1 INFO-VAX	Thu, 22 Aug 2002	Volume 2002 : Issue 461       Contents: Re: .CLD questions.  Re: .CLD questions. # Re: attn: Paddy (re: DCL procedure) . Re: Cheapest way to repair/replace tape drives$ Re: DE101 - can it go to Full Duplex$ Re: DE101 - can it go to Full Duplex1 Re: Determining what .h file a function is in...? 1 RE: Determining what .h file a function is in...? ' Re: Fortune Magazine and a post-VMS rap ' Re: Fortune Magazine and a post-VMS rap ' Re: Fortune Magazine and a post-VMS rap ' Re: Fortune Magazine and a post-VMS rap ' Re: Fortune Magazine and a post-VMS rap  Re: Fortune Magazine wrapper Re: Fortune Magazine wrapper Re: Fortune Magazine wrapperP Re: forwarding email from system to system (was: RE: attn: Paddy (re: DCL procedP Re: Geographically Dispersed Parallel Sysplex and replication (was Re: HP-CompaqP Re: Geographically Dispersed Parallel Sysplex and replication (was Re: HP-CompaqP Re: Geographically Dispersed Parallel Sysplex and replication (was Re: HP-Compaq* Re: How to configure Reflection X-server ?" Re: HP-Compaq Merger Went Smoothly" Re: HP-Compaq Merger Went Smoothly" Re: HP-Compaq Merger Went Smoothly" Re: HP-Compaq Merger Went Smoothly" Re: HP-Compaq Merger Went Smoothly" Re: HP-Compaq Merger Went Smoothly" Re: HP-Compaq Merger Went Smoothly" Re: HP-Compaq Merger Went Smoothly" Re: HP-Compaq Merger Went Smoothly" installation using VMSINSTAL failsG Re: Memo:  Re: Charon-VAX (was: [VAX] VMS to [Alpha] OpenVMS conversion  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP  Re: New missive from HP " Re: OpenVms / Vax -- free download" Re: OpenVms / Vax -- free download" Re: OpenVms / Vax -- free download% RE: oracle/telnet-tcpip 5.3/vms 7.3-1 % Re: oracle/telnet-tcpip 5.3/vms 7.3-1 / Re: Printing barcodes via postscript on OpenVMS ) Re: Problem with DCL procedure using PIPE ) Re: Problem with DCL procedure using PIPE ) Re: Problem with DCL procedure using PIPE ) Re: Problem with DCL procedure using PIPE  Re: Proposal for DECUS US C Re: Proposal for DECUS US (was: Fortune Magazine and a post-VMS rap  Re: Roadshow update  Re: Roadshow update   Re: Size of logical name table ?  Re: Size of logical name table ?  Re: Size of logical name table ?  Re: Size of logical name table ? telnet from *NIX to OpenVMS 7.2 # Re: telnet from *NIX to OpenVMS 7.2 # Re: telnet from *NIX to OpenVMS 7.2 # Re: telnet from *NIX to OpenVMS 7.2 # Re: telnet from *NIX to OpenVMS 7.2   Re: total VMS newbie - pointers?  Re: total VMS newbie - pointers?$ Re: VAX/Alpha Basic example programs$ Re: VAX/Alpha Basic example programs Re: VMS to OpenVMS conversion  Re: VMS to OpenVMS conversion * Re: VMS upgrade with shadowed system disk.* Re: VMS upgrade with shadowed system disk.$ Re: Why C is better than Fortran 95? Re: XFC v2 ECO, negative report  Re: XFC v2 ECO, negative report   F ----------------------------------------------------------------------  % Date: Wed, 21 Aug 2002 13:36:18 -0700 > From: "Kenneth H. Fairfield" <My-Full-Name@intel-com.spam-ban> Subject: Re: .CLD questions.2 Message-ID: <3D63F9C2.1837EE26@intel-com.spam-ban>   Lyndon Bartels wrote:   6 > Have a CLD question.... and no, this isn't homework. > C > I want to create a .CLD so the program has the follwing behavior.  > E > /FILE=filespec    { filespec other than default. Always an option } G > /HTML             { switch to format output in html format. Only good 	 > /LIST } 6 > /CREATE           { switch to create new data file }3 > /OUTPUT=filespec  { filespec to send output to. } J > /LIST[=username]  { to output a user, if only one is specified. Or all }4 >                   { if no username is specified. } > ! > /MODIFY=username/field=newvalue  > J >                   { command to modify a given user's data. "username" isI >                   { required. At least one /field=newvalue is required.  > } H >                   { "field" is a field in the data file, "newvalue" is > } 3 >                   { the new value to set it at. }  > 6 > /MODIFY, /LIST, and /CREATE are mutually exlclusive.' > /CREATE prohibits /HTML, and /OUTPUT.  > 0 > Here's the .CLD file I've come up with so far. >  > DEFINE VERB DBACCESSD >         IMAGE "USER01:[MGR.LBARTELS.PROGRAMMING.DATABASE]DBACCESS"   Comment-out the IMAGE line...    >  >         NOPARAMETERSA >         QUALIFIER FILE, NONNEGATABLE,VALUE(REQUIRED,TYPE=$FILE) C >         QUALIFIER OUTPUT, NONNEGATABLE,VALUE(REQUIRED,TYPE=$FILE) ( >         QUALIFIER CREATE, NONNEGATABLE& >         QUALIFIER HTML, NONNEGATABLE= >         QUALIFIER MODIFY, NONNEGATABLE,SYNTAX=MODIFY_SYNTAX A >         QUALIFIER LIST, NONNEGATABLE,VALUE(TYPE=$QUOTED_STRING) - >         DISALLOW ANY2(CREATE, MODIFY, LIST) $ >         DISALLOW CREATE AND OUTPUT" >         DISALLOW CREATE AND HTML >  > DEFINE SYNTAX MODIFY_SYNTAX A >         QUALIFIER FILE, NONNEGATABLE,VALUE(REQUIRED,TYPE=$FILE)  >         QUALIFIER MODIFY, 2 > NONNEGATABLE,VALUE(REQUIRED,TYPE=$QUOTED_STRING) >         QUALIFIER USERNAME, 2 > NONNEGATABLE,VALUE(REQUIRED,TYPE=$QUOTED_STRING) >         QUALIFIER PASSWORD, 2 > NONNEGATABLE,VALUE(REQUIRED,TYPE=$QUOTED_STRING) >         QUALIFIER EMAIL,2 > NONNEGATABLE,VALUE(REQUIRED,TYPE=$QUOTED_STRING)J >         QUALIFIER PRIV, NONNEGATABLE,VALUE(REQUIRED,LIST,TYPE=PRIV_VALS)G >         QUALIFIER EXPIRE, NONNEGATABLE,VALUE(REQUIRED,TYPE=$DATETIME)  >  > DEFINE TYPE PRIV_VALS " >         KEYWORD ADMIN, NEGATABLE! >         KEYWORD BASE, NEGATABLE # >         KEYWORD CREATE, NEGATABLE $ >         KEYWORD DISABLE, NEGATABLE > % > The datafile is structured as such:  > # > username:       char[some-length] # > password:       char[some-length] # > email:          char[some-length]  > privileges:     bitmap > expire date:    datetime >  > How does it look?  > & > When I wish to create an obj module: > # > $ SET COMMAND/OBJECT DBACCESS.CLD  >  > I get the message: > I > %CDU-E-ROUTREQ, Line 2: A ROUTINE clause is required when generating an 
 > object file   @     VAXMAN answered this, but missed the point that you actually; _don't_ need a MODULE statement.  You do need to REMOVE, or ; comment-out, the IMAGE statement.  That's what's giving the ! conflict with SET COMMAND/OBJECT.   ?     If you omit a MODULE statement, the name of the module will > be the name of the .CLD file, e.g., DBACCESS.  That may or may5 not be desirable.  It's common to use something like,        MODULE DBACCESS_CLD      VERB DBACCESS      ...   = Then if you insert this in an object library, the Module name > will be DBACCESS_CLD (and will coexist with a subroutine named DBACCESS, for example).     D > I don't understand that. I've been reading the manual, and I don't
 > understand. - > It talks about a routine with CLI$DISPATCH:  >  > Quote: >  > ROUTINE routine-name > E >      Symbol that specifies a routine the command calls to create an  > object* > module from the command definition file. > G >      The routine-name provides the name of a routine that is executed  > when > CLI$DISPATCH is called.  > ? >      If you do not specify a routine, no default is provided.  > 
 > Unquote: > H > What routine? A function within my program? I just don't get this one.  B     As others have said, the CDU facility is structured to _allow_B calling CLI$DISPATCH, e.g., id your different syntaxes are handledE by different routines.  If you're like me, and you're going to handle A all the command input in one routine, you don't need CLI$DISPATCH @ nor any Routine clauses.  Now I used to have some nice examples,> but they're all on a TK87 at home...  Let me cook an _inexact_; example (in Fortran, OK? No wise cracks!).  Something like,        PROGRAM DBACCESS     ...      EXTERNAL LIB$GET_INPUTD     EXTERNAL DBACCESS_CLD    ! Automatically loads the compiled .CLD/                              ! from the library      CHARACTER*256 CMD   1     Call Lib$Get_Foreign(CMD, 'DBACCESS> ', LCMD) >     Call Cli$Dcl_Parse('DBACCESS '//CMD(1:lCMD), DBACCESS_CLD,B    >                   Lib$Get_Input, Lib$Get_Input, 'DBACCESS> ')     Call DBA_PARSE     ...   L Of course, I haven't done any checking of return status from Lib$Get_Foreign  J or Cli$Dcl_Parse, and I haven't checked whether the user actually supplied. and input into CMD, but this is just a sketch.  E     Note that you need to _prepend_ the verb string, DBACCESS, to the F beginning of the string you send to Cli$Dcl_Parse.  Once Cli$Dcl_ParseE is done, you call the routine you've written to retrieve command line F parameters and qualifiers, I called it DBA_PARSE here, and within that8 routine you make calls to Cli$Present and Cli$Get_Value.  C     When written as above, the image operates "as expected" whether E invoked by a verb in DCLTABLES (with a .CLD that _includes_ the IMAGE J clause), or via a foreign command ($DBA*ACCESS=="$dev:[dir]DBACCESS.EXE"),F or by $ MCR dev:[dir]DBACCESS, or via DCL$PATH:.  That makes gives theD system manager who will be installing this utility flexibility as to= how s/he wants to deploy it.  (It's also useful for testing.)   C     Caveat emptor: it's been more than a couple of years since I've D                    done any CLD/CLI programming, so there may errors&                    in certain details.       -Ken --6 I don't speak for Intel, Intel doesn't speak for me...  
 Ken Fairfield ! D1C Automation VMS System Support  kenneth.h.fairfield#intel.com    ------------------------------    Date: 21 Aug 2002 15:50:00 -0700, From: colive@technologEase.com (Chris Olive) Subject: Re: .CLD questions.= Message-ID: <b10654c6.0208211450.324c9417@posting.google.com>   c lewis@spyder.mitre.org (Keith A. Lewis) wrote in message news:<ak02ig$i8d$1@newslocal.mitre.org>...  > Lyndon Bartels <lbartels@pressenter.com> writes in article <3D632EB4.4320B097@pressenter.com> dated Wed, 21 Aug 2002 06:09:57 -0500: > >DEFINE VERB DBACCESS E > >        IMAGE "USER01:[MGR.LBARTELS.PROGRAMMING.DATABASE]DBACCESS"  >   $ > >$ SET COMMAND/OBJECT DBACCESS.CLD > >  > >I get the message:  > > J > >%CDU-E-ROUTREQ, Line 2: A ROUTINE clause is required when generating an > >object file > > E > >I don't understand that. I've been reading the manual, and I don't  > >understand.. > >It talks about a routine with CLI$DISPATCH: > I > There are 2 ways to use CLI, and you're using elements of both of them.   D These are 2 *popular* ways to use CLI...  There is another I use allF the time, that I prefer, and which the original poster may prefer.  It@ closely parallels the VAXman explanation (which is always eitherC textbook perfect or, if not, isn't in ANY "textbook"!)...  Read on.   L > 1.  Creating a command at the DCL level, you use "SET COMMAND" without theF > /OBJECT qualifier.  Then inside your program (specified in the IMAGEM > clause), you can use CLI$PRESENT, CLI$GET_VALUE, etc. to read the parameter 	 > values.  > E > 2.  Creating a command to use inside your application, you use "SET L > COMMAND/OBJECT" and link the .OBJ file in with the program.  Then when youM > call CLI$DISPATCH, it "calls" your CLI which in turn calls whatever ROUTINE  > you specify.  E 3. [To original poster]: Use the MODULE clause as VAXman said.  Leave D the IMAGE clause out; you are going to link the ensuing image objectF with the CLI object created from SET COMMAND/OBJECT, so you don't need= the IMAGE clause.  Inside your CLD, set your MODULE clause as B 'DBACCESS'. Inside your program assign an external variable to theB name of your CLD module.  When you run the program, you either getD your DCL foreign command line (via LIB$GET_FOREIGN) or other commandD line (via LIB$GET_INPUT as VAXman said), and append it to the stringF 'DBACCESS '.  (I think the space is necessary, but I can't remember --F I always have it.)  Then call CLI$DCL_PARSE using an external variable? in your image that refers to the module name you chose, and the < appended DCL string.  After that you don't even need to callC CLI$DISPATCH if you don't want to.  I never do.  All your other CLI F calls work (CLI$PRESENT, CLI$GET_VALUE, etc) as if the command were inD the command tables.  You must set your image up as a foreign commandF in order to use this method.  (DBACCESS :== $ddcu:[directory]DBACCESS)  @ This is my preferred method for using CLI/CLD because it doesn'tD require someone (esp. on utilities I post for public consumption) to@ install a command into the command tables (either system-wide orB process specific.)  This method, IMO, is simpler than all the rest> (and much simpler than it sounds) and straight forward with no! CLI$DISPATCHing necessary either.   F I use CLI$DISPATCH only if I am getting input via LIB$GET_INPUT from a, "utility" command prompt.  (Ie. DBACCESS> ).  C I can provide you with an example in your language of choice if you A want...  (Please don't pick COBOL! 8(  And I don't have access to F BASIC or BLISS.  I think I have most everything else available to me.)   Chris  -----  Chris Olive  colive(at)technologEase(dot)com    ------------------------------  % Date: Thu, 22 Aug 2002 14:02:12 +1000 * From: James Cameron <james.cameron@hp.com>, Subject: Re: attn: Paddy (re: DCL procedure)9 Message-ID: <pan.2002.08.22.14.01.23.414878.24834@hp.com>   5 On Thu, 22 Aug 2002 01:12:12 +1000, Alan Greig wrote: D > If you setup IMAP on Exchage it will be readable by Mozilla's IMAPJ > client. Other alternative is to setup Exchange to do local SMTP relay to > the VMS box.   I second that.  @ As far as I've experienced, the following options are available;  G - require IMAP or POP service to be enabled on the Exchange server, and 6 use an IMAP or POP client on OpenVMS, such as Mozilla,  G - as above, but use fetchmail to read the messages and deliver directly J via localhost connection to the SMTP server on the OpenVMS box (untested),  J - set the Exchange identity to forward to SMTP address for the VMS box, ifE one is available in the local domain name service ... something Paddy  might not have.   E Some of our customers find it essential to be able to receive mail on C their OpenVMS box ... because it lets us e-mail them ECOs, reducing D double handling and risk of file transfer mistakes.  That might be a' justification for enabling IMAP or POP.   G (264 matches on google for "fetchmail openvms", so there must be a port  in there somewhere!)   --
 James Cameron    ------------------------------    Date: 21 Aug 2002 21:58:45 -0700, From: iskandar@measat.com (Iskandar Hussein)7 Subject: Re: Cheapest way to repair/replace tape drives < Message-ID: <28a7709.0208212058.3c893737@posting.google.com>  j tadamsmar@yahoo.com (Tom Adams) wrote in message news:<ea44f5a1.0208210709.58bf862d@posting.google.com>...; > Four of our Alphas have external SCSI tape drives, TLZ06.  > < > I have been paying $325 to UNIDATA for a fixed rate repair > or replace service.  > $ > Any suggestions for a cheaper way?  F I got a good deal with Data Storage Distribution in West Covina,CA...IF bought 4 unit of TLZ06-AA for 299 each...you can check the web site at( http://www.datastoragedistribution.com/.   regards    Iskandar iskandar@measat.com    ------------------------------   Date: 21 Aug 2002 18:29:22 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)- Subject: Re: DE101 - can it go to Full Duplex * Message-ID: <ak0m62$iu9$3@web1.cup.hp.com>  q In article <6e2f14f4.0208200836.22267f6f@posting.google.com>, glenmark@utxvms.cc.utexas.edu (Glen Martin) writes: F :I have a couple of old AlphaServer 2100a systems running OpenVMS v6.2E :which appear to have DE101 NICs (at least that is what is stamped on @ :the outside of the cards). CLUE CONFIG lists them only as TulipE :adapters. I've been trying to set them to Full-Duplex mode, both via E :LANCP and via the EWA0_MODE environment variable within SRM, but the E :cards aren't going to full (the switch to which they connect are set E :to Full Duplex). Am I fighting an uphill battle? Do these cards even E :support Full Duplex? Thus far, I've been unable to find this info atA :Compaq's website....s  A   The DE101 is not capable of operating at Fast Ethernet speeds.  @   Without intending offense, why would you even bother trying toA   use a decade-old EISA/ISA NIC for anything other than as a nice ?   electronic-themed wall decoration or as a historical exhibit?o  B   DE500 series or DE600 series are required for Fast Ethernet, and@   specific members of the DE500 series NICs are supported on the   AlphaServer 2100A series.a  A   Per the hardware configuration support search engine (availableeB   at the AlphaServer website; http://www.compaq.com/alphaserver/),C   the DE101 is not supported on this platform; members of the DE400wC   series are the oldest supported Ethernet NICs.  (I'd be surprisedcA   to hear that these DE101 NICs were ever officially supported onw?   this platform, as these NICs even very were old back when then0   AlphaServer 2100A series was itself released.)    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Wed, 21 Aug 2002 22:01:21 -0400 2 From: rdeininger@mindspring.com (Robert Deininger)- Subject: Re: DE101 - can it go to Full DuplexwK Message-ID: <rdeininger-2108022201220001@1cust211.tnt2.nashua.nh.da.uu.net>e  < In article <6e2f14f4.0208210916.96257db@posting.google.com>,2 glenmark@utxvms.cc.utexas.edu (Glen Martin) wrote:  , >mustang@ucc.asn.au.invalid wrote in message& news:<ajvcvm$jg2$1@enyo.uwa.edu.au>... >> I was curious, so I checked:i >>   >> These are an 8-bit ISA card. 2 >> I don't think you'll do 10baseT FDX with them.. >>  0 >> Why are you subjecting your 2100s to them? :P >pC >NICs that came with the boxen. Would love to some DE500 cards into E >them, but it is hard to convince management to spend money upgrading A >these systems, especially since they will likely be taken out ofo >service next year...e  H Are you still paying hardware support on these cards?  If not, you mightJ as well upgrade with some (cheap) used stuff.  DE500s are cheap enough youI can afford a few spares; if you can't get them on a maintenance contract, : you'll just have to swap them yourself if they ever break.  F Look up the cards that are supported on the platform.  Check with yourG favorite used equipement dealer, or watch ebay for a few weeks.  If youaH could get 4 or 5 used NICs, with non-DOA guarantee, for $100, would your management let you buy?f   ------------------------------   Date: 21 Aug 2002 17:30:01 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman): Subject: Re: Determining what .h file a function is in...?* Message-ID: <ak0imp$iu9$1@web1.cup.hp.com>  _ In article <3D6154CC.608B137A@pressenter.com>, Lyndon Bartels <lbartels@pressenter.com> writes:dI :I've been looking at some of the sys$examples:*.c files, and was lookinge :at the database examples. :9D :I tried to compile them, but notice some of the sys$ functions were :declaredimplicitely..  H   The sys$examples:*.c stuff is old and at least some of these are knownI   to have build problems.  I've been meaning to re-write various of these-I   examples over time, but my real jobs just keep intruding on the effort.mG   I've already spent time working on the C examples used in the OpenVMS H   manuals, and various new and various updated examples are available in    recent documentation releases.  * :But this leads me to a bigger question... : G :Given a predefined function... How does a person go about figuring outtF :which include file to.... well.... include.. so that that function is
 :"available."T  @   starlet.h for the system and RMS services, or lib$routines.h, C   str$routines.h and such for the OpenVMS RTL calls, and xyzzydef.h A   for the xyzzy$ symbols.  There are a few common hold-overs fromdB   VAX C, such as descrip.h for the descriptors, in addition to the)   related definitions over in dscdef.h.     A   The generic solution is to search sys$sysroot:[decc$lib...]*.h,    assuming the HP C compiler.d  B   I expect this stuff is somewhere in the C manuals.  Also see the>   OpenVMS programming concepts manual and associated examples.  ;   Also please see the C programming information in the FAQ.   N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Wed, 21 Aug 2002 11:54:10 -0700 # From: "Tom Linden" <tom@kednos.com> : Subject: RE: Determining what .h file a function is in...?9 Message-ID: <CIEJLCMNHNNDLLOOGNJIGEIGFJAA.tom@kednos.com>E  ( or use emacs, it knows how to find them.   >-----Original Message-----L: >From: Hoff Hoffman [mailto:hoffman@xdelta.zko.dec.nospam]* >Sent: Wednesday, August 21, 2002 10:30 AM >To: Info-VAX@Mvb.Saic.Com; >Subject: Re: Determining what .h file a function is in...?  >N >T >E? >In article <3D6154CC.608B137A@pressenter.com>, Lyndon Bartels E" ><lbartels@pressenter.com> writes:J >:I've been looking at some of the sys$examples:*.c files, and was looking >:at the database examples.  >:E >:I tried to compile them, but notice some of the sys$ functions were) >:declaredimplicitely..V >OI >  The sys$examples:*.c stuff is old and at least some of these are known J >  to have build problems.  I've been meaning to re-write various of theseJ >  examples over time, but my real jobs just keep intruding on the effort.H >  I've already spent time working on the C examples used in the OpenVMSI >  manuals, and various new and various updated examples are available ine! >  recent documentation releases.- >g+ >:But this leads me to a bigger question...  >:H >:Given a predefined function... How does a person go about figuring outG >:which include file to.... well.... include.. so that that function isA >:"available." > A >  starlet.h for the system and RMS services, or lib$routines.h,  D >  str$routines.h and such for the OpenVMS RTL calls, and xyzzydef.hB >  for the xyzzy$ symbols.  There are a few common hold-overs fromC >  VAX C, such as descrip.h for the descriptors, in addition to thet* >  related definitions over in dscdef.h.   > B >  The generic solution is to search sys$sysroot:[decc$lib...]*.h, >  assuming the HP C compiler. >eC >  I expect this stuff is somewhere in the C manuals.  Also see the ? >  OpenVMS programming concepts manual and associated examples.E >D< >  Also please see the C programming information in the FAQ. >y2 > ---------------------------- #include <rtfaq.h>  >-----------------------------5 >      For additional, please see the OpenVMS FAQ -- . >www.openvms.compaq.com    ,4 > --------------------------- pure personal opinion  >---------------------------3 >   Hoff (Stephen) Hoffman   OpenVMS Engineering     >hoffman#xdelta.zko.dec.comc >a >---' >Incoming mail is certified Virus Free. ; >Checked by AVG anti-virus system (http://www.grisoft.com).v@ >Version: 6.0.381 / Virus Database: 214 - Release Date: 8/2/2002 >i ---l& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).? Version: 6.0.381 / Virus Database: 214 - Release Date: 8/2/2002n   ------------------------------  # Date: Wed, 21 Aug 2002 19:46:14 GMT * From: "Bill Todd" <billtodd@metrocast.net>0 Subject: Re: Fortune Magazine and a post-VMS rapB Message-ID: <a6S89.118987$2p2.5345869@bin4.nnrp.aus1.giganews.com>  5 "David Froble" <davef@tsoft-inc.com> wrote in messaget& news:3D633D0C.6060802@tsoft-inc.com... > Bill Todd wrote: >r9 > > "David Froble" <davef@tsoft-inc.com> wrote in messagel* > > news:3D62F171.2010802@tsoft-inc.com... > >iC > >>>"Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message  > >>>NK > >>>>I've come to the conclusion that it would take a new OS platform, one L > >>>>designed for modern technology but compatible with the core VMS valuesL > >>>>of careful design effort up-front, engineering excellence, intoleranceH > >>>>of any hint of things like data corruption or security weaknesses,C > >>>>extreme reliability, excellent documentation, seamless upward ; > >>>>compatibility, and so forth, to actually replace VMS.D > >>>>, > >>How would such a system differ from VMS? > >> > >d3 > > a)  It would jettison 25 years' worth of cruft.    ...t  . > Cruft could be worked out of an existing OS.  G Not if you wanted to remain backward-compatible.  That's the differencelJ between Keith's comment about 'compatible with the core VMS values' and an evolution of VMS itself.   ...c  K > It would be much easier to replace such, as required, rather than have to L > replace everything at once.  If you're willing to follow a solid design, aK > modular design, and maybe have to sacrifice backward compatibility to get  rid of) > cruft, it still could be done with VMS.E  E If it's not backward-compatible, it's not VMS.  That's not to say one$J couldn't start with a lot of the VMS code, assuming it were made available for use.   >y > ) > > c) It would be owned by someone else.o >e >   > Jury is still out on that one.  J No, it's not.  The 'someone else' could conceivably still be a corporation+ named 'HP', but not run by the same people.    - bill   ------------------------------  % Date: Wed, 21 Aug 2002 16:55:39 -0400l- From: JF Mezei <jfmezei.spamnot@videotron.ca> 0 Subject: Re: Fortune Magazine and a post-VMS rap, Message-ID: <3D63FE39.4E2801CD@videotron.ca>   Bill Todd wrote:0 > > Cruft could be worked out of an existing OS. > I > Not if you wanted to remain backward-compatible.  That's the difference4L > between Keith's comment about 'compatible with the core VMS values' and an > evolution of VMS itself.  D Does VMS really contain such significant "cruft" because it is old ?  I I seem to recall that much of the so called "cruft" is really a strategic9M decision not to embark on fancy tricks that enhance speed because they do notON garantee data integrity. VMS is VMS because it has the quality, data integrityM and is trustable, not because it is fast. The fact that it ran on the world'sqN fastest chip did compensate to some extent its performance weaknesses, so whenJ it moves to IA64, its performance may look bad versus HP's other products.  N On the other hand, VMS's old cruft does include many nice tricks that other OSL don't have which do make it posisble to have high performance stuff. (global) sections, installed images for instance).,  K OK, perhaps there are drivers that are no longer needed. Paper tape driverspL could be ditched :-) And I guess that would take a fait amount of management effort to decide ton? 	1- allow some of the supported faclities/devices to be removedMJ 	2- go through the inventory of devices/facilities to device which can go.  ; But that wouldn't require a rewrite of the OS from scratch.,   ------------------------------  # Date: Wed, 21 Aug 2002 20:56:10 GMT * From: "Bill Todd" <billtodd@metrocast.net>0 Subject: Re: Fortune Magazine and a post-VMS rapB Message-ID: <K7T89.119449$2p2.5384316@bin4.nnrp.aus1.giganews.com>  5 "Bill Todd" <billtodd@metrocast.net> wrote in message < news:a6S89.118987$2p2.5345869@bin4.nnrp.aus1.giganews.com... >e7 > "David Froble" <davef@tsoft-inc.com> wrote in messageu( > news:3D633D0C.6060802@tsoft-inc.com... > > Bill Todd wrote:   ...s  + > > > c) It would be owned by someone else.r > >  > >o" > > Jury is still out on that one. >eL > No, it's not.  The 'someone else' could conceivably still be a corporation- > named 'HP', but not run by the same people.L  E Perhaps I was a bit terse above, so just in case early (and one hopesIH temporary) senility has dimmed your short-term memory let's review thoseC people to whom you seem at least potentially willing to entrust the 7 development of this new, high-quality operating system:   E They are the Compaq people who systematically starved their own eagledB (Alpha) before administering the coup de grace because they saw anL opportunity to get a good deal on a flying pig that was (more hopefully thanI factually) trailing a banner emblazoned with the Words of Power "Industrym Standard!".i  J They are the Compaq people who systematically starved VMS in precisely theH same manner (and now that they've got access to an 'enterprise' Unix andG still hope for an 'enterprise' Windows, the coup de grace may only be an matter of time).  H They are the HP people who treated PA-RISC just as Compaq treated Alpha,K *even after their own engineers had gotten a good enough look at the flyingiF pig to decide that perhaps PA might look like a better bet after all*.  G They are the people (both HP and Compaq) who subscribe to the view thata1 Windows will 'eviscerate' Unix in the enterprise.   E They are the Compaq people who cheerfully and consistently lied about.K Alpha's future before 6/25/01, then lied about both its technical potential G (and who had advised them about it) and its profit potential afterward.   H They are the HP people who cheerfully lied about their own deterioratingI internal projections for the merger during the months preceding its vote.c  J These are the people you would trust to commit to the considerable expenseH and effort of developing a high-quality, non-industry-standard operatingL system that would compete with their industry-standard vision of the future?- I wouldn't trust them to take out my garbage.    - bill   ------------------------------  # Date: Wed, 21 Aug 2002 21:03:10 GMT * From: "Bill Todd" <billtodd@metrocast.net>0 Subject: Re: Fortune Magazine and a post-VMS rapB Message-ID: <ieT89.119518$2p2.5387332@bin4.nnrp.aus1.giganews.com>  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3D63FE39.4E2801CD@videotron.ca... > Bill Todd wrote:2 > > > Cruft could be worked out of an existing OS. > >eK > > Not if you wanted to remain backward-compatible.  That's the differencebK > > between Keith's comment about 'compatible with the core VMS values' and  an > > evolution of VMS itself. >tF > Does VMS really contain such significant "cruft" because it is old ?  I Yes.  A lot of it results from considering 1 MB to be a typical amount ofkD memory and a few hundred MB to be a typical amount of disk.  More, IG suspect, results from considering a few VUPs to be a typical processor.e   >sK > I seem to recall that much of the so called "cruft" is really a strategic-K > decision not to embark on fancy tricks that enhance speed because they doI nota > garantee data integrity.  H You recall incorrectly.  The combination of significant additional speed5 with uncompromised integrity is eminently achievable.A   ...r  G > On the other hand, VMS's old cruft does include many nice tricks that  other OSF > don't have which do make it posisble to have high performance stuff. (globalo+ > sections, installed images for instance).e  L So what?  No one was suggesting deliberately crippling the new system in theK few areas where other systems may not offer equivalent facilities to VMS's.    - bill   ------------------------------  % Date: Wed, 21 Aug 2002 21:40:27 -0400 2 From: rdeininger@mindspring.com (Robert Deininger)0 Subject: Re: Fortune Magazine and a post-VMS rapK Message-ID: <rdeininger-2108022140270001@1cust211.tnt2.nashua.nh.da.uu.net>1  5 In article <3D63FE39.4E2801CD@videotron.ca>, JF Mezei@% <jfmezei.spamnot@videotron.ca> wrote:4   >Bill Todd wrote:p1 >> > Cruft could be worked out of an existing OS.G >> aJ >> Not if you wanted to remain backward-compatible.  That's the differenceM >> between Keith's comment about 'compatible with the core VMS values' and ano >> evolution of VMS itself.) >LE >Does VMS really contain such significant "cruft" because it is old ?t > J >I seem to recall that much of the so called "cruft" is really a strategicN >decision not to embark on fancy tricks that enhance speed because they do notO >garantee data integrity. VMS is VMS because it has the quality, data integrityrN >and is trustable, not because it is fast. The fact that it ran on the world'sO >fastest chip did compensate to some extent its performance weaknesses, so whennK >it moves to IA64, its performance may look bad versus HP's other products.o  K An illustration of the sort of performance problems that "cruft" can cause:n  B It is a goal of VMS to allow programs built on earlier versions to8 continue working on later versions, without recompiling.  G There are a number of places in VMS where user-written code shares dataeC structures with the OS.  These structures are typically defined vian compile-time symbols.t  J The old structures were designed to be compact; some date back to the daysG of the 256 kB VAX 11/785.  Items that would be quadwords (or larger) ifdF defined today were defined as bytes or words 25 years ago.  This oftenH results in items in the structures being misaligned on an alpha system. & There can be a big performance impact.  C In some cases, new structures can be implemented while the old ones-C continue in  exist for old programs.  Some structures have internal H versioning, so the OS can adapt to various versions automagically.  ThenJ just recompiling with the latest symbols definitions brings a user programI up to date.  In other cases, the user's source code would have to change,mJ setting a new bit or switch or whatever.  Either way, the OS likely has toG deal with interactions between both the old and the new simultaneously,-; since not every user image will be rebuilt on the same day.1  J If you could force a complete rebuild of all VMS code, you could "fix" all@ the old data structures.  But that isn't possible within the VMS
 philosphy.  D Having said all that, I think there are MANY places where VMS can be3 improved a lot while still retaining compatibility.    ------------------------------  # Date: Thu, 22 Aug 2002 01:40:45 GMTT1 From: "David J. Dachtera" <djesys.nospam@fsi.net>o% Subject: Re: Fortune Magazine wrapperm' Message-ID: <3D644618.D987B7E3@fsi.net>0   John Smith wrote:b > > > "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message# > news:3D6306E7.B4497DE1@fsi.net...l > > John Smith wrote:o > > >i6 > > > Copyright 2002, John Smith. ALL RIGHTS RESERVED. > >tF > > If "John Smith" is not your real name, your claim to copyright may > > unenforceable. > J > Well, there's an OT comment. Open a phone book some time and turn to theK > pages that start with the letters 'smith'. You may find that there are atk+ > least several of us in nearly every city.o   "John Smith" <a@nonymous.com>e  F Hardly lends credence to one's "credentials" (notice the similarity of4 the words? Think they may share a common heritage?).  N > HP is creatively bankrupt when it comes to VMS advertising ( there's more toI > it than just that as we all know ), and all I want to do if preserve myjF > ability to relieve them of some cash when they need to use the idea.  G I'd say "opportunity" rather than "ability". Given the way they seem to,= work, although they consitently throw good money after bad onlE consultants, studies, etc. when common sense would do just as well at D vastly lower cost, you may very well get a call if/when someone with8 some moxy on the ball finally gets a shot at pusing VMS.  ; > Now that's taken care of, what did you think of the 'ad'?n   I like it. Good print ad.I  C I'm still scripting TV ads, myself. I'm up to 18 now. I'm gonna tryuB pulling some strings over at local access and see if I can do someD proof-of-concept videos for each of them and maybe show them at (the successor to) HPETS 2003.u   -- e David J. Dachteral dba DJE Systemsc http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/.   ------------------------------  # Date: Thu, 22 Aug 2002 02:16:47 GMTl1 From: "David J. Dachtera" <djesys.nospam@fsi.net>l% Subject: Re: Fortune Magazine wrapperR' Message-ID: <3D644E89.45D0AC4B@fsi.net>o   Paul Repacholi wrote:  > [snip]F > My mail just had a notice for tommorows webcast on T64 and HPUX turn7 > up from DECUS. The *VERY FIRST* line of the overview:d > D > "No operating system is more critical to enterprise customers thanG > VMS, with its ability to deliver robust, mission-critical solutions."   H The second half of that is missing. HP acknowledges the customers' need,G but fails to reflect it. One possible extension of that might be had HPtG said, "Likewise, no operating system is more critical HP's relationshipQD with its enterprise customers than VMS. Thus, no operating system is0 more critical to the future of HP than OpenVMS."  E Without actually using the words, "commitment" and/or "raodmap", this-H would state, albeit ambiguously, an implication of commitment that wouldD be difficult for all but the most skeptical to ignore. (Granted, the2 Gatesian coningent might look at it cross-eyed...)  C Hmmm... Wonder if I could get a job as a junior "word-smith" in theg) OpenVMS marketing org. Sue? Any thoughts?-   -- E David J. DachteraI dba DJE Systemsc http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Thu, 22 Aug 2002 00:16:08 -04000- From: JF Mezei <jfmezei.spamnot@videotron.ca>e% Subject: Re: Fortune Magazine wrappers, Message-ID: <3D646586.E3F2C170@videotron.ca>   "David J. Dachtera" wrote:I > but fails to reflect it. One possible extension of that might be had HP I > said, "Likewise, no operating system is more critical HP's relationship F > with its enterprise customers than VMS. Thus, no operating system is2 > more critical to the future of HP than OpenVMS."   Why not something as simple as:o  M <carly> We *now* recognise the superior technical strengths of VMS as well as L the importance VMS plays at our customer sites and think that VMS gives HP aL key technological differentiator that will move HP ahead of its competitors.K We therefore intend to grow the VMS marketplace which has been neglected inl< the past. VMS is good for our customers, VMS is good for HP. </carly>  M When you see great potential on a product, there is no need to put a timeline M to it by making public commitments to support it for X years etc etc. This is N why you don't see commitments for Windows, HP-UX, SUN Solaris etc. CommitmentsM are only necessary when a product is seen as having a finite lifetime and you F want to ensure that lifetime is long enough to keep current customers.   ------------------------------   Date: 21 Aug 2002 17:50:17 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)Y Subject: Re: forwarding email from system to system (was: RE: attn: Paddy (re: DCL procedP* Message-ID: <ak0jsp$iu9$2@web1.cup.hp.com>  E An unattributed author or a random confluence within the ether asked:tE : If anyone can tell me how to get Exchange to deliver mail to me VMSmJ : account and still appear to be from the person who actually sent it, I'd : be grateful. -  I   You will want to discuss this with the folks maintaining your Microsoft.F   Exchange servers.  This automatic delivery to a foreign SMTP host isH   most definitely possible, and Exchange calls the underlying mechanismsG   a "custom recipient".  Microsoft Exchange will also require the "SMTP 2   connector" be present and configured, of course.  I   I am not in a position to support the configuration of nor the internal G   operation of Exchange, and would encourage continuing this discussionFI   in a newsgroup specializing in Microsoft Exchange.  The identity of the6H   remote system -- OpenVMS, in this case -- is not relevent to the emailG   delivery question.  The use of the SMTP transport and the need for an A   Exchange "custom recipient" entry is, of course, relevent here.V  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------    Date: 21 Aug 2002 16:59:01 -07001 From: keithparris_NOSPAM@yahoo.com (Keith Parris)nY Subject: Re: Geographically Dispersed Parallel Sysplex and replication (was Re: HP-Compaq = Message-ID: <cf15391e.0208211559.290b6a01@posting.google.com>c  t "Bill Todd" <billtodd@metrocast.net> wrote in message news:<SBC89.102848$2p2.4809800@bin4.nnrp.aus1.giganews.com>...C > At noted above, I suspect that the synchronous GDPS approach alsodC > effectively has zero data loss:  if not, I'd be interested in thes$ > explanation of how it can lose it.  F I'd be interested as well.  There must be some real potential for dataC loss, or they wouldn't have to qualify it as "minimal to none" in aaB public document.  And since they defined "minimal" as "less than 2D minutes' worth" of data loss elsewhere in the same document, one has
 to wonder.  M > asynchronous replication, which IIRC VMS does not offer:  if it did, *then* K > I'd start to consider VMS as having a clear advantage, because memory and A > processor advances do keep improving software DLM performance).t  F Improvements in cluster interconnect performance (like Memory Channel,E Galaxy Shared Memory Cluster Interconnect, and Gigabit Ethernet) haveaE helped DLM performance greatly as well.  Technologies like InfiniBand.E could help in the future.  And VMS Engineering keeps cranking out newiF DLM features like the speedier lock remastering in 7.2-2 and above andE the dedicated-CPU lock manager option for reduced spinlock contentionpH in SMP systems (see http://vmsone.com/dfwdays/Presentations/gordon.ppt).  ? And VMS clusters can indeed use asynchronous replication, usinguF StorageWorks DRM, or EMC SRDF or Hitachi (aka HP SureStore) ContinuousC Access XP.  Also, ASCI offers a host-based shadowing product calledu- Shadow which can do asynchronous replication.o  F But most VMS users prefer the strict synchronous replication that HBVS@ provides, where success status for an I/O is not returned to theD application until the data safely lands on both the local and remoteD storage subsystems.  Even some hardware solutions billed as offeringD "synchronous" replication may actually be at risk of losing at leastC one outstanding I/O operation upon a site failure, because some mayiB return success status to an I/O operation before the data actually& lands on the remote storage subsystem.  A Also, all controller-based replication solutions today only allowpD access to a given "mirrorset" (I'll use the DRM term; other vendors'= terms vary) through whichever controller is "primary" for thesF mirrorset at a given time, and replication is one-way, from primary toE secondary copy.  And manually-initiated (manual, or at best scripted) E failover sequences are required, after which you can then only accessaD the mirrorset from the opposite controller, and replication is again( one-way, only in the opposite direction.  D Failover of controller-based mirrorsets is relatively time-consumingE and cumbersome compared with the automatic, essentially-instantaneousiB failover VMS cluster users are accustomed to with HBVS.  Even withC StorageWorks DRM, the minimum time they advise folks to plan for to,$ complete the failover is 15 minutes.  C And if, as is common in a VMS cluster, you want to run a portion ofoB the workload at each site, all the I/Os from systems at the remoteD site to a mirrorset which has its primary member at the primary siteE will have to be either MSCP-served or use remote Fibre Channel accessvF through the controller at the primary site.  With HBVS, systems at theE remote site can do all their read operations from the local shadowset @ member there; it is only writes which must suffer the inter-site latency.  B Also, the potential for data loss is inherent with any asychronousB replication scheme, and in much larger amounts than the single-I/OC risk loss some so-called "synchronous" solutions provide, because apC significant amount of data may still reside in cache on the primary F controller, waiting its turn to be sent across the inter-site link, atD the point in time when a disaster destroys both the primary disk and6 the copies of recent writes in its controller's cache.  A On VMS systems with HBVS, we're so used to the strict-synchronousnC shadowing of HBVS and the careful write ordering of the XQP and RMSu8 that we don't tend to worry about potential data loss orD write-ordering as being potential issues.  We also don't think aboutD things like the fact that no matter how accurate a copy of the localC disk a controller implementation can provide on the remote disk, ifaA the Unix system hasn't flushed the buffer cache to disk, the file - system metadata may be unreadable anyway. ;-)-  B Another problem with asynchronous replication is in terms of write> ordering.  Here, there are two issues.  One is with software'sC assumptions about write ordering and disks' behavior.  The other isoB associated with the issue of whether or not a consistent, readableB copy of the data is maintained at the remote site at all times, to9 protect against the loss of the copy at the primary site.u  8 Some implementations do replication by copying data on aC track-by-track basis, rather than preserving the order of writes.  tF But a database package often contains logic that assumes, for example,E that if it writes to the journal log first, and waits for that I/O toc? be reported as complete, and then writes to the actual databasetD storage area later, it can depend on that write ordering having beenC preserved by the storage subsystem when it needs to do things laterrF like recovery.  Some asynchronous replication products provide specialF solutions to this problem, such as providing write ordering guaranteesD to units at the remote site, even across whole sets of separate diskE units (because you always want the database journal log on a separate9, device than the database itself, of course).  F After a backup site outage or a disruption in the inter-site link, the@ data at the remote site is no longer up-to-date.  If the primaryF controller has recorded what tracks have changed, and copies data over; on a track-by-track basis, this effectively re-orders write @ operations, and the backup copy may not be usable until the copy: operation completes; a loss of the primary site during theD resynchronization would mean there may be no valid, readable copy ofF the data left.  StorageWorks DRM, when using the Mini-Merge capability@ (not to be confused with the HBVS term of the same name; this isB closer to the Mini-Copy functionality in HBVS, which brings a diskE more-rapidly up to date after it has been temporarily absent from theaO shadowset -- see http://vmsone.com/dfwdays/Presentations/MCWBM19April2001.PPT),d< updates only those blocks which have changed, minimizing theE resynchronization time and the inter-site bandwidth required, but notiC providing a consistent copy at the remote site until the Mini-MergeS
 completes.  C (VMS HBVS users should also understand that as soon as they start aaB full-copy or Mini-Copy operation to update the disks at the remoteE site, the remote disks may theoretically be worthless (let's say, fore? the sake of argument, the primary member has been backed up and B restored using BACKUP/IMAGE, and thus the data has been completelyE rearranged, since the time the shadowset was broken) unless and until C the copy operation completes, so it might be wise to save a copy ofrB the remote data, either by starting the copy to a different set ofE disks, or perhaps using a snapshot or clone operation before starting0A the copy operation, so that if they were to lose the primary site4A during the shadow copy, they'd at least have a consistent, albeitnD out-of-date, copy of the data at the remote site, instead of no copy of the data at all.)  A StorageWorks DRM has a creative solution to write ordering duringa@ resynchronization; it keeps track of write operations in a WriteC History Log (not to be confused with the Volume Shadowing Assist of D the same name in MSCP controllers).  Using this log, the remote copy@ of the data can be brought up-to-date while still preserving theB original write ordering, in essence simply speeding up time as the? data is updated, while always retaining a consistent, readable,c2 point-in-time copy of the data at the remote site.  F There will be some performance penalty in any asynchronous replication1 schemes which preserve write ordering, of course. . ----------------------------------------------. Keith Parris | parris at encompasserve dot org   ------------------------------  # Date: Thu, 22 Aug 2002 01:16:37 GMTe* From: "Bill Todd" <billtodd@metrocast.net>Y Subject: Re: Geographically Dispersed Parallel Sysplex and replication (was Re: HP-CompaqsA Message-ID: <VXW89.123974$SS.5457580@bin3.nnrp.aus1.giganews.com>-  > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message7 news:cf15391e.0208211559.290b6a01@posting.google.com...e7 > "Bill Todd" <billtodd@metrocast.net> wrote in message > news:<SBC89.102848$2p2.4809800@bin4.nnrp.aus1.giganews.com>...   ...e  H > > asynchronous replication, which IIRC VMS does not offer:  if it did, *then*I > > I'd start to consider VMS as having a clear advantage, because memorye andoC > > processor advances do keep improving software DLM performance).r > H > Improvements in cluster interconnect performance (like Memory Channel,G > Galaxy Shared Memory Cluster Interconnect, and Gigabit Ethernet) haverG > helped DLM performance greatly as well.  Technologies like InfiniBandnG > could help in the future.  And VMS Engineering keeps cranking out new H > DLM features like the speedier lock remastering in 7.2-2 and above andG > the dedicated-CPU lock manager option for reduced spinlock contentionfJ > in SMP systems (see http://vmsone.com/dfwdays/Presentations/gordon.ppt).  K There are aspects of the CF approach which are likely to remain faster thanaD the DLM approach will ever be (at least when executing on comparableB hardware) - though whether they constitute any kind of significant$ differentiator is subject to debate.   >0A > And VMS clusters can indeed use asynchronous replication, usinguH > StorageWorks DRM, or EMC SRDF or Hitachi (aka HP SureStore) ContinuousE > Access XP.  Also, ASCI offers a host-based shadowing product calledi/ > Shadow which can do asynchronous replication.f  L But (I think it's safe to assume) only under the same constraints that applyK to IBM's asynchronous facilities (in particular, potential data loss at theiI secondary copy plus no use of that copy by software at its location whileO the primary is active).e   >tH > But most VMS users prefer the strict synchronous replication that HBVSB > provides, where success status for an I/O is not returned to theF > application until the data safely lands on both the local and remote > storage subsystems.t  I Most VMS clusters don't operate at distances that would make that awkwarde. (and I suspect most Sysplexes may not either).  1   Even some hardware solutions billed as offering F > "synchronous" replication may actually be at risk of losing at leastE > one outstanding I/O operation upon a site failure, because some mayaD > return success status to an I/O operation before the data actually( > lands on the remote storage subsystem.  I If they do so before the data even arrives at the remote site, that's notrI synchronous.  If they require its arrival at the site but not at the diskbJ platter, that's not a problem if the two sites' failure modes *really* areB independent:  the probability of data loss is likely less than theA probability of undetected deterioration in existing on-disk data.n   >eC > Also, all controller-based replication solutions today only alloweF > access to a given "mirrorset" (I'll use the DRM term; other vendors'? > terms vary) through whichever controller is "primary" for thewH > mirrorset at a given time, and replication is one-way, from primary toG > secondary copy.  And manually-initiated (manual, or at best scripted) G > failover sequences are required, after which you can then only accessuF > the mirrorset from the opposite controller, and replication is again* > one-way, only in the opposite direction. >yF > Failover of controller-based mirrorsets is relatively time-consumingG > and cumbersome compared with the automatic, essentially-instantaneous D > failover VMS cluster users are accustomed to with HBVS.  Even withE > StorageWorks DRM, the minimum time they advise folks to plan for toh& > complete the failover is 15 minutes. > E > And if, as is common in a VMS cluster, you want to run a portion ofaD > the workload at each site, all the I/Os from systems at the remoteF > site to a mirrorset which has its primary member at the primary siteG > will have to be either MSCP-served or use remote Fibre Channel access H > through the controller at the primary site.  With HBVS, systems at theG > remote site can do all their read operations from the local shadowseteB > member there; it is only writes which must suffer the inter-site
 > latency.  H My admittedly limited understanding of 'normal' Sysplex systems suggestsL that none of the above limitations apply to them (because of their locality:E none of the disks is particularly 'remote', and load is shared across J multiple copies).  Hence those three paragraphs only add a bit more detailK to your initial observation that 'normal' (non-GDPS) Sysplex clusters can't 9 be distributed over as large an area as VMS clusters can.:   >uD > Also, the potential for data loss is inherent with any asychronousD > replication scheme, and in much larger amounts than the single-I/OE > risk loss some so-called "synchronous" solutions provide, because amE > significant amount of data may still reside in cache on the primaryhH > controller, waiting its turn to be sent across the inter-site link, atF > the point in time when a disaster destroys both the primary disk and8 > the copies of recent writes in its controller's cache.  3 I can't find any new information in that paragraph.t   >tC > On VMS systems with HBVS, we're so used to the strict-synchronousPE > shadowing of HBVS and the careful write ordering of the XQP and RMSl: > that we don't tend to worry about potential data loss or+ > write-ordering as being potential issues.   L And they aren't in Sysplexes, either.  Write-ordering should not be an issueI even in GDPS Sysplexes, though when asynchronous replication is used datasI loss can occur (just as it can when asynchronous replication is used withm VMS).s     We also don't think aboutdF > things like the fact that no matter how accurate a copy of the localE > disk a controller implementation can provide on the remote disk, if-C > the Unix system hasn't flushed the buffer cache to disk, the fileo/ > system metadata may be unreadable anyway. ;-)n  J Why you're introducing Unix into a discussion about Sysplexes (where AFAIKI no similar problem exists) is not clear to me, but it's worth noting thatnK there are quite a few file systems available on many/most Unixes that don't C suffer from the problem you describe (SGI's XFS, AIX's JFS, Linux's I Reiserfs, Sun's journaled UFS, UFS with Berkeley 'soft updates', Veritas'sK VxFS, Tru64's AdvFS, and Linux's ext3fs come immediately to mind, and I mayi well have missed some others).   >PD > Another problem with asynchronous replication is in terms of write@ > ordering.  Here, there are two issues.  One is with software'sE > assumptions about write ordering and disks' behavior.  The other is.D > associated with the issue of whether or not a consistent, readableD > copy of the data is maintained at the remote site at all times, to; > protect against the loss of the copy at the primary site.a >.: > Some implementations do replication by copying data on aC > track-by-track basis, rather than preserving the order of writes.fH > But a database package often contains logic that assumes, for example,G > that if it writes to the journal log first, and waits for that I/O touA > be reported as complete, and then writes to the actual databaselF > storage area later, it can depend on that write ordering having beenE > preserved by the storage subsystem when it needs to do things later H > like recovery.  Some asynchronous replication products provide specialH > solutions to this problem, such as providing write ordering guaranteesF > to units at the remote site, even across whole sets of separate diskG > units (because you always want the database journal log on a separateD. > device than the database itself, of course).  K It would surprise me if IBM were not at least as conversant with the issues8L you describe as you and I are, but if you have actual reason to believe theyJ screwed up that would be a very legitimate knock on Sysplex's asynchronous? replication facilities (though that would surprise me as well).i   ...'  C > StorageWorks DRM has a creative solution to write ordering duringoB > resynchronization; it keeps track of write operations in a WriteE > History Log (not to be confused with the Volume Shadowing Assist ofuF > the same name in MSCP controllers).  Using this log, the remote copyB > of the data can be brought up-to-date while still preserving theD > original write ordering, in essence simply speeding up time as theA > data is updated, while always retaining a consistent, readable,s4 > point-in-time copy of the data at the remote site.  D That's about as much as a disk-level replication mechanism can do (aI file/database-level mechanism can know which replicated writes can safelytC execute in parallel and where synchronization barriers must occur).    - bill   ------------------------------  % Date: Wed, 21 Aug 2002 21:37:59 -0400a- From: JF Mezei <jfmezei.spamnot@videotron.ca>PY Subject: Re: Geographically Dispersed Parallel Sysplex and replication (was Re: HP-Compaqo+ Message-ID: <3D644075.8F39152@videotron.ca>M   Keith Parris wrote:sD > failover VMS cluster users are accustomed to with HBVS.  Even withE > StorageWorks DRM, the minimum time they advise folks to plan for too& > complete the failover is 15 minutes.  N How come it wouldn't be automatic ? What would take 15 minutes to accomplish ?  I Also, are there storage array products that provide volume shadowing that , really works across buildings and are safe ?  L I am curous as to why VMS can have such a robust host based volume shadowingL while dedicated storage arrays wouldn't be able to provide the same level of service/trust ?n   ------------------------------  % Date: Wed, 21 Aug 2002 20:50:32 +0200 " From: "Hans Vlems" <hvlems@iae.nl>3 Subject: Re: How to configure Reflection X-server ?m6 Message-ID: <ak0ndu$1eb5mj$1@ID-143435.news.dfncis.de>  6 "Amit Sawhney" <amits_77@yahoo.com> schreef in bericht7 news:71939fd6.0208200050.26287378@posting.google.com...a > Dear VMS Friends,  >tH > I m trying to configure reflections-x on my W2kPro.I m able to see theG > reflection X root window but not able to see anything more than that.aC > Do i need to run any application on my VMS server before startinghH > Reflections-X ? or is there any problem in configuration of reflection
 > x suite. >s4 You must run TCP/IP between the VMS and W2K systems.K The VMS system must have DECwindows installed, the part that comes with VMSr as well as the DW-MOTIF kit.> Next copy the startapp.crx file that is part of ReflectionX to sys$manager:startapp.com4 That utility is needed to get the X session started.E An example of what the "Connection settings" box displays on my (WXP)f system:o     Method        TELNET Host name    chloorn User name    system  Password     ********  Command    @startapp %IP% term   Hans   ------------------------------  # Date: Wed, 21 Aug 2002 19:22:22 GMT 5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> + Subject: Re: HP-Compaq Merger Went SmoothlyW1 Message-ID: <OLR89.30$Z5.809012@news.cpqcorp.net>e  6 Andrew Harrison SUNUK Consultancy wrote in message ... >t >  >yA >Well apart from our CY2002/Q1 and Q2 numbers no we show no signse >of revenue growth.s >h> >Remind me how long has the joint Compaq and HP entity been in: >existantance. Come on Freddy boy havn't you FUDDED enough >r  I Hmmm.  Please point me to the quarterly reports I've seem to have missed,sG other than the meager EPS of 0.01 from last quarter?  The last quartersoJ are -.06, -.04, -.02 and then +.01.  The current advice to the street is aE EPS of -.01 for the current quarter, and "perhaps" a +.02 in the nextaI quarter (which may be wishful thinking, but it is also Suns traditionally  strongest quarter of the year).   F Anemic at best.  Certainly not strong growth, just fluctuating between& making a slim profit and losing money.  D Go to money.cnn.com, bring up HPQ and click on earnings estimates toK contrast it.  A *lot* more EPS both in the last year, and estimated for thei next year - than Sun.e  G But it is interesting how you've reverted to your childish name callinge though.y   >m= >> Some selected quotes from your own SEC quarterly report...a >>K >> "Over 90% of our products net revenue in the third quarter as well as ine thenK >> first nine months of both fiscal 2002 and 2001 was generated by ComputerTD >> Systems and Network Storage. Computer Systems and Network Storage consistsL >> primarily of servers, storage and desktop computers. Substantially all of >> theJ >> decrease in products net revenue during the third quarter and the first nineI >> months of fiscal 2002, when compared with the corresponding periods ofr >> fiscal 2001 >  >aD >Freddy you have a terrible terrible comprehension problem. Sun's isC >acheiving quarter on quarter revenue growth. We had however a veryrE >steep decline and we have not yet got to a point where revenues haver >recovered to pre crash levels.r >   J Falling back to quoting quarter-to-quarter growth is an old ploy, resortedL to by everyone when they are in trouble.  Big crash, and then anything (evenH losing money) is a quarter-to-quarter growth and you can point even to aF loss as growth.  Which is pretty much where you are.  You've gone fromH losing -.06 to making +.01, and are projected to slip back to losing .01F before perhaps returning to growth.  Ignore the fact that it is anemic! growth, and slim profits at best.i  E >How difficult is this for you to understand ?? Crash, big fall, slowi >recovery get it ??r >r  K How difficult is it for you.  For gosh sake, at least go check out your owntH companies reports, and try not to personally spin things to make it lookG like something it's not.  I wouldn't call one quarters return to a slim  profit "recovery" yet.  H >In the discussions we have had, you have lurched from benchmarks to CPUC >performance to systems architecture without betraying any apparentu= >understanding of the subjects you post on now you have added B >financial analysis to the list of things that you apparently knowD >nothing about as well, is this a clever front or has some malicious  >co-worker stolen your account ? >t  G Hmmm.  And now, having nothing to back up your FUD or your claims, letsdL change the subject and return your other creative writing assignments.  Like musical benchmarks.    >s >e3 >Well of course we do you are in our markets. Sadlyp5 >the whole tone of your arguments now and in the pasts0 >suggests that you don't take Sun seriously. Sad4 >for you not for us. You have worked for 2 companies5 >Digital and Compaq both of who suffered at the handso8 >of their competitors Sun and IBM as HP is now suffering" >at the hands of Sun IBM and Dell. >,  L Who says I don't take you (well, not *you* but Sun) seriously?  I think theyJ have done some absolutely stellar software in some specific areas.  I alsoH think they have done a great marketing job.  They employ someone or someL people who certainly know how to work the government standards game well.  II personally know and have worked with a half dozen engineers there - who IsL have *great* respect for.  I think it's a mistake not to take Sun seriously.J Nonetheless, Sun has a lot of weaknesses, and from having seen some of theL very same problems at the end of DEC (for example) - I think you should be aK little worried about now.  You are being, and will be even more so - picked J apart by competitors from the low end and high end.  The Solaris and SparcH stories are weak long term, and you haven't come up with the replacement6 strategy - cannabalize or be cannabalized as they say.  5 >If your comments reflect the culture of both defuncte6 >companies then it goes a long way to explain why they6 >are defunct, lets hope for HP's sake that this aspect8 >of the Digital/Compaq way does not become the prevelant >HP culture. >a  F This is, of course, always your last resort.  Start calling names, andL making personal attacks.  It distracts attention from the hollowness of your
 arguments.  K All I've done here is pointed you and anyone else to the information.  EasydL to get to.  Just try money.cnn.com, or quicken.com, or bloomberg.com, or anyI other one-click-away financial site.  The vast majority of what I've said5L here isn't my opinion - it's facts from these pages, or opinions - includingI Suns own opinions - from those sources.  Other conjectures are not unique.I from me - I'm just repeating them - and heck you can even discount them -0H like Sun being a possible takeover target at $4 and a market cap of $13bL (oh, earlier I made a mistake, the HP market cap is something like $44.5b orK over 3 times that of Sun.  It returns far more per share in earnings - likerG 20x more last quarter - and something else that Sun doesn't do - pays arH dividend).  If your stock drops much lower, it'll be really tempting for someone.   ------------------------------  # Date: Wed, 21 Aug 2002 19:31:15 GMTr5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>n+ Subject: Re: HP-Compaq Merger Went Smoothlye1 Message-ID: <7UR89.32$o6.822408@news.cpqcorp.net>h  / God I hate to do this...  Ditto.  What he said.   I The Internet is here to stay.  There are many valid uses for it, and evennF more in the future.  But the dot.bomb was the result of the irrationalJ spending that was occuring both from companies that never had a chance, asK well as good companies who got caught up in it thinking that if they didn'tdJ invest in it - they would be left behind.  Well.  As the smoke clears, youK find a lot of companies that have gone toes up, and a lot of companies thats& have cut their investments and losses.  L When will the Internet in general return to a *sustainable* growth, and what" will that level be?  I don't know.  B But feeding the frenzy were a lot of companies - like Sun, but notF exclusively Sun - who had started to wrap their entire business aroundK unbounded growth in the Internet - Sun is the dot in dot.com - yes?  Should I they then not share in being part of the bomb in dot.bomb.  The crash has L been pretty harsh on those companies, and it will probably take some time toG recover.  It's doubly tough when the entire industry is also being hit.s    = JF Mezei wrote in message <3D63C7EF.37BAF5C2@videotron.ca>...b) >Andrew Harrison SUNUK Consultancy wrote:e6 >> It always amuses me when people make the mistake of5 >> talking about dot.com as a past and failed entity.e >aL >Andrew, would you admit that most of the growth during the dot com era cameI >from companies which had no hope for survival and existed solely becausea> >people were stupid enough to pour oodles of money into them ? >nJ >And another aspect of this were the telecoms who also got oodles of money toL >deploy tons of fibre and routers etc etc, thinking that demand for bandwithL >would never cease to grow exponentially. (and this is what caused companiesC >such as Cisco, Nortel and JDS Uniphase to have their stock grow to  unrealistic L >orbital levels, and Nortel making the big mistake of buying up all sorts of >.COMs in the process).c > H >In the case of Nortel, remember that it was very much a voice equipmentL >company just before the .COM boom, and it was through acquisitions (such asF >bay networks) that all of a sudden, Nortel was able to sell equipment neededI >by the .COMs. Problem is that they went overboard in their purchases and- arefI >not paying dearly for it. JDS Uniphase was a little little-known companys who,L >for some reason, became the darling of Wall Street and all of a sudden, itsK >name became just as well known as Microsoft, also with its stock rocketing  toJ >orbit even though most investors only knew that JDS was involved in fiberG >optics, something that was associated with the success of .COMs at thea time.c >s >tK >Now, if you talk about the serious , established, companies who decided toSI >provide web services to customers, suppliers and employees, then you arenJ >correct, there was spending there. But how big was that spending compared toA >the "irrational exhuberance" spending from the "true" dot coms ?h >t >aD >I have to admit that for a company that was so associated with .COM	 spending,rJ >SUN (and Cisco) seems to have faired much better than its peers at Nortel andFJ >Lucent. SUN didn't go on buying/spending binges, so I guess that now that they2 >bubble has burst, SUN hasn't fallen from as high.   ------------------------------  % Date: Wed, 21 Aug 2002 16:06:15 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>m+ Subject: Re: HP-Compaq Merger Went Smoothlyp, Message-ID: <3D63F2A8.E31A5EB7@videotron.ca>   Fred Kleinsorge wrote:N > When will the Internet in general return to a *sustainable* growth, and what$ > will that level be?  I don't know.  N I think that from now on, the internet will yield sustainable growth. But thatJ growth will essentially follow the economy, it will not drive the ecomomy.  D > But feeding the frenzy were a lot of companies - like Sun, but notH > exclusively Sun - who had started to wrap their entire business aroundM > unbounded growth in the Internet - Sun is the dot in dot.com - yes?  Shoulds< > they then not share in being part of the bomb in dot.bomb.  L Contrary to Nortel, Lucent and others, Sun seems to have weathered very well+ the demise of the .COM exhuberant spending.   K I ask this sincerely: During the .COM exhuberant spending era, apart from aeK deal with AOL with regards to Netscape, did SUN buy anything significant inh terms of other companies ?B Did Sun spend much money increasing its manufacturing capability ?  K It would have been very interesting to have been a fly on the wall of Sun'soI boardroom to find out if Sun was fully aware that the .COM era was just aoL temporary bubble and that while SUN should take  every opportunity it could,M it should only grow its infrastructure based on real economic growth, not then2 fake overinflated one perceived by the .COM world.  I Sun seems to have faired pretty good the downfall of the .COM compared todN companies such as Nortel that had gone nuts trying to buy stuff left and right and is now hurting very bad.   ------------------------------  # Date: Wed, 21 Aug 2002 20:07:51 GMTb5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>e+ Subject: Re: HP-Compaq Merger Went Smoothlyu2 Message-ID: <rqS89.40$3%7.760930@news.cpqcorp.net>  = JF Mezei wrote in message <3D63F2A8.E31A5EB7@videotron.ca>...t >oL >I ask this sincerely: During the .COM exhuberant spending era, apart from aL >deal with AOL with regards to Netscape, did SUN buy anything significant in >terms of other companies ?   L They bought 4 companies I seem to remember in their filings.  Who knows whatJ would have happened if they hadn't bought there way into the server market' and were still relying on Workstations.n   >eJ >Sun seems to have faired pretty good the downfall of the .COM compared toI >companies such as Nortel that had gone nuts trying to buy stuff left andr righto >and is now hurting very bad.   G Well, having your earnings tank isn't exactly doing well.  They've lostaL money in 3 out of the last 4 quarters, and look to return to a mild loss forI the current quarter, with anemic earnings projections for the rest of thei year.    ------------------------------  % Date: Wed, 21 Aug 2002 16:17:34 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>h+ Subject: Re: HP-Compaq Merger Went Smoothly , Message-ID: <3D63F54E.985A1ECD@videotron.ca>   Fred Kleinsorge wrote:F > Go to money.cnn.com, bring up HPQ and click on earnings estimates toM > contrast it.  A *lot* more EPS both in the last year, and estimated for theb > next year - than Sun.o  J Earnings per share is perhaps not too representative since it doesn't take into account the share value.   M However, I do not think that HPQ is old enough as a corportate entity to havet= any indication of where it came from and where it is going.  u  G And the impact of 9/11 may have hidden some of the effects of HP/CompaqaD decisions, including Alphacide,  HP takeover of Compaq announcement,F uncertainty about future architectures. Only once the economy picks upI sufficiently will we see if the Alphacide and uncertainty about IA64 wille- hinder HP's growth compared to other vendors.c  K And to be fair, one should wait until real IA64 systems have been availableWN for a full year with VMS/Tandem/HP-UX (with the Tru64 bits in it) before beingM able to judge all the strategic decisions that Carly took in 2001. Until theneK (2005-2006 timeframe), I consider HP to be in a big void with only dead-end ; products available and nothing long term except wintel PCs.t  N Also remember that HP can claim huge revenus from its PC operations, but thoseK revenus are empty sicne they don't generate profits. Sun, on the other handsH may have smaller revenus, but the products it sells are more profitable.  N Also, for the next little while, HP can get away with lower profits because itH cam easily blame those on merger execution and reorganisation costs. ButH eventually, HP will have to stand on its own and will have runned out ofL excuses and won't be able to hide those numbers. Until that time, it is hard to judge HP's numbers.   ------------------------------  % Date: Wed, 21 Aug 2002 16:45:29 -0400d- From: JF Mezei <jfmezei.spamnot@videotron.ca>3+ Subject: Re: HP-Compaq Merger Went Smoothlys, Message-ID: <3D63FBD8.D9CF6EF2@videotron.ca>   Fred Kleinsorge wrote:I > Well, having your earnings tank isn't exactly doing well.  They've losttN > money in 3 out of the last 4 quarters, and look to return to a mild loss forK > the current quarter, with anemic earnings projections for the rest of the  > year.t    N Sorry, but Sun is nowhere near as bad a shape as Nortel or Lucent (hey, Lucent. was the former employer of your current boss).   ------------------------------  # Date: Wed, 21 Aug 2002 20:56:16 GMTa5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>-+ Subject: Re: HP-Compaq Merger Went Smoothly 1 Message-ID: <Q7T89.44$H9.879668@news.cpqcorp.net>s  = JF Mezei wrote in message <3D63F54E.985A1ECD@videotron.ca>...e >Fred Kleinsorge wrote:1G >> Go to money.cnn.com, bring up HPQ and click on earnings estimates to J >> contrast it.  A *lot* more EPS both in the last year, and estimated for theF >> next year - than Sun. > K >Earnings per share is perhaps not too representative since it doesn't take  >into account the share value. >   L Eh?  I'm not quite sure I understand.  If you want to measure earnings, thisI is the generally accepted way to do it.  We can look at pure revenue - incJ which case we can say they fell 32% in the fiscal year ending 6/30/02.  IfK we want to look at equity/market capitalization, the stock has fallen > 40% J over the last 3 months -- 70% for the last year.  We can look at P/E ratioJ (if we ignore the losing quarters and just use the last quarter) Sun is atI 37 (HP is at 18.5) - which means the stock isn't exactly overpriced at $4l compared to HP.   H The only glimmer (which Andrew wants to cling to) is that if you look atJ quarter-to-quarter numbers ***EVEN THOUGH THEY ARE LOSING MONEY IN ALL BUTL ONE*** they can claim wild things like 180% growth.  Meaning that instead of5 losing 6 cents a share, they are losing 4 cents, etc.:   >oI >Also remember that HP can claim huge revenus from its PC operations, butt thosenL >revenus are empty sicne they don't generate profits. Sun, on the other handI >may have smaller revenus, but the products it sells are more profitable.s >h  J It's conjecture to say the entire PC business is losing money.  It is fairI to say that in the retail PC space there is a profitability problem beingeG worked on by everyone in the PC industry - except perhaps Dell - who ise4 driving the profitability problem for everyone else.  H In the end, the real measure of the company is the return on investment.I You can measure that by stock value, dividends, earnings, revenue growth,r etc.  H Sun has lower total sales, less earnings, less earnings per share, lowerI stock value, pretty much every objective and verifyable measure is lower.tF The only measure they can try to point to is quarterly growth over theK disaster of the last 4-5 quarters - meaning that they are slowly recovering  to break even.  L >Also, for the next little while, HP can get away with lower profits because itI >cam easily blame those on merger execution and reorganisation costs. ButmI >eventually, HP will have to stand on its own and will have runned out ofvH >excuses and won't be able to hide those numbers. Until that time, it is hard >to judge HP's numbers.h  E But our profits are something like 20x larger per share.  We return aSJ dividend.  Our stock price while down, isn't anywhere near as down as Sun.I Our P/E ratio is half that of Sun (for the last quarter - since you can'tsL compute a P/E for quarters and years where you lose money) - which means oneL of us is overpriced, or one of us is underpriced - I tend to think that book< value on Sun is starting to set a lower limit on it's price.  K The analysts are predicting large growth in HP's earnings.  I guess the wayw: we will be measured is if we live up to those predictions.   ------------------------------  # Date: Wed, 21 Aug 2002 21:05:40 GMT * From: "Bill Todd" <billtodd@metrocast.net>+ Subject: Re: HP-Compaq Merger Went SmoothlyhA Message-ID: <EgT89.120906$SS.5311802@bin3.nnrp.aus1.giganews.com>y  @ "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote in message+ news:Q7T89.44$H9.879668@news.cpqcorp.net.... >o? > JF Mezei wrote in message <3D63F54E.985A1ECD@videotron.ca>...s   ...k  H > >Earnings per share is perhaps not too representative since it doesn't take  > >into account the share value. > >i > ' > Eh?  I'm not quite sure I understand.a  J It sounds as if JF is suggesting that P/E ratio would be a better measure.4 As the rest of your response even seemed to suggest.   - bill   ------------------------------  % Date: Wed, 21 Aug 2002 18:06:01 -0400m- From: JF Mezei <jfmezei.spamnot@videotron.ca>r+ Subject: Re: HP-Compaq Merger Went SmoothlyI+ Message-ID: <3D640EB2.962DBAC@videotron.ca>e   Bill Todd wrote:L > It sounds as if JF is suggesting that P/E ratio would be a better measure.6 > As the rest of your response even seemed to suggest.  M Correct. As an investor, I want to have an idea of the return on investment IgK might have should I invest in a stock. This means dividends, but also sharee price increase.   K This is why many .COMs failed: they used the proceeds of the IPO not to buy M assets but rather to spend money on plush offices and tons of advertising. SouK they ended up with no hard assets and hence no book value for their shares.aM Lack of profit is therefore fatal the minute you have spent all the IPO moneyu; since you have no hard assets against which you can borrow.   K As far as SUN vs HP is concerned, as I I said before, one cannot compare HP M against anything since HP doesn't yet "exist" with a track record. There haveIK been far too many traumatic changes at HP to be able to simply combine bothu? Compaq and HP numbers to predict when thew new HP will be like.    Here is another example:  M When Air Canada killed its main rival Canadian Airlines and bought it for $92eN million bucks, AC claimed that it would, by combining the two airlines, becomeD one of the top 10 airlines in the world.  But now that AC has finishM integrating the two airlines and shedding excess planes and employees, it haseC fallendown to about the same ranking as it was before (about 16th).e  K And Air Canada didn't announce such drastic changes as HP/Compaq did at the L same time as the merger. It hasn't announced that it is dropping its jets inM favour of propellor planes because it has heard that propellor planes may oner! day fly faster than today's jets.e  M HP has to contend with not only the compressions/shrinking due to the merger,eK but more importantly with the effects of all the architectural changes thattM have been announced but won't be delivered for quite some time, and this willkJ impact some of its more profitable products such as VMS most because thoseE products weren't very strong to begin with due policies of its owner.r   ------------------------------  % Date: Thu, 22 Aug 2002 11:26:27 +0530 ; From: "aniruddha patwardhan" <aniruddha_patwardhan@bmc.com>o+ Subject: installation using VMSINSTAL fails / Message-ID: <um8v7m9ka9bn64@corp.supernews.com>i   Hi,m  K I am installing a product on a vax system using VMSINSTAL. But installationr fails giving following error  6         Beginning installation of PVP04 V51.0 at 01:43  6 %VMSINSTAL-I-RESTORE, Restoring product save set A ...D %PVP04-S-ARCHITECTURE, Incompatible system type. This package can be installed onA -PVP04-S-ARCHITECTURE, AXP(Alpha) OpenVMS operating systems only.tA %VMSINSTAL-E-INSFAIL, The installation of PVP04 V51.0 has failed.     )         VMSINSTAL procedure done at 01:44t    G The product is build for VAX systems only and I am installing it on VAXe9 system. Any suggestions what couls be the reason of this?y   TIAi	 Aniruddhaf   ------------------------------  % Date: Wed, 21 Aug 2002 21:01:05 -0400-* From: "Stanley F. Quayle" <stan@stanq.com>P Subject: Re: Memo:  Re: Charon-VAX (was: [VAX] VMS to [Alpha] OpenVMS conversion/ Message-ID: <3D63FF91.12285.1724B938@localhost>g  6 On 19 Aug 2002 at 14:25, paul.beaudoin@hsbc.com wrote:K > The only explanation offered was that SRI chose to optimise for Intel/AMDeJ > chips rather than Alpha which, from their perspective is probably a goodJ > choice however from here sort of killed any ambitions we had  to replace > 76xx boxes the easy way.  F Actually, the emulator is written in C++, so it's not optimized for a D platform.  Performance is limited by what a particular C++ compiler D can do.  Just for fun, SRI tried the emulator on Tru64Unix.  It was D about 10% faster than the same code, on the same box, as on OpenVMS.  B Since you can buy some truly incredible Alpha systems, I wouldn't C give up on CHARON-VAX -- the new XL and XM versions are available, rE and they are capable of 40+ VUP's.  I'd bet performance on EV7 would r be even better...n  
 --Stan Quayleu! President, Quayle Consulting Inc.t  
 ----------C Stanley F. Quayle, P.E.   N8SQ   +1 614-868-1363   Fax: +1 614 868-t 16711 8572 North Spring Ct. NW, Pickerington, OH  43147n= Preferred address:  stan@stanq.com       http://www.stanq.come   ------------------------------  # Date: Wed, 21 Aug 2002 17:33:57 GMTF! From: Andy <acs@fcgnet.works.net>7  Subject: Re: New missive from HP> Message-ID: <Xns927189B3B7F34acsfcgnetworksnet@216.166.71.232>  : bill@gw5.cs.uofs.edu (Bill Gunshannon) enlightened us with/ news:ak0fke$1e3n90$2@ID-135708.news.dfncis.de: t  8 > In article <ak0ba3$1ehg1v$1@id-141708.news.dfncis.de>,6 >      "Peter Weaver" <peter.weaver@stelco.ca> writes:  > >> customers? Before the group gets into a fight over the termA >> "customer" (again) then how about we say, "How many earn theirl  >> living by working with VMS?"   8 A number higher than X where X = companies using VMS for@ day-to-day tasks and companies providing OpenVMS-based solutions to those companies.   B > Shouldn't you include in that how many WERE earning their living> > working with VMS before their bosses booted VMS out the door@ > because of the actions, or lack there of, by it's owners??  IsA > this time to bring up the lack of any interest int he education  > market again??  < VMS is still in the educational market. Higher Ed (where the@ company I work for focuses its' efforts) really isn't that largeC of a market (A couple thousand schools in the US?) for the kinds of @ applications that run on VMS. You just don't hear much about it.= In a niche market (like education) VMS probably has as much at< presence as it does in other niche markets. For a company to; survive in that niche it can't be purely VMS based anymore.t  < Of course, that's purely on the business side. Most users ofA business applications on VMS in Education have no idea what theirh; applications are running on. And there's no reason why they  should.   B The mind-share issue (VMS exposure for students).... I would argue< that exposing students to VMS at this point is not going to @ increase its' market share in the "real world" It's too late...   > Unless you are referring to visible signs of support from HP ?; Or past support from Compaq... that's something else again.s     -Andy- -- t   ------------------------------  # Date: Wed, 21 Aug 2002 20:56:52 GMTn$ From: winston@SSRL.SLAC.STANFORD.EDU  Subject: Re: New missive from HP8 Message-ID: <00A12CA3.112AB2B4@SSRL04.SLAC.STANFORD.EDU>  f In article <ak0ba3$1ehg1v$1@ID-141708.news.dfncis.de>, "Peter Weaver" <peter.weaver@stelco.ca> writes:; >"JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message:' >news:3D62F249.1C9D07AC@videotron.ca...- >>...6K >> If your customers have rants against you, what should you do ? Tell yournL >> customers to shut up, go away, or do you listen to your customer and heedB >> their suggestions and change the way you present your product ? >>...i >lK >But how many of the "main ranters" in this group are VMS customers? BeforerL >the group gets into a fight over the term "customer" (again) then how about: >we say, "How many earn their living by working with VMS?"  J I'm not sure that's a fair division either.  I mean, people who find they M can't earn their living by working with VMS any more because of the shrinkingdF VMS presence in their communities or market segments have a legitimateM complaint and ideas that could help HP market VMS.  People who no longer feelnL safe, from a job-security standpoint, in proposing VMS solutions because of K an absence of clear support for the product may be earning their livings byrN working with NT or UNIX, but it's not because they _want_ to, and they too canC tell HP what it needs to do to get that business back.  It's reallylM unreasonable to make whether your paycheck today derives from work with VMS ajN litmus test for whether you're allowed to post cranky remarks about the owners handle it.)l  N (That said, I do get really tired of some of the ranting, but it's all clearlyF coming from people who support VMS, some of whom have gotten burned byL supporting it, and the proprietors of VMS need to hear at least some of it. J They also need to hear Andrew, at least to know what their competitors are0 saying about them and be prepared to refute it.)   -- Alan2   >7   ------------------------------  + Date: Thu, 22 Aug 2002 02:03:36 +0000 (UTC)-  From: mustang@ucc.asn.au.invalid  Subject: Re: New missive from HP* Message-ID: <ak1gpo$gon$1@enyo.uwa.edu.au>  , Peter Weaver <peter.weaver@stelco.ca> wrote:< : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message( : news:3D62F249.1C9D07AC@videotron.ca... :>...dK :> If your customers have rants against you, what should you do ? Tell yourhL :> customers to shut up, go away, or do you listen to your customer and heedB :> their suggestions and change the way you present your product ? :>...s  L : But how many of the "main ranters" in this group are VMS customers? BeforeM : the group gets into a fight over the term "customer" (again) then how aboutt; : we say, "How many earn their living by working with VMS?"y   How is that relevant?uF Groundswell and inertia and mindshare and all those buzzword compliant% touchy-feelies are just as important.t  : If 'new users' think VMS is dying, they'll look elsewhere.   D.   -- n% " I don't get mad.... I get stabby. "- - William "Fat Tony" Williams.   ------------------------------  % Date: Wed, 21 Aug 2002 21:53:29 -0400o2 From: rdeininger@mindspring.com (Robert Deininger)  Subject: Re: New missive from HPK Message-ID: <rdeininger-2108022153290001@1cust211.tnt2.nashua.nh.da.uu.net>n  ; In article <3D630DFF.1F79E551@fsi.net>, "David J. Dachtera"E <djesys.nospam@fsi.net> wrote:   >eG >Question: Is it "our" fault for complaining about the mismanagement of E >VMS or is it HP/Q's fault for trying to kill their own cash-cow (and , >destroying our livelihoods in the process)?  G Consider carefully the difference between "complaining" (your word) andaF "ranting and raving" (the phrase Terry used).  Consider the difference@ between complaining about a product's features, prices, etc, andG complaining about the management of the company that produces it.  Also F consider the difference betweeen criticism and spiteful, rude personal attacks.  J IMHO, a company ought to listen to complaints about products, provided theJ complaints are resonably polite and coherent.  On the other hand, "rantingE and raving" (not polite or coherent), personal insults, and offers toeJ restructure the entire management of the company will likely, and rightly,J cause the folks at the receiving end to terminate the conversation as soon as possible.  E There are a handful of people in this newsgroup who wouldn't be worthmI having as customers, no matter how much money they brought to the table. uF Why would (insert name of HP executive) willingly deal with a hissing,F spitting, insulting, hysterical potential customer, just to marginally increase the company's income?   ------------------------------  # Date: Tue, 20 Aug 2002 02:21:02 GMTe# From: "John Smith" <a@nonymous.com>o  Subject: Re: New missive from HPF Message-ID: <iIh89.14584$KZL.312@news04.bloor.is.net.cable.rogers.com>  < "Terry C. Shannon" <terryshannon@attbi.com> wrote in message( news:fuh89.105244$me6.13237@sccrnsc01...  > >-L > Positive action is indeed necessary, negative tone is contraproductive. InK > fact, I have it on VERY reliable authority that a Level One (read MC) was6E > ready to scuttle the whole damned VMS thing last fall after all they rantingn
 > and raving.x >l7 > Obviously this did not happen, which is a Good Thing.t    L I think what that quite effectively demonstrates is the vacuum Compaq seniorJ management operated in when it came to understanding their customers, VMS,L and its place in the enterprise space. So in a fit of pique, MC comes within+ a whisker of slitting the company's throat?-  J In medical terms, operating this way is called schizophrenia...but there'sF medication for that. By the sounds of it, maybe Carly is a blessing in	 disguise.l  J BTW, haven't heard too much about good old MC these days. What's he up to?8 Hanging out with Dick Cheney in an undisclosed location?   ------------------------------  # Date: Thu, 22 Aug 2002 02:07:19 GMTt# From: "John Smith" <a@nonymous.com>m  Subject: Re: New missive from HPI Message-ID: <rHX89.12192$bu81.10954@news02.bloor.is.net.cable.rogers.com>p  1 <winston@SSRL.SLAC.STANFORD.EDU> wrote in message 2 news:00A12CA3.112AB2B4@SSRL04.SLAC.STANFORD.EDU... >eK > I'm not sure that's a fair division either.  I mean, people who find theyaE > can't earn their living by working with VMS any more because of the 	 shrinkingyH > VMS presence in their communities or market segments have a legitimateJ > complaint and ideas that could help HP market VMS.  People who no longer feelJ > safe, from a job-security standpoint, in proposing VMS solutions because ofJ > an absence of clear support for the product may be earning their livings byL > working with NT or UNIX, but it's not because they _want_ to, and they too can E > tell HP what it needs to do to get that business back.  It's reallymI > unreasonable to make whether your paycheck today derives from work with  VMS aiI > litmus test for whether you're allowed to post cranky remarks about they owners
 > handle it.)c > H > (That said, I do get really tired of some of the ranting, but it's all clearlyhH > coming from people who support VMS, some of whom have gotten burned byI > supporting it, and the proprietors of VMS need to hear at least some off it.eL > They also need to hear Andrew, at least to know what their competitors are2 > saying about them and be prepared to refute it.)    
 Well said.   ------------------------------  # Date: Thu, 22 Aug 2002 02:18:19 GMT # From: "John Smith" <a@nonymous.com>l  Subject: Re: New missive from HPI Message-ID: <LRX89.12656$bu81.10263@news02.bloor.is.net.cable.rogers.com>o  ? "Robert Deininger" <rdeininger@mindspring.com> wrote in messagewE news:rdeininger-2108022153290001@1cust211.tnt2.nashua.nh.da.uu.net...e >hI > Consider carefully the difference between "complaining" (your word) and H > "ranting and raving" (the phrase Terry used).  Consider the differenceB > between complaining about a product's features, prices, etc, andI > complaining about the management of the company that produces it.  AlsocH > consider the difference betweeen criticism and spiteful, rude personal
 > attacks. >nL > IMHO, a company ought to listen to complaints about products, provided theL > complaints are resonably polite and coherent.  On the other hand, "rantingG > and raving" (not polite or coherent), personal insults, and offers to L > restructure the entire management of the company will likely, and rightly,L > cause the folks at the receiving end to terminate the conversation as soon > as possible. >lG > There are a handful of people in this newsgroup who wouldn't be wortheJ > having as customers, no matter how much money they brought to the table.H > Why would (insert name of HP executive) willingly deal with a hissing,H > spitting, insulting, hysterical potential customer, just to marginally  > increase the company's income?    F It's not just kissing off the income today, it's saying goodbye to theB on-going 'annuity' income from support contracts, system upgrades,E word-of-mouth referrals, and purchases made at the next company theset individuals go to work at.  J Digital, Compaq, and now you are implying HP will willingly not attempt toI get to the bottom of a customer complaint and concern no matter how it istF expressed? How quickly can you say "Let's close the Enterprise SystemsK division down because we didn't listen to our last remaining customers, andt now we have none?"  C Let's get real here - customers wouldn't be, as you claim, hissing,eF spitting, insulting, or hysterical, unless DEC/CPQ/HPQ gave them ample reason to be that way.  I DEC/CPQ/HPQ seem to have acted a lot like an alcoholic that doesn't admit-L he's got a drinking problem. HPQ needs to do a LOT of self-reflection on whyE a large number of their customers feel this way when it comes to VMS.e" Perhaps then progress can be made.   ------------------------------  # Date: Tue, 20 Aug 2002 12:57:32 GMTa# From: "John Smith" <a@nonymous.com>t  Subject: Re: New missive from HPH Message-ID: <01r89.20034$KZL.10903@news04.bloor.is.net.cable.rogers.com>  < "Terry C. Shannon" <terryshannon@attbi.com> wrote in message( news:fuh89.105244$me6.13237@sccrnsc01... >nL > Positive action is indeed necessary, negative tone is contraproductive. InK > fact, I have it on VERY reliable authority that a Level One (read MC) waseE > ready to scuttle the whole damned VMS thing last fall after all the  ranting 
 > and raving.o >r7 > Obviously this did not happen, which is a Good Thing.  >  >e  L I think what that quite effectively demonstrates is the vacuum Compaq seniorJ management operated in when it came to understanding their customers, VMS,& and its place in the enterprise space.  I So in a fit of pique, MC comes within a whisker of slitting the company's J throat? In medical terms, operating this way is called schizophrenia...but there'sbF medication for that. By the sounds of it, maybe Carly is a blessing in	 disguise.i  J BTW, haven't heard too much about good old MC these days. What's he up to?8 Hanging out with Dick Cheney in an undisclosed location?   ------------------------------  # Date: Thu, 22 Aug 2002 01:48:46 GMTl1 From: "David J. Dachtera" <djesys.nospam@fsi.net>a  Subject: Re: New missive from HP' Message-ID: <3D6447F8.3A3837B2@fsi.net>    Bill Gunshannon wrote: > @ > In article <Xns927189B3B7F34acsfcgnetworksnet@216.166.71.232>,- >         Andy <acs@fcgnet.works.net> writes:i
 > > [snip]@ > > Of course, that's purely on the business side. Most users ofE > > business applications on VMS in Education have no idea what theiri? > > applications are running on. And there's no reason why theyo > > should.y > E > Bill Gates doesn't agree. he wants everyone to know that everythinge4 > is running on Windows.  And he is winning the war.  H This occurred to me yesterday. I figured it was so obvious, that someoneG at M$ *MUST* have come up with it and made it a rallying call for theirn; employees: "Windows isn't everything, it's the ONLY thing!"0  H I'm sure most of this group could come with completions for the sentenceA fragment: "Yeah, it's the only...", most of them, well, less than8 complimentary.   -- e David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/e   ------------------------------  # Date: Thu, 22 Aug 2002 03:06:01 GMTt! From: Andy <acs@fcgnet.works.net>4  Subject: Re: New missive from HP> Message-ID: <Xns9271EAAEFC8ABacsfcgnetworksnet@216.166.71.232>  ? "David J. Dachtera" <djesys.nospam@fsi.net> enlightened us withL  news:3D6447F8.3A3837B2@fsi.net:    > Bill Gunshannon wrote: >> AA >> In article <Xns927189B3B7F34acsfcgnetworksnet@216.166.71.232>,s. >>         Andy <acs@fcgnet.works.net> writes: >> > [snip] A >> > Of course, that's purely on the business side. Most users of-@ >> > business applications on VMS in Education have no idea whatA >> > their applications are running on. And there's no reason whyh >> > they should.] >> t; >> Bill Gates doesn't agree. he wants everyone to know that4@ >> everything is running on Windows.  And he is winning the war.  > Maybe I'm wrong but I don't see this as a war. It's evolution.B Right now Bill Gates is like all those gigantic prehistoric mamalsA just before the last ice age. He's been lucky and the competitionl? hasn't been as tough. But that isn't the same thing as winning.   : The monolithic "Windows everywhere" mantra that Microsoft 9 sycophants spout at the least chance is not going to maket( it in the long run. It's very unhealthy.  6 We (The VMS community) just need to be flexible enough3 and ready to evolve in the right direction when thep= shift away from the Windows monotheism happens. And I believe,* the signs of that are already in the air.   B Our battles for mindshare won't be fought on the same turf as Unix@ and Windows. By the time people get to college, it's already too= late to change what they believe in. How many Linux fanatics  3 became that way IN college ? Or Windows fanatics ? x  A You (Bill) need to change your focus. Get to them BEFORE college.4> Otherwise (as I've said before & you obviously disagree with),# it's too late once they are there !   8 Be creative... get VMS IN THE HOME !!!! Not just through@ the Hobbyist program but in other ways !! Have them grow up with it!   6 How hard would it be to put together a VMS based home . automation system ? (A small home cluster....)8 Isn't that one of the best places to put something where maximum availability is key ?.  6 Anyone want to start a business ? I have some ideas in that area myself.... s  B > This occurred to me yesterday. I figured it was so obvious, thatB > someone at M$ *MUST* have come up with it and made it a rallying? > call for their employees: "Windows isn't everything, it's thes > ONLY thing!" a  $ "Windows is the LAST thing we need."   -Andy-   -- t   ------------------------------  # Date: Thu, 22 Aug 2002 02:59:18 GMTR1 From: "David J. Dachtera" <djesys.nospam@fsi.net>h  Subject: Re: New missive from HP' Message-ID: <3D645881.FF6A52A3@fsi.net>    John Smith wrote:  > [snip]E > Let's get real here - customers wouldn't be, as you claim, hissing,aH > spitting, insulting, or hysterical, unless DEC/CPQ/HPQ gave them ample > reason to be that way.  F But, like everyone else, they are not responsible for the consequencesF of their actions. (Does that not define something under the law?) TheyG are not at fault for their actions, we are at fault for calling them onl their transgressions.n  K > DEC/CPQ/HPQ seem to have acted a lot like an alcoholic that doesn't admittN > he's got a drinking problem. HPQ needs to do a LOT of self-reflection on whyG > a large number of their customers feel this way when it comes to VMS.d$ > Perhaps then progress can be made.  F I lived with a suicidal, manic-depressive alchy for a good three yearsF (end of '89 to late-summer of '92). Suffice it to say, I'm not holding
 my breath.   -- s David J. Dachterar dba DJE Systemss http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/o   ------------------------------  # Date: Thu, 22 Aug 2002 02:51:59 GMT 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>a  Subject: Re: New missive from HP' Message-ID: <3D6456C9.835C9643@fsi.net>u   Robert Deininger wrote:  > = > In article <3D630DFF.1F79E551@fsi.net>, "David J. Dachtera"a  > <djesys.nospam@fsi.net> wrote: >  > >iI > >Question: Is it "our" fault for complaining about the mismanagement ofiG > >VMS or is it HP/Q's fault for trying to kill their own cash-cow (ando. > >destroying our livelihoods in the process)? > I > Consider carefully the difference between "complaining" (your word) andy0 > "ranting and raving" (the phrase Terry used).   E I'd consider them to be synonymous, and not necessarily incompatible.l   > Consider the differencerB > between complaining about a product's features, prices, etc, andC > complaining about the management of the company that produces it."  H "Chicken-and-egg" problem. If management would make things happen, usersC would not complain about "inertia" (wrong word, really, but I don'tl currently have a better one).o   >  AlsooH > consider the difference betweeen criticism and spiteful, rude personal
 > attacks.  > This typically arises from "messing with someone's rice bowl".  F If I work for you and you ship my job over to the sweatshops in China,A should I blame myself for not diversifying my job skills so I cana? rebound more quickly from the layoff, or should I blame you for  profit-mongering?a  G Money, especially one's livelihood, is by nature a very emotion-chargedb8 issue. I don't see that changing anytime in my lifetime.  L > IMHO, a company ought to listen to complaints about products, provided the0 > complaints are resonably polite and coherent.   B Oxymoron. "Complaints" are, almost by definition, not likely to beG either reasonable, polite or coherent. Differences of opinion, perhaps,oA since the name even implies a degree of thought and intelligence.eH Complaints, on the other hand, are typically purely emotional in nature.   > On the other hand, "rantinggG > and raving" (not polite or coherent), personal insults, and offers to-L > restructure the entire management of the company will likely, and rightly,L > cause the folks at the receiving end to terminate the conversation as soon > as possible.  B You can douse a fire or run away from it. The choice is yours, butE consider which would likely be the more profitable option in the long7 run.  - Courage in the face of adversity is a virtue.a  H Escapism, or taking the easy way out, is tantamount to one of the "sevenD deadly sins": Sloth. Not likely to boost either your public image or your self image.  G > There are a handful of people in this newsgroup who wouldn't be worthrJ > having as customers, no matter how much money they brought to the table.  D Rather depends on your degree of profit-motivation. If I were a pureG capitalist, and as devoid of ethics as those currently making headlineslE around the world, I'd sell fireworks to Ralph Nader's kids until theye% stopped buying or until I got busted.   E If all I wanted were happy, contented customers, I'd get a license to'< sell Lithium and Prozac and go into the psych. pharm. field.  ( They're coming to take me away, hah-hah!  They're coming to take me away,  ho-ho, hee-hee, hah-hah,
 to happy homee with trees and flowers w and chirping birds and basket-weavers who 8 sit and smile and1 twiddle their thumnbs and toes,n, and they're coming to take me away, hah-hah!  H > Why would (insert name of HP executive) willingly deal with a hissing,H > spitting, insulting, hysterical potential customer, just to marginally  > increase the company's income?   Define "marginally".  E If <insert the name of the key person at VMS's largest customer> wereeE slightly less unpleasant than Saddam Hussein, he could throw darts atTG me, put battery acid in my coffee and send his thugs out to overturn my'B rented car every time I met with him, as long as keeps buying, andE especially if I increase his buying by addressing some of his issues.i  G Then again, it's a well-known axiom that irate (ex-)customers will givesF you an order of magnitude more bad publicity by word-of-mouth than the@ amount of good publicity you will get from a satisfied customer.  G Seems to me, the odds would favor taking some heat just in an effort ton= reduce the sniping. Ask yourself the question: how many irate  ex-customers can I afford?  G Then again, how many times have we asked here about who uses VMS? WhereuD is the customer list? You MUST have seen the responses. Some of themE almost sound like Rodney Dangerfield on stage, when the cameras startiF flashing, "Hey, no pictures! No pictures! I don't wanna be seen here!"  F The folks who do continue to use VMS prefer to remain anonymous if youG can believe the postings here. No good recommendations to be had there.0  D So, best to trim your losses and try to silence your critics, not to- mention making a few extra bux along the way.3  , Show me another way to get a deal like that!  2 If you can't take the heat, get outta the kitchen!   -- s David J. Dachtera3 dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/i   ------------------------------  % Date: Thu, 22 Aug 2002 00:23:54 -0400u- From: JF Mezei <jfmezei.spamnot@videotron.ca>   Subject: Re: New missive from HP, Message-ID: <3D646758.1A783E96@videotron.ca>   "David J. Dachtera" wrote:I > Seems to me, the odds would favor taking some heat just in an effort toe? > reduce the sniping. Ask yourself the question: how many irateM > ex-customers can I afford?  N But if easy solutions to make those irate customers happy again  are presented3 to you, why wouldn't you accept those suggestions ?y  L If you insist on continuing the irritating comments or VMS-ignoring that youN are told irritates your customers, doesn't that mean that you enjoy irritating your customers ?   ------------------------------  % Date: Wed, 21 Aug 2002 23:50:35 -0400t- From: JF Mezei <jfmezei.spamnot@videotron.ca>   Subject: Re: New missive from HP, Message-ID: <3D645F8B.6509CBA5@videotron.ca>   John Smith wrote:PN > he's got a drinking problem. HPQ needs to do a LOT of self-reflection on whyG > a large number of their customers feel this way when it comes to VMS.4$ > Perhaps then progress can be made.  J It isn't just self reflection. It is also being sensitive to what has beenL done to custoemrs in the past and ensuring that HP doesn't rub salt in thoseH wounds. This is why I complain about HP's continued use of words such asK "commitment" and "plan of record". They remind of a very sensitive, hurtful ) and un-healed wound and rub salt into it.t  J Whether HP and its top brass like it or not, very unpopular decisions haveE been made in the past by both Digital and Compaq. HP can choose to bebW insensitive to those, or it can choose to acknoledge past mistakes and act accordingly.g   ------------------------------  # Date: Thu, 22 Aug 2002 03:44:06 GMT 1 From: LESLIE@JRLVAX.HOUSTON.RR.COM (Jerry Leslie)i  Subject: Re: New missive from HP; Message-ID: <a6Z89.84701$eK6.2832611@twister.austin.rr.com>   " John Smith (a@nonymous.com) wrote: : L : BTW, haven't heard too much about good old MC these days. What's he up to?: : Hanging out with Dick Cheney in an undisclosed location? :  :  URL wrapped to 2 lines:   2    http://www.siliconvalley.com/mld/siliconvalley/1    business/companies/hewlett_packard/3904283.htme    Mercury News | 08/20/2002 | a4    HP exec sees small, medium firms leading recovery  I   "ASPEN, Colo.(Reuters) - Hewlett-Packard Co. President Michael Capellas@B    Tuesday said small and medium businesses will lead the economicC    recovery in the United States, but that the arrival of a broadere(    improvement remains difficult to peg.  E    ``If you're looking for an indicator that's where I'd look,'' said.F    Capellas, referring to what he called mid-tier, general businesses.  C    Capellas -- who holds the No. 2 spot at the Silicon Valley-basednH    technology giant following its mega merger with computer maker CompaqH    -- said small and medium businesses account for a substantial part of?    the economy and already are showing early signs of recovery.d  E    Nevertheless, he said it remains too difficult to predict when thee     overall economy will pick up.  F    ``I do think we have achieved the bottom,'' said Capellas, who alsoA    predicted that the United States will lead the global economicsB    recovery, partly due to productivity gains stemming from recent    advances in technology.  G    ``You will see a stronger recovery than most people think because of H    those productivity gains,'' Capellas told attendees of the Progress &9    Freedom Foundation's annual summit in Aspen, Colorado.h  C    Capellas' talk came after HP's stock had closed down 49 cents atP*    $14.51 on the New York Stock Exchange."  B The Economic Policy Institute predicts unemployment rising to 6.5%B by the end of 2002 and staying between 6.0% and 6.5% through 2003:  2    http://www.epinet.org/briefingpapers/bp121.html;    It ain't over till it's really over (EPI Briefing Paper)o  7   "Slow growth will lead to rising unemployment in 2002n$    and high unemployment in 2003..."    i  2 --Jerry Leslie   (my opinions are strictly my own)9   Note: leslie@jrlvax.houston.rr.com is invalid for email    ------------------------------  # Date: Thu, 22 Aug 2002 04:28:10 GMT'* From: "Bill Todd" <billtodd@metrocast.net>  Subject: Re: New missive from HPA Message-ID: <uLZ89.126032$SS.5576925@bin3.nnrp.aus1.giganews.com>l  . "Andy" <acs@fcgnet.works.net> wrote in message8 news:Xns9271EAAEFC8ABacsfcgnetworksnet@216.166.71.232...A > "David J. Dachtera" <djesys.nospam@fsi.net> enlightened us witha! > news:3D6447F8.3A3837B2@fsi.net:a   ...'  = > >> Bill Gates doesn't agree. he wants everyone to know thateB > >> everything is running on Windows.  And he is winning the war. >h@ > Maybe I'm wrong but I don't see this as a war. It's evolution.D > Right now Bill Gates is like all those gigantic prehistoric mamalsC > just before the last ice age. He's been lucky and the competitioneA > hasn't been as tough. But that isn't the same thing as winning.o  L I think you may seriously misunderstand the way evolution works, and how (in9 whatever sense it may) that applies to operating systems.e  > Prehistoric animals (and dinosaurs before that) were large andK over-specialized.  As such, they depended on a fairly specific ecosystem touG support their appetites.  This is *very* similar to the manner in which L current higher-end systems depend on the current hardware/economic ecosystemK to survive, and bears little resemblance at all to Windows (which is small,iK weak, and almost pathetic, but extremely pervasive and very adaptable, just,D like the small mammals that out-lasted their far larger and arguably superior compatriots).  I And just as major upheavals occurred in the ecosystems that had supportedlL the dinosaurs, wooly mammoths, etc., so upheavals that severely threaten theJ current high-end systems are occurring in the technical environment.  It'sL not just Itanic supplanting better processors like Alpha and PA-RISC (thoughK they could have fought on successfully for at least the rest of this decade0J had they been more willing to adapt to market changes, and Alpha certainlyH had the potential to extend its reach into those pervasive and adaptableH areas rather than isolate itself increasingly in an ultimately untenableL niche), but very likely Windows supplanting better systems like VMS (much asK Unix already has, to a large extent - and Unix, or at least Linux, seems tosJ have a much better chance of surviving to exist alongside Windows than VMSJ does) - again, in large part because VMS never adapted (because its owners had no interest in doing so).e  K Where are systems like VMS most needed now?  High-capacity/high-performance K environments, for one - and at least in the storage area those environmentsoL are rapidly removing dependencies on large-scale single systems (using DAS -I storage directly attached to servers) and moving toward SAN-style systemssE (where the scaling takes place in the storage farm rather than in the L servers themselves, and will soon move to distributed file-level rather thanJ block-level access that will remove the need for server cooperation of theL form that VMS and its DLM provide).  How about high reliability?  Well, onceJ you've solved the storage part of the problem, then clustered Unix systemsL can do quite well with the rest of it, and even Windows is adequate for someK environments (mostly very carefully bounded ones, but *some* improvement ispC likely to occur over time).  And Unix-style (or even Windows-style) I clustering is adequate to handle most high-availability requirements oncekL (again) you've removed the problem of scalable/reliable/available/manageable storage from the equation.  I That leaves only super-reliable, large-scale SMP/NUMA systems that aren't K amenable to Beowulf-style replacement:  the same niche that true mainframesnE have occupied for so long.  It will be a long time before Windows and'I possibly Linux as well can tackle that niche, but its going to get prettywD crowded there with zOS, VMS, NSK, and the remaining high-end Unixes.  C So the climate has changed and VMS's owners haven't shown a hint ofnJ inclination to help VMS adapt to it in any significant manner.  Windows isL adapting/evolving, Linux is adapting/evolving, storage is adapting/evolving,H networks are adapting/evolvingk, even some Unixes are adapting/evolving,H while VMS just sits with all its magnificence in one place with less and less to eat.  B The lessons of evolution make it pretty clear where VMS is headed.   - bill   ------------------------------    Date: 21 Aug 2002 12:50:40 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)a+ Subject: Re: OpenVms / Vax -- free download 3 Message-ID: <gmKyzOVbKpaK@eisner.encompasserve.org>n  [ In article <3D63B952.466F2705@email.uc.edu>, David Michaels <michaedi@email.uc.edu> writes:e  J > Is it legal to post a copy of the Operating System CD for OpenVms Vax on > a public server???  F Only if you have the permission the copyright owner (Hewlett Packard).B Since the Hobbyist Program does not have such permission, it seems. unlikely that you would be granted permission.   ------------------------------    Date: 21 Aug 2002 20:20:41 +0200a From: holitska_a@removehomo-togetvalide-mailhomo-ludens.elte.hu (Holi - Holitska Andrs)s+ Subject: Re: OpenVms / Vax -- free downloadt! Message-ID: <KLdDcc02GGf8@ludens>t  q In article <20020821171522.57287.qmail@web20210.mail.yahoo.com>, Fabio Cardoso <fabiopenvms@yahoo.com.br> writes:f > FREE DOWNLOAD FOR WHAT ????  > & > OpenVMS / VAX doesnt run in PCs !!!!  5   Well, it would be nice anyway. Many students around 2   here would like to try OpenVMS (I'm working at a5   university), but of course they don't want to spendi5   money for just trying, and when the Hobbyist Licens-3   is free for everyone, why shouldn't be OpenVMS be93   free for everyone (for home use only, of course).i  1   Currently the only way is to give them our CDs, 6   which isn't the best solution (one person at a time,1   and who knows how long they want to play around,   with it).,  7   It would be great, if Compaq/HP would allow HobbyistsC8   to download CD images of OpenVMS VAX and Alpha from an7   official server, or allow the copying of OpenVMS CDs.s7   (I think copying the CDs or distributing CD images onn$   the network is currently illegal.)   > 7 > In 2 weeks I will receive as giff - I hope - an Alphai2 > 2100 and a VAX 4000. I dont know what I will do 6 > with the  VAX. I want just the external disks (SBB)  > to my one disk Alpha 1000.  8   Tell everyone here where they can pick it up. I'm sure*   it will find a new owner very shortly :]   > 	 > Regardsh >  > FC o >  > 3 > --- David Michaels <michaedi@email.uc.edu> wrote:v >> Hi,5 >> Is it legal to post a copy of the Operating System9 >> CD for OpenVms Vax on >> a public server???. >> - - - - 8< - - -   Bye:  <Holi>   ------------------------------   Date: 21 Aug 2002 19:25:49 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)+ Subject: Re: OpenVms / Vax -- free download * Message-ID: <ak0pft$iu9$5@web1.cup.hp.com>  [ In article <3D63B952.466F2705@email.uc.edu>, David Michaels <michaedi@email.uc.edu> writes:i  I :Is it legal to post a copy of the Operating System CD for OpenVms Vax one :a public server???a  E   I would expect that there would be folks that would take a dim viewf>   of this.  Folks from both UC and HP potentially among these.  G :The software could still be licensed the same way it is now (regardingu :hobbyist).      So?o  I :A free download would increase intrest in OpenVMS tremendously... A freeoI :download would bring stabillity and logevity to the operating system andt
 :hardware.  D   Rather than obtaining a licensed copy at US$30 or so?  I doubt it.  H :The hobbist program is great, but it seems like they are allways out of	 :stock...b  C   Currently remastering for V7.3-1 and current LP revisions, AFAIK.g  G :The best thing for the future of OpenVMS/VAX would be for the OS to ben :freely downloadable...      Best for whom?  I :Eventually I think it will happen, it just depends on how dead the thinge5 :gets first.. OpenVMS/Vax doesn't make HP any money...  A   Trolling?  Or assuming?  Or do you know something that I don't?   I :It's not like I'm going to run OpenVMS on my Ultra 30 and cheat them outdI :of a hardware sale. I obviously bought a copy of the operating system ifq$ :I've got a Vax (they came together)  I   Many VAX systems do not have OpenVMS licenses -- ULTRIX was licensed on I   and can operate on various VAX boxes, and then there are the occasional G   and various legal cases where hardware and software must be availabledG   seperately -- and OpenVMS VAX also operates on and can be licensed onx)   software emulators.  But you knew that.i  H   And then there is the simple assumption that folks would be willing toH   download an OpenVMS kit from a potentially untrusted network site -- IJ   am sure that some would, of course.  But for all I know, an unauthorizedH   and uncontrolled OpenVMS download kit could be infected with something   really ugly and nasty.  J   While certainly not an official corporate representative, I would assumeJ   that the simple and obvious omission of (unrestricted) network access toH   these and other kits on the HP corporate OpenVMS website nor on the HPI   OpenVMS hobbyist website could be generalized into a corporate positionrK   on this matter.  (We have already provided targeted on-line distribution,i(   though not unrestricted distribution.)  L   If you are interested in obtaining formal permission (or a formal denial),I   please write up a short proposal and I'll pass it along to the releventfI   folks in the OpenVMS business management organization.  I can also pass K   your current posting along, as well, but I would expect that some of yourg2   statements from the posting would be questioned.    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Wed, 21 Aug 2002 14:16:48 -0400,* From: "Kaledas, Ronald" <RKaledas@dmc.org>. Subject: RE: oracle/telnet-tcpip 5.3/vms 7.3-1L Message-ID: <B39B28A01B62D311B1360090277B3A58070F7833@tro-nt-exchg2.dmc.org>  J This message is in MIME format. Since your mail reader does not understand< this format, some or all of this message may not be legible.  ' ------_=_NextPart_001_01C2493E.EA2B6E30l Content-Type: text/plain;s 	charset="iso-8859-1"g  E I also tried rsh and rlogin.  No luck.  I tried set watch on both theeI sethost session and the telnet session, and compared the output of both -s nothing apparent there.d  D Didn't think about checking opcom, but the operator log doesn't have anything about it.  + It happened on 2 accounts, both privileged.n  J Also, I had to go back to the field test version (!) of 7.3-1 because theyK needed this environment, and everything is working fine now!  (I had backed H up the system disk before the upgrade, of course, and going back to that( system disk made everything work again.)I Of course, I can't stay with the field test version because of bugs in ituC that cause crashes once a month or so.  But, at least oracle works!e   Thanks,f Ron    > -----Original Message-----3 > From: Mike Zarudzki [mailto:mike.zarudzki@hp.com]s+ > Sent: Wednesday, August 21, 2002 11:22 AM  > To: Info-VAX@Mvb.Saic.Comk0 > Subject: Re: oracle/telnet-tcpip 5.3/vms 7.3-1 >  > ? > oracle/telnet-tcpip 5.3/vms 7.3-1Have you tried RSH into it? e > Build a command 0 > procedure that adds the set watch file command* > and see what results, if any get logged. > 2 > I assume your not seeing anything OPCOM related? > ; > I assume this is not just one account, but accounts with i > various levels ofo > privs? > = > Anything wierd on the telnet service itself? File defined? t > Log opts defined?t >  > HTHe
 > -Mike Z. > "All opinions my own..." >  > 7 > "Kaledas, Ronald" <RKaledas@dmc.org> wrote in message H > news:B39B28A01B62D311B1360090277B3A58070F7827@tro-nt-exchg2.dmc.org...A > I'm having a strange problem that I hope someone can help with.e? > In our test cluster, we were running vms E7.3-1 (field test) s > with tcpip 5.3 > and oracleF > 7.3.3.6 (and also oracle 8.1.7.3).  Yesterday, I upgraded vms to the > production version of > >  V7.3-1.  Nothing else changed.  When the cluster came back  > up, everything > worked@ >  fine...except SVRMGRL and SQLPLUS.  When you telnet into the  > node, or comeg > in via= >  reflection X, invoking svrmgrl produces the following two   > lines (after the > normal header):a3 > MGR-11401: input error, unable to read input linet- > MGR-01508: unable to close the current file @ > Invoking sqlplus, without a username/password, is supposed to  > prompt and > wait for the= >  username.  It comes up with the username prompt, and then   > immediately exitsi
 > back to dcls? >  without waiting.  Invoking sqlplus with a username/password   > gets into then > program, but thent= >  the sql> prompt then keeps appearing and scrolling up the   > screen as if you > were holding? >  down the RETURN key!  So, it seems to me these programs are  
 > acting like. > sys$input is not# >  defined, or something like that.t< > Here's the kicker...everything works fine if you SET HOST  > into the node.
 > SVRMGRL andn! >  SQLPLUS both work as expected.s > Argh!cG > Everything else seems to work fine, any other program I run, just not  > oracle.  I've triedn= > logging in with /nocom, in case there was some interaction   > there, I've triede > defining sysH > $input, I've done this in oracle 7.3 and 8i, and I also booted off the > backup copy of the< >  system disk, to see if the problem was in oracle, but it  > worked fine there!G > It seems to be some interaction with TELNET and oracle, but the tcpip. > version DID NOTr= > change!  I also reinstalled tcpip 5.3 and the mup for 5.3, 9 > but that didn'tt > change anything.= > Oh, and I didn't have the mup in place right away, so this a > happened both  > with and without >  the mup. < > I've logged a call, but Colorado support is also not sure  > what else to lookp > at.2" > Does anyone here have any ideas?	 > Thanks,/ > Ronv >  >   ' ------_=_NextPart_001_01C2493E.EA2B6E308 Content-Type: text/html; 	charset="iso-8859-1"x+ Content-Transfer-Encoding: quoted-printable   1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">i <HTML> <HEAD>9 <META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =a charset=3Diso-8859-1">@ <META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
 5.5.2653.12">i4 <TITLE>RE: oracle/telnet-tcpip 5.3/vms 7.3-1</TITLE> </HEAD>  <BODY>  G <P><FONT SIZE=3D2>I also tried rsh and rlogin.&nbsp; No luck.&nbsp; I =gI tried set watch on both the sethost session and the telnet session, and =n@ compared the output of both - nothing apparent there.</FONT></P>  G <P><FONT SIZE=3D2>Didn't think about checking opcom, but the operator =a* log doesn't have anything about it.</FONT> </P>  D <P><FONT SIZE=3D2>It happened on 2 accounts, both privileged.</FONT> </P>  H <P><FONT SIZE=3D2>Also, I had to go back to the field test version (!) =B of 7.3-1 because they needed this environment, and everything is =E working fine now!&nbsp; (I had backed up the system disk before the =cH upgrade, of course, and going back to that system disk made everything = work again.)</FONT></P>u  G <P><FONT SIZE=3D2>Of course, I can't stay with the field test version =pI because of bugs in it that cause crashes once a month or so.&nbsp; But, =F! at least oracle works!</FONT></P>     <P><FONT SIZE=3D2>Thanks,</FONT> <BR><FONT SIZE=3D2>Ron</FONT>w </P>  8 <P><FONT SIZE=3D2>&gt; -----Original Message-----</FONT>1 <BR><FONT SIZE=3D2>&gt; From: Mike Zarudzki [<A =iI HREF=3D"mailto:mike.zarudzki@hp.com">mailto:mike.zarudzki@hp.com</A>]</F=0 ONT>@ <BR><FONT SIZE=3D2>&gt; Sent: Wednesday, August 21, 2002 11:22 =	 AM</FONT>s8 <BR><FONT SIZE=3D2>&gt; To: Info-VAX@Mvb.Saic.Com</FONT>B <BR><FONT SIZE=3D2>&gt; Subject: Re: oracle/telnet-tcpip 5.3/vms = 7.3-1</FONT> <BR><FONT SIZE=3D2>&gt; </FONT>r <BR><FONT SIZE=3D2>&gt; </FONT>eI <BR><FONT SIZE=3D2>&gt; oracle/telnet-tcpip 5.3/vms 7.3-1Have you tried =- RSH into it? </FONT>. <BR><FONT SIZE=3D2>&gt; Build a command</FONT>@ <BR><FONT SIZE=3D2>&gt; procedure that adds the set watch file = command</FONT>G <BR><FONT SIZE=3D2>&gt; and see what results, if any get logged.</FONT>t <BR><FONT SIZE=3D2>&gt; </FONT>AA <BR><FONT SIZE=3D2>&gt; I assume your not seeing anything OPCOM =. related?</FONT>n <BR><FONT SIZE=3D2>&gt; </FONT>FD <BR><FONT SIZE=3D2>&gt; I assume this is not just one account, but = accounts with </FONT>e0 <BR><FONT SIZE=3D2>&gt; various levels of</FONT>% <BR><FONT SIZE=3D2>&gt; privs?</FONT>n <BR><FONT SIZE=3D2>&gt; </FONT>IF <BR><FONT SIZE=3D2>&gt; Anything wierd on the telnet service itself? = File defined? </FONT>f0 <BR><FONT SIZE=3D2>&gt; Log opts defined?</FONT> <BR><FONT SIZE=3D2>&gt; </FONT> " <BR><FONT SIZE=3D2>&gt; HTH</FONT>' <BR><FONT SIZE=3D2>&gt; -Mike Z.</FONT>dA <BR><FONT SIZE=3D2>&gt; &quot;All opinions my own...&quot;</FONT>) <BR><FONT SIZE=3D2>&gt; </FONT>s <BR><FONT SIZE=3D2>&gt; </FONT>n5 <BR><FONT SIZE=3D2>&gt; &quot;Kaledas, Ronald&quot; =A0 &lt;RKaledas@dmc.org&gt; wrote in message</FONT> <BR><FONT SIZE=3D2>&gt; <A =I HREF=3D"news:B39B28A01B62D311B1360090277B3A58070F7827@tro-nt-exchg2.dmc.=e org" =I TARGET=3D"_blank">news:B39B28A01B62D311B1360090277B3A58070F7827@tro-nt-e=r xchg2.dmc.org</A>...</FONT>(B <BR><FONT SIZE=3D2>&gt; I'm having a strange problem that I hope = someone can help with.</FONT>eI <BR><FONT SIZE=3D2>&gt; In our test cluster, we were running vms E7.3-1 =c (field test) </FONT>- <BR><FONT SIZE=3D2>&gt; with tcpip 5.3</FONT>s) <BR><FONT SIZE=3D2>&gt; and oracle</FONT>mB <BR><FONT SIZE=3D2>&gt; 7.3.3.6 (and also oracle 8.1.7.3).&nbsp; =' Yesterday, I upgraded vms to the</FONT>a4 <BR><FONT SIZE=3D2>&gt; production version of</FONT>I <BR><FONT SIZE=3D2>&gt;&nbsp; V7.3-1.&nbsp; Nothing else changed.&nbsp; =c" When the cluster came back </FONT>- <BR><FONT SIZE=3D2>&gt; up, everything</FONT> % <BR><FONT SIZE=3D2>&gt; worked</FONT> H <BR><FONT SIZE=3D2>&gt;&nbsp; fine...except SVRMGRL and SQLPLUS.&nbsp; =  When you telnet into the </FONT>, <BR><FONT SIZE=3D2>&gt; node, or come</FONT>% <BR><FONT SIZE=3D2>&gt; in via</FONT>eG <BR><FONT SIZE=3D2>&gt;&nbsp; reflection X, invoking svrmgrl produces =  the following two </FONT>e/ <BR><FONT SIZE=3D2>&gt; lines (after the</FONT> . <BR><FONT SIZE=3D2>&gt; normal header):</FONT>F <BR><FONT SIZE=3D2>&gt; MGR-11401: input error, unable to read input = line</FONT>g@ <BR><FONT SIZE=3D2>&gt; MGR-01508: unable to close the current = file</FONT>rH <BR><FONT SIZE=3D2>&gt; Invoking sqlplus, without a username/password, = is supposed to </FONT>) <BR><FONT SIZE=3D2>&gt; prompt and</FONT> + <BR><FONT SIZE=3D2>&gt; wait for the</FONT>eD <BR><FONT SIZE=3D2>&gt;&nbsp; username.&nbsp; It comes up with the =! username prompt, and then </FONT>t0 <BR><FONT SIZE=3D2>&gt; immediately exits</FONT>* <BR><FONT SIZE=3D2>&gt; back to dcl</FONT>G <BR><FONT SIZE=3D2>&gt;&nbsp; without waiting.&nbsp; Invoking sqlplus =t  with a username/password </FONT>, <BR><FONT SIZE=3D2>&gt; gets into the</FONT>0 <BR><FONT SIZE=3D2>&gt; program, but then</FONT>G <BR><FONT SIZE=3D2>&gt;&nbsp; the sql&gt; prompt then keeps appearing =e and scrolling up the </FONT>/ <BR><FONT SIZE=3D2>&gt; screen as if you</FONT>s+ <BR><FONT SIZE=3D2>&gt; were holding</FONT> G <BR><FONT SIZE=3D2>&gt;&nbsp; down the RETURN key!&nbsp; So, it seems =a  to me these programs are </FONT>* <BR><FONT SIZE=3D2>&gt; acting like</FONT>/ <BR><FONT SIZE=3D2>&gt; sys$input is not</FONT>eE <BR><FONT SIZE=3D2>&gt;&nbsp; defined, or something like that.</FONT>rF <BR><FONT SIZE=3D2>&gt; Here's the kicker...everything works fine if = you SET HOST </FONT>- <BR><FONT SIZE=3D2>&gt; into the node.</FONT>n* <BR><FONT SIZE=3D2>&gt; SVRMGRL and</FONT>C <BR><FONT SIZE=3D2>&gt;&nbsp; SQLPLUS both work as expected.</FONT>t$ <BR><FONT SIZE=3D2>&gt; Argh!</FONT>G <BR><FONT SIZE=3D2>&gt; Everything else seems to work fine, any other =d program I run, just not</FONT>7 <BR><FONT SIZE=3D2>&gt; oracle.&nbsp; I've tried</FONT>iH <BR><FONT SIZE=3D2>&gt; logging in with /nocom, in case there was some = interaction </FONT>o0 <BR><FONT SIZE=3D2>&gt; there, I've tried</FONT>+ <BR><FONT SIZE=3D2>&gt; defining sys</FONT>tF <BR><FONT SIZE=3D2>&gt; $input, I've done this in oracle 7.3 and 8i, =  and I also booted off the</FONT>1 <BR><FONT SIZE=3D2>&gt; backup copy of the</FONT>-I <BR><FONT SIZE=3D2>&gt;&nbsp; system disk, to see if the problem was in =g oracle, but it </FONT>1 <BR><FONT SIZE=3D2>&gt; worked fine there!</FONT>wI <BR><FONT SIZE=3D2>&gt; It seems to be some interaction with TELNET and =< oracle, but the tcpip</FONT>. <BR><FONT SIZE=3D2>&gt; version DID NOT</FONT>H <BR><FONT SIZE=3D2>&gt; change!&nbsp; I also reinstalled tcpip 5.3 and = the mup for 5.3, </FONT>. <BR><FONT SIZE=3D2>&gt; but that didn't</FONT>/ <BR><FONT SIZE=3D2>&gt; change anything.</FONT>PF <BR><FONT SIZE=3D2>&gt; Oh, and I didn't have the mup in place right = away, so this </FONT> , <BR><FONT SIZE=3D2>&gt; happened both</FONT>/ <BR><FONT SIZE=3D2>&gt; with and without</FONT>- <BR><FONT SIZE=3D2>&gt;&nbsp; the mup.</FONT>(E <BR><FONT SIZE=3D2>&gt; I've logged a call, but Colorado support is =  also not sure </FONT>f0 <BR><FONT SIZE=3D2>&gt; what else to look</FONT>" <BR><FONT SIZE=3D2>&gt; at.</FONT>? <BR><FONT SIZE=3D2>&gt; Does anyone here have any ideas?</FONT>H& <BR><FONT SIZE=3D2>&gt; Thanks,</FONT>" <BR><FONT SIZE=3D2>&gt; Ron</FONT> <BR><FONT SIZE=3D2>&gt; </FONT>  <BR><FONT SIZE=3D2>&gt; </FONT>S </P>   </BODY>a </HTML>r) ------_=_NextPart_001_01C2493E.EA2B6E30--n   ------------------------------    Date: 21 Aug 2002 13:48:10 -0700* From: ken.randell@fortel.com (Ken Randell). Subject: Re: oracle/telnet-tcpip 5.3/vms 7.3-1= Message-ID: <8debc3ff.0208211248.4d7f8dcd@posting.google.com>h  3 Check your value for MAXBUF and tell us what is ...n  @ I think I have found an issue with C (it's in RMS I bet, not the? C-RTL) where if you do something simple (like getc() calls), ita@ appears to be broken if MAXBUF is set to a large (i.e., > 40000)$ value.  I think the default is 8192.   Ken Randelli  ~ "Kaledas, Ronald" <RKaledas@dmc.org> wrote in message news:<B39B28A01B62D311B1360090277B3A58070F7827@tro-nt-exchg2.dmc.org>...A > I'm having a strange problem that I hope someone can help with.  > M > In our test cluster, we were running vms E7.3-1 (field test) with tcpip 5.3 
 > and oracle nF > 7.3.3.6 (and also oracle 8.1.7.3).  Yesterday, I upgraded vms to the > production version ofeL >  V7.3-1.  Nothing else changed.  When the cluster came back up, everything > workedM >  fine...except SVRMGRL and SQLPLUS.  When you telnet into the node, or come  > in viaM >  reflection X, invoking svrmgrl produces the following two lines (after the- > normal header):1 > 3 > MGR-11401: input error, unable to read input linea- > MGR-01508: unable to close the current file  > J > Invoking sqlplus, without a username/password, is supposed to prompt and > wait for theN >  username.  It comes up with the username prompt, and then immediately exits
 > back to dcleL >  without waiting.  Invoking sqlplus with a username/password gets into the > program, but then M >  the sql> prompt then keeps appearing and scrolling up the screen as if youe > were holdingJ >  down the RETURN key!  So, it seems to me these programs are acting like > sys$input is not% >  defined, or something like that.  e > J > Here's the kicker...everything works fine if you SET HOST into the node.
 > SVRMGRL ando! >  SQLPLUS both work as expected.r >  > Argh!r > G > Everything else seems to work fine, any other program I run, just notp > oracle.  I've tried N > logging in with /nocom, in case there was some interaction there, I've tried > defining sysH > $input, I've done this in oracle 7.3 and 8i, and I also booted off the > backup copy of theN >  system disk, to see if the problem was in oracle, but it worked fine there! > G > It seems to be some interaction with TELNET and oracle, but the tcpips > version DID NOT L > change!  I also reinstalled tcpip 5.3 and the mup for 5.3, but that didn't > change anything.J > Oh, and I didn't have the mup in place right away, so this happened both > with and without >  the mup.  > M > I've logged a call, but Colorado support is also not sure what else to look  > at.e > " > Does anyone here have any ideas? > 	 > Thanks,  > Ron  >  > --   ------------------------------  % Date: Wed, 21 Aug 2002 14:58:58 -0400h- From: JF Mezei <jfmezei.spamnot@videotron.ca> 8 Subject: Re: Printing barcodes via postscript on OpenVMS, Message-ID: <3D63E2E8.EB313D30@videotron.ca>   "John.Malmberg" wrote:H > I am sure that they can be purchased from third parties.  However IIRCH > there is an example in one of the Adobe PostScript books that documentK > how to write your own.  The Adobe books are available at many bookstores.o  M There are freely downloadable versions of various barcode fonts.  As you saidtK though, you need to test them. In fact, with any bar code font you get, youuM absolutely need to test them to ensure that they are printed properly by your : printer/paper and readable by whatever equipment you have.  E > Assuming that your printer is a generic PostScript printer, you canoG > either have your application download the font to the printer, or youl? > can put the font in a text library, and use a printer form to C > automatically download the font to the printer as a setup module.i    K The proper way to do this is to create a postscript prologue which contains N not only the font definition, but also all other definitions for the pages youN will be printing. Then the application simply generates simple postscript that/ makes use of the stuff defined in the prologue.    when you print, you simply:i? print/queue=mypostscript  my_prologue.ps,my_application_data.psn  M This method makes it far easier to keep the prologue in the application area,-N and especially less of a hassle to update, especially during the testing stageJ since often the text libraries used by the queue manager are locked by the queue manager.  N The definition of a BAR 39 I obtained on the net takes only 200 lines of text,N including comments. So it isn't a big deal to just include it in the prologue.   ------------------------------  # Date: Wed, 21 Aug 2002 18:14:07 GMTM" From: Guy Peleg <guy.peleg@hp.com>2 Subject: Re: Problem with DCL procedure using PIPE& Message-ID: <3D63D6CA.2F237AC3@hp.com>   Without any commitment !!!!!!o   Rough estimate is V7.3-2   Larry Kilgallen wrote:  y > In article <01KLKGCPC3Q0970ARJ@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:eJ > >> The numerical number has not been assigned.  Schedules are subject toJ > >> change so posting any dates would not be to anyone's benefit.  Things5 > >> are not laid in stone yet, only set in jello ;*)a > > D > > Sure, the usual uncertainties.  But what about a ROUGH ESTIMATE? >,C > My outside rough estimate is that the number will be greater thani > 7.3-1 and less than 9.0 :-)e   ------------------------------  # Date: Thu, 22 Aug 2002 01:53:48 GMTu1 From: "David J. Dachtera" <djesys.nospam@fsi.net>s2 Subject: Re: Problem with DCL procedure using PIPE' Message-ID: <3D64492A.1CD4E9FA@fsi.net>n   Guy Peleg wrote: >  > Just as an FYI,  > N > The next version of OpenVMS (code name Opal) will extend the DCL buffer fromM > 255bytes to 4KB and the expansion buffer to 8KB. This will make the problemp > reported here go away.  C I just hope 4K is enough that it will take some time before someone F "blows the limit" again. With the advent of ODS-5, paths and filenames# can become truly ridiculously long.    -- u David J. Dachteraa dba DJE Systemsa http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/?   ------------------------------  # Date: Thu, 22 Aug 2002 01:54:48 GMTo1 From: "David J. Dachtera" <djesys.nospam@fsi.net>i2 Subject: Re: Problem with DCL procedure using PIPE' Message-ID: <3D644966.7225D810@fsi.net>9   Guy Peleg wrote: >  > Without any commitment !!!!!!r >  > Rough estimate is V7.3-2  - Back ported to ... ? (V5.5-2? Pretty-please?)r   -- l David J. Dachterai dba DJE Systemse http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/e   ------------------------------  # Date: Thu, 22 Aug 2002 05:09:30 GMT " From: Guy Peleg <guy.peleg@hp.com>2 Subject: Re: Problem with DCL procedure using PIPE& Message-ID: <3D647068.614FE748@hp.com>  > To support the change in the DCL image, we had to increase the' supervisor stack as well as modify RMS.   , So the answer is absolutely no back porting.   Guy      "David J. Dachtera" wrote:   > Guy Peleg wrote: > >f! > > Without any commitment !!!!!!  > >a > > Rough estimate is V7.3-2 >r/ > Back ported to ... ? (V5.5-2? Pretty-please?)u >i > -- > David J. Dachtera  > dba DJE Systemsl > http://www.djesys.com/ >w* > Unofficial Affordable OpenVMS Home Page:! > http://www.djesys.com/vms/soho/    ------------------------------  % Date: Wed, 21 Aug 2002 13:44:44 -0500 $ From: "Art Beane" <beane@petris.com>" Subject: Re: Proposal for DECUS US7 Message-ID: <002701c24942$d16c4c80$342810ac@petris.com>y   Bill Gunshannon contributed:  6   Will they revive the old memberships automatically??   Member# 368513    Card still in my wallet!!  :-)  > I just found mine last week:  http://198.170.183.110/decus.htm   ------------------------------  % Date: Wed, 21 Aug 2002 10:39:28 -0700v& From: Greg Cagle <gregc@gregcagle.com>L Subject: Re: Proposal for DECUS US (was: Fortune Magazine and a post-VMS rap, Message-ID: <3D63D050.3020809@gregcagle.com>   Larry Kilgallen wrote:  I > And if the "CETS" symposium can be jointly run by "Encompass" and "CSA" F > and "Compaq", certainly the "HPETS" conference can be jointly run by? > "DECUS" and "<HP successor to CSA>" and "<HP-UX user group>".e  I The <HP-UX user group> (Interex) already has a highly technical HP-UX andh Linux conference.i   -- v
 Greg Cagle gregc at gregcagle dot com   ------------------------------  % Date: Wed, 21 Aug 2002 15:02:16 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>e Subject: Re: Roadshow update, Message-ID: <3D63E3AE.375937C0@videotron.ca>   Sue Skonetski wrote:N > - First, the road show is being put on my the HP Americas U.S. Public Sector > Sales group.  L Sue, could you clarify what the "Public Sector" really means in terms of theA remaining VMS market niches ? Does this really mean "military" ? x  I Or is there a much wider audience targetted such as municipalities, watercJ works, state government, as well as all sectors of federal government from forests to national defense ?D   ------------------------------  % Date: Wed, 21 Aug 2002 20:53:08 +0200 2 From: Didier Morandi <Didier.Morandi@notsoFree.fr> Subject: Re: Roadshow update, Message-ID: <3D63E194.445CEC1F@notsoFree.fr>   JF Mezei wrote:  >  > Sue Skonetski wrote:P > > - First, the road show is being put on my the HP Americas U.S. Public Sector > > Sales group. > N > Sue, could you clarify what the "Public Sector" really means in terms of theB > remaining VMS market niches ? Does this really mean "military" ? > K > Or is there a much wider audience targetted such as municipalities, waterwL > works, state government, as well as all sectors of federal government from > forests to national defense ?m   as well as Europe :-)a   D. -- <1   ===> Remove notso before replying via mail <===i2 --------------------------------------------------2 MORANDI Consultants  http://Didier.Morandi.Free.fr0   19 chemin de la Butte, 31400 Toulouse, France.2 Tel.: +33 (0)6 7983 6418 - Fax: +33 (0)5 6154 19282 OpenVMS, APPLE, Computer Security, Migration plans2 --------------------------------------------------   ------------------------------    Date: 21 Aug 2002 12:06:07 -07002 From: lee.gleason@halliburton.com (Lee K. Gleason)) Subject: Re: Size of logical name table ? = Message-ID: <a74ee7e6.0208211106.68e223b2@posting.google.com>F  U Jan-Erik Sderholm <aaa@aaa.com> wrote in message news:<3D636592.E45A4F23@aaa.com>...B  : > I'v been searching on what system or process resourse or4 > quota limits the size of, or the number of entries5 > in, a logical name table. I'v seen that you can useN7 > the /QUOTA switch to limit the size, but if not used,25 > the docs says that the size is "unlimited", well...< > ' > Anyone seeing any problem with this ?r > Or any other drawback ?/  <   I wrote an article about the storage of logical names, and@ a utility to analyze how effective the hashing was, for the goodC old VAX Professional magazine. I'd be glad to email it along to ya,  if you like.   Lee K. Gleason N5ZMR Control-G ConsultantsF lgleason@houston.rr.com    ------------------------------  % Date: Wed, 21 Aug 2002 21:12:33 +020029 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>h) Subject: Re: Size of logical name table ?/& Message-ID: <3D63E621.67E1131@aaa.com>   Yes, that would be interesting.20 Or, why not just tell us, was it effective ? :-)6 And, do you think that your results rom then still are actual now ?  ? Control-G Consultants, hmm wait....., no, didn't ring a bell.../ :-)s   Jan-Erik Sderholm   "Lee K. Gleason" wrote:E > > >   I wrote an article about the storage of logical names, andB > a utility to analyze how effective the hashing was, for the goodE > old VAX Professional magazine. I'd be glad to email it along to ya,A > if you like. >  > Lee K. Gleason N5ZMR > Control-G Consultantse > lgleason@houston.rr.comO   ------------------------------    Date: 22 Aug 2002 05:23:32 +0800, From: Paul Repacholi <prep@prep.synonet.com>) Subject: Re: Size of logical name table ?r- Message-ID: <87hehngb3f.fsf@prep.synonet.com>>  ( Jan-Erik Sderholm <aaa@aaa.com> writes:  D > Yes, I saw no problem in actualy *using* the method, just checkingA > there is no problem with a large number of logicals in the sameo
 > table...  ; There used to be some nit if you got over half a million orN! so. How do you define `large'? ;)      -- l< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.E@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Thu, 22 Aug 2002 04:35:41 GMTn- From: "Lee Gleason" <lgleason@houston.rr.com> ) Subject: Re: Size of logical name table ? ; Message-ID: <xSZ89.85288$eK6.2842131@twister.austin.rr.com>F  3 "Jan-Erik Sderholm" <aaa@aaa.com> wrote in messagen  news:3D63E621.67E1131@aaa.com...! > Yes, that would be interesting.:2 > Or, why not just tell us, was it effective ? :-)8 > And, do you think that your results rom then still are > actual now ? >e  K   Let's see, I recall, that it was usually middlin' effective, but the sizet of theH tables could be sized too small if you had a large number of names, thusK causing excess collisions. Also, due to the fact that all shareable logicaliH names are located by the same hash table, the logicals that are job wideK for everyone all wind up on the same entry, so the chain off the index  form? things like sys$scratch wind up being very long, and degenerate D into a linear search. I just ran this on an Alpha, and it looks like2 things haven't changed in this area o' VMS lately.  K   Since you'll have a goodly slug of names, upping the value of LNMSHASHTBLsG in sysgen might be good...(I realize, in these days of monstrously fastbG processors & memory, this ain't an area to worry about much, but, heck,<L we should always try to make things as efficient as possible...it's a way of life.)   Lee K. Gleason N5ZMR Control-G Consultantse lgleason@houston.rr.comt   ------------------------------  % Date: Wed, 21 Aug 2002 18:29:24 -0000f$ From: joshua@www.iocc.com (Franklin)( Subject: telnet from *NIX to OpenVMS 7.2, Message-ID: <slrnam7n06.i1q.joshua@iocc.com>  L OK. there's an OpenVMS 7.2 machine set up running a library card catalog for< a couple of universities. The interface is all telnet-based.K It appears to work relatively well from Windows (HyperTerm, cmd.exe telnet,NH TeraTerm, etc.) with the exception of sending a form-feed to the default	 printer. w  G However, I'm having a problem with a telnet from linux hosts, I believe3H because the interface sends a ^O control character. Has anyone run into G this kind of problem before? I'm not the admin of the machine, but if I,J could get some clear steps I could forward them to the appropriate person.K Alternatively, if anyone knows of some workarounds on the linux end I couldm put together an FAQ for it.    Thanks!F -- I Joshua Daniel Franklin Network AdministratorN IOCC.COM   ------------------------------    Date: 21 Aug 2002 16:45:44 -0400& From: fdc@columbia.edu (Frank da Cruz), Subject: Re: telnet from *NIX to OpenVMS 7.21 Message-ID: <ak0u5o$2ii$1@watsol.cc.columbia.edu>=  , In article <slrnam7n06.i1q.joshua@iocc.com>,% Franklin <joshua@www.iocc.com> wrote:IN : OK. there's an OpenVMS 7.2 machine set up running a library card catalog for> : a couple of universities. The interface is all telnet-based.M : It appears to work relatively well from Windows (HyperTerm, cmd.exe telnet,OJ : TeraTerm, etc.) with the exception of sending a form-feed to the default : printer.   : I : However, I'm having a problem with a telnet from linux hosts, I believe&J : because the interface sends a ^O control character. Has anyone run into I : this kind of problem before? I'm not the admin of the machine, but if I&L : could get some clear steps I could forward them to the appropriate person.M : Alternatively, if anyone knows of some workarounds on the linux end I could- : put together an FAQ for it.  : F Why is it sending a Ctrl-O character?  To "shift out" from ASCII to anC 8-bit character set such as the Right Half of ISO 8859-1 on a 7-bitt> connection?  Then you need a Unix telnet program that supports/ (a) character sets, and (b) Shift-In/Shift-Out:U  -   http://www.columbia.edu/kermit/ckermit.htmld   - Franki   ------------------------------  % Date: Wed, 21 Aug 2002 22:39:44 -0000l$ From: joshua@www.iocc.com (Franklin), Subject: Re: telnet from *NIX to OpenVMS 7.2, Message-ID: <slrnam85li.r48.joshua@iocc.com>  F In article <ak0u5o$2ii$1@watsol.cc.columbia.edu>, Frank da Cruz wrote:. > In article <slrnam7n06.i1q.joshua@iocc.com>,' > Franklin <joshua@www.iocc.com> wrote:nO >: OK. there's an OpenVMS 7.2 machine set up running a library card catalog fore? >: a couple of universities. The interface is all telnet-based. N >: It appears to work relatively well from Windows (HyperTerm, cmd.exe telnet,K >: TeraTerm, etc.) with the exception of sending a form-feed to the defaultc >: printer.  >:  J >: However, I'm having a problem with a telnet from linux hosts, I believeK >: because the interface sends a ^O control character. Has anyone run into gJ >: this kind of problem before? I'm not the admin of the machine, but if IM >: could get some clear steps I could forward them to the appropriate person. N >: Alternatively, if anyone knows of some workarounds on the linux end I could >: put together an FAQ for it. u >: h) > Why is it sending a Ctrl-O character?  e  G Very good question. Perhaps I was unclear... the telnet session begins /6 normally, I get the VMS banner. Then the card catalog L ("Information Gateway  --  V2.5-2 Copyright 2000, Data Research Associates, I Inc.  All Rights Reserved.") starts up.  It is this card catalog program nA that is sending control characters such as ^O. I don't know why.  O It is supposed to be an ANSI interface, I think, though TERM settings of VT100, . ANSI, etc. don't seems to make any difference)  ! > To "shift out" from ASCII to an E > 8-bit character set such as the Right Half of ISO 8859-1 on a 7-bit @ > connection?  Then you need a Unix telnet program that supports1 > (a) character sets, and (b) Shift-In/Shift-Out:e > / >   http://www.columbia.edu/kermit/ckermit.htmlf > 	 > - Frankc  ; Thanks for the tip, but ckermit exhibits the same behavior.,   --   Joshua Daniel Franklin Network Administratorn IOCC.COM   ------------------------------  % Date: Wed, 21 Aug 2002 22:53:12 -0400s2 From: rdeininger@mindspring.com (Robert Deininger), Subject: Re: telnet from *NIX to OpenVMS 7.2K Message-ID: <rdeininger-2108022253130001@1cust211.tnt2.nashua.nh.da.uu.net>t  @ In article <slrnam85li.r48.joshua@iocc.com>, joshua@www.iocc.com (Franklin) wrote:s  o* >> Why is it sending a Ctrl-O character?   > H >Very good question. Perhaps I was unclear... the telnet session begins 7 >normally, I get the VMS banner. Then the card catalog iM >("Information Gateway  --  V2.5-2 Copyright 2000, Data Research Associates,  J >Inc.  All Rights Reserved.") starts up.  It is this card catalog program B >that is sending control characters such as ^O. I don't know why. P >It is supposed to be an ANSI interface, I think, though TERM settings of VT100,/ >ANSI, etc. don't seems to make any difference)r  I Perhaps the VMS system is executing a SET TERMINAL/INQUIRE command?  This H command sends some escape sequences, which a standard-compliant terminalG would answer with some identifying information.  Few, if any, PC-hosted " terminal emulators get this right.  B I just did a SET TERM/INQUIRE on my VMS 7.3 system, with the VT420I terminal set to display the control characters.  The output was something  like <esc>[c<esc>\<esc>Z<esc>[0c<cr> G then VMS timed out, since the terminal didn't respond.  Dunno if any ofn! that matches what you are seeing.i  C Bottom line, if VMS is sending odd control characters, something isp telling it to do so.   ------------------------------  # Date: Thu, 22 Aug 2002 03:21:53 GMTi! From: Andy <acs@fcgnet.works.net>f, Subject: Re: telnet from *NIX to OpenVMS 7.2> Message-ID: <Xns9271ED5F28AF4acsfcgnetworksnet@216.166.71.232>  2 joshua@www.iocc.com (Franklin) enlightened us with% news:slrnam85li.r48.joshua@iocc.com: t  A > In article <ak0u5o$2ii$1@watsol.cc.columbia.edu>, Frank da Cruze	 > wrote: e/ >> In article <slrnam7n06.i1q.joshua@iocc.com>,e( >> Franklin <joshua@www.iocc.com> wrote:? >>: OK. there's an OpenVMS 7.2 machine set up running a library ? >>: card catalog for a couple of universities. The interface isi= >>: all telnet-based. It appears to work relatively well fromf@ >>: Windows (HyperTerm, cmd.exe telnet, TeraTerm, etc.) with the= >>: exception of sending a form-feed to the default printer.   >>: A >>: However, I'm having a problem with a telnet from linux hosts,cA >>: I believe because the interface sends a ^O control character..@ >>: Has anyone run into this kind of problem before? I'm not the? >>: admin of the machine, but if I could get some clear steps I3@ >>: could forward them to the appropriate person. Alternatively,A >>: if anyone knows of some workarounds on the linux end I could    >>: put together an FAQ for it.   A > Very good question. Perhaps I was unclear... the telnet sessione? > begins normally, I get the VMS banner. Then the card catalog *A > ("Information Gateway  --  V2.5-2 Copyright 2000, Data ResearchHA > Associates, Inc.  All Rights Reserved.") starts up.  It is this A > card catalog program that is sending control characters such as-A > ^O. I don't know why. It is supposed to be an ANSI interface, IdA > think, though TERM settings of VT100, ANSI, etc. don't seems toC > make any difference) e  A (I didn't know they (DRA) were still doing VMS-based card catalogt@ software though from the looks of it ( http://www.dra.com ) theyB appear to have merged with someone else and the DRA product is now called DRA Classic.)  A Have you tried DRA support ? This software has been around a longn@ time so they are bound to have run into similar problems before.; Or asked the libraries that you are having problems with toe5 contact DRA support ? That should be the first step. r   -Andy- -- t   ------------------------------    Date: 22 Aug 2002 05:27:02 +0800, From: Paul Repacholi <prep@prep.synonet.com>) Subject: Re: total VMS newbie - pointers?u- Message-ID: <87bs7vgaxl.fsf@prep.synonet.com>   * Nic Clews <sendspamhere@127.0.0.1> writes:  I > Seriously, in the UK, CPC sell the flat 6 way cable, MMJs and crimpers.u  e > www.cpc.co.uk   , Nic, does BT still use MMJs for phone lines?   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.y@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Wed, 21 Aug 2002 22:28:02 -040032 From: rdeininger@mindspring.com (Robert Deininger)) Subject: Re: total VMS newbie - pointers?eK Message-ID: <rdeininger-2108022228090001@1cust211.tnt2.nashua.nh.da.uu.net>o  5 In article <3D63C018.AC785684@videotron.ca>, JF Mezeie% <jfmezei.spamnot@videotron.ca> wrote:t   >re: finding MMJ connectorss  J I expect you can still buy MMJ stuff from HP, since they are still selling  new alphaservers with MMJ ports.  N Getting prices and part numbers is likely just as hard as it was under Compaq.   ------------------------------   Date: 21 Aug 2002 18:37:03 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)- Subject: Re: VAX/Alpha Basic example programsP* Message-ID: <ak0mkf$iu9$4@web1.cup.hp.com>  | In article <79DE14DAA1A8D3119B6E00805FEB980BAE418E@zakary.turpin-distribution.com>, David Gray <GrayD@turpinltd.com> writes: :b :Greeting all, d :iJ :I find myself in the position of needing to write a routine in VAX-BASIC.L :Quite honestly, it's been sometime since I have done this and was wonderingH :if anyone knows of a good site with some example programs.  The OpenVMSK :documentation site gives small code snippets but not what I'm looking for..  =   From the OpenVMS Frequently Asked Questions (FAQ) document:"  D           __________________________________________________________4           10.1  Can I have an example of calling...?  J                    Please use the available SEARCH command on OpenVMS, andE                    please remember to search the available resources,hD                    including the support databases and the newsgroupI                    archives. Please also realize that most OpenVMS system J                    services use similar calling sequences, meaning that anJ                    example of calling sys$getjpi can be used as an exampleI                    for sys$getsyi and sys$getdvi. Students: please do not F                    expect folks to write your homework for you. As for$                    search resources:  4                    o  SEARCH SYS$EXAMPLES:*.* target  6                    o  SEARCH TCPIP$EXAMPLES:*.* target  -                    o  http://askq.compaq.com/l  =                    o  http://www.openvms.compaq.com/freeware/e  ,                    o  http://www.google.com/  C                    OpenVMS programming documentation, including thenH                    numerous example programs found in recent versions ofI                    the OpenVMS Programming Concepts manual, is available:i  9                    o  http://www.openvms.compaq.com:8000/b  8                    o  http://www.openvms.compaq.com/doc/  ?                    o  http://www.openvms.compaq.com/commercial/M  J                    As for details of argument passing, most OpenVMS systemE                    services and RTL routines pass string arguments byeD                    descriptor. Languages which support native stringE                    data types create descriptors automatically; those>E                    which do not (eg., C) require that you set them up-G                    explicitly. For further details on using descriptors.H                    and particularly for using descriptors from C, please%                    see Section 10.12.,   		...o      N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Wed, 21 Aug 2002 16:39:53 -0700>2 From: "Randy Park" <rjpark@mindspring.nospaam.com>- Subject: Re: VAX/Alpha Basic example programst3 Message-ID: <ak18cf$jbk$1@nntp9.atl.mindspring.net>   1 David Gray <GrayD@turpinltd.com> wrote in messagetL news:79DE14DAA1A8D3119B6E00805FEB980BAE418E@zakary.turpin-distribution.com.. .< >  > Greeting all,  > K > I find myself in the position of needing to write a routine in VAX-BASIC.tC > Quite honestly, it's been sometime since I have done this and wase	 wonderinggI > if anyone knows of a good site with some example programs.  The OpenVMSaL > documentation site gives small code snippets but not what I'm looking for. >sH > I'm trying to include an .RRD file in the basic program as output fromI > Oracle RMU/UNLOAD.  We have BASIC but nobody has ever used it, hence noc8 > source code to look through.  We also do NOT have CDD. >  > XDS_APACHE> typ cw.rrd5 > DEFINE FIELD LAST_NAME DATATYPE IS TEXT SIZE IS 14.t6 > DEFINE FIELD FIRST_NAME DATATYPE IS TEXT SIZE IS 10.0 > DEFINE FIELD DBKEY DATATYPE IS TEXT SIZE IS 8. > DEFINE RECORD CANDIDATES_VW. >    LAST_NAME . >    FIRST_NAME .  > END CANDIDATES_VW RECORD.K >l > Thanks in advance, > David. >a > OpenVMS 7.1 (alpha)e > DEC/COMPAQ/HP BASIC V1.3 >l  : Since you seem to be interested in BASIC with Rdb, here is5 a couple of include files I have used for 10 years toa: declare and define SQLCA and SQLDA.  I'm assuming you know" what SQLDA and SQLCA are used for.  5 The RECORD - END RECORD pair and the lines in between 7 are analogous to the DEFINE RECORD - END RECORD pair ine your .RRD file.u  9 The MAP statement statically allocates memory and assignsg7 it a PSECT name to whatever is between the parenthesis..; The COMMON statement is similar to the MAP, read the manuale$ for the differences between the two.  = To use a data item defined in a record, such as the SQLCA_REC  record, do something like this:u!         DECLARE LONG SQL_ERR_CODEo%         SQL_ERR_CODE = SQLCA::SQLCODE < A more complex example such as retrieving all the data types: and all data lengths, of a recently prepared/described SQL statement might be:          DECLARE LONG LOOPu<         DIM WORD MY_TYPE(SQL_MAX_COLS), MY_LEN(SQL_MAX_COLS)#         FOR LOOP = 1 TO SQLDA::SQLD 6           MY_TYPE(LOOP) = SQLDA::SQLVAR(LOOP)::SQLTYPE5           MY_LEN(LOO)   = SQLDA::SQLVAR(LOOP)::SQLLENm         NEXT LOOPl  
 Good Luck.    =         !****************************************************x'         !               X U _ S Q L D Ac=         !****************************************************e)  DECLARE LONG CONSTANT SQL_MAX_COLS = 500 %         RECORD RDB$MESSAGE_VECTOR_RECu+                 LONG   RDB$LU_NUM_ARGUMENTS $                 LONG   RDB$LU_STATUS/                 LONG   RDB$LU_ARGUMENT(1 TO 18)c)         END RECORD RDB$MESSAGE_VECTOR_REC,         RECORD SQLDA_REC"                 STRING SQLDAID = 8                 LONG   SQLDABC                 WORD   SQLN.                 WORD   SQLDl/                 GROUP SQLVAR(1 TO SQL_MAX_COLS)m$                         WORD SQLTYPE#                         WORD SQLLENh$                         LONG SQLDATA#                         LONG SQLINDd'                         WORD SQLNAMELENo+                         STRING SQLNAME = 30,                  END GROUP SQLVAR         END RECORD SQLDA_REC  >   MAP (SQLDA) SQLDA_REC SQLDA, WORD SQLNULL(1 TO SQL_MAX_COLS)?  COMMON (RDB$MESSAGE_VECTOR) RDB$MESSAGE_VECTOR_REC RDB$MSG_VECe    ;         !**************************************************a'         !               X U _ S Q L C As;         !**************************************************'         RECORD SQLCA_REC"                 STRING SQLCAID = 8                 LONG   SQLCABC                 LONG   SQLCODE                 WORD   SQLERRML2$                 STRING SQLERRMC = 70&                 LONG   SQLERRD(1 TO 6)*                 STRING SQLWARN(0 TO 7) = 1!                 STRING SQLEXT = 8V         END RECORD SQLCA_REC    MAP (SQLCA) SQLCA_REC SQLCA   ------------------------------    Date: 22 Aug 2002 05:16:18 +0800, From: Paul Repacholi <prep@prep.synonet.com>& Subject: Re: VMS to OpenVMS conversion- Message-ID: <87lm6zgbfh.fsf@prep.synonet.com>f  + "James M. Knox" <jknox@trisoft.com> writes:e  E > I think part of it is "terminology."  In the lab we can do anything A > we want.  However, we would have to buy several old VAX-8600's,bF > complete with disk clusters and star couplers, in order to replicateD > the key parts of the existing system.  The existing system can not= > be touched until any replacement has been shown to be fully ; > operational - that is a good year's effort all by itself.   E But you do not replicate the system! Step one, is to get all the code-B to current OS and current *compilers*. When it is clean, and looksF good on your regression tests, then port the clean code to Alpha. ThisA way, problem with moving to up to date compilers and runtimes arejE sepereated from Vax -> Alpha problem. Change one thing at a time, esp 2 if the documentation is less than totaly complete!  C > So we *could* upgrade the VAX systems to 7.3, but it would take aa? > year or more, much money, (would blow our schedule), and justmE > doesn't appear to make sense -- spending all that time and money too6 > validate something that we have no real interest in.  @ You are not out to put the new vax code into production. It is a? launching point to get first stage done. Doing it this way willtB probably save you time and money. It may seem the long way around,@ but it is the quickest, unless you score a totally no-problem at all port...e   -- .< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.h@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Wed, 21 Aug 2002 22:38:48 -0400p2 From: rdeininger@mindspring.com (Robert Deininger)& Subject: Re: VMS to OpenVMS conversionK Message-ID: <rdeininger-2108022238480001@1cust211.tnt2.nashua.nh.da.uu.net>u  < In article <87lm6zgbfh.fsf@prep.synonet.com>, Paul Repacholi <prep@prep.synonet.com> wrote:  , >"James M. Knox" <jknox@trisoft.com> writes: > F >> I think part of it is "terminology."  In the lab we can do anythingB >> we want.  However, we would have to buy several old VAX-8600's,G >> complete with disk clusters and star couplers, in order to replicateeE >> the key parts of the existing system.  The existing system can noty> >> be touched until any replacement has been shown to be fully< >> operational - that is a good year's effort all by itself. >pF >But you do not replicate the system! Step one, is to get all the codeC >to current OS and current *compilers*. When it is clean, and lookscG >good on your regression tests, then port the clean code to Alpha. ThisdB >way, problem with moving to up to date compilers and runtimes areF >sepereated from Vax -> Alpha problem. Change one thing at a time, esp3 >if the documentation is less than totaly complete!p  H I agree.  Even if the software is never _validated_ on VAX 7.3, it mightF be worthwhile to build it and test it on VAX with recent compilers and< OS.  Then finish by porting to Alpha, and do the validation.  I Debugging might be more efficient on VAX, even if the final validation is  done on Alpha.  H In a all-at-once port to alpha, you'll be changing the hardware, the OS,F and the compiler(s).  It won't work at first, and you'll have to worry about everything at once.n  E Updating the software on VAX might really be a bad idea, but it's notf obvious to me at first glance.   ------------------------------    Date: 21 Aug 2002 12:44:26 -06002 From: cochrane@encompasserve.org (Arthur Cochrane)3 Subject: Re: VMS upgrade with shadowed system disk. 3 Message-ID: <9djtaRGBixxH@eisner.encompasserve.org>m   > For instance:m > O > you have two machines. Each is set to boot from their own drive which happensrJ > to form a shadowset with each other. Machine 1 would boot first, not seeP > machine 2's system drive and thus form a one member DSA10 shadowset. Machine 2M > would boot second, would it immediately see machine 1's drive (AND existingSP > DSA10) or would MSCP served devices appear much later in the boot process ? IfP > it doesn't see the machine 1 ,s drives right away, it would form its own DSA10G > shadowset with its own drive, and what happens later ? You'd have thee( > equivalent of a partitioned shadowset. > M > I guess in such a case, would the only safe method be to always have one ofrO > the 2 machines boot from machine 1 (via ethernet) ? This way, it would ensurelI > that system drives are not partitioned into two independent shadowsets.s  '     The way the above to work would be:'  H     Node 1 boots from its local disk (VOTE=1, EXPECT=1) and forms single"     member shadow set system disk.  H     Node 2 boots as satellite (VOTE=0) from NODE 1 and in the startup itE     adds a local disk to the system disk shadow set and a shadow copy,     follows.  F     Node 1 crashes or shutdowns the Node 2 hangs until Node 1 reboots,7     system disk may have to merge copy (not sure here)..  M     In disaster crash Node 2 conversationally and set cluster=0 or vote=1 andsE     boot from local disk (root that Node 2 used when it was satellite      booted).  I     I think system disk shadow set mounting is done way before MSCP couldu     serve the disks.   ------------------------------  % Date: Wed, 21 Aug 2002 22:14:19 -0400c2 From: rdeininger@mindspring.com (Robert Deininger)3 Subject: Re: VMS upgrade with shadowed system disk.2K Message-ID: <rdeininger-2108022214190001@1cust211.tnt2.nashua.nh.da.uu.net>*  5 In article <3D63CAAE.98FD20F5@videotron.ca>, JF Mezeit% <jfmezei.spamnot@videotron.ca> wrote:i  O >Not quite related to shadowing itself, but on alpha, when you boot from CD and M >execute the upgrade from it, is it able to do an autogen of your real system> >disk ?    No, not as far as I know..  B >Or is that accomplished as a second phase of an upgrade when yourM >system is rebooted from its real system disk with a different /STARTUP whichu) >executes the remainder of the upgrade ? n  G Yes.  The first boot is a special "upgrade" boot, which does an Autogen D among other things.  On a system disk with multiple roots that bootsH multiple nodes, each node does the upgrade boot in turn, adjusting stuff in its own root.  N >In the later case, having the right settings in modparams and sysgen in terms4 >of shadowed sytstem drives would be very important.  G Yes.  The upgrade instructions explain how to disable shadowing for thep/ duration of the upgrade, and then re-enable it.e  I JF, don't you have a hobbyist alpha system?  If not, you should hunt downd a nice inexpensive one on ebay.A   ------------------------------  % Date: Thu, 22 Aug 2002 01:41:41 -0000s/ From: Michael Zarlenga <zarlenga@conan.ids.net>e- Subject: Re: Why C is better than Fortran 95?-/ Message-ID: <um8galrdo9d480@corp.supernews.com>i  . Didier Morandi <Didier.Morandi@Free.fr> wrote:# : So, why is C having such success?    More skilled programmers.   D What Fortran can do only now, and only on some platforms, C has been able to do for a decade.    A : To be able to execute embedded macro code? Who does this today?e  G Anyone who wants to use the RPCC on a Alpha box to accurately benchmarkt code.o   ------------------------------  % Date: Wed, 21 Aug 2002 13:01:54 -0700 , From: "Scott Stark" <starkh@saic.com_nospam>( Subject: Re: XFC v2 ECO, negative report' Message-ID: <3d638f43$1@cpns1.saic.com>a  L Any update?  Based upon the original post, does this seem to be SMP related?L I've successfully installed on a two-node DS10 cluster (not SMP) and haven't# seen any of the problems as listed.A, <LBohan@dbc.spam_less..com> wrote in message2 news:646olu8o9ti2qb2jktdt3ve5r0tvv2mqjp@4ax.com...F > On 15 Aug 2002 15:48:00 -0600, young_r@encompasserve.org (Rob Young) > wrote:8 > > Out of curiousity, you did first apply the mandatory > > VMS73_UPDATE-V0100 ? > > Robi > yes." > on this 'ticular system, so far: >,C > DEC AXPVMS VMS73_XFC V2.0      Patch Install 13-AUG-2002 21:09:37MC > DEC AXPVMS VMS73_PTHREAD V3.0  Patch Install 13-AUG-2002 21:08:56 C > DEC AXPVMS VMS73_DDTM V1.0     Patch Install 28-APR-2002 08:07:42lC > DEC AXPVMS VMS73_SYS V4.0      Patch Install 28-APR-2002 08:03:02 C > DEC AXPVMS VMS73_RMS V3.0      Patch Install 28-APR-2002 07:58:58tC > DEC AXPVMS VMS73_MANAGE V1.0   Patch Install 28-APR-2002 07:53:44-C > DEC AXPVMS VMS73_INIT V1.0     Patch Install 28-APR-2002 07:49:01dC > DEC AXPVMS VMS73_RTPAD V1.0    Patch Install 28-APR-2002 07:44:28wC > DEC AXPVMS VMS73_ACRTL V2.0    Patch Install 28-APR-2002 07:40:37nC > DEC AXPVMS VMS73_DCL V2.0      Patch Install 28-APR-2002 07:28:54aC > DEC AXPVMS VMS73_BACKUP V1.0   Patch Install 28-APR-2002 07:23:10MC > DEC AXPVMS VMS73_CLUSTER V2.0  Patch Install 19-JAN-2002 21:21:44 C > DEC AXPVMS VMS73_APB V1.0      Patch Install 19-JAN-2002 21:17:41rC > DEC AXPVMS VMS73_SYSINI V1.0   Patch Install 19-JAN-2002 21:14:10 C > DEC AXPVMS VMS73_DRIVER V2.0   Patch Install 19-JAN-2002 21:10:24rC > DEC AXPVMS VMS73_UPDATE V1.0   Patch Install 19-JAN-2002 21:06:31pC > DEC AXPVMS VMS73_LIBRTL V1.0   Patch Install 19-JAN-2002 20:59:33>   ------------------------------   Date: 21 Aug 2002 21:41:02 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)( Subject: Re: XFC v2 ECO, negative report* Message-ID: <ak11de$npk$1@web1.cup.hp.com>  V In article <3d638f43$1@cpns1.saic.com>, "Scott Stark" <starkh@saic.com_nospam> writes:M :Any update?  Based upon the original post, does this seem to be SMP related?OM :I've successfully installed on a two-node DS10 cluster (not SMP) and haven'tf$ :seen any of the problems as listed.  F   The OpenVMS engineer most familiar with XFC is directly working withC   the reporting customer site, and the engineer is actively pursing 3   details of the configuration and local operation.v  H   There is presently an attempt underway to replicate this error here inE   OpenVMS Engineering, having built a configuration with the releventpF   OpenVMS components and particularly having downloaded and configuredG   the other locally-relevent privileged-mode software component that isi   apparently relevent here.g  F   To date, we have not managed to replicate the reported error here inF   OpenVMS Engineering.  There is, however, a working theory as to what<   might be occuring within this site-specific configuration.  G   This does NOT presently appear to be a generic problem with XFC.  The I   OpenVMS V7.3-1 systems in the main engineering clusters as well as "my"T>   local V7.3-1 systems are all still running with XFC enabled.  K   As relevent details are known and particularly when the guilty (software) J   component (XFC or otherwise) is confirmed, more information will follow.   	--i  J   This may provide yet more fodder for my upcoming OpenVMS Hints and Kinks   presentations, too. :-)       N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   End of INFO-VAX 2002.461 ************************yeD                    descriptor. Languages which support native stringE                    data types create descriptors automatically; those>E                    which do not (eg., C) require that you set them up-G                    explicitly. For further details on using descriptors.H                    and particularly for using descriptors from C, please%                    see Section 10.12.,   		...oN    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N    N     O    O    O    O    O    O    O    O    O    	O    
O    O    O    
O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O     O    !O    "O    #O    $O    %O    &O    'O    (O    )O    *O    +O    ,O    -O    .O    /O    0O    1O    2O    3O    4O    5O    6O    7O    8O    9O    :O    ;O    <O    =O    >O    ?O    @O    AO    BO    CO    DO    EO    FO    GO    HO    IO    JO    KO    LO    MO    NO    OO    PO    QO    RO    SO    TO    UO    VO    WO    XO    YO    ZO    [O    \O    ]O    ^O    _O    `O    aO    bO    cO    dO    eO    fO    gO    hO    iO    jO    kO    lO    mO    nO    oO    pO    qO    rO    sO    tO    uO    vO    wO    xO    yO    zO    {O    |O    }O    ~O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    O    