  # Date: Wed, 15 May 2002 17:23:55 GMTd* From: "Bill Todd" <billtodd@metrocast.net>6 Subject: Re: Forced migration to HPHUX - Storm WarningB Message-ID: <LQwE8.148960$M7.14504213@bin7.nnrp.aus1.giganews.com>  , "Roy Omond" <Roy@Omond.net> wrote in message# news:3CE256EB.2A084B2C@Omond.net...d* > Andrew Harrison SUNUK Consultancy wrote: >h > > Roy Omond wrote: > >w. > > > Andrew Harrison SUNUK Consultancy wrote: > > >  > > >hF > > >>No the upgrade to 7.3 went badly and they lost their environment6 > > >>completely for more than an hour more than once. > > >>D > > >>This has not made them happy bunnies hence the desire to scrap > > >>OpenVMS. > > >> > > >nG > > > I can think of only one scenario where this could have pertained,rF > > > namely gross incompetence, both at the technical level ("upgrade5 > > > to VMS 7.3") and at the (PHB) managerial level.i > > > G > > > I am a bit suspicious of your claim.  Sounds like you made it up.h > > >a > >BB > > I didn't. The upgrade to 7.3 was done so that it would make it3 > > easier to migrate off the GS140's onto GS320's.: > > + > > The failures were caused by bugs in 7.3  >j: > Now I'm convinced you made it up.  I'll call your bluff. >tC > I'll bet we don't get to know which customer this is meant to be.t? > If it were true (and I'm convinced it's not), then I'd expectnA > we'd get to hear of this in the meeja.  Why not drop Mike Mageel > an anonymous hint ?  >l# > Ha, didn't think you would.  FUD.e  K Hmmm.  Didn't I recently see the comment that XFC was by default enabled insC 7.3?  If they didn't think to disable it, that might explain a lot.k   - bill   ------------------------------  + Date: Wed, 15 May 2002 17:31:09 +0000 (UTC)a5 From: "Bill Sticker" <NOSPAMPLEASE@SPAMSTOPPER.CO.UK>s6 Subject: Re: Forced migration to HPHUX - Storm Warning1 Message-ID: <abu60t$ao5$1@knossos.btinternet.com>s  K A decision to scrap all of their Alphas because of an upgrade failure showseG incompetency at the highest level. The upgrades are well documented andeG tested, and if the environment was lost then it is clearly the fault oftL whoever performed the upgrade. Where was the contingency part of the upgradeF plan. Surely a large investment bank would not perform such an upgradeE without testing it first and having a recovery plan if it went wrong.I  Would put my money in that bank.G And if they let it get out into the "meeja" I suspect no-one else will.l@ In over 20 years of Admin on VMS systems, I have never found theI documentation and dicumented procedures to be at fault and never had that" kind of a disaster.iK Serves them right for hiring cheap incompetent kids with university degrees  instead of experience.      # "Andrew Harrison SUNUK Consultancy" > <andrew_nospam.harrison_remove_this@sun#.com> wrote in message* news:abu2u2$o5f$1@new-usenet.uk.sun.com... >w >e >d >n > Roy Omond wrote: >t, > > Andrew Harrison SUNUK Consultancy wrote: > >e > >t > >>Roy Omond wrote: > >> > >>- > >>>Andrew Harrison SUNUK Consultancy wrote:p > >>>n > >>>h > >>>aF > >>>>No the upgrade to 7.3 went badly and they lost their environment6 > >>>>completely for more than an hour more than once. > >>>>D > >>>>This has not made them happy bunnies hence the desire to scrap > >>>>OpenVMS. > >>>> > >>>>F > >>>I can think of only one scenario where this could have pertained,E > >>>namely gross incompetence, both at the technical level ("upgrade 4 > >>>to VMS 7.3") and at the (PHB) managerial level. > >>> F > >>>I am a bit suspicious of your claim.  Sounds like you made it up. > >>>e > >>>eB > >>I didn't. The upgrade to 7.3 was done so that it would make it3 > >>easier to migrate off the GS140's onto GS320's.u > >>+ > >>The failures were caused by bugs in 7.3a > >> > >t< > > Now I'm convinced you made it up.  I'll call your bluff. > >pE > > I'll bet we don't get to know which customer this is meant to be.uA > > If it were true (and I'm convinced it's not), then I'd expectwC > > we'd get to hear of this in the meeja.  Why not drop Mike Magee  > > an anonymous hint ?  > >  >  > > > Its a large investment bank. I really cannot say who or what? > they are doing with the Alphas. I am under non disclosure andl> > while it would be nice to drop an anonymous hint it would be > unethical. >-? > You can choose to believe what you want but as you know there ? > is a longish snagging list with 7.3 and my customer fell foul 
 > of that. >r	 > Regardso > Andrew HarrisonW >i >    ------------------------------    Date: 15 May 2002 05:31:57 -07003 From: samuel.linjer@nilsongroup.com (Samuel Linjer)e# Subject: FTP mget on latest files ?m= Message-ID: <d60cd868.0205150431.47a18d40@posting.google.com>.   Hi,>  D I have a folder with many files (+1000). New files are created every day in this folder.tB From the folder i want to copy only the latest via ftp - the files+ that has been created within the last week.   # Is this possible with "mget" in FTPr   Any ideas ?s   //SAM>   ------------------------------    Date: 15 May 2002 07:55:30 -0600 From: briggs@encompasserve.org' Subject: Re: FTP mget on latest files ?e3 Message-ID: <uZ9toGKSjxIo@eisner.encompasserve.org>   s In article <d60cd868.0205150431.47a18d40@posting.google.com>, samuel.linjer@nilsongroup.com (Samuel Linjer) writes:m > Hi,n > F > I have a folder with many files (+1000). New files are created every > day in this folder..D > From the folder i want to copy only the latest via ftp - the files- > that has been created within the last week.  > % > Is this possible with "mget" in FTPr  A No, it is not.  mget is typically implemented using NLST and NLST6 doesn't do dates.r  
 > Any ideas ?   B Get a GUI client that lets you sort the directory listing by date.F Left-click to first file you want, shift-left click the last and press8 the little arrow thingie to transfer them to the client.  ? Or use FTP> quote site spawn dir /since=dd-mmm-yyyy and get thei list of file names from there.  @ Or arrange for the file names to contain the date and do an mget$ based on that sort of pattern match.   	John Briggs   ------------------------------  % Date: Wed, 15 May 2002 21:38:22 +0800e- From: David B Sneddon <dbsneddon@bigpond.com> ' Subject: Re: FTP mget on latest files ? " Message-ID: <5084371@MVB.SAIC.COM>  : On Wed, 15 May 2002 21:55, briggs@encompasserve.org wrote:@ > In article <d60cd868.0205150431.47a18d40@posting.google.com>, 5 samuel.linjer@nilsongroup.com (Samuel Linjer) writes:n > > Hi,i > >UH > > I have a folder with many files (+1000). New files are created every > > day in this folder. F > > From the folder i want to copy only the latest via ftp - the files/ > > that has been created within the last week.  > >n' > > Is this possible with "mget" in FTPc >rC > No, it is not.  mget is typically implemented using NLST and NLSTt > doesn't do dates.v >( > > Any ideas ?S >3D > Get a GUI client that lets you sort the directory listing by date.H > Left-click to first file you want, shift-left click the last and press: > the little arrow thingie to transfer them to the client. > A > Or use FTP> quote site spawn dir /since=dd-mmm-yyyy and get the.  > list of file names from there. >-B > Or arrange for the file names to contain the date and do an mget& > based on that sort of pattern match. >c > 	John Briggs  0 or try C-Kermit, it does selections by date/time   Regards, Dave.n --  I David B Sneddon (dbs)  OpenVMS Systems Programmer   dbsneddon@bigpond.comnI Sneddo's quick guide ...          http://www.users.bigpond.com/dbsneddon/.I DBS freeware at ...   http://www.users.bigpond.com/dbsneddon/software.htmaI "Life is what happens to you while you're busy making other plans" Lennone   ------------------------------  # Date: Wed, 15 May 2002 13:43:17 GMTnL From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr")' Subject: Re: FTP mget on latest files ? 8 Message-ID: <00A0DF64.4222F976@SSRL04.SLAC.STANFORD.EDU>  T In article <uZ9toGKSjxIo@eisner.encompasserve.org>, briggs@encompasserve.org writes:t >In article <d60cd868.0205150431.47a18d40@posting.google.com>, samuel.linjer@nilsongroup.com (Samuel Linjer) writes: >> Hi, >> rG >> I have a folder with many files (+1000). New files are created everyy >> day in this folder.E >> From the folder i want to copy only the latest via ftp - the filesu. >> that has been created within the last week. >> d& >> Is this possible with "mget" in FTP > B >No, it is not.  mget is typically implemented using NLST and NLST >doesn't do dates. >  >> Any ideas ? > C >Get a GUI client that lets you sort the directory listing by date.0G >Left-click to first file you want, shift-left click the last and presso9 >the little arrow thingie to transfer them to the client.e >r@ >Or use FTP> quote site spawn dir /since=dd-mmm-yyyy and get the >list of file names from there.M >iA >Or arrange for the file names to contain the date and do an mget>% >based on that sort of pattern match.4  L Or if what you're really trying to do is to maintain a copy on  VMS  of thatE directory which has  the same thousands of files, you might really beeL interested in an FTP_MIRROR.COM procedure.  I think you can find that in the usual VMS freeware locations.    -- Alanw      O ===============================================================================c0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056dM  Physical mail to: SSRL -- SLAC BIN 69, PO BOX 4349, STANFORD, CA  94309-0210 O ===============================================================================    ------------------------------  % Date: Wed, 15 May 2002 16:34:45 +0000t  From: Steve.Spires@yellgroup.com8 Subject: Irish Railways [was Re: Stallards smoking gun!]: Message-ID: <OF1A3F1946.2C7830D5-ON00256BBA.005AF703@btyp>  K Not an Urban Myth - I contacted Irish Railways a year or so ago about this.  Feel free to do so yourself.   Cheers   Steve Sn        ! Andrew Harrison SUNUK ConsultancygG <andrew_nospam.harrison_remove_this@sun#.com> on 05/14/2002 05:09:25 PM     To:        Info-VAX@Mvb.Saic.Com cc:., From:      Andrew Harrison SUNUK ConsultancyK            <andrew_nospam.harrison_remove_this@sun#.com>, 14 May 2002, 5:09o            p.m.g   Re: Stallards smoking gun!    3 Urban myth along with the hairy handed hitch hiker.    Regardsr Andrew Harrison9   Main, Kerry wrote:  @ >>>>>Neither have I. Only started using VMS 20 years ago. ;-}<<< >>>>>n >tF > Well, having spent time in a prior lifetime doing CS support, I haveD > seen lots of crash dumps, but the point is that there are numerous5 > examples of extra ordinary uptimes for VMS systems.  >sI > Case in point - the Irish National Railway application ran CONTINUOUSLY0A > (no scheduled or unscheduled system downtime which impacted themG > application availability) for 17 years before being retired some time  > last year. >e	 > Regardse >A > Kerry Main > Senior Consultant  > HP Global Services > HP Canada Ltd. > Voice: 613-592-4660i > Fax  :  819-772-7036 > Email: Kerry.Main@hp.com >s >U > -----Original Message-----8 > From: Terry C. Shannon [mailto:terryshannon@attbi.com] > Sent: May 10, 2002 8:09 AM > To: Info-VAX@Mvb.Saic.Come% > Subject: Re: Stallards smoking gun!  >t >n >w. > "Roy Omond" <Roy@Omond.net> wrote in message% > news:3CDB8BE8.FE82DBDF@Omond.net...  >  >>Carl Karcher wrote:s >> >>C >>>In a previous article, bob@instantwhip.com (Bob Ceculski) wrote:  >>>iF >>>->I could care less about anyone else ... as long as I have vms and >>>->I >>>i > have >uC >>>->support, let everyone else get shut down ... I will be up, the  >>>->company >>>  > Ii >eG >>>->work for will be up 99.9999, and I will be just fine ... for thoset >>>a >b >>>->who >>>  > fail >s >>>->... >>>dE >>>Come on Bob, while I have a similar mantra, that's six nines or 31eC >>>seconds of downtime a year if I did the math correctly. Even VMSn, >>>would be hard pressed to accomplish that. >>> @ >>Come on Carl, you haven't been reading Good Old Bob Ceculski'sC >>postings. He's never seen a VMS crash in the last 32,767 *years*.y >> >?: > Neither have I. Only started using VMS 20 years ago. ;-} >l >o >s >m            F ______________________________________________________________________     [Information] -- PostMaster:D This transmission is intended solely for the addressee(s) and may beG confidential. If you are not the named addressee, or if the message hasaG been addressed to you in error, you must not read, disclose, reproduce,n$ distribute or use this transmission.  H Delivery of this message to any person other than the named addressee isG not intended in any way to waive confidentiality.  If you have receivedtK this transmission in error please contact the sender or delete the message.g  
 Thank you.  D Yell Limited, Queens Walk, Oxford Road, Reading, Berkshire, RG1 7PT.; Registered in England and Wales, registered number 4205228.   I Yellow Pages Sales Limited, Queens Walk, Oxford Road, Reading, Berkshire, D RG1 7PT. Registered in England and Wales, registered number 1403041.   ------------------------------  + Date: Wed, 15 May 2002 16:51:01 +0000 (UTC)n5 From: "Bill Sticker" <NOSPAMPLEASE@SPAMSTOPPER.CO.UK>dY Subject: Re: Let go of VMS for the good of your career? (was "Re: System manager  availab / Message-ID: <abu3lk$69c$1@helle.btinternet.com>   I I would agree with you to a certain extent, especially from a programmersdJ point of view. However after 20 years as a VMS administrator, who is goingG to give me a job as a Unix administrator. I have already been through a8K course program (although all versions of Unix seem so simple and lacking in F operating system functionality that I keep thinking that I have missedG something) but still do not have the experience required to get even an3& entry level job in the current market. Will you give me a job?c J. Oxley    ? "Atlant Schmidt" <atlantnospam@mindspring.com> wrote in messageb( news:3CE28E16.27828080@mindspring.com...8 > Chris Bardell wrote (in Re: System Manager avaliable): >tI > > An earlier reponse was 'get into Unix'. A valid point, but to to jumptJ > > across into any other technical field often means an 'entry-level' job >wG > > unless you have some proper commercial experience of the new field. I > > Entry-level UK salaries are dreadful when you consider cost of livingv@ > > here. Barely adequate for a single person, let alone someone > > supporting kids, etc.- > >-J > > Sorry, wish I could suggest something more valid other than continuing >sI > > to search around. Also, I'm about to configure from scratch an old PC E > > with Linux (& Apache, PHP, etc) to try to add *some* skills to mytJ > > portfolio. If you have an old PC knocking about, it's cheap/free & the > " > > experience might prove useful. >AD > As I've been reading c.o.v. the last few days, I've been wonderingH > about this general topic. A lot of you sound like you're sticking with@ > VMS simply because that's what you know. I wonder if this is a4 > wise choice for the long-term good of your career? >MI > I can't speak for system administration (because I was never a big-timeeF > VMS administrator nor am I a big-time Unix administrator, although IA > could mostly run my own systems in both cases), but for systems>? > programming, a journeyman VMS programmer already knows nearly.E > all the "good stuff" that they need to know to be a journeyman UnixmD > programmer. There just isn't that much difference any more between= > *ANY* of the full-functionality, serious operating systems.e >oC > Okay, shells are different than DCL, and VMS still doesn't fork()k> > or grep, and Unix doesn't have EDT (or any emulation worth aA > damn), but beyond that, it's mostly just a matter of sticking ah@ > bunch of PostIts! (tm) on your monitor to remind you of what's > what.  > < > Maybe my view is jaundiced by the fact that, for years, my; > LOGIN.COM file had a zillion symbols covering things likehA > "rm", "more", "head", "tail", etc., but if I were giving careera: > advice, I'd certainly say that a move to Unix/Linux puts; > your career on a much safer footing than being one of the 6 > last few hold-outs on VMS. A change to Windows would< > probably be even better, career-wise, but I understand why% > that doesn't appeall to many of us.r >s< > Obviously, I'm interested in your opinions about this (and: > if we stay focused on the "career" aspects, maybe we can' > even avoid a Unix vs. VMS flame war).r > 9 > What do you think? Stay with VMS? Or let it go and moveH> > on, if only for the sake of the Salary Continuation Program? >  > Atlant >r >o9 > (Full disclosure: I made this move about two years ago,g1 > moving from Compaq's VMS Engineering group to ae > Solaris-based OEM.)r >r >h   ------------------------------  % Date: Wed, 15 May 2002 11:56:37 -0500i1 From: "Dave Gudewicz" <david.gudewicz@abbott.com>>Y Subject: Re: Let go of VMS for the good of your career? (was "Re: System manager  availaba1 Message-ID: <abu453$27a$1@fizban.pprd.abbott.com>l  H You raise an interesting question.  And it just so happens that about anL hour ago, I was given some evidence (perhaps not convincing to all) that VMS! isn't all that bad a place to be.   C UNIX guy comes by and asked if I used outbound telnet via a gateway K recently.  I said yes, as recently as this very morning.  Been doing it forjE years.  To node eisner in fact, the DECUS now Encompass Notes server.l  K He said that was odd, cause my password on that gateway had expired, but heuI did notice that I had gotten "out" this very day.  He walked away lookings? somewhat confused.  I raised my eyebrows and went back to work.    -- Dave...t  ) Adam and Eve had many advantages, but thee- principle one was that they escaped teething.e -----Mark Twain   ? "Atlant Schmidt" <atlantnospam@mindspring.com> wrote in message ( news:3CE28E16.27828080@mindspring.com...8 > Chris Bardell wrote (in Re: System Manager avaliable): >cI > > An earlier reponse was 'get into Unix'. A valid point, but to to jumpHJ > > across into any other technical field often means an 'entry-level' job >iG > > unless you have some proper commercial experience of the new field.nI > > Entry-level UK salaries are dreadful when you consider cost of livingt@ > > here. Barely adequate for a single person, let alone someone > > supporting kids, etc.p > >.J > > Sorry, wish I could suggest something more valid other than continuing > I > > to search around. Also, I'm about to configure from scratch an old PC E > > with Linux (& Apache, PHP, etc) to try to add *some* skills to mynJ > > portfolio. If you have an old PC knocking about, it's cheap/free & the > " > > experience might prove useful. >eD > As I've been reading c.o.v. the last few days, I've been wonderingH > about this general topic. A lot of you sound like you're sticking with@ > VMS simply because that's what you know. I wonder if this is a4 > wise choice for the long-term good of your career? >eI > I can't speak for system administration (because I was never a big-time@F > VMS administrator nor am I a big-time Unix administrator, although IA > could mostly run my own systems in both cases), but for systemsr? > programming, a journeyman VMS programmer already knows nearly E > all the "good stuff" that they need to know to be a journeyman Unix D > programmer. There just isn't that much difference any more between= > *ANY* of the full-functionality, serious operating systems.n >tC > Okay, shells are different than DCL, and VMS still doesn't fork()o> > or grep, and Unix doesn't have EDT (or any emulation worth aA > damn), but beyond that, it's mostly just a matter of sticking at@ > bunch of PostIts! (tm) on your monitor to remind you of what's > what.p >o< > Maybe my view is jaundiced by the fact that, for years, my; > LOGIN.COM file had a zillion symbols covering things likemA > "rm", "more", "head", "tail", etc., but if I were giving career : > advice, I'd certainly say that a move to Unix/Linux puts; > your career on a much safer footing than being one of the 6 > last few hold-outs on VMS. A change to Windows would< > probably be even better, career-wise, but I understand why% > that doesn't appeall to many of us.t >e< > Obviously, I'm interested in your opinions about this (and: > if we stay focused on the "career" aspects, maybe we can' > even avoid a Unix vs. VMS flame war).a >d9 > What do you think? Stay with VMS? Or let it go and movel> > on, if only for the sake of the Salary Continuation Program? >a > Atlant >  > 9 > (Full disclosure: I made this move about two years ago,s1 > moving from Compaq's VMS Engineering group to ae > Solaris-based OEM.)r >  >a   ------------------------------  % Date: Wed, 15 May 2002 12:34:30 -0400 2 From: Atlant Schmidt <atlantnospam@mindspring.com>Y Subject: Let go of VMS for the good of your career? (was "Re: System manager available") 2. Message-ID: <3CE28E16.27828080@mindspring.com>  6 Chris Bardell wrote (in Re: System Manager avaliable):  G > An earlier reponse was 'get into Unix'. A valid point, but to to jump H > across into any other technical field often means an 'entry-level' job  E > unless you have some proper commercial experience of the new field.-G > Entry-level UK salaries are dreadful when you consider cost of livings> > here. Barely adequate for a single person, let alone someone > supporting kids, etc.i >rH > Sorry, wish I could suggest something more valid other than continuing  G > to search around. Also, I'm about to configure from scratch an old PC C > with Linux (& Apache, PHP, etc) to try to add *some* skills to myeH > portfolio. If you have an old PC knocking about, it's cheap/free & the    > experience might prove useful.  B As I've been reading c.o.v. the last few days, I've been wonderingF about this general topic. A lot of you sound like you're sticking with> VMS simply because that's what you know. I wonder if this is a2 wise choice for the long-term good of your career?  G I can't speak for system administration (because I was never a big-timehD VMS administrator nor am I a big-time Unix administrator, although I? could mostly run my own systems in both cases), but for systemse= programming, a journeyman VMS programmer already knows nearlytC all the "good stuff" that they need to know to be a journeyman Unix B programmer. There just isn't that much difference any more between; *ANY* of the full-functionality, serious operating systems.   A Okay, shells are different than DCL, and VMS still doesn't fork()S< or grep, and Unix doesn't have EDT (or any emulation worth a? damn), but beyond that, it's mostly just a matter of sticking al> bunch of PostIts! (tm) on your monitor to remind you of what's what.i  : Maybe my view is jaundiced by the fact that, for years, my9 LOGIN.COM file had a zillion symbols covering things likee? "rm", "more", "head", "tail", etc., but if I were giving careerh8 advice, I'd certainly say that a move to Unix/Linux puts9 your career on a much safer footing than being one of thet4 last few hold-outs on VMS. A change to Windows would: probably be even better, career-wise, but I understand why# that doesn't appeall to many of us.t  : Obviously, I'm interested in your opinions about this (and8 if we stay focused on the "career" aspects, maybe we can% even avoid a Unix vs. VMS flame war).e  7 What do you think? Stay with VMS? Or let it go and move < on, if only for the sake of the Salary Continuation Program?   Atlant    7 (Full disclosure: I made this move about two years ago,t/ moving from Compaq's VMS Engineering group to a, Solaris-based OEM.)y   ------------------------------  % Date: Wed, 15 May 2002 10:46:21 -0400s From: TuCaN <Tuc@n.com>i& Subject: Newbie using Openvms and C.../ Message-ID: <3ce27466$1@dnewserver.firstcom.cl>B  ! Hi dudes...i have this porblem...bG I'm new using openvms and i installed COMPAQ C V6.5...but when i try tou use it..it says...  ; %LICENSE-F-NOAUTH, DEC C use is not authorized on this node,D -LICENSE-F-NOLICENSE, no license is active for this software product1 -LICENSE-I-SYSMGR, please see your system managerP  < ****** COMPAQ C V6.5 Installation Verification FAILED ******   What can i do ?...  = How can i active the license?...Do i have to buy the license?a   Thanks in advnace...   Jorged   ------------------------------    Date: 15 May 2002 09:58:13 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) * Subject: Re: Newbie using Openvms and C...3 Message-ID: <ZPp7K$Lpg8mC@eisner.encompasserve.org>   I In article <3ce27466$1@dnewserver.firstcom.cl>, TuCaN <Tuc@n.com> writes:a# > Hi dudes...i have this porblem...eI > I'm new using openvms and i installed COMPAQ C V6.5...but when i try to  > use it..it says... > = > %LICENSE-F-NOAUTH, DEC C use is not authorized on this node F > -LICENSE-F-NOLICENSE, no license is active for this software product3 > -LICENSE-I-SYSMGR, please see your system managere > > > ****** COMPAQ C V6.5 Installation Verification FAILED ****** >  > What can i do ?... > ? > How can i active the license?...Do i have to buy the license?a  I Yes, you have to buy the license, unless you qualify for the VMS Hobbyistd" Program (non-commercial home use).   ------------------------------  % Date: Wed, 15 May 2002 11:15:57 -0400p From: TuCaN <Tuc@n.com>t* Subject: Re: Newbie using Openvms and C.../ Message-ID: <3ce27b56$1@dnewserver.firstcom.cl>.  : On Wed, 15 May 2002 11:58:13 -0400, Larry Kilgallen wrote:  C > In article <3ce27466$1@dnewserver.firstcom.cl>, TuCaN <Tuc@n.com>=	 > writes:=$ >> Hi dudes...i have this porblem...J >> I'm new using openvms and i installed COMPAQ C V6.5...but when i try to >> use it..it says...S >> P> >> %LICENSE-F-NOAUTH, DEC C use is not authorized on this nodeG >> -LICENSE-F-NOLICENSE, no license is active for this software product=4 >> -LICENSE-I-SYSMGR, please see your system manager >> -? >> ****** COMPAQ C V6.5 Installation Verification FAILED ******m >> v >> What can i do ?...  >> c@ >> How can i active the license?...Do i have to buy the license? > B > Yes, you have to buy the license, unless you qualify for the VMS- > Hobbyist Program (non-commercial home use).h   % VMS Hobbyist Program?...what is that?s   thanks   ------------------------------    Date: 15 May 2002 15:48:25 -0000= From: Doc.Cypher <Use-Author-Supplied-Address-Header@[127.1]> * Subject: Re: Newbie using Openvms and C...6 Message-ID: <20020515154825.14947.qmail@gacracker.org>  - On Wed, 15 May 2002, TuCaN <Tuc@n.com> wrote:0; >On Wed, 15 May 2002 11:58:13 -0400, Larry Kilgallen wrote:  >hD >> In article <3ce27466$1@dnewserver.firstcom.cl>, TuCaN <Tuc@n.com> >> writes: e% >>> Hi dudes...i have this porblem... K >>> I'm new using openvms and i installed COMPAQ C V6.5...but when i try to  >>> use it..it says... >>> ? >>> %LICENSE-F-NOAUTH, DEC C use is not authorized on this node H >>> -LICENSE-F-NOLICENSE, no license is active for this software product5 >>> -LICENSE-I-SYSMGR, please see your system manageri >>> @ >>> ****** COMPAQ C V6.5 Installation Verification FAILED ****** >>>  >>> What can i do ?... >>> A >>> How can i active the license?...Do i have to buy the license?y >>  C >> Yes, you have to buy the license, unless you qualify for the VMSe. >> Hobbyist Program (non-commercial home use). > & >VMS Hobbyist Program?...what is that?  K The VMS Hobbyist program is for private users of a VMS system. You can findoI it at http://www.montagar.com/hobbyist/ You will also find links there toSH DECUS/Encompass which you would need to join before you would be able toH get Hobbyist licenses. (The license application process requires a valid membership number.)w  K Where did you get hold of a VMS system, and what have you got? Or is this aoJ case of someone at work has let you loose on an old machine? If you've gotI hold of an old system for your own use then technically you can't use thedF licenses on it, but I don't think anyone will complain if you are justI learning it while waiting to get Hobbyist licenses (the whole process canw take a little time).     Doc. --  6 The bigger the humbug, the better people will like it.K ~ Phineas Taylor Barnum.                             https://vmsbox.cjb.nete   ------------------------------  + Date: Wed, 15 May 2002 15:53:08 +0000 (UTC)a From: david20@alpha2.mdx.ac.uk* Subject: Re: Newbie using Openvms and C...* Message-ID: <abu094$80$1@aquila.mdx.ac.uk>  I In article <3ce27b56$1@dnewserver.firstcom.cl>, TuCaN <Tuc@n.com> writes: ; >On Wed, 15 May 2002 11:58:13 -0400, Larry Kilgallen wrote:_ >_D >> In article <3ce27466$1@dnewserver.firstcom.cl>, TuCaN <Tuc@n.com>
 >> writes:% >>> Hi dudes...i have this porblem... K >>> I'm new using openvms and i installed COMPAQ C V6.5...but when i try ton >>> use it..it says... >>> ? >>> %LICENSE-F-NOAUTH, DEC C use is not authorized on this nodeaH >>> -LICENSE-F-NOLICENSE, no license is active for this software product5 >>> -LICENSE-I-SYSMGR, please see your system managert >>> @ >>> ****** COMPAQ C V6.5 Installation Verification FAILED ****** >>>  >>> What can i do ?... >>> A >>> How can i active the license?...Do i have to buy the license?B >> iC >> Yes, you have to buy the license, unless you qualify for the VMS2. >> Hobbyist Program (non-commercial home use). > & >VMS Hobbyist Program?...what is that? >g >thanksn  L Free VMS licenses and licenses for over 100 layered products (compilers etc)7 for the use of home hobbyist (ie non-commercial) users.B  % See http://www.montagar.com/hobbyist/.    L This requires that you be a member of DECUS however you can use the HobbyistH program if you become an Associate rather than a full member which costs nothing.  = To get Associate membership you can fill in a form online at    / http://www.encompassus.org/membership/join.htmlt     " encompass is the US Decus chapter.D Information on other chapters is available from www.compaqusers.org.  O (looking on www.compaqusers.org I couldn't see anything on associate membershipaN or the hobbyist program - perhaps someone else can comment on whether everyone? can/should be getting associate memberships through encompass).r    
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  % Date: Wed, 15 May 2002 08:06:48 -0400n' From: "Main, Kerry" <Kerry.Main@hp.com>  Subject: RE: No new Alpha salestT Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF402664115@kaoexc01.americas.cpqcorp.net>  D >>> The forementioned  HP statement refers only to their decision of HP-UX vs. Tru64,<<  ; Very good response. Could not have stated it any better.=20-  C As previous replies have stated, to many people are micro-analyzinge9 every single word that comes out of marketing these days.   F For those doubting "show-me" readers, if no new Alpha business were to@ be allowed, how does one explain the following new VMS Customer:5 http://www.mgex.com/news/news_releases/release28.htmli  > Or do you think Customers with these types of mission criticalH environments did not receive the assurances from the powers on high that  they had nothing to worry about?     Regards     
 Kerry Main Senior Consultantt Hewlett-Packard Canada# Consulting and Integration Services  Voice: 613-592-4660  Fax  :  819-772-7036 Email: Kerry.Main@hp.com     -----Original Message-----+ From: John Smith [mailto:a@nonymous.com]=20o Sent: May 13, 2002 3:01 PM To: Info-VAX@Mvb.Saic.Comr Subject: Re: No new Alpha salest      A "Howard Taylor" <Howard.Taylor@pacificcoast.net> wrote in messageh news:3cdedfee$1@nubby2.... >y? > "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message=20a( > news:3CDECE4C.2B1F0FEF@videotron.ca... > G > >  Rationale: We want to reinforce our commitment to our customers bya > followinguJ > > the roadmaps we had already established. We're leading with PA-RISC=20 > > forr > neweI > > business opportunities for two reasons: First, the PA-RISC systems=20d	 > > will,( > in> > > most cases, be upgradeable in the box to future Itanium=20C > > microprocessors. Second, HP-UX is the long-term UNIX for HP. ##u >  > This statement says to me:B > - HP will *not* recommend OpenVMS for new business opportunitiesG > - Because OpenVMS is not HP-UX (or UNIX, for that matter), OpenVMS iss *not*n > long-term for HP > B > Does this sound correct, or have I mis-interpreted your posting? >' > Howard Taylorh     Sort of right.  F The forementioned  HP statement refers only to their decision of HP-UXG vs. Tru64, and in my opinion only because HP-UX has larger market sharebD than Tru64, and hence less distruptive to the combined unix customerH base. If they had made the decision of which to chose based on technicalH superiority, then they would have chosen Tru64, but that also would haveG meant Alpha, and imagine the screams from HP-UX users who would then berH in a position to have to buy Alpha boxes only to then have to buy Itanic boxes.  @ The bit about the boxes being field upgradable to Itanic is niceF marketing, but virtually none of the customers will do it. They'll buyC new boxes with newer bus architecutes and interconnects before theye# upgrade an existing 3-year old box.r    F Alpha will be sold until IA64 is ready, and probably for a short whileG after that. But if you are a new (to HP) customer, don't expect them to-F ever mention VMS to you. It'll be HP-UX or Windows that they'll try to stuff down your throat.h    G About VMS - you got that right too. There have previously been verbatimcG quotes taken from HP material about this, although some posts here seemsH to suggest that HP (Stallard in particular) still don;t have their mindsF made up on this. If you want VMS to live, write Carly, write Stallard, ditto Marcello et al.   6 If you don't care about VMS anymore, call IBM for AIX.   ------------------------------    Date: 15 May 2002 07:48:46 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)d Subject: RE: No new Alpha sales.3 Message-ID: <h0y4208DSPJ1@eisner.encompasserve.org>   ~ In article <BE56C50EA024184DAF48F0B9A47F5CF402664115@kaoexc01.americas.cpqcorp.net>, "Main, Kerry" <Kerry.Main@hp.com> writes: >  > E >>>> The forementioned  HP statement refers only to their decision of0 > HP-UX vs. Tru64,<< > = > Very good response. Could not have stated it any better.=20n > E > As previous replies have stated, to many people are micro-analyzingc; > every single word that comes out of marketing these days.e  F Consider that VMS Development releases software that still has defectsI in it.  Were that not the case, nobody would be interested in patch kits.   I Why would one expect HP Marketing to release collateral that is perfect ?   2 Which one deserves a higher degree of perfection ?  E Sloppy marketing (just so it is there) works a lot better than sloppy 	 software.o   ------------------------------  # Date: Wed, 15 May 2002 14:43:13 GMTe. From: Michael Rice <MichaelARice@adelphia.net> Subject: Re: No new Alpha sales ) Message-ID: <3CE2746D.30303@adelphia.net>    Main, Kerry wrote: > D >>>>The forementioned  HP statement refers only to their decision of >>>w > HP-UX vs. Tru64,<< > ; > Very good response. Could not have stated it any better. M > E > As previous replies have stated, to many people are micro-analyzing ; > every single word that comes out of marketing these days.o > H > For those doubting "show-me" readers, if no new Alpha business were toB > be allowed, how does one explain the following new VMS Customer:7 > http://www.mgex.com/news/news_releases/release28.htmlg > @ > Or do you think Customers with these types of mission criticalJ > environments did not receive the assurances from the powers on high that" > they had nothing to worry about? >  >     H Did I miss something?  Where in that article did it mention HP, Compaq,  Alpha, or VMS?   ------------------------------  + Date: Wed, 15 May 2002 17:04:22 +0100 (MET)'9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>h Subject: Re: No new Alpha saless; Message-ID: <01KHRI7JM6JE95MSV4@sysdev.deutsche-boerse.com>d  7 > http://www.mgex.com/news/news_releases/release28.htmlb  J > Did I miss something?  Where in that article did it mention HP, Compaq,= =20n > Alpha, or VMS?  G It didn't have to.  It mentioned OM, a well known Swedish company whicheG runs stock exchanges.  It is a well known fact that OM, like, err, some F other stock-exchange companies I am aware of :-) , use VMS, these daysB on ALPHA; as you know, both were available from Compaq and are now available from HP.=20t  ) Go to, for example, http://www.om.com/=20m      offerings      financial      SECUR      technical  6 to see some technical details (for example: VMS, Rdb).  A Note that this is an article about a grain exchange.  I doubt the F internals of the technology involved interest the average Minnesota=20 farmer.g     --=20 H Phillip Helbig              Email ... helbphi@sysdev.Deutsche-Boerse.comJ Deutsche B=F6rse Systems AG   Email ... Phillip.Helbig@Deutsche-Boerse.co= m H Xetra/Eurex Operations      Tel. ...................... +49 69 2101 4921H 60485 Frankfurt am Main     Fax ....................... +49 69 2101 3411  H My opinions are not necessarily those of my employer and, no, I don't=20 have any stock tips for you.   ------------------------------  # Date: Wed, 15 May 2002 16:50:38 GMTe# From: "John Smith" <a@nonymous.com>n Subject: Re: No new Alpha salesoE Message-ID: <ylwE8.20195$t8_.53@news01.bloor.is.net.cable.rogers.com>e   Did I just hear a 'rimshot'?    < "Dave Gudewicz" <david.gudewicz@abbott.com> wrote in message+ news:abrcbo$gk8$1@fizban.pprd.abbott.com...sB > If the beast was a python, was its nickname Monty by any chance? >m > --	 > Dave...g >d+ > Adam and Eve had many advantages, but theo/ > principle one was that they escaped teething.s > -----Mark Twain  > 0 > "John Smith" <a@nonymous.com> wrote in messageE > news:9tYD8.89584$GLp1.20086@news01.bloor.is.net.cable.rogers.com...oK > > You forgot the bit about cleaning out the 'dust bunnies' from the powernI > > supplies, and if in Australia, the odd python that finds large serverv > boxesn# > > nice warm places to curl up in.o > >yK > > Wouldn't catch me putting my arm under a raised floor in Australia. OneI ofL > > my relatives just found the skin of some sort of constrictor (least theyI > > didn't think it was poisonous) in their home near Brisbane this week.i > >a > > Nice place to visit though.n > >t > >lB > > "Bill Gunshannon" <bill@triangle.cs.uofs.edu> wrote in message* > > news:abp3ce$1f2e$1@info.cs.uofs.edu... > > > In article@ > <K3UD8.83446$GLp1.15577@news01.bloor.is.net.cable.rogers.com>,, > > >  "John Smith" <a@nonymous.com> writes: > > > |>I > > > |> The bit about the boxes being field upgradable to Itanic is niceg > > marketing,H > > > |> but virtually none of the customers will do it. They'll buy new boxesi > > withH > > > |> newer bus architecutes and interconnects before they upgrade an > > existing > > > |> 3-year old box. > > > |> > > >rE > > > Actually, I ssupect the "upgrade" will be about the same as theb upgradet1 > > > for the 3B2 family when NCR took that over.s > > > Step 1.  Open case0 > > > Step 2.  Dump entire contents into garbage1 > > > Step 3.  Put whole new computer in old casea > > >o > > > Upgrade complete.r > > >a > > >a
 > > > bill > > >i > > > --I > > > Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Threen > wolvesJ > > > bill@cs.scranton.edu     |  and a sheep voting on what's for dinner.  > > > University of Scranton   |D > > > Scranton, Pennsylvania   |         #include <std.disclaimer.h> > >e > >  >n >o   ------------------------------  # Date: Wed, 15 May 2002 17:03:53 GMTt# From: "John Smith" <a@nonymous.com>  Subject: Re: No new Alpha salesrB Message-ID: <ZxwE8.23$ah_.22@news01.bloor.is.net.cable.rogers.com>  ; "Michael Rice" <MichaelARice@adelphia.net> wrote in messageo# news:3CE2746D.30303@adelphia.net...- >-I > Did I miss something?  Where in that article did it mention HP, Compaq,m > Alpha, or VMS?  J I believe that all of OM Group's applications run on OpenVMS only (I standC to be corrected), and it is OM software that is at the core of thisl	 exchange.C  L OM Group is one of the most significant vendors of  'stock exchange' trading platforms around.    ------------------------------  # Date: Wed, 15 May 2002 17:00:07 GMTp# From: "John Smith" <a@nonymous.com>  Subject: Re: No new Alpha salesMB Message-ID: <ruwE8.19$ah_.10@news01.bloor.is.net.cable.rogers.com>   Kerry,  I a) This was probably negotiated and contracted for prior to June 25, 2001n anyway.   J b) The app was the one they wanted an OM software only runs VMS right now,G if my memory serves me correctly. This could change with time, and HP's  pronouncements.   J c) Since the app runs on VMS and 'time-to-market' was most likely criticalH for a variety of business reasons, the choice of Alpha-VMS was a lay-up.  K d) Again, the 3-5 year replacement cycle come into play, so if in 3-5 yearshI HP doesn't really sell VMS anymore, count on OM to port it to unix, if ito isn't already there.  J e) Most importantly - If these kinds of customers merited comforting wordsF from 'on high', then surely the rest of us 'great unwashed mass of VMSJ users' do too. We're still the bulk of the customer base and we get pissedJ off too when our legitimate need to know is ignored. It is stuff like this that makes us mistrust HP.  I What's Carly's telephone number? I want to call her and talk to her about. VMS.   Thanks.-      2 "Main, Kerry" <Kerry.Main@hp.com> wrote in messageL news:BE56C50EA024184DAF48F0B9A47F5CF402664115@kaoexc01.americas.cpqcorp.net. ..    D >>> The forementioned  HP statement refers only to their decision of HP-UX vs. Tru64,<<  8 Very good response. Could not have stated it any better.  C As previous replies have stated, to many people are micro-analyzing>9 every single word that comes out of marketing these days.w  F For those doubting "show-me" readers, if no new Alpha business were to@ be allowed, how does one explain the following new VMS Customer:5 http://www.mgex.com/news/news_releases/release28.html-  > Or do you think Customers with these types of mission criticalH environments did not receive the assurances from the powers on high that  they had nothing to worry about?     Regards     
 Kerry Main Senior Consultant, Hewlett-Packard Canada# Consulting and Integration Services  Voice: 613-592-4660o Fax  :  819-772-7036 Email: Kerry.Main@hp.com     -----Original Message-----( From: John Smith [mailto:a@nonymous.com] Sent: May 13, 2002 3:01 PM To: Info-VAX@Mvb.Saic.Comr Subject: Re: No new Alpha sales       A "Howard Taylor" <Howard.Taylor@pacificcoast.net> wrote in message. news:3cdedfee$1@nubby2.... >n< > "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message( > news:3CDECE4C.2B1F0FEF@videotron.ca... > G > >  Rationale: We want to reinforce our commitment to our customers by  > following G > > the roadmaps we had already established. We're leading with PA-RISCn > > forO > newVF > > business opportunities for two reasons: First, the PA-RISC systems	 > > will,C > in; > > most cases, be upgradeable in the box to future ItaniumlC > > microprocessors. Second, HP-UX is the long-term UNIX for HP. ##e >  > This statement says to me:B > - HP will *not* recommend OpenVMS for new business opportunitiesG > - Because OpenVMS is not HP-UX (or UNIX, for that matter), OpenVMS iso *not*  > long-term for HP >>B > Does this sound correct, or have I mis-interpreted your posting? >e > Howard Taylorg     Sort of right.  F The forementioned  HP statement refers only to their decision of HP-UXG vs. Tru64, and in my opinion only because HP-UX has larger market share/D than Tru64, and hence less distruptive to the combined unix customerH base. If they had made the decision of which to chose based on technicalH superiority, then they would have chosen Tru64, but that also would haveG meant Alpha, and imagine the screams from HP-UX users who would then beoH in a position to have to buy Alpha boxes only to then have to buy Itanic boxes.  @ The bit about the boxes being field upgradable to Itanic is niceF marketing, but virtually none of the customers will do it. They'll buyC new boxes with newer bus architecutes and interconnects before theye# upgrade an existing 3-year old box.l    F Alpha will be sold until IA64 is ready, and probably for a short whileG after that. But if you are a new (to HP) customer, don't expect them tosF ever mention VMS to you. It'll be HP-UX or Windows that they'll try to stuff down your throat.r    G About VMS - you got that right too. There have previously been verbatimsG quotes taken from HP material about this, although some posts here seemeH to suggest that HP (Stallard in particular) still don;t have their mindsF made up on this. If you want VMS to live, write Carly, write Stallard, ditto Marcello et al.u  6 If you don't care about VMS anymore, call IBM for AIX.   ------------------------------  # Date: Wed, 15 May 2002 17:26:04 GMTt* From: "Bill Todd" <billtodd@metrocast.net> Subject: Re: No new Alpha sales A Message-ID: <MSwE8.81909$eV5.6421876@bin2.nnrp.aus1.giganews.com>c  2 "Main, Kerry" <Kerry.Main@hp.com> wrote in messageL news:BE56C50EA024184DAF48F0B9A47F5CF402664115@kaoexc01.americas.cpqcorp.net. ..    D >>> The forementioned  HP statement refers only to their decision of HP-UX vs. Tru64,<<  8 Very good response. Could not have stated it any better.  C As previous replies have stated, to many people are micro-analyzingn9 every single word that comes out of marketing these days.   F For those doubting "show-me" readers, if no new Alpha business were to@ be allowed, how does one explain the following new VMS Customer:5 http://www.mgex.com/news/news_releases/release28.htmlb   ***   I A reasonable explanation would be that this customer came on board beforei* the acquisition (and new policy) occurred.   - bill   ------------------------------  % Date: Wed, 15 May 2002 14:05:38 -0400l- From: JF Mezei <jfmezei.spamnot@videotron.ca>s Subject: Re: No new Alpha sales , Message-ID: <3CE2A370.C437F205@videotron.ca>   "Main, Kerry" wrote:H > For those doubting "show-me" readers, if no new Alpha business were toB > be allowed, how does one explain the following new VMS Customer:7 > http://www.mgex.com/news/news_releases/release28.html   H Oh come on. It is too soon for last week's announcements to have been inK effect at the time such deals would have been decided. Also, I didn't see a-H mention of VMS in there. Perhaps the whole sale was done by the softwareK company, at which point Compaq would have sold the hardware and licenses toCQ the software company who would have then sold a turn key package to the exchange.e    @ > Or do you think Customers with these types of mission criticalJ > environments did not receive the assurances from the powers on high that" > they had nothing to worry about?  L Well, nobody claims that HP intends to stop supporting VMS on Alpha any timeO soon. (Has anyone noticed the total lack of mention of VAX support for VMS ?). o  L The problem is that HP stated very clearly that new sales were to go PA-RISCM and not Alpha. How many new customers will fight against HP to gain the righteI to buy a new alpha ? It was bad enough during Compaq era to get VMS salesr, stuff, imagine what it will be like with HP.  L Even if HP is forced to eventually relax its rules in order to save revenus,E the fact remains that HP announced its colours last week and its trueaN intentions are not kind to VMS. They may temporarily have to bite their tongue8 and accept VMS, but it is clear that they don't want to.   ------------------------------    Date: 15 May 2002 09:44:11 -0700, From: srp336@getcoactive.com (Steve Pfister)% Subject: Re: OpenVMS Hobbyist License3< Message-ID: <45126e60.0205150844.a3006cc@posting.google.com>  E > If you are the impatient sort - you are never going to enjoy VMS...  > Good Luck.  < I can be patient... It just concerns me a bit when I send inF applications and emails and there's no response... nothing to indicate= that my mail was even received or read. I realize that a freer? membership is hardly on anybody's top priority list, and if the D application takes two weeks, or a month before anyone sees it, I canE understand that. But as it is, it's kind of indefinite. I did get ones4 message back last Friday saying "in a day or two"...  C sorry for venting...I think I've got too much going with too little( progress... :-)    Thanks!e   ------------------------------    Date: 15 May 2002 05:20:55 -0700. From: fabiopenvms@yahoo.com.br (Fabio Cardoso) Subject: Re: OpenVMS SNMP TRAP= Message-ID: <f30679fb.0205150420.42dbacd4@posting.google.com>o   Try      http://www.comtekservices.coml  % They have SNMP Agents for OpenVMS !!!      Regardsn  
 Fabio Cardosor   ------------------------------  / Date: Wed, 15 May 2002 08:58:49 +0200 (MET DST) & From: Rudolf Wingert <win@fom.fgan.de>5 Subject: Problem with the internal clock of an XP1000a6 Message-ID: <200205150658.IAA00653@sinet1.fom.fgan.de>   Hello,  C an user reports, that the clock of an XP1000 with OpenVMS 7.1-2 AXPtA slows down if the CPU is heave loaded. Does anybody see the same?b- Is there any patch out to solve this problem?n   TIA and regards Rudolf Wingert   ------------------------------  # Date: Wed, 15 May 2002 15:01:43 GMTe5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> 9 Subject: Re: Problem with the internal clock of an XP1000w9 Message-ID: <rLuE8.23$913.495707@cacnews.cac.cpqcorp.net>n  L What graphics are you using?  There is a problem in the ELSA server that wasK fixed recently that fixes a time loss problem.  It seems that there is some J bug in the hardware where a memory barrier in a spin loop reading the fifoI counter was putting the PCI into a disconnect/retry sequence, and causingl timer interrupts to be lost.   _Fredu   Rudolf Wingert wrote in message - <200205150658.IAA00653@sinet1.fom.fgan.de>...  >Hello,f >uD >an user reports, that the clock of an XP1000 with OpenVMS 7.1-2 AXPB >slows down if the CPU is heave loaded. Does anybody see the same?. >Is there any patch out to solve this problem? >l >TIA and regards Rudolf WingertS >@   ------------------------------  % Date: Wed, 15 May 2002 16:07:17 +02002, From: "Schild Niklaus" <schin@hta-bi.bfh.ch>9 Subject: problems setting socket nonblocking with fcntl()n* Message-ID: <abtq2g$1uu$1@rex.ip-plus.net>   Hi thereG I'm working on DEC C V6.0-001 on OpenVMS Alpha V7.3 and I need to set aa0 socket to non-blocking mode. The way I do it is:   ....)  11382:   opts = fcntl(socket, F_GETFL) ;i  11383:   if (opts LT ZERO)a
  11384:     {a(  11385:       perror("fcntl(F_GETFL)") ;  11386:       return (-1) ;d
  11387:     }  ...t@ then I get the error "fcntl(F_GETFL): function not implemented "? I the web then I found a peace of code which does it like this:f
 ..........,  11382:   opts = fcntl(socket, F_GETFL, 0) ;  11383:   if (opts LT ZERO)s
  11384:     {w(  11385:       perror("fcntl(F_GETFL)") ;  11386:       return (-1) ;-
  11387:     }e .....a1 an here I get "fcntl(F_GETFL): invalid argument".i  H Because the the third argument is declared as optional (help cc run-timeL fcntl) I do not understand both error messages. I use the following compiler flags:L /DECC/STAND=VAXC/MEMBER_ALIGN/INCLUDE_DIRECTORY=(DECC$USER_INCLUDE,C$INCLUDE5 )/WARNING=DISABLE=ADDRCONSTEXT/NOOPTIMIZE/DEBUG/OBJECp T=OBJ:/LIST=LISa  I Has anyone experiences wiht setting sockets to non-blocking on VMS or hass$ anyone an idea what I'm doing wrong?  
 Thanks Nik   ------------------------------    Date: 15 May 2002 04:49:29 -0700# From: tadamsmar@aol.com (Tom Adams)n Subject: Proxy question/; Message-ID: <793af3df.0205150349.a04646@posting.google.com>    If I add a proxy like this:t   add/proxy *::doe doe/d  C Then what determines what nodes can access the doe account and whathC node can't access it?   Can any node that can do a set host to thisi3 system access it.  What limit the set host command?    ------------------------------  % Date: Wed, 15 May 2002 14:16:31 +01002* From: "Rainer Giese" <giese@volkswerft.de> Subject: Re: Proxy questionr5 Message-ID: <abtjiv$ku8e5$1@ID-138444.news.dfncis.de>   6 "Tom Adams" <tadamsmar@aol.com> schrieb im Newsbeitrag5 news:793af3df.0205150349.a04646@posting.google.com...  > If I add a proxy like this:  >t > add/proxy *::doe doe/d >oE > Then what determines what nodes can access the doe account and whataE > node can't access it?   Can any node that can do a set host to this 5 > system access it.  What limit the set host command?n  J A user doe from _all_ other nodes can access to the local nodes via FAL orK task-to-task without an explicit authorization. He will mapped to the localf! user doe by default. For example:    $ DIR remote_node::n  ; will show the default directory of user doe at remote_node.aL If you create a proxy like "add/proxy *::doe other" (without /default), thenD he must give the username, that he want to use remote, for example :   $ DIR remote_node"OTHER"::  H otherwise he will mapped to the default decnet account, if you have one.  G It has nothing to do with set host, after that you are always asked forh username and password.   -- Regards, Rainer Giese   ------------------------------  % Date: Wed, 15 May 2002 08:36:33 -0700., From: "James Gessling" <jgessling@yahoo.com>Y Subject: Re: Reads vs. Writes (was Re: What is good model for disk i/o w        /shadowino4 Message-ID: <abtva3$kplqs$1@ID-46415.news.dfncis.de>  > on the hsXNN controlers vtdpy gives a good indication of this.   Jim:  - <norm.raphael@jamesbury.com> wrote in message_. news:C2256BB9.006021CE.00@jklh22.valmet.com... >  >t0 > How does one do this on a running application? >c >b >t >mG > andrew_nospam.harrison_remove_this@sun#.com on 05/13/2002 11:09:26 AMe >a? > Please respond to andrew_nospam.harrison_remove_this@sun#.com  >a > To:   Info-VAX@mvb.saic.comm > cc:a< > Subject:  Re: What is good model for disk i/o w/shadowing? >s >  >- >rH > ..the origional poster should be able to determine what % of reads and writes > their[cq] system has.... >. >l	 > Regardsm > Andrew Harrisont >e >t >o >c >n >  >  >    ------------------------------  % Date: Tue, 14 May 2002 06:12:20 -0700p6 From: "Howard Taylor" <Howard.Taylor@pacificcoast.net>% Subject: Re: System manager availablen  Message-ID: <3ce10bae$1@nubby2.>  @ "Bill Sticker" <NOSPAMPLEASE@SPAMSTOPPER.CO.UK> wrote in message) news:abqd23$rir$1@helle.btinternet.com...hA > Anyone need a good VMS system manager with 20 years experience?t > Nationality British. > info@johnoxley.com >n    I did see this post on Yahoo ...   Descriptionc  L ---------------------------------------------------------------------------- ----   HELP!! - COME OUT OF RETIRMENT) URGENTLY WANTED VMS SYSTEMS ADMINISTRATOR    IN PERTH  I Do you have extensive previous systems administration experience in a VMSf environment?  E Our Client urgently requires the skills of an experienced VMS Systemsc8 Administrator, will consider both CONTRACT or PERMANENT.  K For further information or confidential chat please contact Tony Jackson ons (08) 9212 1018 ASAP       
 Howard Taylor    ------------------------------  # Date: Wed, 15 May 2002 12:24:33 GMTh( From: "Mark E. Levy" <mlevy70@attbi.com>% Subject: Re: System manager availableO+ Message-ID: <5ssE8.5853$gD6.8618@sccrnsc01>   < "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message! news:3CE1D332.9592C5D2@fsi.net...  > Bill Sticker wrote:i > > 1 > > I'd switch to anything if I could find a job.e >cB > I've often said that I wished I'd gotten that job on the garbage
 > truck...  I But David, is that really any different from working with Windows? Either ( way, you're slinging garbage all day....  	 Mark Levys SMAe   ------------------------------  % Date: Wed, 15 May 2002 09:12:52 -0400 ! From: Jim Agnew <jpagnew@vcu.edu> % Subject: Re: System manager availableN' Message-ID: <3CE25ED4.868BBAB5@vcu.edu>$  H Garbage smells better, don't cha think???  At least the most you have toM worry about is biotoxic waste, dogs, etc...  and you see old garbage slingersi all day long, too.   "Mark E. Levy" wrote:d  > > "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message# > news:3CE1D332.9592C5D2@fsi.net...V > > Bill Sticker wrote:o > > >)3 > > > I'd switch to anything if I could find a job.  > >nD > > I've often said that I wished I'd gotten that job on the garbage > > truck... >lK > But David, is that really any different from working with Windows? Either1* > way, you're slinging garbage all day.... >m > Mark LevyR > SMAe   ------------------------------    Date: 15 May 2002 08:33:22 -0700- From: tessier-ashpool@usa.net (Chris Bardell)n% Subject: Re: System manager availableu= Message-ID: <9f261edc.0205150733.74b7a1e0@posting.google.com>   l "Bill Sticker" <NOSPAMPLEASE@SPAMSTOPPER.CO.UK> wrote in message news:<abr8un$djf$1@paris.btinternet.com>.../ > I'd switch to anything if I could find a job.a > 6 > "Howard S Shubs" <howard@shubs.net> wrote in message8 > news:howard-0C489C.06555514052002@enews.newsguy.com...3 > > In article <abqd23$rir$1@helle.btinternet.com>,o; > >  "Bill Sticker" <NOSPAMPLEASE@SPAMSTOPPER.CO.UK> wrote:e > >nE > > > Anyone need a good VMS system manager with 20 years experience?o > >l  B Know what you mean mate - the UK sucks for OpenVMS sysmgr/techsuppC jobs (contract & permie) at the mo. I'm in exactly the same boat as  you since end of Feb-2002.  E An earlier reponse was 'get into Unix'. A valid point, but to to jump F across into any other technical field often means an 'entry-level' jobC unless you have some proper commercial experience of the new field.uE Entry-level UK salaries are dreadful when you consider cost of livingm< here. Barely adequate for a single person, let alone someone supporting kids, etc.   F Sorry, wish I could suggest something more valid other than continuingE to search around. Also, I'm about to configure from scratch an old PCaA with Linux (& Apache, PHP, etc) to try to add *some* skills to myiF portfolio. If you have an old PC knocking about, it's cheap/free & the experience might prove useful.  ! Check out a recent similar threadw http://groups.google.com/groups?dq=&hl=en&threadm=87lmapdwzm.fsf%40prep.synonet.com&prev=/groups%3Fdq%3D%26num%3D25%26hl%3Den%26group%3Dcomp.os.vms%26start%3D50  F In that thread, you'll find that the reason you & I are out of work is< that we're "deadwood". OK, so it was perhaps only implied...   ------------------------------  % Date: Wed, 15 May 2002 08:00:27 -0500a% From: "J. Tang" <jaltang@hotmail.com>  Subject: TCP/IP for VMS...HELP!1A Message-ID: <3ce25c26$0$19585$4c41069e@reader1.ash.ops.us.uu.net>-  G I am in the process of upgrading a VMS system to an Windows 2000 server F based software system.  I need access to the VMS legacy box during theI process and for inquiry purposes after go live.  We are currently running@I Windows 98SE and using LAT to make connections into the VMS box.  Since IeD cannot use LAT with the Windows 2000 PC's, I am looking for a TCP/IPJ solution for the VMS box that is compatable with VMS version 6.2 (I cannotG upgrade the VMS operating system because of incompatibility issues with H software on the VMS box).  Any suggestions would be greatly appreciated.   Thanks,i JT   ------------------------------  % Date: Wed, 15 May 2002 06:06:16 -0700i# From: "Tom Linden" <tom@kednos.com>o# Subject: RE: TCP/IP for VMS...HELP!s9 Message-ID: <CIEJLCMNHNNDLLOOGNJIIENCEPAA.tom@kednos.com>o  D migrating would be a more descriptive and neutral term.  Do you needD more than telnet or ftp access?  for telnet I use Putty, its a free 	 download.    >-----Original Message-----t+ >From: J. Tang [mailto:jaltang@hotmail.com]o& >Sent: Wednesday, May 15, 2002 6:00 AM >To: Info-VAX@Mvb.Saic.Com  >Subject: TCP/IP for VMS...HELP! >  >oH >I am in the process of upgrading a VMS system to an Windows 2000 serverG >based software system.  I need access to the VMS legacy box during theeJ >process and for inquiry purposes after go live.  We are currently runningJ >Windows 98SE and using LAT to make connections into the VMS box.  Since IE >cannot use LAT with the Windows 2000 PC's, I am looking for a TCP/IPyK >solution for the VMS box that is compatable with VMS version 6.2 (I cannoteH >upgrade the VMS operating system because of incompatibility issues withI >software on the VMS box).  Any suggestions would be greatly appreciated.u >  >Thanks, >JTt >e >e >e >  >---' >Incoming mail is certified Virus Free. ; >Checked by AVG anti-virus system (http://www.grisoft.com).i@ >Version: 6.0.360 / Virus Database: 199 - Release Date: 5/7/2002 >h ---s& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).? Version: 6.0.360 / Virus Database: 199 - Release Date: 5/7/2002n   ------------------------------  + Date: Wed, 15 May 2002 15:16:39 +0100 (MET)e9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> # Subject: Re: TCP/IP for VMS...HELP! ; Message-ID: <01KHRESBJM8A95MSV4@sysdev.deutsche-boerse.com>c   Sorry, couldn't resist!t  I > I am in the process of upgrading a VMS system to an Windows 2000 server "                          ^^^^^^^^^  I think you meant "downgrading".  H > based software system.  I need access to the VMS legacy box during the9                                                    ^^^^^^rH I'll give you the benefit of the doubt.  A Mercedes-Benz (been around a A long time, still being developed and produced today) is a legacy   automobile.   K > process and for inquiry purposes after go live.  We are currently runninglK > Windows 98SE and using LAT to make connections into the VMS box.  Since IpF > cannot use LAT with the Windows 2000 PC's, I am looking for a TCP/IP!   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^g Sounds like a downgrade again.  L > solution for the VMS box that is compatable with VMS version 6.2 (I cannotI > upgrade the VMS operating system because of incompatibility issues with J > software on the VMS box).  Any suggestions would be greatly appreciated.  G Slightly more serious answer: there are several TCPIP stacks available -F for VMS, including one from the vendor of VMS.  In recent versions of 7 VMS, this has been included on the operating-system CD.e   ------------------------------  # Date: Wed, 15 May 2002 13:48:16 GMT L From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr")# Subject: Re: TCP/IP for VMS...HELP!l8 Message-ID: <00A0DF64.F4515976@SSRL04.SLAC.STANFORD.EDU>  i In article <3ce25c26$0$19585$4c41069e@reader1.ash.ops.us.uu.net>, "J. Tang" <jaltang@hotmail.com> writes:a  H >I am in the process of upgrading a VMS system to an Windows 2000 serverG >based software system.  I need access to the VMS legacy box during thenJ >process and for inquiry purposes after go live.  We are currently runningJ >Windows 98SE and using LAT to make connections into the VMS box.  Since IE >cannot use LAT with the Windows 2000 PC's, I am looking for a TCP/IPlK >solution for the VMS box that is compatable with VMS version 6.2 (I cannot H >upgrade the VMS operating system because of incompatibility issues withH >software on the VMS box).  Any suggestions would be greatly appreciated  J 1) Why can't you use LAT on Windows 2000?  I run Reflection 4 on a W2k box and it supports LAT.  D 2) You're looking for either Multinet or TCPware to run on VMS 6.2. , www.process.com should have links to either.   -- Alan?  O =============================================================================== 0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056cM  Physical mail to: SSRL -- SLAC BIN 69, PO BOX 4349, STANFORD, CA  94309-0210yO ===============================================================================    ------------------------------    Date: 15 May 2002 08:55:49 -0600- From: koehler@encompasserve.org (Bob Koehler)t# Subject: Re: TCP/IP for VMS...HELP!b3 Message-ID: <cHsfnuC4Rtp1@eisner.encompasserve.org>6  w In article <01KHRESBJM8A95MSV4@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:p > Sorry, couldn't resist!m > J >> I am in the process of upgrading a VMS system to an Windows 2000 server$ >                          ^^^^^^^^^" > I think you meant "downgrading". > I >> based software system.  I need access to the VMS legacy box during thet; >                                                    ^^^^^^ J > I'll give you the benefit of the doubt.  A Mercedes-Benz (been around a C > long time, still being developed and produced today) is a legacy .
 > automobile.g  +    An a Subaru Legacy is a pretty fine car.    ------------------------------    Date: 15 May 2002 08:54:54 -0600- From: koehler@encompasserve.org (Bob Koehler)r# Subject: Re: TCP/IP for VMS...HELP!S3 Message-ID: <7mWOd83BUT8L@eisner.encompasserve.org>@  i In article <3ce25c26$0$19585$4c41069e@reader1.ash.ops.us.uu.net>, "J. Tang" <jaltang@hotmail.com> writes: I > I am in the process of upgrading a VMS system to an Windows 2000 server,H > based software system.  I need access to the VMS legacy box during theK > process and for inquiry purposes after go live.  We are currently running0K > Windows 98SE and using LAT to make connections into the VMS box.  Since I F > cannot use LAT with the Windows 2000 PC's, I am looking for a TCP/IPL > solution for the VMS box that is compatable with VMS version 6.2 (I cannotI > upgrade the VMS operating system because of incompatibility issues with-J > software on the VMS box).  Any suggestions would be greatly appreciated.  N    No, you are in the process of downgrading  a VMS system to an Windows 2000 
    server.  C    Suggestion:  don't do it, your cost of ownership will skyrocket.    ------------------------------  + Date: Wed, 15 May 2002 14:18:02 +0000 (UTC)  From: david20@alpha2.mdx.ac.uk# Subject: Re: TCP/IP for VMS...HELP!r+ Message-ID: <abtqmq$rdv$2@aquila.mdx.ac.uk>n   In article <00A0DF64.F4515976@SSRL04.SLAC.STANFORD.EDU>, winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr") writes:nj >In article <3ce25c26$0$19585$4c41069e@reader1.ash.ops.us.uu.net>, "J. Tang" <jaltang@hotmail.com> writes: >eI >>I am in the process of upgrading a VMS system to an Windows 2000 server H >>based software system.  I need access to the VMS legacy box during theK >>process and for inquiry purposes after go live.  We are currently running-K >>Windows 98SE and using LAT to make connections into the VMS box.  Since IiF >>cannot use LAT with the Windows 2000 PC's, I am looking for a TCP/IPL >>solution for the VMS box that is compatable with VMS version 6.2 (I cannotI >>upgrade the VMS operating system because of incompatibility issues withcI >>software on the VMS box).  Any suggestions would be greatly appreciatedT > K >1) Why can't you use LAT on Windows 2000?  I run Reflection 4 on a W2k box( >and it supports LAT.  > E >2) You're looking for either Multinet or TCPware to run on VMS 6.2. F- >www.process.com should have links to either.I >u    N Or UCX version 4.2 with the latest eco patches. DEC TCPIP Services (UCX) 5 and above won't run on VMS 6.2. O Look around and see if you already have the license (either a UCX license or a 3L NAS license) and the software on a CD. If not then you will probably have to) goto process and get Multinet or Tcpware.h  
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  % Date: Wed, 15 May 2002 09:38:13 -0500D% From: "J. Tang" <jaltang@hotmail.com>E# Subject: Re: TCP/IP for VMS...HELP!/A Message-ID: <3ce2730f$0$19587$4c41069e@reader1.ash.ops.us.uu.net>I  H I apologize for the terminalogy used in my post.  I love the VMS system.8 This was not my idea, I am merely stuck implementing it.   JT.-    : "Bob Koehler" <koehler@encompasserve.org> wrote in message- news:7mWOd83BUT8L@eisner.encompasserve.org...oG > In article <3ce25c26$0$19585$4c41069e@reader1.ash.ops.us.uu.net>, "J. # Tang" <jaltang@hotmail.com> writes:dK > > I am in the process of upgrading a VMS system to an Windows 2000 servercJ > > based software system.  I need access to the VMS legacy box during theE > > process and for inquiry purposes after go live.  We are currently- running-K > > Windows 98SE and using LAT to make connections into the VMS box.  Since" IiH > > cannot use LAT with the Windows 2000 PC's, I am looking for a TCP/IPG > > solution for the VMS box that is compatable with VMS version 6.2 (I  cannotK > > upgrade the VMS operating system because of incompatibility issues withpL > > software on the VMS box).  Any suggestions would be greatly appreciated. >oJ >    No, you are in the process of downgrading  a VMS system to an Windows 2000 >    server. >hE >    Suggestion:  don't do it, your cost of ownership will skyrocket.  >s   ------------------------------    Date: 14 May 2002 23:52:04 -07006 From: andrew.rycroft@intrinsitech.com (Andrew Rycroft)< Subject: Telecom fix-line billing systems running on OpenVMS= Message-ID: <58ba0101.0205142252.4c7da4ac@posting.google.com>    Hi,r  @ Does anyone know of a fix line telecom billing system running on3 OpenVMS ? We have a potential opportunity for this.a   Thanks Andrew   ------------------------------  + Date: Wed, 15 May 2002 16:24:53 +0000 (UTC) 5 From: "Bill Sticker" <NOSPAMPLEASE@SPAMSTOPPER.CO.UK>n@ Subject: Re: Telecom fix-line billing systems running on OpenVMS/ Message-ID: <abu24l$a1n$1@paris.btinternet.com>e  K I know one. it is called CIS and will comfortably bill upto 400000 accounts K per month. If you want a copy, let me know and I'll try to get one for you.w J. OxleyC "Andrew Rycroft" <andrew.rycroft@intrinsitech.com> wrote in message$7 news:58ba0101.0205142252.4c7da4ac@posting.google.com...r > Hi,  >eB > Does anyone know of a fix line telecom billing system running on5 > OpenVMS ? We have a potential opportunity for this.l >. > Thanks > Andrew   ------------------------------  # Date: Wed, 15 May 2002 12:06:11 GMTs2 From: "William Hymen" <t18_pilot@hotmail.spam.com>! Subject: UNIX Security a good betnD Message-ID: <TasE8.2150$l82.238240@newsread1.prod.itd.earthlink.net>  3 I think a move to unix security administration is ar6 good bet for job security.  Our bank requires 4 layers4 of aftemarket software to make unix as secure as our7 vms system is "out of the box". Our unix security staffm0 is vastly larger than our vms security staff is.   Bill    > "I haven't lost my mind; It's backed up on tape somewhere ..."    < "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message! news:3CE1CAD9.57A28872@fsi.net...e > "Stuart, Ed" wrote:u > >>F > > The "Special Edition of the TRU64 UNIX and OpenVMS Times" has been	 modified.i+ > > The OpenVMS/HP-UX migration Q&A is now:  > >V* > > Q: For OpenVMS customers who have made- > > a firm decision to move to UNIX, will you $ > > offer a migration path to HP-UX? > >e5 > > A: Yes. For our OpenVMS customers who have made a-6 > > decision to move to UNIX, we believe that HP-UX is8 > > an excellent choice, and we will work with them on a6 > > migration plan and provide tools and services that( > > can help ensure a smooth transition. > >mJ > > Looks like the folks at the new HP read those emails that were sent by > > OpenVMS advocates. >s? > Well, here's some questions and answers that I'd like to see:n >eG > Q: For OpenVMS customers who have made a firm commitment to stay withoH > OpenVMS, does HP have a plan to provide hardware support until a fully; > viable, commercial-quality IPF processor finally appears?cI > A: Yes - HP will continue to ensure the availability of Alpha Processor1F > Family systems until IPF is ready for prime time, no matter how long
 > that takes.5 >fB > Q: For former OpenVMS customers, ISVs and VARs who are exploringH > returning to OpenVMS, does HP have a plan to court them, support their; > decision to return and help them migrate back to OpenVMS?lD > A: Of course! OpenVMS is unsurpassed for versatility, security andI > stability and we'll do everything humanly possible - and then some - to-4 > help bring the former faithful back into the fold! >eH > Q: For OpenVMS customers who need applications not currently availableG > for OpenVMS, does HP have a plan to work with ISVs to get more app.'s ' > written for and/or ported to OpenVMS?uE > A: Oh, yes! Going forward, we're budgeting unprecedented amounts tol@ > recruit and retain ISVs, VARs and OEMs for OpenVMS systems andC > applications. This includes OpenVMS advertising in the mainstreamfI > publications (not just those visible to CIOs and other executive types)iH > as well as adjustments to OpenVMS's pricing structure so resellers canI > finally compete against alternate platform proposals dollar-for-dollar! J > We're serious this time! OpenVMS is our most lucrative product and we'reD > going full-throttle after a market that has been too-long ignored! > 7 > (Earth calling David J.!! Come back to reality, son!)@ >. > -- > David J. Dachterac > dba DJE Systemse > http://www.djesys.com/ >2* > Unofficial Affordable OpenVMS Home Page:! > http://www.djesys.com/vms/soho/S >S   ------------------------------  + Date: Wed, 15 May 2002 14:46:17 +0100 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> ! Subject: UNIX Security a good betm; Message-ID: <01KHRDUZU2LG95MSV4@sysdev.deutsche-boerse.com>.  F > I think a move to unix security administration is a good bet for jobF > security.  Our bank requires 4 layers of aftemarket software to makeI > unix as secure as our vms system is "out of the box". Our unix securitye9 > staff is vastly larger than our vms security staff is. o  F In the "Security" chapter of the Unix Hater's Handbook, it says "when B your sysadmin mentions security, he's talking about his job".  :-)   ------------------------------  # Date: Wed, 15 May 2002 11:59:54 GMTn- From: Jack Patteeuw <jjpatteeuw@peoplepc.com>  Subject: Version of the KZPBAw, Message-ID: <3CE24DD5.726541CD@peoplepc.com>   I am officially confused !  J There appears to be several variation of the KZPBA PCI - SCSI Differential> adapter.  I have seen references to -CA, -CB and -CC variants.  = Can some one explain the differences between these versions ?i  # Do they all use the same driver ?  x  M I have various different vintages of Alphas, all with PCI I/O bus. Is there aaB simple way I can find which variant is compatible with which box ?  9 Is there any significance between the DS- an 3x- prefix ?p  2 Is there a newer PCI - SCSI Differential adapter ?  - (Don't say talk to sales, they are clueless.)% -- 3  
 Jack Patteeuwp   ------------------------------  % Date: Wed, 15 May 2002 08:25:18 -0500 C From: "Craig A. Berry" <craig.berry@nospam.SignalTreeSolutions.com> ! Subject: Re: Version of the KZPBA-H Message-ID: <craig.berry-3AD81C.08251715052002@news.directvinternet.com>  , In article <3CE24DD5.726541CD@peoplepc.com>,/  Jack Patteeuw <jjpatteeuw@peoplepc.com> wrote:1  L > There appears to be several variation of the KZPBA PCI - SCSI Differential@ > adapter.  I have seen references to -CA, -CB and -CC variants. > ? > Can some one explain the differences between these versions ?e  D I think CA is single-ended and CB is differential.  CX is the field E replacement version of CA and CY is the field replacement version of P CB.  I've never heard of CC.  % > Do they all use the same driver ?  a  @ Possibly, but that does not mean they are interchangeable.  The F KZPBA-CB is HVD and thus cannot deal with any single-ended devices on ! the bus the way LVD adapters can.c  O > I have various different vintages of Alphas, all with PCI I/O bus. Is there a0D > simple way I can find which variant is compatible with which box ?  H You need to consult both the VMS SPD and the hardware documentation for  supported options.    ; > Is there any significance between the DS- an 3x- prefix ?a > 4 > Is there a newer PCI - SCSI Differential adapter ?  H There's the KZPEA, but I think it costs a thousand bucks and only works = on VMS 7.3 or 7.2-2 (plus ECO in either case) on EV6 systems.I  G If you're poking around on the resale market, also keep an eye out for eG the KZPCM which has SCSI and ethernet on one card (but I don't know if w it's differential).c  / > (Don't say talk to sales, they are clueless.)d   ------------------------------   Date: 15 MAY 2002 13:51:36 GMT4 From: karcher@thuria.waisman.wisc.edu (Carl Karcher)! Subject: Re: Version of the KZPBAi6 Message-ID: <15MAY02.13513655@thuria.waisman.wisc.edu>  E In a previous article, Jack Patteeuw <jjpatteeuw@peoplepc.com> wrote:e  L ->There appears to be several variation of the KZPBA PCI - SCSI Differential@ ->adapter.  I have seen references to -CA, -CB and -CC variants. ->  ? ->Can some one explain the differences between these versions ?f -> i  G They are all Ultra-SCSI. From the info I have (the manuals), the -CA isVE single ended, the -CB is differential, the -CC is the replacement fors the -CB (which is eol'ed).  % ->Do they all use the same driver ?  y  D I believe they do. Same firmware too. They all have "Qlogic ISP1020" in the SCSI ID string.  O ->I have various different vintages of Alphas, all with PCI I/O bus. Is there a D ->simple way I can find which variant is compatible with which box ?  D You'll have to see the supported options list for each box. Or try a google search.   --G -- Carl Karcher, Waisman Computing Services, Waisman Center, UW-Madison  --   ------------------------------  + Date: Wed, 15 May 2002 12:07:55 +0000 (UTC)w From: david20@alpha1.mdx.ac.uk' Subject: Re: Who cares about marketing!o+ Message-ID: <abtj2r$p60$1@aquila.mdx.ac.uk>   \ In article <3CE172B7.A24BEC0C@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: >Bill Todd wrote:i >>K >I don't think you can pin it to a single fault. It was a systematic fault.nK >Prices kept too high, dropping of education as an important focus, lack of8M >advertising. And another important one was the use of proprietary techniquesLL >to prevent loss of revenus (such as expoxy in a bus) which not only angeredM >customers, but also gave Sun et al great ammunution to call DEC "proprietarya >and expensive". >V  N Although the epoxy was rather blatant I don't think DEC were the first or lastJ company to produce two versions of a product (a fully functional one and aN cheaper crippled version). I don't see whats proprietary about that unless you2 mean the specific brand of epoxy which was used :)  
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  % Date: Wed, 15 May 2002 09:08:11 -0500 1 From: "Dave Gudewicz" <david.gudewicz@abbott.com>N' Subject: Re: Who cares about marketing!g0 Message-ID: <abtqib$kr$1@fizban.pprd.abbott.com>  J From what I understand, Oracle up to v6 was first on VMS, then on everyoneJ else.  Then the pendulum swung in Sun's direction.  And AFAIK the pendulum never stops swinging.e   -- Dave...r  ) Adam and Eve had many advantages, but the - principle one was that they escaped teething.M -----Mark Twainn  5 "Bill Todd" <billtodd@metrocast.net> wrote in message < news:M8kE8.140133$M7.13915207@bin7.nnrp.aus1.giganews.com... >n% > "Andrew Harrison SUNUK Consultancy"n@ > <andrew_nospam.harrison_remove_this@sun#.com> wrote in message, > news:abs935$7p4$1@new-usenet.uk.sun.com... > >  > >e > > Bill Todd wrote: > >a; > > > "Nic Clews" <sendspamhere@127.0.0.1> wrote in message ) > > > news:3CE12BDA.47C2B852@127.0.0.1...s > > >o	 > > > ...t > > >e > > > J > > >>History. He we are ~15 years later and where is VMS? Technically farK > > >>superior than it was yet  not where it should be. This is a different  > > >>world2 > > >> > > >.@ > > > Why?  What evidence do you have to offer that the world is
 significantly=J > > > different in terms of high-tech than it was during the 30 years that DECA9 > > > succeeded - wildly - without traditional marketing?L > > >S > >P > > B > > 30 years ago the SW market was not dominated by a small numberE > > of vendors because applications suites like SAP, PeopleSoft, JDE,=E > > Oracle, Baan Siebel etc did not exist. If you wanted a payroll or H > > general  ledger app you rolled your own or payed someone to roll one > > for you. >:K > You need to read more carefully:  the period under discussion ended abouto 15J > years ago, not 30.  At that point in time, there *was* major third-partyL > software available (certainly Oracle), and during the intervening period aH > lot of it (certainly SAP, I think PeopleSoft and Baan, likely others - it'sJ > not an area I keep up with) *has* been available on VMS (though a lot of? > them dropped support as VMS's fortunes continued to dwindle).d >n > >tB > > DEC was a leader in the mini-computer market without garneringD > > any support from these vendors (they didn't exist in that form). > I > DEC certainly garnered support from Oracle, even while it was competingoK > against it with Rdb.  For a very long time VMS was the *only* environment- > where OPS ran. >5 > >2D > > 30 years on its mostly about applications with a minimal ammountA > > of customisation that grabs customers very few people want tor > > roll their own.  > >aD > > Digital and Compaq for Ultrix, Tru64 and OpenVMS were singularly< > > unsucessfull at getting ISV's to support their platform. >iF > Only after DEC had already declined:  they had no problem whatsoever% > attracting ISV support before that.g >e > >r- > > Ultrix was less well supported than SunOSW< > > Tru64 is less well supported than Solaris, AIX and HP-UX > > ditto for OpenVMS. >oH > And all of this post-dates DEC's decline, rather than was in any way a cause/ > of it. >n > - bill >b >t >    ------------------------------  # Date: Wed, 15 May 2002 15:17:58 GMT # From: "John Smith" <a@nonymous.com>u' Subject: Re: Who cares about marketing!yG Message-ID: <G_uE8.19959$t8_.6036@news01.bloor.is.net.cable.rogers.com>s  5 "Bill Todd" <billtodd@metrocast.net> wrote in messager; news:3clE8.20443$Ze4.2056295@bin5.nnrp.aus1.giganews.com...i >s >oK > Sun took over because  1) its products were a lot cheaper and  2) DEC was0I > both complacent and arrogant.  Neither of these has anything much to doE with > marketing. >m  H Pricing IS part of the marketing process. It is just as much an issue as whether,K you do advertising or not, have a decent educational licencing plan or not,sI take you customers to basketball games or not, make presentations to sellT your gear or not, etc....w  K Pricing determines who you get to talk to. If it's a high priced item, thens you J effectively limit the number of people you get to talk to and who can signA the purchase order - the converse is true for lower priced items.      >-I > If you continue to concentrate on what has happened since the mid-'80s, / > you'll continue to completely miss the point.I  # SSDD. (same shit, different decade)r   ------------------------------  # Date: Wed, 15 May 2002 15:18:04 GMT # From: "John Smith" <a@nonymous.com>n' Subject: Re: Who cares about marketing! H Message-ID: <M_uE8.19960$t8_.18714@news01.bloor.is.net.cable.rogers.com>  5 "Bill Todd" <billtodd@metrocast.net> wrote in message < news:6QfE8.178251$q8.18022723@bin3.nnrp.aus1.giganews.com... >t > L > *After* DEC fell from grace, *then* the kinds of things you describe aboveJ > *did* become problems, and marketing could have helped (though not saved DECvK > unless the other, more basic, internal problems had been solved as well).r >y  L Wrong. These were exactly the things that CAUSED the 'fall from grace'.  TheI crap Sun was selling at the time couldn't hold a technical candle to what I DEC could and was delivering technically. But Sun ate DEC's lunch becauserI they did ALL the things that constitute marketing well - and in my books,0J marketing starts with the CEO and the Board of Directors and permeates the whole organization  L It's making sure that calls get returned; it means that ordering new gear isL easy and painless; it means that executives get schmoozed; it means that youB offer to bring experts in to solve peculiar problems for free fromG time-to-time; it means advertising; it means getting lots of young kidseH experienced with your stuff for no cost; it means that your products are priced competitively.g  J If mentioned this before and I think virtually everyone will agree that inK the mid-late 80's when DEC 'owned' the mid-range market, all they had to dolG to seriously cripple unix increasing its market share was to drop their K prices. Nobody I knew who was sane *wanted* to voluntarily port from VMS tooL unix at the time, but they did so because DEC made it financially attractiveH to do so (and that was part of their flawed marketing strategy). At thatK time it was Sun that benefited. And dealing with Sun was a whole lot easiercL to do than dealing with Digital when it came to paperwork and billing. I mayL get some disagreement here when I say this, but in my estimation, Sun didn'tH have any hardware that could effectively compete with the larger systemsE from DEC until around 1995, yet people put up with all Sun's relative-L deficiencies in hardware gear and software for a long time because they WERE marketed to at all levels.   ------------------------------  % Date: Wed, 15 May 2002 10:40:27 -0500 1 From: "Dave Gudewicz" <david.gudewicz@abbott.com>o' Subject: Re: Who cares about marketing! 1 Message-ID: <abtvm9$1gt$1@fizban.pprd.abbott.com>M  : The story that John tells is what went on where I sit now.   -- Dave...-  ) Adam and Eve had many advantages, but the2- principle one was that they escaped teething.O -----Mark TwainO  . "John Smith" <a@nonymous.com> wrote in messageB news:M_uE8.19960$t8_.18714@news01.bloor.is.net.cable.rogers.com... >t7 > "Bill Todd" <billtodd@metrocast.net> wrote in messagel> > news:6QfE8.178251$q8.18022723@bin3.nnrp.aus1.giganews.com... > >n > >dH > > *After* DEC fell from grace, *then* the kinds of things you describe abovesL > > *did* become problems, and marketing could have helped (though not saved > DECrF > > unless the other, more basic, internal problems had been solved as well). > >p >MI > Wrong. These were exactly the things that CAUSED the 'fall from grace'.- The-K > crap Sun was selling at the time couldn't hold a technical candle to whatmK > DEC could and was delivering technically. But Sun ate DEC's lunch becausesK > they did ALL the things that constitute marketing well - and in my books, L > marketing starts with the CEO and the Board of Directors and permeates the > whole organization >aK > It's making sure that calls get returned; it means that ordering new gearo isJ > easy and painless; it means that executives get schmoozed; it means that youlD > offer to bring experts in to solve peculiar problems for free fromI > time-to-time; it means advertising; it means getting lots of young kidsCJ > experienced with your stuff for no cost; it means that your products are > priced competitively.d >eL > If mentioned this before and I think virtually everyone will agree that inJ > the mid-late 80's when DEC 'owned' the mid-range market, all they had to doI > to seriously cripple unix increasing its market share was to drop their J > prices. Nobody I knew who was sane *wanted* to voluntarily port from VMS toC > unix at the time, but they did so because DEC made it financiallye
 attractiveJ > to do so (and that was part of their flawed marketing strategy). At thatF > time it was Sun that benefited. And dealing with Sun was a whole lot easierJ > to do than dealing with Digital when it came to paperwork and billing. I may G > get some disagreement here when I say this, but in my estimation, Suna didn'tJ > have any hardware that could effectively compete with the larger systemsG > from DEC until around 1995, yet people put up with all Sun's relativelI > deficiencies in hardware gear and software for a long time because theyh WERE > marketed to at all levels. >F >    ------------------------------  # Date: Wed, 15 May 2002 16:51:14 GMTr0 From: prune@ZAnkh-Morpork.mv.com (Paul Winalski)' Subject: Re: Who cares about marketing!e9 Message-ID: <3ce29125.2843516750@proxy.news.easynews.com>a  F On Wed, 15 May 2002 04:11:29 GMT, "Bill Todd" <billtodd@metrocast.net> wrote:  J >Sun took over because  1) its products were a lot cheaper and  2) DEC wasM >both complacent and arrogant.  Neither of these has anything much to do with  >marketing.O  E Pricing is a vitally important part of marketing.  So is knowing yourn3 competition, and not being complacent and arrogant.e   >>C >> From the perspective of someone who watched the whole thing playwE >> out from start to finish from within, the internal bickering was aVG >> side issue.  The main problem was that the old, traditional, pre-VAXpC >> DEC never "got it" when it came to marketing and its importance.v >bL >They didn't 'get it' because they didn't need to - until *DEC* changed (not >the external circumstances).c  A No, the circumstances changed.  DEC grew too large for the little-F niche it grew up in, didn't realize it had to actively sell and marketE its products rather than just take orders, never learned either skilln. properly, and as a consequence got blown away.  . >> I'm afraid that you, Bill, are one of them. >r0 >Damn right.  And I'm afraid you are clueless.    B We'll have to agree to disagree on this issue, then, since we both& consider each other clueless about it.   >However, IIRC you do have theM >excuse of having been cocooned within the VMS development group at the time,cM >where what was going wrong with the rest of the company was far less visiblevE >(since VMS itself seemed to be moving along as well as it ever had).i  A Wrong--I was in the Software Tools Group.  I've never been in VMSa Develpment.t  M >Or do you think that Cutler located his team as far away from New England ascL >was physically possible without crossing an international border because of >marketing deficiencies?  D I'm not saying DEC didn't have internal political problems, which it9 certainly did.  Cutler's arrogance, BTW, was one of them.-  
 ---------- Remove 'Z' to reply by email.n   ------------------------------  # Date: Wed, 15 May 2002 17:45:21 GMT * From: "Bill Todd" <billtodd@metrocast.net>' Subject: Re: Who cares about marketing! = Message-ID: <R8xE8.348$Yg7.88890@bin4.nnrp.aus1.giganews.com>i  = "Paul Winalski" <prune@ZAnkh-Morpork.mv.com> wrote in messagev3 news:3ce29125.2843516750@proxy.news.easynews.com...eH > On Wed, 15 May 2002 04:11:29 GMT, "Bill Todd" <billtodd@metrocast.net> > wrote: >tL > >Sun took over because  1) its products were a lot cheaper and  2) DEC wasJ > >both complacent and arrogant.  Neither of these has anything much to do with
 > >marketing.p >iG > Pricing is a vitally important part of marketing.  So is knowing yourt5 > competition, and not being complacent and arrogant.u  C See other post:  the old DEC had no deficiencies in any such areas.    >  > >>E > >> From the perspective of someone who watched the whole thing playcG > >> out from start to finish from within, the internal bickering was a0I > >> side issue.  The main problem was that the old, traditional, pre-VAX,E > >> DEC never "got it" when it came to marketing and its importance.o > >eI > >They didn't 'get it' because they didn't need to - until *DEC* changede (not > >the external circumstances).t > C > No, the circumstances changed.  DEC grew too large for the little H > niche it grew up in, didn't realize it had to actively sell and marketG > its products rather than just take orders, never learned either skill 0 > properly, and as a consequence got blown away.  = Wrong.  As I said before, DEC managed the move from primarilyeH engineering/scientific/academic sales to primarily commercial sales justL fine.  What it outgrew was its britches, not its niche:  had it continued toL concentrate on providing what customers needed, it could have continued just fine.h   - bill   ------------------------------  # Date: Mon, 06 May 2002 14:51:47 GMTa3 From: sy18889@rabbit.fmr.com (Bradford J. Hamilton).0 Subject: Re: WIS??? Did it die with the HP deal?. Message-ID: <7MwB8.15$O3.39@news-srv1.fmr.com>  	 Hi Brian,i  * Works OK from here as of 11:00, Eastern...    P In article <00A0D86F.422C593B@SendSpamHere.ORG>, system@SendSpamHere.ORG writes:K >I've been trying to check on issues I've reported via WIS.  However, thereeK >doesn't seem to be a WIS since the start of the weekend.  Is this just thed+ >beginning of things to come under HP rule?n >s >--lP >VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM >           6 >  "Well my son, life is like a beanstalk, isn't it?"  >  Bradford J. Hamilton& braMdhamAilPtoSn@aMtAtPbi.cSom		(home)& sMy1A88P89S@rabMbit.fAmPr.coSm		(work)  ; "All opinions that I express are my own, not my employer's"  "Lose the MAPS"    ------------------------------  # Date: Tue, 07 May 2002 14:29:01 GMT 3 From: sy18889@rabbit.fmr.com (Bradford J. Hamilton) - Subject: Re: www.compaq.com is link to hp.comt. Message-ID: <NwRB8.16$O3.38@news-srv1.fmr.com>   Just for fun, I typed in,    www.openvms.compaq.com  4 and came to the familiar VMS page.  However, typing,   www.openvms.hp.com  $ brings up a "502 Bad Gateway" error.   Just an oversight, I'm sure...  c In article <ZRIFM7Vtdmk9@eisner.encompasserve.org>, koehler@encompasserve.org (Bob Koehler) writes:o >dG >   Well, there it is.  Enter www.compaq.com and end up a hp.com.  Evenw# >   www.digital.com gets you there.i > A >   I could find VMS by following the Software and OS link, but Io6 >   couldn't find Alpha by following the servers link. >, Bradford J. Hamilton& braMdhamAilPtoSn@aMtAtPbi.cSom		(home)& sMy1A88P89S@rabMbit.fAmPr.coSm		(work)  ; "All opinions that I express are my own, not my employer's"m "Lose the MAPS":   ------------------------------   End of INFO-VAX 2002.268 ************************