1 INFO-VAX	Wed, 16 Apr 2003	Volume 2003 : Issue 210       Contents:& Re: Day Light Savings for VMS and UNIX) DCL question - null byte in string symbol - Re: DCL question - null byte in string symbol - Re: DCL question - null byte in string symbol - Re: DCL question - null byte in string symbol - Re: DCL question - null byte in string symbol - Re: DCL question - null byte in string symbol  Re: DECnet hacker?% Re: Excellent new OpenVMS Testimonial  Re: Fiber channel question Re: Fiber channel question HP in the financial news RE: HP in the financial news Re: HP in the financial news Re: HP in the financial news HSZ50 	 Re: HSZ50 	 Re: HSZ50 	 Re: HSZ50 	 Re: HSZ50  HTML favourite editor? Re: HTML favourite editor? Re: HTML favourite editor? Re: HTML favourite editor? Re: HTML favourite editor? Re: HTML favourite editor? Re: HTML favourite editor? Re: HTML favourite editor? Re: HTML favourite editor? Re: HTML favourite editor? Re: HTML favourite editor? Re: HTML favourite editor? Re: HTML favourite editor?' Re: Need to buy 8 x VAX 4000 or similar I RE: Newsgroup posting conventions, was: RE: VMS Software Product & Online ' Re: OpenVMS and Security - REALLY !!!!! P Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha RetaiP Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha RetaiP Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha RetaiP Re: OpenVMS.org: Marvel article and HP's press release for MarvelandAlpha Retain Re: Process Quota Tuning Q - Printer Problems# setting timeout on socket_read (MU) ' Re: setting timeout on socket_read (MU)  software maintenance RE: software maintenance Re: software maintenance& Re: TCPIP Services and IP masquerading& Re: TCPIP Services and IP masquerading& Re: TCPIP Services and IP masquerading Re: The value of documentation Re: TPU: SHOW BUFFER code  Re: What is a VMS Cluster  RE: What is a VMS Cluster  Re: What is a VMS Cluster  Re: What is a VMS Cluster  Re: What is a VMS Cluster  Re: What is a VMS Cluster  Re: What is a VMS Cluster  Re: What is a VMS Cluster  Re: What is a VMS Cluster  Re: What is the downside?  Re: What is the downside? P Re: Why did Digital go broke? (Was: Re: VMS Source Listing CD - Table of content Re: XDM problems (TCPIP 5.3)1 [OpenVMS Alpha] How to get image version in DCL ? 5 Re: [OpenVMS Alpha] How to get image version in DCL ? 5 Re: [OpenVMS Alpha] How to get image version in DCL ? 5 Re: [OpenVMS Alpha] How to get image version in DCL ? 5 Re: [OpenVMS Alpha] How to get image version in DCL ? 5 Re: [OpenVMS Alpha] How to get image version in DCL ? 5 Re: [OpenVMS Alpha] How to get image version in DCL ?   F ----------------------------------------------------------------------  % Date: Wed, 16 Apr 2003 09:45:46 +0100 ( From: Nic Clews <sendspamhere@127.0.0.1>/ Subject: Re: Day Light Savings for VMS and UNIX ) Message-ID: <3E9D183A.273C5EBA@127.0.0.1>    "David J. Dachtera" wrote:  J > The full hour adjustment doesn't really make sense, but it may have been? > the best political compromise ol' Ben could manage, given his H > contemporaries. A thirty-minute shift, once, permanently may make more3 > sense, but I've not analyzed it in global detail.   B Isn't it because there's a "skew", and when you revert back to nonE daylight savings, you actually gain more light at the end of the day, H certainly at latitude 53.5N I recall "when the clocks went back" you got) a little longer daylight in the evenings.   G In North Scotland, some locals are less happy with the current state of  controversial clock shifting.   H I suppose today with current technology capability, we could have clocksF which can track the tropical movement, and "midday" 12:00 could alwaysH be when the sun is at its highest for you in your locality. However they9 would skew at different rates dependent on your location.    Just musing. --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot com    ------------------------------  % Date: Wed, 16 Apr 2003 11:13:09 -0400  From: norm.raphael@metso.com2 Subject: DCL question - null byte in string symbol? Message-ID: <OFF16CBE7A.94AB9BF2-ON85256D0A.00519FD8@metso.com>    Is this expected behavior?  I Note: I am having some difficulties in general with the fact that when an J expression used as one term of a relation in an IF test fails to evaluate, the IF statement fails "ugly".   $ type null_test.com $ null1[0,8]=0 $ show symbol null1 A $ if "''null1'" .eqs. "0"  then  write sys$output "null1 is zero"    $ @null_test
   NULL1 = "." @ %DCL-E-INVIFNEST, invalid IF-THEN-ELSE nesting structure or data
 inconsistency  $    $ set verify $ @null_test $ null1[0,8]=0 $ sho sym null1 
   NULL1 = "."  $ if "@ %DCL-E-INVIFNEST, invalid IF-THEN-ELSE nesting structure or data
 inconsistency  $    What is the real error here?  E [Note: in practice, the variable "NULL1" in my case was the result of  reading data from a G file where the data value is not known ahead of time; this example just  demonstrates.]   ------------------------------  % Date: Wed, 16 Apr 2003 10:17:38 -0500  From: brandon@dalsemi.com 6 Subject: Re: DCL question - null byte in string symbol1 Message-ID: <03041610173805@dscis6-0.dalsemi.com>   L Why do you want to force the translation?  I think that is the problem and IJ would suspect that the %DCL-E-INVIFNEST is correct behavior because of the forced translation...      $ @null_test.com;  $ null1[0,8]=0 $ show symbol null1 
   NULL1 = "." < $ if null1 .eqs. "0"  then  write sys$output "null1 is zero" $ if "N %DCL-E-INVIFNEST, invalid IF-THEN-ELSE nesting structure or data inconsistency     John Brandon VMS Systems Administrator  Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fx    ------------------------------  % Date: Wed, 16 Apr 2003 11:39:37 -0400  From: norm.raphael@metso.com6 Subject: Re: DCL question - null byte in string symbol? Message-ID: <OFDBA4E3CD.CF4D6903-ON85256D0A.0055A48B@metso.com>   ( I don't quite get your response, either.: I expected a string, but in the file, the string was null, giving this error on execution. 1 My question is around what IF does or does not do 9 when an expression fails to evaluate.  Here it exits with 2 an IF error, without any other status.  Perhaps an/ Invalid expression error with a "-" would help?   A [I had to do some "stuff" to find out that the value was actually  a null string.]     1 From:  brandon@dalsemi.com on 04/16/2003 11:17 AM   % Please respond to brandon@dalsemi.com    To:    Info-VAX@Mvb.Saic.Com cc:   9 Subject:    Re: DCL question - null byte in string symbol     J Why do you want to force the translation?  I think that is the problem and I J would suspect that the %DCL-E-INVIFNEST is correct behavior because of the forced translation...      $ @null_test.com;  $ null1[0,8]=0 $ show symbol null1 
   NULL1 = "." < $ if null1 .eqs. "0"  then  write sys$output "null1 is zero" $ if "@ %DCL-E-INVIFNEST, invalid IF-THEN-ELSE nesting structure or data
 inconsistency      John Brandon VMS Systems Administrator  Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fx    ------------------------------  % Date: Wed, 16 Apr 2003 11:04:16 -0500  From: brandon@dalsemi.com 6 Subject: Re: DCL question - null byte in string symbol1 Message-ID: <03041611041677@dscis6-0.dalsemi.com>   * > I don't quite get your response, either.  3 Oh well.  Not the first time and it could be worse.   < > I expected a string, but in the file, the string was null,! > giving this error on execution. 3 > My question is around what IF does or does not do ; > when an expression fails to evaluate.  Here it exits with 4 > an IF error, without any other status.  Perhaps an1 > Invalid expression error with a "-" would help?     N If you look at the output and the associated error.  The forced translation ofJ the "''null1'" is terminating the line immediately after the <$ if ">.  MyN question is why force the translation?  Forcing the translation is terminatingO the line in mid-stream thus the error.  Why not just check the variable without  forced translation?      For example:   BRANDON$ type null_test.com; $ null1[0,8]=0 $ show symbol null1 ; $ if null1 .eqs. ""  then  write sys$output "null1 is NULL" < $ if null1 .eqs. "0"  then  write sys$output "null1 is zero"A $ if "''null1'" .eqs. "0"  then  write sys$output "null1 is zero"    BRANDON$ @NULL_TEST.COM  $ null1[0,8]=0 $ show symbol null1 
   NULL1 = "." ; $ if null1 .eqs. ""  then  write sys$output "null1 is NULL" 
 null1 is NULL < $ if null1 .eqs. "0"  then  write sys$output "null1 is zero" $ if "N %DCL-E-INVIFNEST, invalid IF-THEN-ELSE nesting structure or data inconsistency     John Brandon VMS Systems Administrator  Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fx    ------------------------------  % Date: Wed, 16 Apr 2003 09:14:30 -0700 + From: "Barry Treahy, Jr." <Treahy@MMaz.com> 6 Subject: Re: DCL question - null byte in string symbol' Message-ID: <3E9D8166.4070900@MMaz.com>    norm.raphael@metso.com wrote:    >Is this expected behavior?  > J >Note: I am having some difficulties in general with the fact that when anK >expression used as one term of a relation in an IF test fails to evaluate,  >the IF statement fails "ugly".  >  >$ type null_test.com  >$ null1[0,8]=0  >$ show symbol null1B >$ if "''null1'" .eqs. "0"  then  write sys$output "null1 is zero" > 
 >$ @null_test  >  NULL1 = "."A >%DCL-E-INVIFNEST, invalid IF-THEN-ELSE nesting structure or data  >inconsistency >    > G Null is most frequently used as a termination character for a sequence  D of ASCII characters.  Because you are asking DCL to create a symbol H containing a string content but are prematurely terminating that string G with a null, is most likely the problem...  If you use any other ASCII  9 character, besides null, you will not have this problem.    D I'm not a C coder, but isn't this the normal convention in C?  Null K terminated strings?  Anyone out there that does grind C can confirm this...    Barry    --    @ Barry Treahy, Jr  *  Midwest Microwave  *  Vice President & CIO   A E-mail: Treahy@mmaz.com * Phone: 480/314-1320 * FAX: 480/661-7028    ------------------------------  # Date: Wed, 16 Apr 2003 17:23:06 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond) 6 Subject: Re: DCL question - null byte in string symbol0 Message-ID: <_jgna.930$9R3.788@news.cpqcorp.net>  ^ In article <OFDBA4E3CD.CF4D6903-ON85256D0A.0055A48B@metso.com>, norm.raphael@metso.com writes:   ..> >I expected a string, but in the file, the string was null, ..  M No, it was not a null string; it was a string that contained a null charcter. 2 Sounds like I'm splitting hairs, and perhaps I am.  0 In DCL, you can represent a null string with "".  : The string created by <var>[0,8]=0 is not a null string.  : It is a one character string containing the character %x0.  G Now I will conceed that DCL's truncation of the line is a bit "flakey", G but some things are the way they are.  Come to think of it, most things J are the way they are!  Personally, I kinda hope it stays that way. <smile>    	 Try this:    $ null1 = "" $ show symbol null1  $on error then continue A $ if "''null1'" .eqs. "0"  then  write sys$output "null1 is zero"  $ write sys$output "next line"   This also works    $ null1[0,8] = 0 $ show symbol null1  $on error then continue < $ if null1 .eqs. "0"  then  write sys$output "null1 is zero" $ write sys$output "next line"  D Note that neither the null character nor the null string is equal to* the character "0" (which I think is %x31).    G Is there a particular problem you are trying to solve, Norm, or is this  just an intellectual exercise.   --  J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Wed, 16 Apr 2003 07:40:52 +0200 1 From: PRSTSC::DTL <Didier.Morandi.nospam@Free.fr>  Subject: Re: DECnet hacker? 4 Message-ID: <3e9cecdb$0$28769$626a54ce@news.free.fr>   John Santos wrote:  E > BTW, Dave - it is a common convention for anonymous FTP, but for no E > other protocol I've ever encountered, to send your email address as E > the password, which isn't otherwise used.  I think the FTP protocol E > doesn't treat "anonymous" any differently than any other user name, F > so you have to send *something* for the password, but the FTP server< > ignores it if it allows anonymous logins.  I don't know ifA > the common VMS FTP servers record the passwords for "anonymous" A > anywhere (so you can tell who has been using your system.)  And % > of course, the bad guys always lie!   D I became a "bad guy" after I started to receive dozen of spam a day.   D.   ------------------------------    Date: 16 Apr 2003 09:38:12 -07001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) . Subject: Re: Excellent new OpenVMS Testimonial= Message-ID: <cf15391e.0304160838.77355891@posting.google.com>   d "Bill Todd" <billtodd@metrocast.net> wrote in message news:<gdWdnVUD9-t7OAGjXTWcqg@metrocast.net>... > the customer's presentation I > makes it clear that their needs will be *expanding* throughout the next M > decade, which means they'll be needing to purchase *additional* systems for I > that period, not just until HP stops selling them 4 - 5 years from now.   A Just because HP stops manufacturing them doesn't mean folks can't C procure additional systems to meet their needs.  I note on HP's web F site at http://h18002.www1.hp.com/alphaserver/vax/ that HP tells folksA even VAX systems are still available for sale today to meet their % needs (through remarketing channels).   E To put this into perspective, Fraport's previous platform was VAXft.  B IIRC, that platform hasn't been manufactured since sometime in theD early-to-mid '90s, and yet they have been able to continue to use it@ successfully until they were ready to migrate to Alpha.  They'llE continue to be able to use Alpha for many, many years, until they are ( ready to migrate to their next platform.  @ The important thing for mission-critical businesses is support. = Hardware support for Alpha has been promised through at least D 2012-2013.  Software support will likely last much longer.  (The webB page above states support for OpenVMS VAX extends through at leastF 2010.  Since last-sale of VAX hardware was in 2000, one could estimateE software support for OpenVMS Alpha would likely also extend a minimum F of a decade after last-sale, and thus extend through about 2017-2018. B And with common source code between Alpha and Itanium, support forD Alpha will be much easier and less-costly and thus less likely to be! terminated than support for VAX.)    ------------------------------  % Date: Wed, 16 Apr 2003 03:50:17 -0500  From: brandon@dalsemi.com # Subject: Re: Fiber channel question 1 Message-ID: <03041603501760@dscis6-0.dalsemi.com>   9 > 	You can only quickset one device.  You have 4 slots so D > 	Bart is right.  The slots in use and how to do more than one are 6 > 	all in the wwidmgr manual found on the firmware CD.   -snip-& Page 17 (2.7) Limitations of -quickset 1. only one deviceO 2. more than one device if those devices share access through the same ports on 	 the HSG80  ..N For installations with topologies where more than one disk must be set up, andM those disks not on accessed through common HSG80 ports, the -set port command L must be used.  Installations that would need to use -set port would include:  O 1. A VMS installation where the boot disk and the dump disk do not share common  paths.  O 2. An installation where more than one boot disk needs to be set up for console 2 use and those boot disks are not on a common path. .. Page 31 (3.6.1) set port ..M The -mode qualifier selects one of 4 non-volatile positions to store away the  ..	 -endsnip-   M So, yes Bart is right.  I am right in my own mind!  ;-)  I set up my server 2 K years ago and I do not recall the set port function.  I did not need use of L multiple boot devices so I guess I was not looking for it.  Maybe it was not! even available then... oh bother.   8 However this brings to mind a response by John Travell -  F > Yes, but the fun comes when your system disk is a shadow set, with aI > potential 3 members, and 4 paths to each member, you have a total of 12  > paths.B > Now consider what happens when it is time to write a bugcheck...N > The console (SRM) writes the dump to the first active path that it finds. ItI > has no way of knowing whether the disk that path goes to is the current I > master member of the shadow set, or even a current member at all!. This H > leaves you with the interesting situation that dumpfile writes fail at	 > random.   O If you have a shared dump drive - specifically for dump and nothing else - that N is hardware RAID only, then your problem is solved.  That is IF you have a SAN solution...      John Brandon VMS Systems Administrator  Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fx    ------------------------------  + Date: Wed, 16 Apr 2003 10:08:22 +0000 (UTC) + From: david20@alpha2.mdx.ac.uk (David Webb) # Subject: Re: Fiber channel question + Message-ID: <b7ja2m$8k6$1@aquila.mdx.ac.uk>   M In article <03041515412066@dscis6-0.dalsemi.com>, brandon@dalsemi.com writes: L >> This number will turn up as disk number on VMS, so if you used identifier
 >> number N >> 1234, you will see disk $1$dga1234 on VMS. No difficult translations here ! > M >If you happen to dismount the device from the VMS server(s) and  reconfigure O >that device at the controller level you will need to reboot all the servers to - >see that device - if you keep the same ID #.  >    Why ?   F I've never worked with Fiber channel controllers on VMS but with otherJ controllers (HSZ50 etc) I've never had trouble just dismounting the deviceK from all systems , reconfiguring and then remounting it. It's a while since L I've done this so it's possible I might have needed to do an io autoconfig - but I don't think so.     L On TRU64 the disconnect between the device number and the Unit number on the controller is a REAL pain.  M On Tru64 to find the underlying storage associated with a partition you need   all of these steps :-   	 1) df -k    $    shows you your mounted partitions  @ 2) pick a partition then assuming using the Advanced file system      showfdmn   domain_name   K    this will tell you the disk name (and disk partition) as Tru64 sees it       eg dsk7c    3) hwmgr -view devices    H to show the mapping between the dsk device and the lun - something like    .  .  . 6 ... /dev/disk/dsk7c         ...     bus-5-tar-0-lun-2  .  .  .   D You can then look on the controller to see what unit D2 consists of.    
 David Webb VMS and Unix team leader CCSS Middlesex University  
 >John Brandon  >VMS Systems Administrator >Dallas Semiconductor  >john.brandon@dalsemi.com  >972.371.4172 wk >972.371.4003 fx   ------------------------------  # Date: Wed, 16 Apr 2003 14:06:32 GMT # From: "John Smith" <a@nonymous.com> ! Subject: HP in the financial news G Message-ID: <Irdna.47960$BQi.3953@news04.bloor.is.net.cable.rogers.com>   2 IBM weak hardware sales bode poorly for HP-analyst   Tuesday April 15, 4:54 pm ET$ By Caroline Humer and Duncan Martell  C NEW YORK/SAN FRANCISCO, April 15 (Reuters) - International Business C Machines Corp.'s decline in hardware sales during the first quarter C does not bode well for its next largest competitor, Hewlett-Packard * Co., according to one Wall street analyst.  D IBM reported on Monday that its hardware sales fell about 1 percent,E but that higher revenues in its large computer services division more D than offset the decline, enabling its first-quarter revenues to rise 11 percent to $20.06 billion.   B For the first time, services accounted for more than 50 percent of IBM's revenues.   E HP garners only about 17 percent of its revenue from services and the C bulk from personal computers and large systems for corporations, as E well as from its lucrative imaging and printing division. HP does not / have the large services business that IBM does.   C Merrill Lynch said it is not changing its estimates for HP revenues ; for its first quarter ended in April, but they are at risk.   > "We are more nervous about HP making particularly our top lineE estimate," analyst Steven Milunovich said in a research note in which 7 he downgraded his investment rating on HP to "neutral."   C HP shares fell about 2.6 percent on Tuesday, while investors backed % IBM, pushing its shares up 3 percent.   E Some analysts and investors said that, while a slowdown in technology E spending will affect HP much the same way it held back new technology B orders at IBM, it is not possible to extrapolate directly from the/ performance of IBM's hardware business to HP's.   D For instance, IBM's personal systems group, which is mostly personalF computers, was only about 12 percent of revenues in the first quarter,< while HP gets about 29 percent of its revenues from personal5 computers, Lehman Brothers analyst Daniel Niles said.   F "They obviously have a huge PC business and they're battling with DellE for the No. 1 position there. For IBM, that's a business that they've . been continually de- emphasizing," Niles said.  A Dell is the world's second largest personal computer maker, right 
 behind HP.  A Milunovich also wrote in his research note that he believes HP is B getting squeezed between Dell Computer Corp. on PCs and IBM on its= most expensive systems comprised of itshardware, software and 	 services.   > Peter Blackmore, who runs HP's enterprise systems group, which% includes computer servers, disagreed.   C "It's very simplistic to say we get squeezed between IBM and Dell," E Blackmore said in an interview on Tuesday. "Customers like the choice ! between Windows, Linux and Unix."   E HP, like IBM, offers all three operating systems, while Dell does not ? sell super high-end systems that use the Unix operating system.   @ Blackmore said Dell and HP compete only on one HP product in the@ server market, which is its Intel Pentium and Xeon-based serversD versus Dell's servers. He added that, in the fourth calendar quarterF of 2002, HP grew one percentage point faster than Dell in that market.  D In the Unix market, where HP competes with IBM and Sun Microsystems,E Blackmore said HP remains strong, noting HP enjoys No. 1 market share A in the mid-range Unix servers and the No. 1 position for high-end / servers, which includes HP's Superdome product.   A Blackmore added HP's enterprise systems business is on tack to be * profitable in the second half of the year.2 "We feel pretty good about where we are," he said.  F HP does not report financial results until mid-May, so it is not known/ how it is faring on the most expensive systems.   A But IBM's sales of mainframes fell 16 percent during the quarter, E while the sales of its powerful computers based on its own microchips A and running the popular Unix operating system were up 15 percent.   C Its iSeries computer, which was previously called the AS400 and was A once the workhorse of most data centers, saw sales increase by 22  percent.  F Sales of xSeries computers, or those that are based on microprocessors* made by Intel Corp., increased 20 percent.  E IBM's year-on-year gains, however, come on the back of a particularly A weak year for IBM, analysts cautioned. For instance, sales of the E Unix-based computers -- which compete against HP's Unix-based system, + had fallen about 35 percent a year earlier.   C "It's good to see that decline stop, but it didn't increase a lot," ! Gartner analyst Tom Bittman said.    -----------   E Interesting that IBM manages to increase 'legacy' AS400 sales by 22%.    ------------------------------  % Date: Wed, 16 Apr 2003 07:34:27 -0700 # From: "Tom Linden" <tom@kednos.com> % Subject: RE: HP in the financial news 9 Message-ID: <CIEJLCMNHNNDLLOOGNJIAECCHAAA.tom@kednos.com>    >-----Original Message----- ) >From: John Smith [mailto:a@nonymous.com] ( >Sent: Wednesday, April 16, 2003 7:07 AM >To: Info-VAX@Mvb.Saic.Com" >Subject: HP in the financial news >  > 3 >IBM weak hardware sales bode poorly for HP-analyst  >  >Tuesday April 15, 4:54 pm ET % >By Caroline Humer and Duncan Martell  > D >NEW YORK/SAN FRANCISCO, April 15 (Reuters) - International BusinessD >Machines Corp.'s decline in hardware sales during the first quarterD >does not bode well for its next largest competitor, Hewlett-Packard+ >Co., according to one Wall street analyst.  > E >IBM reported on Monday that its hardware sales fell about 1 percent, F >but that higher revenues in its large computer services division moreE >than offset the decline, enabling its first-quarter revenues to rise  >11 percent to $20.06 billion. > C >For the first time, services accounted for more than 50 percent of  >IBM's revenues. > F >HP garners only about 17 percent of its revenue from services and theD >bulk from personal computers and large systems for corporations, asF >well as from its lucrative imaging and printing division. HP does not0 >have the large services business that IBM does. > D >Merrill Lynch said it is not changing its estimates for HP revenues< >for its first quarter ended in April, but they are at risk. > ? >"We are more nervous about HP making particularly our top line F >estimate," analyst Steven Milunovich said in a research note in which8 >he downgraded his investment rating on HP to "neutral." > D >HP shares fell about 2.6 percent on Tuesday, while investors backed& >IBM, pushing its shares up 3 percent. > F >Some analysts and investors said that, while a slowdown in technologyF >spending will affect HP much the same way it held back new technologyC >orders at IBM, it is not possible to extrapolate directly from the 0 >performance of IBM's hardware business to HP's. > E >For instance, IBM's personal systems group, which is mostly personal G >computers, was only about 12 percent of revenues in the first quarter, = >while HP gets about 29 percent of its revenues from personal 6 >computers, Lehman Brothers analyst Daniel Niles said. > G >"They obviously have a huge PC business and they're battling with Dell F >for the No. 1 position there. For IBM, that's a business that they've/ >been continually de- emphasizing," Niles said.  > B >Dell is the world's second largest personal computer maker, right >behind HP.  > B >Milunovich also wrote in his research note that he believes HP isC >getting squeezed between Dell Computer Corp. on PCs and IBM on its > >most expensive systems comprised of itshardware, software and
 >services. > ? >Peter Blackmore, who runs HP's enterprise systems group, which & >includes computer servers, disagreed. > D >"It's very simplistic to say we get squeezed between IBM and Dell,"F >Blackmore said in an interview on Tuesday. "Customers like the choice" >between Windows, Linux and Unix." > F >HP, like IBM, offers all three operating systems, while Dell does not@ >sell super high-end systems that use the Unix operating system. > A >Blackmore said Dell and HP compete only on one HP product in the A >server market, which is its Intel Pentium and Xeon-based servers E >versus Dell's servers. He added that, in the fourth calendar quarter G >of 2002, HP grew one percentage point faster than Dell in that market.  > E >In the Unix market, where HP competes with IBM and Sun Microsystems, F >Blackmore said HP remains strong, noting HP enjoys No. 1 market shareB >in the mid-range Unix servers and the No. 1 position for high-end0 >servers, which includes HP's Superdome product. > B >Blackmore added HP's enterprise systems business is on tack to be+ >profitable in the second half of the year. 3 >"We feel pretty good about where we are," he said.  > G >HP does not report financial results until mid-May, so it is not known 0 >how it is faring on the most expensive systems. > B >But IBM's sales of mainframes fell 16 percent during the quarter,F >while the sales of its powerful computers based on its own microchipsB >and running the popular Unix operating system were up 15 percent. > D >Its iSeries computer, which was previously called the AS400 and wasB >once the workhorse of most data centers, saw sales increase by 22	 >percent.  > G >Sales of xSeries computers, or those that are based on microprocessors + >made by Intel Corp., increased 20 percent.  > F >IBM's year-on-year gains, however, come on the back of a particularlyB >weak year for IBM, analysts cautioned. For instance, sales of theF >Unix-based computers -- which compete against HP's Unix-based system,, >had fallen about 35 percent a year earlier. > D >"It's good to see that decline stop, but it didn't increase a lot,"" >Gartner analyst Tom Bittman said. >  >----------- > F >Interesting that IBM manages to increase 'legacy' AS400 sales by 22%.  K They also didn't mention mainframe sales, which elsewhere has been reported  as on the increase for IBM.    >  >  >---' >Incoming mail is certified Virus Free. ; >Checked by AVG anti-virus system (http://www.grisoft.com). @ >Version: 6.0.467 / Virus Database: 266 - Release Date: 4/1/2003 >  --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).? Version: 6.0.467 / Virus Database: 266 - Release Date: 4/1/2003    ------------------------------   Date: 16 Apr 2003 14:43:13 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)% Subject: Re: HP in the financial news 5 Message-ID: <b7jq60$1midr$2@ID-135708.news.dfncis.de>   G In article <Irdna.47960$BQi.3953@news04.bloor.is.net.cable.rogers.com>, & 	"John Smith" <a@nonymous.com> writes: > @ > Peter Blackmore, who runs HP's enterprise systems group, which' > includes computer servers, disagreed.  > E > "It's very simplistic to say we get squeezed between IBM and Dell," G > Blackmore said in an interview on Tuesday. "Customers like the choice # > between Windows, Linux and Unix."   F Ignoring for the moment the bad financial news (which made most of theD TV financial news programs I saw yesterday) what is missing from the statement above?    Looks like business as usual!!!!   bill     --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  # Date: Wed, 16 Apr 2003 15:13:30 GMT " From:   VAXman-  @SendSpamHere.ORG% Subject: Re: HP in the financial news 0 Message-ID: <00A1E791.ADFDA53C@SendSpamHere.ORG>  ` In article <b7jq60$1midr$2@ID-135708.news.dfncis.de>, bill@cs.uofs.edu (Bill Gunshannon) writes:H >In article <Irdna.47960$BQi.3953@news04.bloor.is.net.cable.rogers.com>,' >	"John Smith" <a@nonymous.com> writes:  >>  A >> Peter Blackmore, who runs HP's enterprise systems group, which ( >> includes computer servers, disagreed. >>  F >> "It's very simplistic to say we get squeezed between IBM and Dell,"H >> Blackmore said in an interview on Tuesday. "Customers like the choice$ >> between Windows, Linux and Unix." > G >Ignoring for the moment the bad financial news (which made most of the E >TV financial news programs I saw yesterday) what is missing from the  >statement above?    Duhwa, I dunno...  ;(     ! >Looks like business as usual!!!!    Yup. --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM              5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------  # Date: Wed, 16 Apr 2003 13:10:04 GMT # From: "John N." <JNixon@cfl.rr.com>  Subject: HSZ50= Message-ID: <MCcna.160630$j8.3504548@twister.tampabay.rr.com>   L What is the largest disk I can use in an HSZ50 attached to an AS2100 running VMS? Where can I find this info?   K I would love to replace the HSZ50, but that is not in the cards this fiscal  year.    ------------------------------  % Date: Wed, 16 Apr 2003 12:55:33 -0400 + From: "Martin O'Connor" <moconnor@dvfs.com>  Subject: Re: HSZ505 Message-ID: <b7k1u7$1rmci$1@ID-118202.news.dfncis.de>   9 "Vinit Adya" <vinit.adya@mizuhocbus.com> wrote in message 7 news:eb8f4d7b.0304160848.417663d2@posting.google.com... @ : To the best of my knowlege, 18GB disks can be supported on theC : controller. However u could create larger storage sets using 18GBs : disks.F : But i use HSZ50 with HPUX not with VMS systems. So maybe VMS version- : of HSZ firmware could support larger disks.n :   Z I have HSJ 40s and 50s with 36 GB disks so I would think that HSZ50s would take them also.   Marty    ------------------------------   Date: 16 APR 2003 15:52:44 GMT4 From: karcher@thuria.waisman.wisc.edu (Carl Karcher) Subject: Re: HSZ506 Message-ID: <16APR03.15524445@thuria.waisman.wisc.edu>  ; In a previous article, "John N." <JNixon@cfl.rr.com> wrote:iN ->What is the largest disk I can use in an HSZ50 attached to an AS2100 running" ->VMS? Where can I find this info?  F In the SPD (61.36) for HSOF software, where V5.7 is the final release,F the largest disk listed is a 36.4GB DS-RZ1FC-VW. The minimum microcode< (for the disk) is 3B07 and minimum hardware revision is A01.   --G -- Carl Karcher, Waisman Computing Services, Waisman Center, UW-Madisone6 --               karcher.nomorespam@waisman.wisc.edu     ------------------------------    Date: 16 Apr 2003 09:48:07 -0700, From: vinit.adya@mizuhocbus.com (Vinit Adya) Subject: Re: HSZ50= Message-ID: <eb8f4d7b.0304160848.417663d2@posting.google.com>e  > To the best of my knowlege, 18GB disks can be supported on theA controller. However u could create larger storage sets using 18GBe disks.D But i use HSZ50 with HPUX not with VMS systems. So maybe VMS version+ of HSZ firmware could support larger disks.r  h "John N." <JNixon@cfl.rr.com> wrote in message news:<MCcna.160630$j8.3504548@twister.tampabay.rr.com>...N > What is the largest disk I can use in an HSZ50 attached to an AS2100 running > VMS? > Where can I find this info?a > M > I would love to replace the HSZ50, but that is not in the cards this fiscale > year.e   ------------------------------    Date: 16 Apr 2003 12:44:00 -0500+ From: kuhrt@encompasserve.org (Marty Kuhrt)U Subject: Re: HSZ503 Message-ID: <Wx2NPktI3xqM@eisner.encompasserve.org>h  c In article <MCcna.160630$j8.3504548@twister.tampabay.rr.com>, "John N." <JNixon@cfl.rr.com> writes: N > What is the largest disk I can use in an HSZ50 attached to an AS2100 running > VMS? > Where can I find this info?h > M > I would love to replace the HSZ50, but that is not in the cards this fiscal  > year.f >   I I had thirty-six 36G disks hooked up to my HSZ50 working just fine until i! we shut it down a short time ago.o   ------------------------------  % Date: Wed, 16 Apr 2003 14:06:25 +0200i1 From: PRSTSC::DTL <Didier.Morandi.nospam@Free.fr>i Subject: HTML favourite editor? 4 Message-ID: <3e9d4736$0$28769$626a54ce@news.free.fr>  - What is to you the greatest HTML code editor?S! I say editor, not visual dev env.    edt  Notepad.exe-	 WebExpert- ?-   D.   ------------------------------  % Date: Wed, 16 Apr 2003 13:08:27 +01000( From: Nic Clews <sendspamhere@127.0.0.1># Subject: Re: HTML favourite editor?r) Message-ID: <3E9D47BB.CE3BB711@127.0.0.1>E   PRSTSC::DTL wrote: > / > What is to you the greatest HTML code editor?d# > I say editor, not visual dev env.h  $ EDT, followed by a DCL command file.     --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences, nclews at csc dot comm   ------------------------------  # Date: Wed, 16 Apr 2003 13:38:48 GMTt1 From: Michael Austin <maustin@firstdbasource.com> # Subject: Re: HTML favourite editor?y2 Message-ID: <3E9D4F5A.601EA5C1@firstdbasource.com>   PRSTSC::DTL wrote: > / > What is to you the greatest HTML code editor?n# > I say editor, not visual dev env.N >  > edtn
 > Notepad.exea > WebExpert  > ?t >  > D.  E TPU/EVE  -- split screen with 2 or more files at the same time can bet really nice.F notepad is a good choice IF you NEED to be on a Wxx system.  But sinceD my web server is on the Alpha, I don't NEED to do if from a Wxx box.  C I you really aren't sure about the html syntax you can search for adG freeware gui that can help until you can <html> without having to thinkrF too hard or look it up :)   Google is very good about "html howto... "   -- - Regards,  6 Michael Austin            OpenVMS User since June 19847 First DBA Source, Inc.    Registered Linux User #261163s   ------------------------------    Date: 16 Apr 2003 07:45:18 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)o# Subject: Re: HTML favourite editor?a3 Message-ID: <Mw2HarcZtv7L@eisner.encompasserve.org>l  h In article <3e9d4736$0$28769$626a54ce@news.free.fr>, PRSTSC::DTL <Didier.Morandi.nospam@Free.fr> writes:/ > What is to you the greatest HTML code editor?0# > I say editor, not visual dev env.e  4    I prefer TPU, but I'll use EDT from time to time.   ------------------------------    Date: 16 Apr 2003 08:07:00 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)h# Subject: Re: HTML favourite editor? 3 Message-ID: <i1$bpx2mu2Jx@eisner.encompasserve.org>u  h In article <3e9d4736$0$28769$626a54ce@news.free.fr>, PRSTSC::DTL <Didier.Morandi.nospam@Free.fr> writes:/ > What is to you the greatest HTML code editor? # > I say editor, not visual dev env.o >  > edt 
 > Notepad.exep > WebExperto > ?t   BBedit   ======  7 But for W3.org compliance, I really prefer Dreamweaver.    ------------------------------  % Date: Wed, 16 Apr 2003 09:09:56 -0400n! From: VAXVMS <bounce@notmail.com>0# Subject: Re: HTML favourite editor?uK Message-ID: <BA52530E3149734A9BAABDBBFA808E4903027BE9@rlghncst964.usps.gov>p   Hi, Michael-   As for editors:r  = There's a graphical freeware HTML editor for VMS called ASHE._  7 I have it on a machine that isn't up at the moment so I  can't tell you where I got it.  E In the past I've used something called PFE (Programmer's File Editor)TD because you can use it to globally make changes to all HTML files in a given directory.   As for references:  7 webreference.com serves as a constant reminder of just 0, how very little I know about doing webstuff;  3 htmlhelp.com has two very nice Windows help files, l one for HTML 4 and one for CSS.t   WWWebb  2 Michael Austin [maustin@firstdbasource.com] wrote: >O >PRSTSC::DTL wrote:t >> m0 >> What is to you the greatest HTML code editor?$ >> I say editor, not visual dev env. >> i >> edt >> Notepad.exe >> WebExpert >> ? >> y >> D.o >sF >TPU/EVE  -- split screen with 2 or more files at the same time can be
 >really nice.iG >notepad is a good choice IF you NEED to be on a Wxx system.  But sincerE >my web server is on the Alpha, I don't NEED to do if from a Wxx box.n >uD >I you really aren't sure about the html syntax you can search for aH >freeware gui that can help until you can <html> without having to thinkG >too hard or look it up :)   Google is very good about "html howto... "- >l >-- 	 >Regards,s >o7 >Michael Austin            OpenVMS User since June 1984k8 >First DBA Source, Inc.    Registered Linux User #261163 ========================  William W. Webb - EMS Operations) OpenVMS Systems Support - USPS DSSC Annex[, 4730 Hargrove Road, Raleigh, NC 27616-2874  > 919.325.7500 x4186  <FirstInitialLastNameAtEmailDotUSPSDotGov>   ------------------------------  % Date: Wed, 16 Apr 2003 09:12:53 -0400o& From: Ken Robinson <kenrbnsn1@rcn.com># Subject: Re: HTML favourite editor?r< Message-ID: <5.1.0.14.2.20030416091145.048fa138@pop.rcn.com>  ' At 08:07 AM 4/16/2003 -0500, you wrote: B >In article <3e9d4736$0$28769$626a54ce@news.free.fr>, PRSTSC::DTL ( ><Didier.Morandi.nospam@Free.fr> writes:1 > > What is to you the greatest HTML code editor?i% > > I say editor, not visual dev env.o > >e > > edt  > > Notepad.exe 
 > > WebExperti > > ?    On the PC: Homesite v5C On VMS: any editor is fine. Has anyone customized LSE for HTML? :-)    Kenu   ------------------------------  % Date: Wed, 16 Apr 2003 14:13:23 +0100r( From: Nic Clews <sendspamhere@127.0.0.1># Subject: Re: HTML favourite editor?r) Message-ID: <3E9D56F3.DF3E8FF7@127.0.0.1>e   Larry Kilgallen wrote: > j > In article <3e9d4736$0$28769$626a54ce@news.free.fr>, PRSTSC::DTL <Didier.Morandi.nospam@Free.fr> writes:1 > > What is to you the greatest HTML code editor? % > > I say editor, not visual dev env.  > >n > > edtn > > Notepad.exea
 > > WebExperti > > ?t >  > BBedit >  > ====== > 9 > But for W3.org compliance, I really prefer Dreamweaver.o  1 I'm disappointed, I was _certain_ you'd say TECO.,   --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciencesn nclews at csc dot coml   ------------------------------  % Date: Wed, 16 Apr 2003 09:50:58 -0400 ! From: Jim Agnew <jpagnew@vcu.edu>i# Subject: Re: HTML favourite editor? ' Message-ID: <3E9D5FC2.225A8FDA@vcu.edu>U  B we are big into Dreamweaver... It's quite well written, very good,C expensive unless you're at an education site, and you can start offl simple..  % and it cleans up Word html nicely!!!!E   jimn   PRSTSC::DTL wrote: > / > What is to you the greatest HTML code editor?i# > I say editor, not visual dev env.b >  > edte
 > Notepad.exe. > WebExpert. > ?0 >  > D.   -- -F "4,000 years ago I made a mistake."  Elrond Half-Elven, in "Fellowship of the Ring"   ------------------------------   Date: 16 Apr 2003 13:55:43 GMT2 From: Thierry Dussuet <thierry@squeeeez.no-ip.com># Subject: Re: HTML favourite editor? 1 Message-ID: <slrnb9qo70.3u4.thierry@andro.family>i  J In article <3e9d4736$0$28769$626a54ce@news.free.fr>, PRSTSC : : DTL wrote:/ > What is to you the greatest HTML code editor?v# > I say editor, not visual dev env.l >  > edt$
 > Notepad.exe. > WebExpert1  C Personnally I use vim; it's available on Wxx, un*x, vms, so this is 6 quite nice :-) (and it has a nice syntax highlighting)   ThierryC   ------------------------------    Date: 16 Apr 2003 09:48:47 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)d# Subject: Re: HTML favourite editor?u3 Message-ID: <VH2e36v1EPJv@eisner.encompasserve.org>D  T In article <3E9D56F3.DF3E8FF7@127.0.0.1>, Nic Clews <sendspamhere@127.0.0.1> writes: > Larry Kilgallen wrote: >> lk >> In article <3e9d4736$0$28769$626a54ce@news.free.fr>, PRSTSC::DTL <Didier.Morandi.nospam@Free.fr> writes: 2 >> > What is to you the greatest HTML code editor?& >> > I say editor, not visual dev env. >> > >> > edt >> > Notepad.exe >> > WebExpert >> > ? >> a	 >> BBedit  >> a	 >> ======  >>  : >> But for W3.org compliance, I really prefer Dreamweaver. > 3 > I'm disappointed, I was _certain_ you'd say TECO.   3 BBedit has a lot of features specifically for HTML.33 With TECO one would have to program those features,a1 and I know of no suitable existing set of macros.t  < My advice to the original poster would still be Dreamweaver,1 but the problem statement seems to preclude that.0   ------------------------------  % Date: Wed, 16 Apr 2003 10:58:03 -0400i! From: Jim Agnew <jpagnew@vcu.edu> # Subject: Re: HTML favourite editor?i' Message-ID: <3E9D6F7B.2EE75E03@vcu.edu>   3 That's right, it does!!!  I'm trying to learn it...s   Thierry Dussuet wrote: > L > In article <3e9d4736$0$28769$626a54ce@news.free.fr>, PRSTSC : : DTL wrote:1 > > What is to you the greatest HTML code editor?n% > > I say editor, not visual dev env.e > >a > > edtr > > Notepad.exei
 > > WebExperta > E > Personnally I use vim; it's available on Wxx, un*x, vms, so this isi8 > quite nice :-) (and it has a nice syntax highlighting) > 	 > Thierry<   --  F "4,000 years ago I made a mistake."  Elrond Half-Elven, in "Fellowship of the Ring"   ------------------------------    Date: 16 Apr 2003 17:39:18 +0200' From: huber@mppmu.mpg.de (Joseph Huber) # Subject: Re: HTML favourite editor?u+ Message-ID: <NwlNmQo2ym6d@vms.mppmu.mpg.de>q  h In article <3e9d4736$0$28769$626a54ce@news.free.fr>, PRSTSC::DTL <Didier.Morandi.nospam@Free.fr> writes:/ > What is to you the greatest HTML code editor?h# > I say editor, not visual dev env.  >  > edte
 > Notepad.exer > WebExpert  > ?g  < Being in the vms ng, there was an LSE module for HTML, just  don't have a pointer.m   There are also VMS versions of    nedit  vim
  xhtml (ashe)   @ All have syntax highlighting and macros for HTML tag generation.C Find some of the docs on my site http://wwwvms.mppmu.mpg.de/vmsdoc/.    -- -N Joseph "Sepp" Huber   mailto:joseph.huber@web.de   http://www.huber-joseph.de/   ------------------------------  % Date: Wed, 16 Apr 2003 09:14:11 -0400g. From: "Rob Lyons" <rob.lyons@resilientsys.com>0 Subject: Re: Need to buy 8 x VAX 4000 or similar+ Message-ID: <b7jl0u$hn3$1@bob.news.rcn.net>   \ David J. Dachtera <djesys.nospam@fsi.net> wrote in message news:3E9CCAF4.2CF02C4E@fsi.net... > Rob Lyons wrote:
 > > [snip]> > > You can't expect HP to give away the Alpha OS on the host, >y7 > Well, technically, yes I can. Think: double taxation.  > H > As a compromise, perhaps a limited OpenVMS-ALPHA license would be more: > fair, something that only runs as a dedicated appliance.  G An interesting idea but unlikely to happen since I don't think businessu? finance works that way.  Unless SRI or HP can expect additionala9 revenue, they wouldn't invest the resources to create it.a  G > Most folks don't hack around the way I do, so many don't know how the7H > stand-alone environment of the OpenVMS-Alpha (and VAX, really) boot-CDJ > works. Perhaps, instead of running SA_STARTUP.COM, it ran something thatD > does a full CONFIG, then enters the Charon-VAX environment, so theJ > underlying OpenVMS-Alpha is not running in an interactive mode, and goes> > immediately into a shutdown procedure when Charon-VAX exits.  K That would not be sufficient during installation.  I still need to edit theiF Charon-VAX config file. I still need INIT and MOUNT to build containerN files or volumes.  I need authorize to set quotas for the CHARON user account.E I need to autogen the host system to adjust system parameters for thenI Charon application.  Having TCPIP on the host to facilitate transfer froms' the old to the new is also very useful.b  J Once everything is running, and if there never is a need to change things,C then a reduced feature "canned" environment could work.  But then I D would want to be able to use monitor to look at resource usage and I5 would want to check the error logs from time to time.e  	 Rob Lyonsu   ------------------------------    Date: 16 Apr 2003 11:47:53 -0500B From: clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley)R Subject: RE: Newsgroup posting conventions, was: RE: VMS Software Product & Online3 Message-ID: <ZBJHCZKHQFBN@eisner.encompasserve.org>s  ~ In article <BE56C50EA024184DAF48F0B9A47F5CF403FB5C7F@kaoexc01.americas.cpqcorp.net>, "Main, Kerry" <Kerry.Main@hp.com> writes: > Peter, > I > Thanks for the tip .. How does the attached look to your reader client?0 > 	 > Regardss >  > Kerry Main >    That's much better Kerry.k  H The only remaining things would be to put the reply at the bottom of the% message and trim the quoted material.    Simon.   -- eB Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP       L VMS advocate: One who makes a Mac advocate look like a beginner at advocacy.   ------------------------------    Date: 16 Apr 2003 07:36:09 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)o0 Subject: Re: OpenVMS and Security - REALLY !!!!!3 Message-ID: <dNQ7g5oH9iHe@eisner.encompasserve.org>e  o In article <BA52530E3149734A9BAABDBBFA808E4903027BE8@rlghncst964.usps.gov>, VAXVMS <bounce@notmail.com> writes:u > H >    But I did once put an HP 1000 in a fairly quite room.  Sounded like+ >    a 707 taking off when I powered it up.o >  > Two words:  ESDI drives.  D    You must be speaking Sun.  The HP 1000 had a drive in a cast ironE    casing, something like 19" x 28" x 16", wieghed a ton, and sat on      a GPIB bus.   ------------------------------  % Date: Wed, 16 Apr 2003 11:21:24 +0100 ' From: Andrew Harrison SUNUK Consultancy.Y Subject: Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retai.. Message-ID: <3E9D2EA4.8050008@nospamn.sun.com>   jlsue wrote:G > On Mon, 14 Apr 2003 17:50:15 +0100, Andrew Harrison SUNUK Consultancya0 > <Andrew_No.Harrison_No@nospamn.sun.com> wrote: >  >  >> >>jlsue wrote: >  > G >>>Bullshit it doesn't.  I don't care *how many* times you tell me.  IttF >>>says specifically, in an EXACT QUOTE, that the vms upgrade gave theH >>>28% performance increase.  I have provided proof, within the article,I >>>to support my claim.  Your position is only one of conjecture based on." >>>your own self-serving opinions. >>>mI >>>You can keep repeating your lies, but you can't support it with actuale% >>>data/info/quotes from the article.t >>>  >>E >>So tell us what was the impact of moving from Turbolaser to GS160 !o >  > D > I don't have to find or present that information because it is notC > germane to my side of this discussion:  They are happy customers.aH > The 28% improvement from the VMS upgrade is just a side issue that youB > bring out because you can't counter it with anything meaningful. >   # Huu are you really this challenged.u  @ I asked you for references showing what performance improvements4 people got when moving from 8400/GS140 to GS160/320.  ? So what are you now claiming again ?????? You are claiming that 9 the reference is for how much faster their GS system wenty when they upgraded to 7.3.  : Fine, lets for the sake of totally ruining your day assume> that you are correct and that is what the reference refers to.  8 I say ruining because if you are correct then this isn't7 a reference for how much their 8400->GS upgrade boosted 8 their throughput. But thats what you origionally claimed0 to have provided with the Bank Austia reference.  7 So you are now claiming that the reference you providedb isn't a reference at all, fine.i  7 Do you get it yet. This is a rock and a hard place withS you in the middle.  4 So what  was the performance improvement from moving to the GS ?S  F > You give it a rest.  There is nothing in my argument that depends onF > these numbers, so it's not necessary for me to show them.  There are@ > some possible answers to your inquiry, including these two for
 > example: > G > 1.  The GS160 upgrade gave them zero improvement, but after upgradingi& > to VMS 7.3 they got 28% improvement. >  > 	ORi > H > 2.  The GS160 upgrade gave them great performance improvement, but the* > VMS upgrade gave them an additional 28%. > E > Both of us are only guessing which it may be, but the difference isn4 > that my argument doesn't depend on either outcome. > E > In any case, they are HAPPY.  Do you have anything to counter that?o >   H No I don't need to bother, you either clarify which of the two scenariosD you refer to and then quantify them or the reference you supplied is useless.    6 >>And you have the idiocity to claim that there are no6 >>performance problems, without yourself being able to5 >>provide on concrete example that stands scrutiny ton >>support your conjecture. >  > D > You would have a point if you didn't lie about my claims.  I claimE > that, in the real world, the benchmarks don't tell the whole story,oH > and that there are happy customers.   But YOU claim that there can notG > be ANY happy customers based on generic information.  I did NOT claim C > that there are no performance problems.  And I gave TWO "concrete F > examples" to support my claim.  And you have nothing to counter them1 > except bland, experimental, generic benchmarks.c > 6 Bullshit, I have never claimed that there are no happy1 GS customers and I defy you to prove that I have.m  6 Why not spend a little more time thinking through your= response and a little less time attempting to put words in myr mouth, you will benefit.  G > The thing is, my position is much easier to prove, and your is easiereF > to disprove.  I only have to find ONE counter argument to support myC > logical argument and disprove yours.  That's why your claim is sol > hopeless.s >   7 Well except that you have totally failed to prove it so 2 far, suggesting that its a bit more difficult than	 it seems.i  G > Why do you spend so much time on this?  You haven't provided anythingeC > to prove that my real claim is wrong.  If you have something thateD > shows these two articles are no longer valid - e.g., the customersD > have switched to something else because they were unhappy with theG > GS160 - then show it.  If not, then you have nothing to counter with.a >   @ Well except for your own customers sending dissapointed messages to this group.   Regardst Andrew Harrison    ------------------------------  % Date: Wed, 16 Apr 2003 11:25:44 +0100t' From: Andrew Harrison SUNUK ConsultancyaY Subject: Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retaia. Message-ID: <3E9D2FA8.7060203@nospamn.sun.com>   Bill Todd wrote:6 > "David Svensson" <icerq4a@spray.se> wrote in message9 > news:734da31c.0304151154.682d0c33@posting.google.com...c > # >>Andrew Harrison SUNUK ConsultancytD >><Andrew_No.Harrison_No@nospamn.sun.com> wrote > I challenge you to- >>trawl back through all the public benchmarka >>A >>>results for the 8400-GS140-GS320, compare them with Sun/HP/IBMt? >>>results for those tests at the same time, you will find thatsB >>>except for period between the introduction of the 8400 and when< >>>Sun introduced the E4500-E10K there is no time when the 37 >>>servers held a performance lead over Sun/HP and IBM.o >>E >>I haven't seen that Sun have been any good on benchmarks the last 4eB >>years. When they are better they have lots of more CPU's and are8 >>compared and still slower to machines with less CPU's. >  > I > Andrew's point likely being that Sun has supported systems sufficientlylN > larger than Alpha's largest systems to maintain an overall, though certainly( > not a per-processor, performance lead. >    No not my point at all.   9 If you take the GS320 as the case in point on most of theo= commercial benchmarks for the system except SPECint and SPECfe< the Sun's at the time had better total throughput and better throughput per CPU.s  ( Oracle Apps/SAP and TPC-H all show this.   Regards  Andrew Harrisone   ------------------------------  % Date: Wed, 16 Apr 2003 15:35:55 +0100u' From: Andrew Harrison SUNUK Consultancy,Y Subject: Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retai . Message-ID: <3E9D6A4B.5020006@nospamn.sun.com>   David Svensson wrote:c# > Andrew Harrison SUNUK ConsultancynD > <Andrew_No.Harrison_No@nospamn.sun.com> wrote > I challenge you to- > trawl back through all the public benchmarkd > @ >>results for the 8400-GS140-GS320, compare them with Sun/HP/IBM> >>results for those tests at the same time, you will find thatA >>except for period between the introduction of the 8400 and whenH; >>Sun introduced the E4500-E10K there is no time when the 3e6 >>servers held a performance lead over Sun/HP and IBM. >  > E > I haven't seen that Sun have been any good on benchmarks the last 4eB > years. When they are better they have lots of more CPU's and areF > compared and still slower to machines with less CPU's. Before Power4  > IBM were not that good either. >   B Well you haven't been looking much further than SPECint and SPECfp then.s  7 4 years ago Sun led for TPC-C, TPC-D, SAP and a raft ofu, other commercial throughput type benchmarks.  8 Sun currently leads TPC-H non clustered, SAP, PeopleSoft! and a number of other benchmarks.e  9 IBM has the leading TPC-C result, it also had the leadingy$ result before Power 4 with the P680.  : The vendor absent from any leadership position in the last- 4 years in any of these benchmarks is Compaq.a   > > >>There is a history around Alpha servers of hugely optimistic; >>performance claims which have never been supported by anyf
 >>collateral.  >  > F > From what I have seen and experienced Alpha has been a very fast CPUG > and many benchmarks have also shown that. Digital/Compaq never reallyaF > advertised the Alpha so the "hugely optimistic performance" are from# > real users solving real problems.t >   9 Ahh you are refering to the Alpha syndrome. This is wheret8 people extrapolate the performance of a large SMP server= from the performance of single CPU running a micro benchmark.2  ; There are plenty of SPECint and SPECfp results that do show < that Alpha is/was a fast CPU. There were also a large number: of benchmarks that showed that this single CPU performance9 never managed to translate itself into throughput for SMPi severs.3  < There were lots of good reasons for this. The Turbolaser was8 bandwidth, memory capacity and I/O limited when compared with most of its competitors.l  > The GS160/320 has a memory subsystem that has a higher average9 memory latency for a lot of scenarios than the Turbolaseri? that it replaced, this coupled with the system being introducedtA with a CPU that had 50% of the eCache when compared with the laste@ GS140 model ment that people experienced lack luster performanceA until they added a significantly larger number of CPU's than they  had in the previous system.i   Regardsm Andrew Harrisonv   ------------------------------  # Date: Wed, 16 Apr 2003 15:56:21 GMT # From: "John Smith" <a@nonymous.com>oY Subject: Re: OpenVMS.org: Marvel article and HP's press release for MarvelandAlpha RetaineH Message-ID: <F2fna.59733$Vzu.30751@news02.bloor.is.net.cable.rogers.com>  # "Andrew Harrison SUNUK Consultancy" 8 <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message( news:3E9D2EA4.8050008@nospamn.sun.com... >n
 >   <snip> >aB > Well except for your own customers sending dissapointed messages > to this group. >m   Andrew,   E It isn't so much a case of technically disappointed customers posting D here as it is a case of marketing and promotion disappointment being posted here.  ? Perhaps there are a large group of technically disappointed SundF customers, but we don't see them in c.o.v, however I'm sure that if weC went searching through newsgroups for evidence of such we'd find itr too.  > Every platform/model from every manufacturer has strengths andC weaknesses. I can point out a Sun financial sector customer who waseE more pleased with an AIX box for a particular application than with at# comparable Sun box. And so it goes.    ------------------------------  % Date: Wed, 16 Apr 2003 18:13:36 +0200 $ From: Michael Unger <unger@decus.de>! Subject: Re: Process Quota Tuningf+ Message-ID: <00A1E7CC.5CE94A05.21@decus.de>e  ( "Main, Kerry" <Kerry.Main@hp.com> wrote:   > Jeff,m > I > Another good option is to use the new Availability Manager (AM) that is J > free with OpenVMS and can be downloaded from the HP OpenVMS web site at: >  > @ > http://h71000.www7.hp.com/openvms/products/availman/index.html > J > You can also dynamically and interactively increase quota's on a running7 > process which is not something DCL procedures can do.  > I > If interested, I can send you some sample screen shots from a large VAXeF > to Alpha project that I just completed last weekend. For performance' > monitoring, we used AM a great deal. P    I What about putting these on the web page mentioned? That would make them t available to all.e   Michaelt   ------------------------------  % Date: Wed, 16 Apr 2003 12:15:06 +0100e' From: "David Gray" <police@spamcop.net>e Subject: Q - Printer Problemsn1 Message-ID: <ZvOcnUaM2-oTpgCjXTWcpg@giganews.com>l   Hi all,r  
 OpenVMS 7.3-1i5 Compaq TCP/IP Services for OpenVMS Alpha Version V5.3  HP LaserJet 4000 - PCL6.  E Recently added a reset module to a VMS print queue for the purpose of & removing blank pages, now if more than7 one print is sent the second and subsequent jobs stall.d  D The printer queue uses the TCPIP$TELNETSYM symbiont and is set up as follows:  G Printer queue ITSERVPR1, idle, on JERRY::"itservpr1:9100", mounted forms WORD11$FORM 2  /BASE_PRIORITY=4 /DEFAULT=(FEED,FORM=WORD11$FORM)A  /LIBRARY=HP Lowercase /OWNER=[SYSTEM] /PROCESSOR=TCPIP$TELNETSYMe  /PROTECTION=(S:M,O:D,G:R,W:S)     Example:8 XTS_JERRY> print /que=itservpr1 /setup=HPLJII_PORT 1.txt* Job 1 (queue ITSERVPR1, entry 161) pending1      pending status caused by queue stopped state 8 XTS_JERRY> print /que=itservpr1 /setup=HPLJII_PORT 1.txt* Job 1 (queue ITSERVPR1, entry 162) pending1      pending status caused by queue stopped stater8 XTS_JERRY> print /que=itservpr1 /setup=HPLJII_PORT 1.txt* Job 1 (queue ITSERVPR1, entry 163) pending1      pending status caused by queue stopped states XTS_JERRY> show entrys   XTS_JERRY> start itservpr14   Entry  Jobname         Username     Blocks  Status4   -----  -------         --------     ------  ------5     162  1               GRAYD             1  Stallede+          On stalled printer queue ITSERVPR1S  5     163  1               GRAYD             1  Pendingr+          On stalled printer queue ITSERVPR1t  I The jobs do eventually print after 5-10 minutes and during this 'Stalled'y( period I'm getting the following message from the OPS console.t  L     TCPIP$TELNETSYM - (ITSERVPR1) open_socket_ast invoked with bad IOSB 660:" connect to network object rejected  K Investigation on the net shows this to be caused by print jobs from Windows 8 not releasing the port after a successful print, and theI workaround being to define the timeout logicals.  I have done this but amh3 still unable to get rid delay due to jobs stalling.r   (LNM$SYSTEM_TABLE)  2   "OPENVMS$TELNET" = "SYS$SYSTEM:TCPIP$TELNET.EXE"#   "TCPIP$TELNETSYM_ENABLE" = ".1.."e2   "TCPIP$TELNETSYM_IDLE_TIMEOUT" = "0 00:00:30.00"!   "TCPIP$TELNETSYM_RAW_TCP" = "1"a,   "TCPIP$TELNETSYM_SUPPRESS_FORMFEEDS" = "1"    "TCPIP$TELNET_ENABLE" = ".1.."0   "UCX$TELNETSYM_IDLE_TIMEOUT" = "0 00:00:30.00"   "UCX$TELNETSYM_RAW_TCP" = "1"/*   "UCX$TELNETSYM_SUPPRESS_FORMFEEDS" = "1"  D Has anyone come across this before and if so do you have a solution?  G in case it matters the setup and reset modules contain the following...n   ******************************% JD1:[GEN_USER.GRAYD]HPLJII_PORT.TXT;1e  $ <ESC>P<ESC>E<ESC>&l0O<ESC>&l0S<ESC>\   ******************************" JD1:[GEN_USER.GRAYD]HP_RESET.TXT;4  # <ESC>]VMS;2<ESC>\<ESC>P<ESC>E<ESC>\s   and the form is defined as  9 Form name                            Number   Descriptioni9 ---------                            ------   -----------r9 WORD11$FORM                              12   WORD11$FORMcF     /LENGTH=72 /MARGIN=(BOTTOM=6) /STOCK=WORD11$FORM /WIDTH=1024 /WRAP   Thanks in advance,  	     Dave.    ------------------------------  # Date: Wed, 16 Apr 2003 15:28:00 GMTt2 From: "frank brown" <frank.brown@ci.seattle.wa.us>, Subject: setting timeout on socket_read (MU)1 Message-ID: <4Eena.15$qW2.4579@news-west.eli.net>s  E I'm running VMS 5.5-2 with MultiNet TCP/IP.  I'm writing an app whichnG creates a socket via socket(AF_INET,SOCK_STREAM,0) and eventually callsdE socket_read(). If no data comes across...the app waits forever at thep socket_read call.   G Is there a way to set a timeout value on the socket?  This is a fragile L state of affairs if the process I'm writing to across the socket doesn't ACK in response.   -Frank Brown Seattle Fire Dept. http://www.inwa.net/~frog/   ------------------------------  % Date: Wed, 16 Apr 2003 19:23:34 +0200h, From: Albrecht Schlosser <ajs856@tiscali.de>0 Subject: Re: setting timeout on socket_read (MU), Message-ID: <li3k7b.6fq.ln@news.hus-soft.de>   frank brown wrote: > G > I'm running VMS 5.5-2 with MultiNet TCP/IP.  I'm writing an app whicheI > creates a socket via socket(AF_INET,SOCK_STREAM,0) and eventually callsnG > socket_read(). If no data comes across...the app waits forever at thew > socket_read call.  > I > Is there a way to set a timeout value on the socket?  This is a fragilesN > state of affairs if the process I'm writing to across the socket doesn't ACK > in response.  F Call select() with your timeout before doing the read (set your socketC in the READFDS argument). Select will time out, or there will be at  least one byte to read.-   ------------------------------  # Date: Wed, 16 Apr 2003 13:49:41 GMTJ# From: "John N." <JNixon@cfl.rr.com>e Subject: software maintenance5= Message-ID: <Vbdna.160639$j8.3507529@twister.tampabay.rr.com>u  L Can anyone shed any light on some sort of new software maintenance scheme HP has, or is planning to have?  K In the past, ES40s, ES45s, DS10s, MVAXes,  were among systems classified as- "workgroup" class.  8 AS2100s,  VAX 4705, GS80 were classified as departmental  ? VAX 7000s,  AS8400, GS140, GS160 were Enterprise class systems.$  : Software maintenance was priced according to class system.  F Now, from what I understand, they want to classify the way you use theH system.  The way I understand it, if you use a DS10 as part of a missionE critical application ,they are going to charge you "Mission critical"G maintenance prices.   J I was planning to upgrade an AS2100 to an ES40 because the maintenance forJ the ES40 would be reduced enough for me to justify the trade up, including1 the purchase cost of the ES40.  Now, it does not.s   ------------------------------  % Date: Wed, 16 Apr 2003 07:18:09 -07004# From: "Tom Linden" <tom@kednos.com>s! Subject: RE: software maintenanceH9 Message-ID: <CIEJLCMNHNNDLLOOGNJIGECBHAAA.tom@kednos.com>>  D Is there an URL reference for this licensing?  Does Mission critical entail a faster response time?   >-----Original Message-----0) >From: John N. [mailto:JNixon@cfl.rr.com]?( >Sent: Wednesday, April 16, 2003 6:50 AM >To: Info-VAX@Mvb.Saic.Com >Subject: software maintenance >3 >GC >Can anyone shed any light on some sort of new software maintenancem
 >scheme HP >has, or is planning to have?  >cL >In the past, ES40s, ES45s, DS10s, MVAXes,  were among systems classified as >"workgroup" class.a >a9 >AS2100s,  VAX 4705, GS80 were classified as departmentalA >2@ >VAX 7000s,  AS8400, GS140, GS160 were Enterprise class systems. >j; >Software maintenance was priced according to class system.t >.G >Now, from what I understand, they want to classify the way you use thenI >system.  The way I understand it, if you use a DS10 as part of a missioniF >critical application ,they are going to charge you "Mission critical" >maintenance prices. > K >I was planning to upgrade an AS2100 to an ES40 because the maintenance for:K >the ES40 would be reduced enough for me to justify the trade up, includingi2 >the purchase cost of the ES40.  Now, it does not. >  >M >  >h >---' >Incoming mail is certified Virus Free.t; >Checked by AVG anti-virus system (http://www.grisoft.com).e@ >Version: 6.0.467 / Virus Database: 266 - Release Date: 4/1/2003 >4 ----& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).? Version: 6.0.467 / Virus Database: 266 - Release Date: 4/1/2003w   ------------------------------  % Date: Wed, 16 Apr 2003 10:06:32 -0500h From: brandon@dalsemi.com ! Subject: Re: software maintenancei1 Message-ID: <03041610063227@dscis6-0.dalsemi.com>   C >Can anyone shed any light on some sort of new software maintenancea
 >scheme HP >has, or is planning to have?s >hL >In the past, ES40s, ES45s, DS10s, MVAXes,  were among systems classified as >"workgroup" class.- >-9 >AS2100s,  VAX 4705, GS80 were classified as departmentalm >R@ >VAX 7000s,  AS8400, GS140, GS160 were Enterprise class systems. >d; >Software maintenance was priced according to class system.  >hG >Now, from what I understand, they want to classify the way you use thevI >system.  The way I understand it, if you use a DS10 as part of a mission-F >critical application ,they are going to charge you "Mission critical" >maintenance prices. >jK >I was planning to upgrade an AS2100 to an ES40 because the maintenance for K >the ES40 would be reduced enough for me to justify the trade up, including 2 >the purchase cost of the ES40.  Now, it does not.  G New?  I just went through the maintenance contracts on our VMS servers. O There are a number of ways to configure the contract - according to your needs.    My experience:  M HP expects that your hardware and software support should match.  If you have:M 7x24 hardware support on your server you should have 7x24 software support on> that server.  I HP expects that storage attached to your hardware should reflect the samerN support as your server.  Which is kind of insulting.  You typically build yourN storage array to have failover redundancy with online spares, dual power, etc.  C HP expects that if your O/S is 7x24, then your LPS support is 7x24.B  L HP expects that your SSS (anchor node) be on your largest machine.  And that- SNS (remote node) be the same level service. x  N HP will not bid against its own VAR;s.  You either have HP bid (solely) or youG have VAR's & Re-sellers bid against one another.  But no HP and VAR's. mF Requires a signed letter on comany letterhead stating that you want HP directly.  9  H Lots of hooks involved.  If you have this, then you have to have that...  O I suggest that you talk with at least 3 VAR's/re-sellers about the mainteenance.L contract.  You will get a different story from each, however you will get in the ball park.  3 Hope this helps - hope this is not to off the mark.0     GOLD HW & SW.% 30 minute response.  Follow up calls.nK Gives you a TAM (some guy who will hold your hand), dedicated team, monthly L reports, ECO notifications, etc.  If you are solid VMS admin you should have. no problem doing the same job - if not better!   SILVER	 Software.0 1 hour response(?).gO Gives you a dedicated team, some reports, etc.  Again, solid VMS admin!  If yougG make a habit of tracking your own calls, etc. you are on top of things!2   BRONZE	 Software.eN Basic support.  No dedicated nothing.  First in the queue, first out, etc.  No; bells, no whistles, just support.  WIS, FTP downloads, etc.e- Can be same day service or next business day.   
 DECservice	 Hardware. $ 7x24 x4 (day x hour x hour-response) 7x24 x9e 5x24 x4  5x24 x9-* Most any combination you can come up with.       John Brandon VMS Systems Administratort Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk5 972.371.4003 fxB   ------------------------------    Date: 16 Apr 2003 06:24:47 -0700) From: P.Young@unsw.EDU.AU (Patrick Young)M/ Subject: Re: TCPIP Services and IP masqueradingm= Message-ID: <55f85d77.0304160524.5b61b463@posting.google.com>I   "John Gemignani, Jr." <jon-nope@thiswontworkossc.net> wrote in message news:<zgmha.21912$TW2.3367433@news1.news.adelphia.net>...N > I've got a cable modem and am using the linksys BEFSR41 as well and it worksM > VERY well.  I have three VMS systems, one NT, 5 Windoz (two are wireless) awL > LAN color laser printer and an HP 50X print server for the inkjet/scanner. >   E Many thanks to everyone for the advice. I'm now typing this using the G Linksys BEFSR41W to my wireless Linux Laptop. Authentication done usingM the OpenVMS box.  C The reason I didn't really want to get into this is (a) Obviously Ib= don't know much about the products available (b) Am too cheap2B (c) They tend to change every 10 seconds (d) The fact that my ISP,< Telstra are a "you *will* use M$ Window(tm), you *will* only: authenticate a single address, Apple Macs maybe used as anA afterthought - and we use a strange authentication protocol" type A camp .... so did not want to buy something that would be useless.w  C They have a good DHCP service though - can get as many IP addressesd as you want.  J > The cable modem is a Motorola -- and an important thing that you need to  < Mine too - "CyberSURFR Wave". Had it for a few years now and? used the BPALOGIN (BIDS2 client) I ported (dragged) to OpenVMS.t  6 The Linksys is pretty sweet and working well, although> I have not worked out if I can get the Linsys to authenticate,= regardless of ftp://ftp.linksys.com.sg/telstra_cable_linksys.-  ; The only thing I didn't like is that I had to buy a Linksysi9 wireless network card at A$106 when the laptop Avaya cardr	 was A$80.-  @ Yes, this is mobile phone stuff - my leg is starting to hurt the@ same way my face does when I go *near* a mobile phone (currently= have laptop at weird angle away from leg - shoved *way* over.>   ------------------------------  + Date: Wed, 16 Apr 2003 15:39:43 +0100 (MET)o9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>e/ Subject: Re: TCPIP Services and IP masqueradingS; Message-ID: <01KUSTDPJG4WA9QUZV@sysdev.deutsche-boerse.com>   H > I've got a cable modem and am using the linksys BEFSR41 as well and itG > works VERY well.  I have three VMS systems, one NT, 5 Windoz (two areoH > wireless) a LAN color laser printer and an HP 50X print server for the > inkjet/scanner.   F I also have a linksys BEFSR41.  It works OK.  However, I haven't been D able to get a DECwindows application running on a remote machine to D display on a VMS machine behind the Linksys.  Since this works with I another router (a re-badged Zyxel) and also works with an ISDN router, I uF am sure that this must be some sort of configuration problem with the  Linksys router.n  D Originally, it wouldn't allow OUTgoing X-windows connections either.D After some emails and a while on the phone with Linksys support (whoF also seem to assume that OF COURSE one is using Windows---actually theC Linksys FAQ in the docu says that it will work with other operating9> systems but is not supported with them, whatever that means) aH suggestion came to change the MTU value to 1492.  Actually, that is whatG it was set at, and is also the default (I think).  I changed it to 1500!@ or something and NOTHING worked anymore, then changed it back toE 1492---apparently what it had been before---and OUTgoing connections e/ started working, but incoming ones still don't.   C I would be interested in hearing from ANYONE using the BEFSR41 (or w? similar) router who can display incoming X-windows connections.A  B Again, it works with two other routers so I am sure that there is ? something wrong with (the configuration of) the Linksys router..   ------------------------------   Date: 16 Apr 2003 16:02:46 GMT1 From: JONESD@er6.eng.ohio-state.edu (David Jones)y/ Subject: Re: TCPIP Services and IP masqueradingr: Message-ID: <b7jur6$2qj$1@charm.magnus.acs.ohio-state.edu>  ; In message <01KUSTDPJG4WA9QUZV@sysdev.deutsche-boerse.com>,,=   Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:tF >I also have a linksys BEFSR41.  It works OK.  However, I haven't beenD >able to get a DECwindows application running on a remote machine toD >display on a VMS machine behind the Linksys.  Since this works withI >another router (a re-badged Zyxel) and also works with an ISDN router, IsF >am sure that this must be some sort of configuration problem with the >Linksys router.  J I'd think you'd have to set the port map on the LinkSys to direct incomingK requests on port 6000 (server 0) to the VMS machine.  Another, more secure,vL way to do it would be to use an SSH connection between the 2 machines with X forwarding enabled.>      < David L. Jones               |      Phone:    (614) 292-6929- Ohio State University        |      Internet: L 140 W. 19th St. Rm. 231a     |               jonesd@er6s1.eng.ohio-state.edu: Columbus, OH 43210           |               vman+@osu.edu  1 Disclaimer: I'm looking for marbles all day long.    ------------------------------    Date: 16 Apr 2003 07:41:29 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)-' Subject: Re: The value of documentation03 Message-ID: <me9rj3LjKPK5@eisner.encompasserve.org>:  b In article <3E9CE29C.7A5809D6@vl.videotron.ca>, JF Mezei <jfmezei.spamnot@vl.videotron.ca> writes:M > One thing that had set VMS apart in its heydays was its complete, accurate,7M > reliable documentation. When you needed to do a task, you opened the manuals5 > and knew that your task would work without a hitch.e >  > L > My latests attempts are getting XDM to work has made me start to wonder if# > that was really the case anymore.h  G    DEC stopped providing complete, accurate, reliable documentation foreF    all things X related a while ago.  They seemed to think that if allC    other vendors can survive with the poor excuse for documentation 0    that's readily available for X, they can too.   ------------------------------  % Date: Wed, 16 Apr 2003 11:55:30 -0400m& From: David M Smith <dsmit115@csc.com>" Subject: Re: TPU: SHOW BUFFER code8 Message-ID: <45vq9vkiprsbd6epmlc5l0qv8u7s8jvb7o@4ax.com>  N On Sun, 13 Apr 2003 04:49:10 -0400, JF Mezei <jfmezei.spamnot@vl.videotron.ca> wrote:  ) >And Flist might prove most useful too...g  O If you've not tried it, then try it out! I downloaded it several years ago, nottN having a clue I would find it so useful, and I used it literally tens of times each day. It is very handy!cI -------------------------------------------------------------------------sI David M. Smith 302.391.8533                       dsmit115 at csc dot com@I Computer Sciences Corporation     (Opinions are those of the writer only)aI -------------------------------------------------------------------------t   ------------------------------  % Date: Wed, 16 Apr 2003 10:51:12 +0100j' From: Andrew Harrison SUNUK ConsultancyD" Subject: Re: What is a VMS Cluster. Message-ID: <3E9D2790.6030602@nospamn.sun.com>   Main, Kerry wrote: > John,  > , > The following article may be of interest.  > I > It discusses shared everything vs shared nothing cluster architectures. B > OpenVMS and MVS (now z/OS) are examples of shared everything (orI > sometimes called shared disk) cluster architectures while UNIX, Windowsr> > and NSK are considered shared nothing cluster architectures. > J > http://archive.infoworld.com/articles/fe/xml/01/12/17/011217feclustertca > .xml >     6 Appart from the fact that its not very accurate thanks for sending it.I  4 The statement that UNIX only supports shared nothing clustering is wrong.  = UNIX (Solaris/AIX/HP-UX/Tru64) all support global filesystemsg2 and device access across the nodes in the cluster.  ? Repeating the well worn OpenVMS mantra that UNIX cannot do thati/ is boring, it has been able to do it for years.   < In addition the article completely ignores what is problably= the largest installed base of HA clusters which is the simplei: failover cluster with A-B nodes where both access the same< storage but only with only one node being able to access the storage at a given time.  : It also confuses the issues with shared nothing and shared( everything with respect to partitioning.  9 Partitioning is required for shared nothing clusters liket/ the high end TPC-C results published by Compaq.   ; Partitioning is also highly desirable for shared everything 8 clusters since it minimises the effects of having a some$ form of distributed locking/caching.  8 The issues with respect to how you partition clients and7 data either for shared nothing or shared everything arei; very similar. The only difference being that you have to do 9 it for shared nothing for it to work at all while you mayd> have to do it for shared everything to get decent performance.  6 Apart from that its a very well researched and written5 article, thanks for sending it out. (did you actuallyn read it ??)6     RegardsF Andrew HarrisonB        	 > Regardsc >  > Kerry Main > Senior Consultanta > Hewlett-Packard (Canada) Co.# > Consulting & Integration Servicese > Voice: 613-592-4660r > Fax   : 613-591-4477 > Email: kerryDOTmain@hpDOTcom/ >     (remove the DOT's and replace with "."'s)t! > OpenVMS DCL - the original .COMt >  >  > -----Original Message-----+ > From: John N. [mailto:JNixon@cfl.rr.com] l > Sent: April 13, 2003 8:55 PM > To: Info-VAX@Mvb.Saic.ComM  > Subject: What is a VMS Cluster >  > H > Of course, I know what a VMS Cluster is.  I have been working with VMSH > since 1981 and with Clusters since 1986.  I guess my problem is that IH > understand them so well that it is difficult for me to explain to PHMsD > what they are, and what their advantage are over other "so-called" > clusters.D > E > Can you guys help me come up with a simple definition of what a VMSs6 > Cluster is, and why it is better than the imposters? >  >    ------------------------------  % Date: Wed, 16 Apr 2003 07:21:24 -0400o' From: "Main, Kerry" <Kerry.Main@hp.com>k" Subject: RE: What is a VMS ClusterT Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF4040ECFAD@kaoexc01.americas.cpqcorp.net>   Ahhh Andrew ..  & Feeling a little touchy these days eh?  B First - where did I say one strategy was better than the other?=20  4 As the article indicated, there are places for both.  D >> UNIX (Solaris/AIX/HP-UX/Tru64) all support global filesystems and1 device access across the nodes in the cluster.>>.0  H So, in this scenario, are you saying all nodes can write directly to theB same block of data at the same time on the same disk (using a lockC manager) or are you saying one node acts as a co-ordinator to serve ! write access to other members?=20U  J As far as data partitioning goes, there are pro's-n-con's here as well.=20  F As the article mentioned, you increase local caching capability at theG expense of having to deal with the potential of one node becoming a hotmH spot while other nodes are idle. You also need to better understand yourB workloads before going live as repartitioning a database to better; spread the loads around can be messy and requires downtime.    But - you knew all this ..  A By the way - while locking does provide some overhead on a sharedhD everything cluster, one needs to put this in perspective with todays hardware.=20  G On a new, but very active 3 node OpenVMS ES45 cluster that went live onwA Monday (hundred's of IO's per second on several RAID devices), weuG weasured the lock traffic at approx 5Mb/sec per port (load balanced) onpH a pair of 100Mb DE602's. So even if one assumes the real throughput of aD 100Mb ethernet device to be in the range of 60Mb, we were still only- using about 10% of the cluster interconnects.r   Regards,  
 Kerry Main Senior Consultantt Hewlett-Packard (Canada) Co.! Consulting & Integration Servicesa Voice: 613-592-4660s Fax   : 613-591-4477 Email: kerryDOTmain@hpDOTcom-     (remove the DOT's and replace with "."'s)0 OpenVMS DCL - the original .COMo       > -----Original Message-----, > From: Andrew Harrison SUNUK Consultancy=203 > [mailto:Andrew_No.Harrison_No@nospamn.sun.com]=20f > Sent: April 16, 2003 5:51 AM > To: Info-VAX@Mvb.Saic.Comy$ > Subject: Re: What is a VMS Cluster >=20 >=20 >=20 >=20 > Main, Kerry wrote:	 > > John,r > >=200 > > The following article may be of interest.=20 > >=20? > > It discusses shared everything vs shared nothing cluster=20u > architectures.D > > OpenVMS and MVS (now z/OS) are examples of shared everything (or@ > > sometimes called shared disk) cluster architectures while=20 > UNIX, Windowse@ > > and NSK are considered shared nothing cluster architectures. > >=20 > >=20@ > http://archive.infoworld.com/articles/fe/xml/01/12/17/011217fe > clustertca > > .xml > >=20 >=20 >=208 > Appart from the fact that its not very accurate thanks > for sending it.e >=206 > The statement that UNIX only supports shared nothing > clustering is wrong. >=20? > UNIX (Solaris/AIX/HP-UX/Tru64) all support global filesystems 4 > and device access across the nodes in the cluster. >=20A > Repeating the well worn OpenVMS mantra that UNIX cannot do thatn1 > is boring, it has been able to do it for years.e >=20> > In addition the article completely ignores what is problably? > the largest installed base of HA clusters which is the simple < > failover cluster with A-B nodes where both access the same> > storage but only with only one node being able to access the > storage at a given time. >=20< > It also confuses the issues with shared nothing and shared* > everything with respect to partitioning. >=20; > Partitioning is required for shared nothing clusters likel1 > the high end TPC-C results published by Compaq.d >=20= > Partitioning is also highly desirable for shared everythinge: > clusters since it minimises the effects of having a some& > form of distributed locking/caching. >=20: > The issues with respect to how you partition clients and9 > data either for shared nothing or shared everything aret= > very similar. The only difference being that you have to dog; > it for shared nothing for it to work at all while you may @ > have to do it for shared everything to get decent performance. >=208 > Apart from that its a very well researched and written7 > article, thanks for sending it out. (did you actuallyu
 > read it ??)  >=20 >=20	 > Regardsl > Andrew Harrisonv >=20 >=20 >=20 >=20 > > Regards, > >=20 > > Kerry Main > > Senior Consultant   > > Hewlett-Packard (Canada) Co.% > > Consulting & Integration Servicesd > > Voice: 613-592-4660t > > Fax   : 613-591-4477  > > Email: kerryDOTmain@hpDOTcom1 > >     (remove the DOT's and replace with "."'s),# > > OpenVMS DCL - the original .COMm > >=20 > >=20 > > -----Original Message-----/ > > From: John N. [mailto:JNixon@cfl.rr.com]=20t  > > Sent: April 13, 2003 8:55 PM > > To: Info-VAX@Mvb.Saic.Coms" > > Subject: What is a VMS Cluster > >=20 > >=20< > > Of course, I know what a VMS Cluster is.  I have been=20 > working with VMS; > > since 1981 and with Clusters since 1986.  I guess my=20u > problem is that Ig= > > understand them so well that it is difficult for me to=20a > explain to PHMs9F > > what they are, and what their advantage are over other "so-called"
 > > clusters.e > >=20G > > Can you guys help me come up with a simple definition of what a VMS 8 > > Cluster is, and why it is better than the imposters? > >=20 > >=20 >=20 >=20   ------------------------------    Date: 16 Apr 2003 07:44:32 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)S" Subject: Re: What is a VMS Cluster3 Message-ID: <eDV7mKZC2GHc@eisner.encompasserve.org>o  X In article <3E9D2790.6030602@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy writes: > ? > UNIX (Solaris/AIX/HP-UX/Tru64) all support global filesystems/4 > and device access across the nodes in the cluster.  A    Wrong.  All these UNIX support NFS, which is not a shared filey+    system, it's a client/server filesystem.a  A > Repeating the well worn OpenVMS mantra that UNIX cannot do thatt1 > is boring, it has been able to do it for years.i  8    It's still true, unless you load TruCluster on Tru64.  = > Partitioning is also highly desirable for shared everythinga: > clusters since it minimises the effects of having a some& > form of distributed locking/caching.  2    Without a DLM, you don't really have a cluster.   ------------------------------  % Date: Wed, 16 Apr 2003 14:27:21 +0100g' From: Andrew Harrison SUNUK Consultancy-" Subject: Re: What is a VMS Cluster. Message-ID: <3E9D5A39.7070200@nospamn.sun.com>   Main, Kerry wrote: > Ahhh Andrew .. > ( > Feeling a little touchy these days eh? > B > First - where did I say one strategy was better than the other?  > 6 > As the article indicated, there are places for both. >  > D >>>UNIX (Solaris/AIX/HP-UX/Tru64) all support global filesystems and >>3 > device access across the nodes in the cluster.>>.o > J > So, in this scenario, are you saying all nodes can write directly to theD > same block of data at the same time on the same disk (using a lockE > manager) or are you saying one node acts as a co-ordinator to served! > write access to other members?    
 The former   > J > As far as data partitioning goes, there are pro's-n-con's here as well.  > H > As the article mentioned, you increase local caching capability at theI > expense of having to deal with the potential of one node becoming a hotsJ > spot while other nodes are idle. You also need to better understand yourD > workloads before going live as repartitioning a database to better= > spread the loads around can be messy and requires downtime.g >  > But - you knew all this .. > C > By the way - while locking does provide some overhead on a sharedlF > everything cluster, one needs to put this in perspective with todays > hardware.  > I > On a new, but very active 3 node OpenVMS ES45 cluster that went live on C > Monday (hundred's of IO's per second on several RAID devices), wehI > weasured the lock traffic at approx 5Mb/sec per port (load balanced) onhJ > a pair of 100Mb DE602's. So even if one assumes the real throughput of aF > 100Mb ethernet device to be in the range of 60Mb, we were still only/ > using about 10% of the cluster interconnects.  >   ? It isn't the bandwidth of the lock traffic thats the real issuev4 its the latency, shame you measured the wrong thing.   Regardso Andrew Harrison   
 > Regards, >  > Kerry Main > Senior Consultantr > Hewlett-Packard (Canada) Co.# > Consulting & Integration Servicest > Voice: 613-592-4660g > Fax   : 613-591-4477 > Email: kerryDOTmain@hpDOTcom/ >     (remove the DOT's and replace with "."'s)n! > OpenVMS DCL - the original .COM  >  >  >  >  >>-----Original Message-----* >>From: Andrew Harrison SUNUK Consultancy 1 >>[mailto:Andrew_No.Harrison_No@nospamn.sun.com] a >>Sent: April 16, 2003 5:51 AM >>To: Info-VAX@Mvb.Saic.Come$ >>Subject: Re: What is a VMS Cluster >> >> >> >> >>Main, Kerry wrote: >> >>>John, >>>w- >>>The following article may be of interest.   >>>S< >>>It discusses shared everything vs shared nothing cluster  >> >>architectures. >>C >>>OpenVMS and MVS (now z/OS) are examples of shared everything (or = >>>sometimes called shared disk) cluster architectures while   >> >>UNIX, Windowse >>? >>>and NSK are considered shared nothing cluster architectures.n >>>n >>>h >>@ >>http://archive.infoworld.com/articles/fe/xml/01/12/17/011217fe >>clustertca >> >>>.xmld >>>n >> >>8 >>Appart from the fact that its not very accurate thanks >>for sending it.M >>6 >>The statement that UNIX only supports shared nothing >>clustering is wrong. >>? >>UNIX (Solaris/AIX/HP-UX/Tru64) all support global filesystems54 >>and device access across the nodes in the cluster. >>A >>Repeating the well worn OpenVMS mantra that UNIX cannot do thats1 >>is boring, it has been able to do it for years.p >>> >>In addition the article completely ignores what is problably? >>the largest installed base of HA clusters which is the simplew< >>failover cluster with A-B nodes where both access the same> >>storage but only with only one node being able to access the >>storage at a given time. >>< >>It also confuses the issues with shared nothing and shared* >>everything with respect to partitioning. >>; >>Partitioning is required for shared nothing clusters likeg1 >>the high end TPC-C results published by Compaq.v >>= >>Partitioning is also highly desirable for shared everythinge: >>clusters since it minimises the effects of having a some& >>form of distributed locking/caching. >>: >>The issues with respect to how you partition clients and9 >>data either for shared nothing or shared everything are = >>very similar. The only difference being that you have to do ; >>it for shared nothing for it to work at all while you may9@ >>have to do it for shared everything to get decent performance. >>8 >>Apart from that its a very well researched and written7 >>article, thanks for sending it out. (did you actuallye
 >>read it ??)m >> >>	 >>Regards  >>Andrew Harrison  >> >> >> >> >>
 >>>Regards >>>d
 >>>Kerry Mainn >>>Senior Consultant >>>Hewlett-Packard (Canada) Co.e$ >>>Consulting & Integration Services >>>Voice: 613-592-4660 >>>Fax   : 613-591-4477o >>>Email: kerryDOTmain@hpDOTcomn0 >>>    (remove the DOT's and replace with "."'s)" >>>OpenVMS DCL - the original .COM >>>l >>>e >>>-----Original Message-----o, >>>From: John N. [mailto:JNixon@cfl.rr.com]  >>>Sent: April 13, 2003 8:55 PMm >>>To: Info-VAX@Mvb.Saic.Com! >>>Subject: What is a VMS Cluster  >>>- >>>-9 >>>Of course, I know what a VMS Cluster is.  I have been 0 >> >>working with VMS >>8 >>>since 1981 and with Clusters since 1986.  I guess my  >> >>problem is that Ip >>: >>>understand them so well that it is difficult for me to  >> >>explain to PHMsa >>E >>>what they are, and what their advantage are over other "so-called"o >>>clusters. >>>fF >>>Can you guys help me come up with a simple definition of what a VMS7 >>>Cluster is, and why it is better than the imposters?e >>>u >>>s >> >    ------------------------------  % Date: Wed, 16 Apr 2003 14:37:11 +0100,' From: Andrew Harrison SUNUK Consultancyd" Subject: Re: What is a VMS Cluster. Message-ID: <3E9D5C87.8080408@nospamn.sun.com>   Bob Koehler wrote:Z > In article <3E9D2790.6030602@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy writes: > ? >>UNIX (Solaris/AIX/HP-UX/Tru64) all support global filesystems14 >>and device access across the nodes in the cluster. >  > C >    Wrong.  All these UNIX support NFS, which is not a shared file - >    system, it's a client/server filesystem.   < Ohh dear of dear. Ever heard of Sun pxfs ?? Obviously not or, Sun QFS or the Tru64 TruCluster Filesystem ?  , I guess much much more research is in order.  7 I would have been less rude had you been less emphatic.    >  > A >>Repeating the well worn OpenVMS mantra that UNIX cannot do thatl1 >>is boring, it has been able to do it for years.  >  > : >    It's still true, unless you load TruCluster on Tru64. >   7 Ahh now we get to it, unbundled software doesn't exist.   6 In which case OpenVMS wasn't able to talk TCP/IP until3 Compaq started bundling the stack with the OS whichl was fairly recently.  9 Fact is Solaris Tru64 etc all support cluster filesystemso as OpenVMS supported TCP/IP.     > = >>Partitioning is also highly desirable for shared everythingh: >>clusters since it minimises the effects of having a some& >>form of distributed locking/caching. >  > 4 >    Without a DLM, you don't really have a cluster. >   7 Well it depends entirely on what kind of cluster we areo talking about.  : Sadly the article left out about 4 other widely used kinds6 of cluster some of which would benefit from a DLM some of which would not.P   Regards" Andrew HarrisonS   ------------------------------   Date: 16 Apr 2003 14:33:05 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)" Subject: Re: What is a VMS Cluster5 Message-ID: <b7jpj0$1midr$1@ID-135708.news.dfncis.de>E  3 In article <eDV7mKZC2GHc@eisner.encompasserve.org>,R> 	koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:Z > In article <3E9D2790.6030602@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy writes: >> p> >> Partitioning is also highly desirable for shared everything; >> clusters since it minimises the effects of having a someC' >> form of distributed locking/caching.* > 4 >    Without a DLM, you don't really have a cluster. >   G And DLM is available for AIX and Linux (that I know of, maybe more) andmF with the release of the code by IBM, on every other Unix very shortly.   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   ,   ------------------------------  % Date: Wed, 16 Apr 2003 16:13:48 +0100u( From: Nic Clews <sendspamhere@127.0.0.1>" Subject: Re: What is a VMS Cluster) Message-ID: <3E9D732C.CDCC0134@127.0.0.1>.  ( Andrew Harrison SUNUK Consultancy wrote: >   > > Ohh dear of dear. Ever heard of Sun pxfs ?? Obviously not or. > Sun QFS or the Tru64 TruCluster Filesystem ? > . > I guess much much more research is in order.  G I searched Sun's web site in UK and USA. I found QFS but no overview ofg, pxfs (but I did find the documentation set).  
 Any pointers?    -- d? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciencesp nclews at csc dot com    ------------------------------  % Date: Wed, 16 Apr 2003 17:17:42 +0100 ' From: Andrew Harrison SUNUK Consultancyd" Subject: Re: What is a VMS Cluster. Message-ID: <3E9D8226.8050209@nospamn.sun.com>   Nic Clews wrote:* > Andrew Harrison SUNUK Consultancy wrote: >  > > >>Ohh dear of dear. Ever heard of Sun pxfs ?? Obviously not or. >>Sun QFS or the Tru64 TruCluster Filesystem ? >>. >>I guess much much more research is in order. >  > I > I searched Sun's web site in UK and USA. I found QFS but no overview ofi. > pxfs (but I did find the documentation set). >  > Any pointers?  >   / pxfs is the layer that globalises UFS and otherK/ filesystems. We are currently qualifying Oracleu" using UFS and VxFS and this layer.   Regardse Andrew Harrison)   ------------------------------  % Date: Wed, 16 Apr 2003 13:26:07 -0400-* From: "Bill Todd" <billtodd@metrocast.net>" Subject: Re: What is a VMS Cluster2 Message-ID: <4PCdnfjXoZmyDwCjXTWcqw@metrocast.net>  H "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in message- news:eDV7mKZC2GHc@eisner.encompasserve.org...tF > In article <3E9D2790.6030602@nospamn.sun.com>, Andrew Harrison SUNUK Consultancy writes:  > >sA > > UNIX (Solaris/AIX/HP-UX/Tru64) all support global filesystems,6 > > and device access across the nodes in the cluster. >0C >    Wrong.  All these UNIX support NFS, which is not a shared filem- >    system, it's a client/server filesystem..  L Wrong, Bob.  AIX (GPFS) and Tru64 (Cluster File System) *definitely* supportL true cluster file systems (that don't use NFS or NFS-style facilities in anyJ manner), and I'm pretty sure that Sun now does as well (though it's been aF while since I checked and back then there wasn't sufficiently-detailedF information to be absolutely certain).  For that matter, so does Linux (GFS).  Don't know about PH-UX.c   >fC > > Repeating the well worn OpenVMS mantra that UNIX cannot do thata3 > > is boring, it has been able to do it for years.e ><: >    It's still true, unless you load TruCluster on Tru64. >D? > > Partitioning is also highly desirable for shared everythingR< > > clusters since it minimises the effects of having a some( > > form of distributed locking/caching. >s4 >    Without a DLM, you don't really have a cluster.  I If you won't take my word that that's rubbish, you should take it up withe' Greg ("In Search of Clusters") Pfister.y   - bill   ------------------------------  % Date: Wed, 16 Apr 2003 10:17:05 +0100 * From: "John Travell" <john@travell.uk.net>" Subject: Re: What is the downside?5 Message-ID: <b7j7c8$1gh29$1@ID-120847.news.dfncis.de>s  4 "Howard S Shubs" <howard@shubs.net> wrote in message6 news:howard-798DD0.23505515042003@enews.newsguy.com...4 > In article <7FidncgYCLLDOgGjXTWcog@metrocast.net>,. >  "Bill Todd" <billtodd@metrocast.net> wrote: >aJ > > Ancient internal code that assumes that a (512-byte) block number will fit,L > > in 32 bits (and that might, or might not, have trouble handling a numberJ > > that appeared to be 'negative', hence limiting the nominally-supported rangeg
 > > to 1 TB).t >nH > Ah, so it's not recent.  Guess it's time to change it.  ODS-6, anyone? >tG Nope, I'd hazard a guess that this is at least as old as V3, possibly a L day-1 limitation, in the same vein as the "how could anyone conceivably wantJ more than 512Mb of memory" assumption built into the early versions of VAX VMS.B Regrettably I no-longer have my inside track to engineering... :-(   -- John Travell  VMS crashdump expertise for hire john@travell.uk.neti http://www.travell.uk.net/       ---t& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.443 / Virus Database: 248 - Release Date: 10/01/2003n   ------------------------------    Date: 16 Apr 2003 03:09:19 -0700) From: jbrankin@ntlworld.com (Jim Brankin)f" Subject: Re: What is the downside?= Message-ID: <863f19d6.0304160209.7a9b2bf8@posting.google.com>0  B Howard S Shubs <howard@shubs.net> wrote in message news:<howard-> H > Ah, so it's not recent.  Guess it's time to change it.  ODS-6, anyone?  # They seem to be working on it. See s  I http://groups.google.com/groups?selm=3e3ff9c2%241_2%40hpb10302.boi.hp.comt   ------------------------------  % Date: Wed, 16 Apr 2003 18:52:50 +0100s4 From: "Kevin Playford" <k.playford@nospamvirgin.net>Y Subject: Re: Why did Digital go broke? (Was: Re: VMS Source Listing CD - Table of contentc4 Message-ID: <b7k595$frt$1$8302bc10@news.demon.co.uk>   Richard,  D Contrary to what you may believe LMCP does NOT make use of the LM$SS
 routines thate= you have identified as being part of the DTI$SHARE shareable.   H LMCP was part of DECdtm's own log manager facility and is fully aware of what the on-diskL format of the DECdtm transaction log is. It does not need to make use of any services providedpB through DTI$SHARE in order to format up the contents of the DECdtm transaction log.  D It does make use of $GETDTI and $SETDTI itself as part of the repair functionality provided3 by the LMCP command line interface, but that is it.r  I I take it that your beef concerns the fact that GETDTI does not return ton you the full RM participant I name in the transaction information block that it returns to you during at wildcard GETDTI.  J However, I don't believe that that is your only concern. What you'd really  like are functional enhancementsC that let you get at the full RM name that could be stored in the RM. participant chunk of a DECdtmtJ transaction record. I guess what you're hoping is that that is anything up to 255 characters.   Is this correct?   regards  Kevin Playford Software Exploration Limited    > "Richard Maher" <maher_rj@hotspamnotmail.com> wrote in message* news:b74hol$l08$1@sparta.btinternet.com... > Hi Brian,s >hJ > > Let me know if you really need the argument calling format (that ought to > be0 > > worth a fortnight stay in a public house ;). > K > I've got three different people over here, who swear that they've got one0K > somewhere, going through their basements and attics as we speak. How much  do > they cost these days?h >)J > But to answer your question directly - "Yes please! I really do need the > argument calling format."  >pK > It's a *very* long story (I'll attach some of it at the end of this note)oI > but to cut a long story short VMS engineering are treating me like shita and I > refusing to even countenance the idea of fixing a bug with $getdti thath willJ > allow me to bring some much needed functionality to VMS that will enable any,F > application written with *any* middleware to obtain bullet-proof 2PC bewteeneJ > Windows2000 MTS/DTC and VMS DECdtm, TIER3 hotTIP and (Rdb, DBMS, RMS) or, > other RM like Oracle 9i via the XA veneer. >xH > So I'm trying to bypass the snootiness of VMS engineering by hopefullyL > calling the API that LMCP uses to access the DECdtm log file directly. (ie6 > *not* $getdti) My best guess is the LM$SS* routines. >-J > Now while I'm more than happy to pay up (as long as I can QA the payment :-)tJ > and I'm also happy to take advantage of your kind offer, I have to point out5K > that this will not be a 15 minute job. So I won't feel bad if you have top
 > decline. >sI > I need the argument formats and passing mechanisms and lm$ssread_by_tidnL > should be asynchronous (or have an async form) and there should be a macroH > (that I couldn't find) for the record layout that looks something like > this:- >w@ > LMCP> dump/active SYS$COMMON:[SYSEXE]SYSTEM$TIER3.LM$JOURNAL;1+ > ****HEADER BLOCK? Block 1? Block ALQ?****oF > Dump of transaction log SYS$COMMON:[SYSEXE]SYSTEM$TIER3.LM$JOURNAL;1' > End of file block 402 / Allocated 402s > Log Version 1.0o= > Transaction log UID:   33595820-07D8-11D3-947E-08002B9AA3FCb. > Penultimate Checkpoint:    000000018A39 0039. > Last Checkpoint:           000000018E78 0078 > ****BY TID STUFF****" > Transaction state (1):  PREPAREDC > Transaction ID: BF73B3A0-4DB1-11D7-A016-08002B9AA3FC ( 3-MAR-2003l > 19:53:04.73)! > DECdtm Services Log Format V1.1tJ > Type ( 3): LOCAL RM         Log ID: 0000011F-0001-0000-0000-000000000000/ > Name (32): "Y..USER_DISK   ...G............." F >      (00000000 00010001 00000005 02470000 00202020 4B534944 5F524553 > 550200DD) J > Type ( 5): FOREIGN TM       Log ID: 00000000-0000-0000-0000-000000000000
 > Name (111):y >SL "T3$Exceptionally_Large_string_name_geater_than_32_characters***************' > ************************************"kD >      (2A2A2A 2A2A2A2A 2A2A2A2A 2A2A2A2A 2A2A2A2A 2A2A2A2A 2A2A2A2A 2A2A2A2A: > 2A2A2A2A 2A2A2A2A 2A2A2A2A 2A2A2A2A 2A2A2A2A 73726574 63G > 617261 68635F32 335F6E61 68745F72 65746165 675F656D 616E5F67 6E697274t7 > 735F6567 72614C5F 796C6C61 6E6F6974 70656378 45243354w > = > Total of 1 transactions active, 1 prepared and 0 committed.  >fI > As I said, it's probably not gonna be that easy to glean but if you cana helpE > at all that would be great. But if not I still owe you those beers!w >  > Cheers Richard >-. > <VAXman- @SendSpamHere.ORG> wrote in message, > news:00A1E0C6.53836E81@SendSpamHere.ORG...C > > In article <b6socc$j7a$1@titan.btinternet.com>, "Richard Maher"m' > <maher_rj@hotspamnotmail.com> writes:Q > > >Hi Brian, > > >r > > >Thanks very much for that!d > > >a: > > >More than happy to pay up next time you're over here. > > >a > > >Cheers Richard. > >TL > > Marillion are planning a world tour next year which will likely rule out > therK > > yearly weekend convention in 2004, so I might not be able to collect my  > com- > > pensation until 2005. :) > >wJ > > Let me know if you really need the argument calling format (that ought to > be0 > > worth a fortnight stay in a public house ;). > >e > >e > > --6 > > VAXman- OpenVMS APE certification number: AAA-0001 > VAXman(at)TMESIS(dot)COM > >a8 > >   "Well my son, life is like a beanstalk, isn't it?" > >d >  > The story thus far:- >n >t > ----- Original Message ----- > From: Richard's Hotmailt > To: oraclerdb@jcc.coml) > Sent: Wednesday, March 26, 2003 2:29 PMn. > Subject: Another 2PC bug - I need your help! >a >t > Hi,w > J > [I'm probably not best known for singing the praises of Rdb engineers inI > this forum (although I hope my devotion to, and awe of, the product has K > never been brought into question) so I'd just like to take this moment to J > say thanks to those engineers who in recent months have responded to theL > bugs (or queries) I have highlighted here. Sometimes they responded with aD > fix but almost always with an investigation and for that I am very	 grateful!oL > My problem is that I have always taken this for granted. Assuming that theJ > bugs that affect me were the most important and that I was in a position toG > demand engineering resources be immediately made available due to ther sheer-J > righteousness of my cause. (As if they'd have anything more important toF > do?) But that's enough gushing for this year. Back to the plot. . .] >mI > This is not a bug that Rdb engineering can fix (so the above was not an K > attempt to curry favour) but please read on anyway so that one of you outgK > there with a VMS support agreement may see my frustration levels peaking,1L > take pity on me and raise an IPMT. (I assume that acronym is not a wind upA > and refers to some form you have to fill in to raise a VMS bug)t >L. > SYS$GETDTIW - The root cause of the problem. >rF > In the output "itmlst" argument of a successful call to $getdti, the; > dti$b_part_name_len and the dti$t_part_name fields of thesI > dti$_transaction_information record are populated *NOT* with the valuesr thatJ > have been retrieved from the DECdtm log file but instead with the values9 > that were specified for their counterparts in the inputt= > dti$_search_resolved_state record of the "search" argument.. >nF > When I call this service I specify "T3$" as the first three wildcardL > characters of my participant name because I'm only interested in TIER3 TIPI > transactions. The problem is that I *need* to see the other 29 bytes of+L > lovely participant name info. Once again, If I only specify three bytes asJ > input then I only get three bytes output regardless of the fact that the3 > participant name is 26 bytes long. THIS IS A BUG!: >eJ > What the system service documentation says for these two output fields:- > C > DTI$B_PART_NAME_LEN A byte containing the length of the followingp* > participant name field  DTI$T_PART_NAME.B > DTI$T_PART_NAME A character field containing DTI$B_PART_NAME_LEN
 characters- > that specifies a resource manager instance.  >nK > Now sys$getdtiw lives in sys$share:dti$share.exe so as far as patches go, F > what could be easier than "Anyone who wants TIP functionality has to install H > a new version of this shareable." Rdb engineering doesn't put anything other,I > than the facility code in the participant name so they shouldn't have aeL > problem with regression testing. Can anyone honestly conceive of an easier > VMS bug to fix or release? >lJ > Why do I need this? (and why you should be interested!) - *NO ADDITIONALI > LOG/JOURNAL FILE* required for TIP functionality! Unlike the XA Gatewayu you:G > would not have to configure and manage another journal file and *most7J > importantly* (given read-check and write-check) you'd be saving four I/O perdI > transaction. How do I achieve this minor miracle? Well, part of the TIPvL > standard stipulates that if the link to the MTS/DTC coordinator dies afterL > TIER3 and Rdb have prepared a 2PC txn (and the outcome of that transactionJ > was abort) then MTS/DTC is under no obligation to get back in touch withK > TIER3's TIP Transaction Manager and tell us. Therefore, I have to attempt  toK > contact MTS/DTC periodically until a connection can be re-established. IffL > W2K no longer knows about the transaction then the outcome was Abort, else I-L > have to wait for the transaction to be reconnected. With "T3$" + OleTxId(1K > byte) + Binary MTS TID (16 Bytes) + IP Addr(4 bytes) + Port# (2 bytes) weVL > have plenty of change left over from 32 bytes. (probably enough to stretch
 > to IPv6) > L > You're right. It *is* bloody clever! The one snag at the moment is that it > just doesn't work :-)c >tI > So it was at this point that I approached VMS engineering with the samerH > gusto and zeal that you've all become accustomed to, only to have them lookL > at me like I was a skid-mark on a hotel towel, and send me on my way. WhatL > would you expect? After all I am self-funding and developing software *on*F > VMS exclusively *for* VMS. They must have thought I was some sort of nutter!t > = > So do I spend the next few weeks implementing and testing aiL > Prepared-Transactions journal or just go into Beta with my software in theK > hope that someone has to see sense sooner or later? Or is it too late andr noI > one really cares anymore? (Everything else is fully tested and ready tot go!) >s > Regards Richard Maher. > K > FYI. There is latent support in DECdtm, and the log file, for participantrL > names greater than 32 bytes. See example below. (Maybe even enough for the > XA_SWITCH structure?)t >nL > If someone has the VMS source listing CD (and $getdti is on it) could theyK > check if there is an alternative to dti$_search_resolved_state that mightrI > return the *full* participant name with the current version of $getdti?hL > Please? The $dtidef macro mentions dti$_search_td and dti$_search_rm_name.9 > If these functions are active, what are the parameters?e >hK > I think a new item code would be very desirable but if I can't even get as& > bug fixed then it's just fantasy :-( >a >  > ----- Original Message ----- > From: Norman Lastovica > To: oraclerdb@jcc.comf) > Sent: Wednesday, March 26, 2003 4:07 PMs2 > Subject: RE: Another 2PC bug - I need your help! >e > L > I've entered a PTR for OpenVMS.  I hope that this was to be reported based+ > on VMS V7.3-1 with all patches installed.t >k > ----- Original Message -----6 > From: Norman Lastovica <norman.lastovica@oracle.com># > To: oraclerdb <oraclerdb@jcc.com> ) > Sent: Wednesday, April 02, 2003 1:05 AMn3 > Subject: SYS$GETDTI return field lengths and datac >6 >54 > > The PTR that I entered for OpenVMS in regards to2 > > Richard's observations about the functionality2 > > and behavior of the $GETDTI system service has< > > been closed.  Here is the text from OpenVMS engineering.6 > > If anyone would like this to behave in a different8 > > fashion, I'm afraid that you'll need to get in touch< > > with  HP directly and make your business case with them. > >n > > Dialog text: > >1H > > I have researched the code and determined that this behavior is veryL > > deliberate; the code was designed to do this, it is not a bug. Since the? > > original designers/coders are no longer available, why thisD
 functionality  > > was L > > designed like this cannot be answered. I am closing this because it is aJ > > request for an enhancement, which may be considered for inclusion in a > later1 > > release. > >  >0L > -------------------------------------------------------------------------- > --& > > Problem attributes for 75-13-1038:9 > >    Abstract: SYS$GETDTI return field lengths and data.? > >    Severity:     MED       Status: CL        Resolution: UI ? > >    Problem type: QAR       Origin: WEB       Workaround: Noi? > >    Reproducible: Yes       Public: Yes       Waiting:    NoX+ > >    Submitted on 26-MAR-2003 08:52:54.20e/ > >    Last modified on 26-MAR-2003 15:03:01.50  > >    Domain:  OPENVMSr > >    Folder:  PRIOR_RELEASES > >    Product: OPENVMSo > >    Version: V7.3-1 > >    Component:    DDTM  > >    Architecture: ALPHA > >    Number of attachments: 0n > >t > > Notification Events: > >    Problem report is closeda > >r > >n >  > ----- Original Message -----0 > From: Richard's Hotmail <maher_rj@hotmail.com> > To: <oraclerdb@jcc.com>l& > Sent: Monday, April 07, 2003 9:42 PMK > Subject: Brigadoon Syndrome (Was: Re: SYS$GETDTI return field lengths and. > data)  >d >  > > Hi Norm, > >oL > > For a fleeting moment there the world was on my side and people listened > to,aD > > and agreed with, what I was saying and offers of assistance were arrivingH > > with the bouquets. Oh well, it was magical while it lasted. Maybe in	 > anotherr > > hundred years eh?  > >r- > > > Sent: Wednesday, April 02, 2003 1:05 AM  > >eH > > Thanks again for that! (And if anyone ever says that you hate DECdtm > around > > me, I'll fight 'em :-) > >s8 > > > If anyone would like this to behave in a different: > > > fashion, I'm afraid that you'll need to get in touch> > > > with  HP directly and make your business case with them. > >tK > > Ah, that famous Norman "Pit Bull" Lastovica tenacity that I've heard son > muchI > > about. As long as you didn't give up without a fight, that's the main  > thing! > >t- > > >    Submitted on 26-MAR-2003 08:52:54.20 1 > > >    Last modified on 26-MAR-2003 15:03:01.50r > > K > > See? That's exactly the look I was telling you about! (And then the PTR  > wasi2 > > sat on for a couple of days for good measure?) > >aJ > > > I have researched the code and determined that this behavior is veryD > > > deliberate; the code was designed to do this, it is not a bug. > >S= > > Sorry? You WHAT? Let me just pause for a moment.  .  .  .n > > J > > > I have researched the code and determined that this behavior is veryD > > > deliberate; the code was designed to do this, it is not a bug. > > + > > Nope, that's definitely what she wrote.a > >vH > > I am familiar with the concepts of "My Country, Right or Wrong!" and > "PapalG > > Infallibility" but this is the first time that I've come across theu
 > argumentL > > "It's not a bug, because we all thought it was a bloody good idea at theF > > time!". Let that be a lesson to all the kiddies out there. If your reallyC > > gonna stuff things up then as long as you do it with convictionI
 everything > > will be all right. > >iL > > (Blown the budgie's head off with Dad's gun? "Did you do it deliberately5 > > son?" "Sure did Ma!" "That's ok then. Run along.)s > > L > > "Oh it's so easy for Mr. Maher to turn up with his revisionist year 2003H > > morality and try and pigeon hole what we did with judgemental labels such > asJ > > "BUGS" but in the heat of coding it's not always as simple as right orH > > wrong, black or white. We've lost some good men and women since this codeL > > was written and I for one am not about to try and second guess what theyK > > were thinking when they pulled the trigger. So Maher can take his 20/20 & > > hindsight and shove it I tell ya!" > >n > > Why did Digital go broke?e > >dI > > Anyway let's leave the Twilight Zone for a moment and get back to theb > plot.d > > . .  > >tK > > I  fear  I've muddied the waters somewhat by telling everyone when, whyi > and J > > what  I  want to store in the participant name. So can we please leave > thatH > > to  one  side  for  a  minute  and  look specifically at the $getdti > system6 > > service,  and  its  documentation,  in  isolation? > >c: > > Here is a snippet from the documentation for the inputF > > dti$_search_resolved_state record for the "search" argument to the > > sys$getdti service: -= > > 3 > > | DTI$T_PART_NAME  A character field containingn6 > > | DTI$B_PART_NAME_LEN characters that specifies  |6 > > | a resource manager name. When the resource     |6 > > | manager name string is supplied, a wildcard    |, > > | search can be specified. The left-most6 > > | characters supplied in this string will be     |6 > > | matched against all resource managers with the |6 > > | same leftmost characters. If the string        |6 > > | entered has a length of 0, all resource        |  > > | managers will be selected. > >-B > > If I am writing a transaction reporting tool (or an unresolved transactioneI > > repair utility (or whatever you choose)) please tell me exactly HOW IaA > > obtain the participant name to show the user so that they canS investigate L > > whether or not to commit or abort? If, just like the above documentationI > > suggested, I have entered a length of 0 (so I can select ALL resourceiK > > managers) then HOW can I tell which ones I've selected? Maybe I want to J > > sort on participant name and do a level break with a transaction count foru' > > each facility. HOW would I do that?o > >wL > > If I want to retrieve all Rdb and RMS transactions I will enter "R" as aJ > > wildcard participant name and set the length field to 1. HOW do I tell if/ > > the transaction returned is for Rdb or RMS?v > > L > > What if my software is multi-version (like Rdb) and there is a change inL > > behaviour regarding how my software deals with transactions between V1.0I > > and V2.0? I could set up participant names of fac$v10 and fac$v20 and6 whenF > > I'm doing a $getdti I can use its lovely wildcard functionality to lookupG > > just fac$ and then I can choose a code path based on whether it's a 	 > versioni4 > > 1.0 or 2.0 transaction. Sounds good! Doesn't it? > >rK > > Please tell me "What is the point of wild carding if there is no way tod! > > obtain the resultant name???"r > >nF > > Imagine a database where you can select all employees with surname
 > startingJ > > with 'SM'. The problem here is this RDBMS may well return 100 rows butG > > instead of getting a list of surnames like SMITH and SMYTHE you aree simply" > > told that there all called SM. > >-F > > If you look at the latest version of the DECdtm Manual as recently revisedlF > > (from  the  old  V1.2  Functional  Spec)  by Jim Johnson's crew at
 > SoftwareI > > Exploration (or the VMS documentation web site) you can see that theyhG > > deliberately changed  the  documentation for the PART_NAME argumenthA > > to both the $JOIN_RM and  $DECLARE_RM system services to readrD > > "The string must be no longer than 32 characters" down from 256. > > (See comments fo 6-Jun-01) > >cI > > I  can  live  with this limit of 32 bytes; please tell me that it wasi > neverdK > > their  intention  for  the  participant  name to be a Write Once - Readm > Zero > > construct. > >e > > Cheers Richard Maher.d > >sJ > > PS. Has anyone ever seen Ian Smith and Carey Donat in the same room at the K > > same time? (See, I just don't make any friends with comments like that!  > But ( > > hands up anyone who didn't smile :-) > >c >g >p >l   ------------------------------  % Date: Wed, 16 Apr 2003 10:57:59 +0100m* From: "Richard Brodie" <R.Brodie@rl.ac.uk>% Subject: Re: XDM problems (TCPIP 5.3)o+ Message-ID: <b7j9f7$k72@newton.cc.rl.ac.uk>i  = "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in messagey) news:3E9BCB62.46515EF5@vl.videotron.ca...n  L > The MI/X server on the mac, upon receiving a "willing" offer from the vax,: > makes an official request, and in that request there is: >c >     0 connection adresses   A 0 addresses! No surprise that you get a "No valid address" reply.n  0 > Oh, and I had a hell of a time using TCPTRACE.  = Yes. Do yourself a favour and use a proper protocol analyser.o   ------------------------------  # Date: Wed, 16 Apr 2003 07:19:39 GMTn6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER): Subject: [OpenVMS Alpha] How to get image version in DCL ?2 Message-ID: <fu7na.63959$UR.505049@news.chello.at>  C Before I wade through a decade of google postings, I'd like to ask:   F Does anyone have a good/simple (means F$CVUI, F$FAO, ...) DCL solutionG for getting Image Identification Information of .EXE files ? On Alpha ?k  K Do you know of a freeware tool (MACRO, FORTRAN, PASCAL, C, ...) to get such I information (image name, image file identification, ...) in DCL symbols ?O  K I know there was such a discussion (with examples/solutions) about a decade K ago, but it was for VAX. Did anyone update the mentioned tools since then ?>   TIA:   -Peter  M PS: Yes, I'll check freeware&google, too, but usually it is faster to ask ;-)b -- . Peter "EPLAN" LANGSTOEGERs% Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  % Date: Wed, 16 Apr 2003 08:46:37 +0100>0 From: Chris Sharman <chris.sharman@sorry.nospam>> Subject: Re: [OpenVMS Alpha] How to get image version in DCL ?4 Message-ID: <b7j1p0$duq$1$8300dec7@news.demon.co.uk>    Peter 'EPLAN' LANGSTOEGER wrote:E > Before I wade through a decade of google postings, I'd like to ask:c > H > Does anyone have a good/simple (means F$CVUI, F$FAO, ...) DCL solutionI > for getting Image Identification Information of .EXE files ? On Alpha ?   " $ open/read i 'f$parse(p1,".exe")'
 $ read i h	 $ close is/ $ if f$length(h).ne.512 then $ goto novms_imager $ if f$cvui(0,32,h).lt.16e $ then $	arch = "Alpha" $	debug_byte = 80s) $	hi = f$extract(f$cvui(24*8,32,h),104,h)w
 $	imgtimo = 8t- $	imgnam = f$extract(17,f$cvui(16*8,8,hi),hi)a, $	imgid = f$extract(57,f$cvui(56*8,8,hi),hi)- $	linkid = f$extract(73,f$cvui(72*8,8,hi),hi)o $ else $	arch = "VAX" $	debug_byte = 32c' $	hi = f$extract(f$cvui(6*8,16,h),80,h)c) $	imgnam = f$extract(1,f$cvui(0,8,hi),hi) , $	imgid = f$extract(41,f$cvui(40*8,8,hi),hi) $	imgtimo = 56- $	linkid = f$extract(65,f$cvui(64*8,8,hi),hi)s $ endifs" $ debug = f$cvui(debug_byte*8,1,h)$ $ share = f$cvui(debug_byte*8+6,1,h)  6 $ say arch," ",imgnam," ",imgid," by linker ",linkid,-< 	" at ",f$fao("!%D",imgtimo+f$cvui(32,32,f$fao("!AD",8,hi)))C $!^^^^^^^^^^^^^^^^^^^^^^ TIME DCL MAGIC ^^^^^^^^^^^^^^^^^^^^^^^^^^^e? $! Explanation: hi is a string & therefore passed by descriptoroG $!	f$fao("!AD",n,s) expects a string by reference, & therefore returns  / the binary descriptor of hi as an 8 byte stringo@ $!	f$cvui(32,32,desc) returns the actual address of hi from the  descriptor (2nd longword)u2 $!	add whatever offset into the record is required@ $!	f$fao("!%D",a) expects a quadword by reference, and gets the 0 carefully constructed address of a quadword time   ------------------------------  % Date: Wed, 16 Apr 2003 09:48:44 +0200d: From: Karl Rohwedder <extern.karl.rohwedder@volkswagen.de>> Subject: Re: [OpenVMS Alpha] How to get image version in DCL ?/ Message-ID: <b7j1sh$oeg1@doiweb4.volkswagen.de>a  " Peter 'EPLAN' LANGSTOEGER schrieb:E > Before I wade through a decade of google postings, I'd like to ask:e > H > Does anyone have a good/simple (means F$CVUI, F$FAO, ...) DCL solutionI > for getting Image Identification Information of .EXE files ? On Alpha ?> > M > Do you know of a freeware tool (MACRO, FORTRAN, PASCAL, C, ...) to get such:K > information (image name, image file identification, ...) in DCL symbols ?  > M > I know there was such a discussion (with examples/solutions) about a decade>M > ago, but it was for VAX. Did anyone update the mentioned tools since then ?p >  > TIAd >  > -Peter > O > PS: Yes, I'll check freeware&google, too, but usually it is faster to ask ;-)    What about a simple PIPE, e.g.  1 $ pipe ana/ima sys$System:copy.exe/sele=build | -yB   (read sys$input rec1 ; read sys$input rec2 ; def/job item &rec2) $ id = F$trnlnm("item"     -- e  + mit freundlichen Gren | with best regards-    Karl Rohwedder		iT-Ingenieurteam$ Ellernbruch 11		D-38112 Braunschweig  E mailto:it-ingteam(at)t-online.de | mailto:rohwedder(at)decus.decus.de-- mailto:extern.karl.rohwedder(at)volkswagen.deo   ------------------------------  % Date: Wed, 16 Apr 2003 09:53:58 +0100s( From: Nic Clews <sendspamhere@127.0.0.1>> Subject: Re: [OpenVMS Alpha] How to get image version in DCL ?) Message-ID: <3E9D1A26.F222E5EF@127.0.0.1>e    Peter 'EPLAN' LANGSTOEGER wrote:  H > Does anyone have a good/simple (means F$CVUI, F$FAO, ...) DCL solutionI > for getting Image Identification Information of .EXE files ? On Alpha ?o  6 On systems with DECwindows installed, I frequently use  " @sys$update:decw$get_image_version  1 OK, I know it doesn't dump it to a symbol, but...m   --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciencesn nclews at csc dot comn   ------------------------------  % Date: Wed, 16 Apr 2003 09:56:03 +0100e* From: "Richard Brodie" <R.Brodie@rl.ac.uk>> Subject: Re: [OpenVMS Alpha] How to get image version in DCL ?+ Message-ID: <b7j5r3$kva@newton.cc.rl.ac.uk>l  C "Peter 'EPLAN' LANGSTOEGER" <peter@langstoeger.at> wrote in messagen, news:fu7na.63959$UR.505049@news.chello.at...    H > Does anyone have a good/simple (means F$CVUI, F$FAO, ...) DCL solutionI > for getting Image Identification Information of .EXE files ? On Alpha ?   M For later versions: anal/image/select=ident is probably simplest. If you have 8 DCPS installed, DCPS$GET_VERSION.COM is a good template.   ------------------------------  % Date: Wed, 16 Apr 2003 09:38:10 -0400a From: norm.raphael@metso.com> Subject: Re: [OpenVMS Alpha] How to get image version in DCL ?? Message-ID: <OFB27ED335.BCEAB2D5-ON85256D0A.004A8762@metso.com>i   Chris,  K Thanks for this, although it is obviously a section cut from a larger wholeo (say! is not defined within this code).a   You might want to add:  I $                bldid = f$extract(89,f$cvui(88*8,8,hi),hi)     !build ide (Alpha only)  D and adjust the output since AFAIK this is not part of the VAX image.   -Normn  H From:  Chris Sharman <chris.sharman@sorry.nospam> on 04/16/2003 03:46 AM  < Please respond to Chris Sharman <chris.sharman@sorry.nospam>   To:    Info-VAX@Mvb.Saic.Com cc:t  A Subject:    Re: [OpenVMS Alpha] How to get image version in DCL ?i      Peter 'EPLAN' LANGSTOEGER wrote:E > Before I wade through a decade of google postings, I'd like to ask:d >nH > Does anyone have a good/simple (means F$CVUI, F$FAO, ...) DCL solutionI > for getting Image Identification Information of .EXE files ? On Alpha ?   " $ open/read i 'f$parse(p1,".exe")'
 $ read i h	 $ close ic/ $ if f$length(h).ne.512 then $ goto novms_imager $ if f$cvui(0,32,h).lt.16a $ then $            arch = "Alpha"  $            debug_byte = 804 $            hi = f$extract(f$cvui(24*8,32,h),104,h) $            imgtimo = 88 $            imgnam = f$extract(17,f$cvui(16*8,8,hi),hi)7 $            imgid = f$extract(57,f$cvui(56*8,8,hi),hi)18 $            linkid = f$extract(73,f$cvui(72*8,8,hi),hi) $ else $            arch = "VAX"o $            debug_byte = 322 $            hi = f$extract(f$cvui(6*8,16,h),80,h)4 $            imgnam = f$extract(1,f$cvui(0,8,hi),hi)7 $            imgid = f$extract(41,f$cvui(40*8,8,hi),hi)  $            imgtimo = 56i8 $            linkid = f$extract(65,f$cvui(64*8,8,hi),hi) $ endifN" $ debug = f$cvui(debug_byte*8,1,h)$ $ share = f$cvui(debug_byte*8+6,1,h)  6 $ say arch," ",imgnam," ",imgid," by linker ",linkid,-H              " at ",f$fao("!%D",imgtimo+f$cvui(32,32,f$fao("!AD",8,hi)))C $!^^^^^^^^^^^^^^^^^^^^^^ TIME DCL MAGIC ^^^^^^^^^^^^^^^^^^^^^^^^^^^.? $! Explanation: hi is a string & therefore passed by descriptortH $!           f$fao("!AD",n,s) expects a string by reference, & therefore returnsF/ the binary descriptor of hi as an 8 byte stringuI $!           f$cvui(32,32,desc) returns the actual address of hi from thet descriptor (2nd longword)S< $!           add whatever offset into the record is requiredI $!           f$fao("!%D",a) expects a quadword by reference, and gets then0 carefully constructed address of a quadword time   ------------------------------  % Date: Wed, 16 Apr 2003 13:21:40 -0400 & From: David M Smith <dsmit115@csc.com>> Subject: Re: [OpenVMS Alpha] How to get image version in DCL ?8 Message-ID: <f24r9v4t149c0btn573rvgfd83jprvjep5@4ax.com>  N On Wed, 16 Apr 2003 08:46:37 +0100, Chris Sharman <chris.sharman@sorry.nospam> wrote:  # >$ open/read i 'f$parse(p1,".exe")'V >$ read i hn
 >$ close i 	...  M Thanks! A few minor omissions in the code you posted (SAY symbol, NOVMS_IMAGEoK label not defined) but otherwise looks good. What are the symbols DEBUG anda6 SHARE all about? You get them, but don't display them.  5 I had also used the program version of IMGID found atF  ; 	http://vms.process.com/scripts/fileserv/fileserv.com?IMGIDe  L but your version has the advantage that it detemines the image type from theN .EXE file itself, rather than from the architecture on which IMGID is running.  O One limitation of this code (due to use of F$PARSE(P1,".EXE")) is that it won'teF "find" an image in the second (or later) directory of a search list...I -------------------------------------------------------------------------xI David M. Smith 302.391.8533                       dsmit115 at csc dot comMI Computer Sciences Corporation     (Opinions are those of the writer only)tI -------------------------------------------------------------------------.   ------------------------------   End of INFO-VAX 2003.210 ************************