1 INFO-VAX	Sat, 19 Oct 2002	Volume 2002 : Issue 578       Contents:( Re: 3 node cluster w/shadowing questions Re: a couple of things Re: a couple of things Re: Acrobat Reader?  Re: Acrobat Reader? ) Apache upgrade gives wrong file ownership - Re: Apache upgrade gives wrong file ownership + Re: Are there TK50 images for VMS avalable? - Callable mail: personal name and forging FROM 1 Re: Callable mail: personal name and forging FROM 1 Re: Callable mail: personal name and forging FROM 1 Re: Callable mail: personal name and forging FROM 1 Re: Callable mail: personal name and forging FROM 1 Re: Callable mail: personal name and forging FROM  Re: CTRL-Y and VMS BACKUP  Re: CTRL-Y and VMS BACKUP  Re: Dell retakes #1 PC spot  Re: Dell retakes #1 PC spot  Re: Dell retakes #1 PC spot  Re: Dell retakes #1 PC spot  Re: Dell retakes #1 PC spot  Re: Dell retakes #1 PC spot $ Fun: First patch for VMS 7.5 sighted( Re: Fun: First patch for VMS 7.5 sighted Re: graphics card problem 7 Re: HP tells VMS users "Relax, you are sitting pretty"! 7 Re: HP tells VMS users "Relax, you are sitting pretty"! 0 Re: installing the Adobe Acrobat Viewer for Java  Re: java on VMS - help requested  Re: java on VMS - help requested Re: Linker tip of the day  Re: Multia help needed+ RE: Originator of detached process from DCL ! OT - we're back on old cars again ) Re: OT: Oracle's take on Microsoft's .NET  Re: TLZ09 on Alpha?  Re: TLZ09 on Alpha? . Re: VMS tabletop without Office(tm): no chance. Re: VMS tabletop without Office(tm): no chance. Re: VMS tabletop without Office(tm): no chance. Re: VMS tabletop without Office(tm): no chance. Re: VMS tabletop without Office(tm): no chance  F ----------------------------------------------------------------------  % Date: Sat, 19 Oct 2002 13:01:02 +0100 ) From: Witchy <news@sruasonidyranib.co.uk> 1 Subject: Re: 3 node cluster w/shadowing questions 8 Message-ID: <tsh2ru8ts205pmfjs8go9igim0jv4br57i@4ax.com>  B On 17 Oct 2002 09:33:13 -0700, keithparris_NOSPAM@yahoo.com (Keith Parris) wrote:     <snip of excellent post>  D Thanks Peter and Keith - exactly the information I needed. I've beenC round VMS since 3.7 and simple clusters since 4.5, but shadowing is  relatively new to me :)    cheers   Witchy aka Adrian @ home   ------------------------------   Date: 19 Oct 02 17:13:02 +0200) From: p_sture@elias.decus.ch (Paul Sture)  Subject: Re: a couple of things ) Message-ID: <6P6YPqXX1ngR@elias.decus.ch>   h In article <Xns92AB8B9ABA182kenrbnsnrbnsncom@199.184.165.241>, Ken Robinson <kenrbnsn@rbnsn.com> writes:@ > "Brian Tillman" <tillman_brian@notnoone.notnohow.com> wrote in > news:3db04693$1@news.si.com:   > E >>>In Mark's session I was sitting next to an HP person who had never 	 >>>heard   >> about2 >>>VMS and she said she could not wait to sell it. >>  E >> Oops.  There's the next HP person to lose her job, if she tries to  >> sell it to a new client.  > J > Until proven otherwise by current and future events, let's take this as C > "good news" instead of immediately trashing it as "it will never  
 > happen". > J > Quite a few of us who were at the conference could sense that there was F > something different this year. I think it is the feeling that HP is J > genuinely interested in VMS as a product, not as something to be tossed  > away.  > D I didn't go the the conference, but I am getting positive vibes from those who did.  F Looking at the latest roadmap, there are serious plans afoot, which isB a massive difference from the "maintenance only" mode I feared VMS" was going through a few years ago.   --  
 Paul Sture Switzerland    ------------------------------   Date: 19 Oct 02 17:05:03 +0200) From: p_sture@elias.decus.ch (Paul Sture)  Subject: Re: a couple of things ) Message-ID: <UnKpQ+DYuz$y@elias.decus.ch>   b In article <3db04693$1@news.si.com>, "Brian Tillman" <tillman_brian@notnoone.notnohow.com> writes:J >>In Mark's session I was sitting next to an HP person who had never heard > about 1 >>VMS and she said she could not wait to sell it.  > L > Oops.  There's the next HP person to lose her job, if she tries to sell it > to a new client.  F Cynic! I could believe that from Compaq days, but I sincerely hope you are wrong now.   --  
 Paul Sture Switzerland    ------------------------------  % Date: Sat, 19 Oct 2002 10:40:29 +0200  From: Dirk Munk <munk@home.nl> Subject: Re: Acrobat Reader?& Message-ID: <3DB11A7D.8050004@home.nl>  I True, there isn't. There is a Java version of Acrobat Reader, however it  E only works with a rather old version of the Java suite. It seems you  H need a small piece of Java software that is in the Mac (!?!) Java suite E to use it with the newer Java suites if I remember correctly. Anyway  H that Adobe Acrobat Java Reader is most likely a Acrobat 3 reader. Adobe 2 doesn't maintain it, so there are no new versions.  F The XPDF reader is a good alternative. I'm using it with Mozilla, and  I'm quite satisfied.  
 DTL wrote:  I > Er... I do not find VMS in the list of platforms for the Adobe Acrobat   > Reader utility...  >  > D. >    ------------------------------   Date: 19 Oct 02 17:41:50 +0200) From: p_sture@elias.decus.ch (Paul Sture)  Subject: Re: Acrobat Reader?) Message-ID: <y68T28KJwMSY@elias.decus.ch>   Y In article <newscache$14x64h$f19$1@news.tiscali.fr>, DTL <Didier.Morandi@Free.fr> writes: I > Er... I do not find VMS in the list of platforms for the Adobe Acrobat   > Reader utility...  >   @ No you won't. However, at work we have GV/Ghostscript working onE Alpha VMS and IMHO, it's far superior to the Adobe standard offering.   = Let me know if you have problems setting that combination up.    --  
 Paul Sture Switzerland    ------------------------------   Date: 19 Oct 02 11:10:42 +0200) From: p_sture@elias.decus.ch (Paul Sture) 2 Subject: Apache upgrade gives wrong file ownership) Message-ID: <EQrnNg$$eWIL@elias.decus.ch>   @ I recently upgraded CSWS/Apache to the latest version on my home@ system and found it would not start due to having the wrong fileA protections (various files in the APACHE$ROOT directory tree were  owned by SYSTEM).   ? I just did it again this week on a test machine at work and the = same happened - this was the T1.3 beta, with the same result.   	 Solution:   < $ set file/owner=parent on the top level APACHE directories.   --  
 Paul Sture Switzerland    ------------------------------  % Date: Sat, 19 Oct 2002 12:04:34 +0200  From: Dirk Munk <munk@home.nl>6 Subject: Re: Apache upgrade gives wrong file ownership& Message-ID: <3DB12E32.4040408@home.nl>   A very familiar problem  :-) .F After very upgrade of Apache you should run the configuration utility 6 again. That will set the appropriate file protections.   Paul Sture wrote:   A >I recently upgraded CSWS/Apache to the latest version on my home A >system and found it would not start due to having the wrong file B >protections (various files in the APACHE$ROOT directory tree were >owned by SYSTEM). > @ >I just did it again this week on a test machine at work and the> >same happened - this was the T1.3 beta, with the same result. > 
 >Solution: > = >$ set file/owner=parent on the top level APACHE directories.  >  >    >    ------------------------------  % Date: Sat, 19 Oct 2002 12:58:38 +0100  From: Roy Omond <Roy@Omond.net> 4 Subject: Re: Are there TK50 images for VMS avalable?) Message-ID: <3DB148EF.7D3D8C07@Omond.net>    "Zane H. Healy" wrote:  0 > JF Mezei <jfmezei.spamnot@videotron.ca> wrote:4 > >> >But be warned, it [a TK50] takes AGES to boot. > P > > Yeah, but if you think of the TK50 as a first generation DLT drive, then the > > wait is not so bad anymore.  > K > Interesting point, I've seen a restore from a single Compact IV tape take  > DAYS!    Eh ???  ; What were you restoring to ?  RK07 ? RD53 ?  DECtape :-) ??   	 Roy Omond  Blue Bubble Ltd.   ------------------------------  % Date: Sat, 19 Oct 2002 02:16:19 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca> 6 Subject: Callable mail: personal name and forging FROM, Message-ID: <3DB0F8A5.9017681D@videotron.ca>  K I am able to use the MAIL$_SEND_FROM_LINE item for mail$send_add_attribute, ; and know I have to do this prior to the first add_address.    J However, when i do this, it seems to erase the MAIL$_SEND_PERS_NAME I have) specified in the mail$send_begin routine.   L I *think* I may have found the way to do this, but would like a confirmation2 that the following would be the right thing to do:  I In the MAIL$_SEND_FROM_LINE, I provide a string which contains the return M address followed by a single space, followed by the personal name enclosed in  double quotes.  C eg:   smtp%tom.paris@starfleet.com "Lieutenent Thomas Eugene Paris"    Is the correct way ?????   ----  K Also, when building the FROM_LINE and CC_LINE (max 256 chars each), can one  specify personal names ?  1 For instance, could one safely use the following:   N TO: smtp%k.janeway@voyager.com "Katherine Janeway", smtp%tuvok@voyager.com "Mr+ Tuvok", neelix@voyager.com "Mister Neelix"    P ??? Or must one provide only valid email adresses in the TO and CC  text lines ?     ----  @ The documentation states that in order to be able to specify theH MAIL$_SEND_FROM_LINE, one needs to have sysprv, AND the outbound connect0 privileges for the MAIL object need SYSPRV also.  M Is this for the node that is sending a message to another one, or is that for  the target nodes ?  K What exactly happens when my process on node BIKE uses the callable mail to N send mail to a user on node VELO ? Why would the object on node BIKE be used ?E Wouldn't my process simply connect to the MAIL object on node VELO  ?   J Also, I had not set this privilege. Yet. it seems that the TCPIP Services'N SMTP software is perfectly able to deliver emails to VELO with a forged FROM. ( How come SMTP would be able to do this ?  J What are the implications of giving the MAIL object the SYSPRV privilege ?   ------------------------------    Date: 19 Oct 2002 09:29:02 -0000= From: Doc.Cypher <Use-Author-Supplied-Address-Header@[127.1]> : Subject: Re: Callable mail: personal name and forging FROM6 Message-ID: <20021019092902.26128.qmail@gacracker.org>  C On Sat, 19 Oct 2002, JF Mezei <jfmezei.spamnot@videotron.ca> wrote: L >I am able to use the MAIL$_SEND_FROM_LINE item for mail$send_add_attribute,< >and know I have to do this prior to the first add_address.  > K >However, when i do this, it seems to erase the MAIL$_SEND_PERS_NAME I have * >specified in the mail$send_begin routine. > M >I *think* I may have found the way to do this, but would like a confirmation 3 >that the following would be the right thing to do:  > J >In the MAIL$_SEND_FROM_LINE, I provide a string which contains the returnN >address followed by a single space, followed by the personal name enclosed in >double quotes.  > D >eg:   smtp%tom.paris@starfleet.com "Lieutenent Thomas Eugene Paris" >  >Is the correct way ?????   E RFC 822 is your friend. http://www.ietf.org/rfc/rfc0822.txt. It shows D examples with this the other way round, and the actual email address enclosed in <>.      Doc. --  6 The bigger the humbug, the better people will like it.K ~ Phineas Taylor Barnum.                             https://vmsbox.cjb.net    ------------------------------  % Date: Sat, 19 Oct 2002 05:06:52 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca> : Subject: Re: Callable mail: personal name and forging FROM, Message-ID: <3DB12099.68D761F6@videotron.ca>   "Doc.Cypher" wrote: G > RFC 822 is your friend. http://www.ietf.org/rfc/rfc0822.txt. It shows F > examples with this the other way round, and the actual email address > enclosed in <>.   J Nop, I am talking about an argument passed to the mail$send_add_attribute - routines when building a VMS mail message....    ------------------------------    Date: 19 Oct 2002 16:08:38 -0000= From: Doc.Cypher <Use-Author-Supplied-Address-Header@[127.1]> : Subject: Re: Callable mail: personal name and forging FROM5 Message-ID: <20021019160838.2475.qmail@gacracker.org>   C On Sat, 19 Oct 2002, JF Mezei <jfmezei.spamnot@videotron.ca> wrote:  >"Doc.Cypher" wrote:H >> RFC 822 is your friend. http://www.ietf.org/rfc/rfc0822.txt. It showsG >> examples with this the other way round, and the actual email address  >> enclosed in <>. > K >Nop, I am talking about an argument passed to the mail$send_add_attribute  . >routines when building a VMS mail message....  K Why doesn't the RFC apply? You're passing a string of up to 255 characters, I shouldn't it be a RFC 822 compliant from line to ensure the recipient can  reply?  E If I read the help correctly, the "From: " part will be automatically @ supplied, you just have to supply the address and that would be:  ) "JF Mezei <jfmezei.spamnot@videotron.ca>"   I That is what I interpret as being the preferred format from my reading of  the RFC.     Doc. --  6 The bigger the humbug, the better people will like it.K ~ Phineas Taylor Barnum.                             https://vmsbox.cjb.net    ------------------------------  % Date: Sat, 19 Oct 2002 18:05:42 +0200 6 From: Arne =?ISO-8859-1?Q?Vajh=F8j?= <arne@vajhoej.dk>: Subject: Re: Callable mail: personal name and forging FROM) Message-ID: <3DB182D6.7060500@vajhoej.dk>    Doc.Cypher wrote:   E > On Sat, 19 Oct 2002, JF Mezei <jfmezei.spamnot@videotron.ca> wrote:  >>"Doc.Cypher" wrote: H >>>RFC 822 is your friend. http://www.ietf.org/rfc/rfc0822.txt. It showsG >>>examples with this the other way round, and the actual email address  >>>enclosed in <>.    L >>Nop, I am talking about an argument passed to the mail$send_add_attribute / >>routines when building a VMS mail message....     M > Why doesn't the RFC apply? You're passing a string of up to 255 characters, K > shouldn't it be a RFC 822 compliant from line to ensure the recipient can  > reply? > G > If I read the help correctly, the "From: " part will be automatically B > supplied, you just have to supply the address and that would be: > + > "JF Mezei <jfmezei.spamnot@videotron.ca>"  > K > That is what I interpret as being the preferred format from my reading of 
 > the RFC.  8 RFC 821 and 822 descibe how to send mail via two systems) using the SMTP protocol on top of TCP/IP.   3 MAIL$SEND_ADD_ATTRIBUTE is part of the VMS callable 6 mail interface. That mail interface is an API. Without4 any addon software it will use the Mail-11 protocol.  7 Those two things has absolutely nothing with each other  to do.  ; If you install an SMTP package on your VMS system (probably < very common today), then theese API calls may be transformed into an SMTP email.   6 But even in that case there are two layers of software8 between the question and your response. And those layers do convert addresses.   + That is why RFC 821 and 822 are irrelevant.    Arne   ------------------------------    Date: 19 Oct 2002 18:09:50 -00004 From: Doc.Cypher <Use-Author-Address-Header@[127.1]>: Subject: Re: Callable mail: personal name and forging FROM5 Message-ID: <20021019180950.5240.qmail@gacracker.org>   ; On Sat, 19 Oct 2002, Arne Vajh=F8j <arne@vajhoej.dk> wrote:  >Doc.Cypher wrote: > F >> On Sat, 19 Oct 2002, JF Mezei <jfmezei.spamnot@videotron.ca> wrote: >>>"Doc.Cypher" wrote:I >>>>RFC 822 is your friend. http://www.ietf.org/rfc/rfc0822.txt. It shows H >>>>examples with this the other way round, and the actual email address >>>>enclosed in <>.  >  > L >>>Nop, I am talking about an argument passed to the mail$send_add_attribut= e=200 >>>routines when building a VMS mail message.... >  > L >> Why doesn't the RFC apply? You're passing a string of up to 255 characte= rs, L >> shouldn't it be a RFC 822 compliant from line to ensure the recipient ca= n 	 >> reply?  >>=20 H >> If I read the help correctly, the "From: " part will be automaticallyC >> supplied, you just have to supply the address and that would be:  >>=20 , >> "JF Mezei <jfmezei.spamnot@videotron.ca>" >>=20 L >> That is what I interpret as being the preferred format from my reading o= f  >> the RFC.  > 9 >RFC 821 and 822 descibe how to send mail via two systems * >using the SMTP protocol on top of TCP/IP. > 4 >MAIL$SEND_ADD_ATTRIBUTE is part of the VMS callable7 >mail interface. That mail interface is an API. Without 5 >any addon software it will use the Mail-11 protocol.  > 8 >Those two things has absolutely nothing with each other >to do.  > < >If you install an SMTP package on your VMS system (probably= >very common today), then theese API calls may be transformed  >into an SMTP email. > 7 >But even in that case there are two layers of software 9 >between the question and your response. And those layers  >do convert addresses. > , >That is why RFC 821 and 822 are irrelevant.  J Thanks for explaining that. Sorry I was being a little dense about it. :-)     Doc. --=20 6 The bigger the humbug, the better people will like it.L ~ Phineas Taylor Barnum.                             https://vmsbox.cjb.net=   ------------------------------    Date: 19 Oct 2002 08:05:15 -0700. From: spamsink2001@yahoo.com (Alan E. Feldman)" Subject: Re: CTRL-Y and VMS BACKUP= Message-ID: <b096a4ee.0210190705.23833d68@posting.google.com>   h Kilgallen@SpamCop.net (Larry Kilgallen) wrote in message news:<d0iMKfNJ5fHd@eisner.encompasserve.org>...p > In article <b096a4ee.0210180540.2c091e69@posting.google.com>, spamsink2001@yahoo.com (Alan E. Feldman) writes: > > "Gotfryd Smolik, VMS lists" <gotfryd@stanpol.com.pl> wrote in message news:<Pine.LNX.4.21.0210180234010.16963-100000@irys.stanpol.com.pl>...+ > >> On 16 Oct 2002, Alan E. Feldman wrote: 
 > >> [...]J > >> >+There are other time-wasters. Why does DIRECTORY/DATE/TOTAL take soJ > >> >+much longer than DIRECTORY/TOTAL? The DIRECTORY command is probablyF > >> >+looking up the dates in the file headers. But it's not going toM > >> >+display them anyway! Why is this bad? Because many people have symbols . > >> >+like DIR:==DIRECTORY/SIZE=ALL/DATE/PROT > >>  0 > >>  What is "bad" at first - IMO the users :] A > >>  Sustitution of the original DCL commands are the first step  > >> to fit some problems. > > H > > Yes, but some users might assume that /TOTAL will have the effect ofG > > telling DIRECTORY to not bother gathering information that won't be H > > written to an output device. Sort of like DIREC/SIZE/NOSIZE in which) > > /NOSIZE *does* properly negate /SIZE.  > J > Some might prefer that absent an explicit negation of /DATE, the commandJ > DIRECTORY/DATE/TOTAL continue to provide the features it always had, and > detect any missing files.   @ What features? The feature that it looks up dates that never getD written to any output device? Why not throw in a few null megacycles while your at it?   7 Detect missing files? Isn't that the job of ANAL/DISK?    B What I'm saying is that based on some timing experiments, it seemsF that DIREC/DATE/TOTAL actually looks up date information from the fileF headers even though the *only* resultant output is the total number ofD files. The date information is read and discarded. It is not used atF all. I`t is completely irrelevant to the problem at hand. So why waste! time and resources looking it up?   : Try this on a directory with a fair number of files in it:  ' $ DIRECTORY:=         ! just to be sure  $ DIRECTORY/TOTAL     ! time it    $ DIR:=DIRECTORY/DATE  $ DIR/TOTAL           ! time it   E The output of the two DIRECTORY commands above is identical. Yet, the E second takes much longer. Please tell me *why* anyone would want that + as a "feature"? And what files are missing?    > > So even in the case of   > >  > > $ DEFINE XXX *.COM,*.TXT > > J > > why can't DIRECTORY/VERSIONS=n just look for the first match to *.COM,H > > list the first n versions of that file by checking the first n pairsE > > of version-no. and file-id, and then just go to the next matching G > > file? It can go to the next file when you use ;, so why can't it do ' > > that after finding the nth version?  > 3 > Because it depends on the RMS wildcard interface.   ! Could you be a little more vague?    Disclaimer: JMHO Alan E. Feldman    ------------------------------    Date: 19 Oct 2002 08:56:06 -0700. From: spamsink2001@yahoo.com (Alan E. Feldman)" Subject: Re: CTRL-Y and VMS BACKUP= Message-ID: <b096a4ee.0210190756.5080b2c2@posting.google.com>   j whohe@whoever.com (DL Phillips) wrote in message news:<af0dc2ea.0210181120.5e471ee3@posting.google.com>...u > spamsink2001@yahoo.com (Alan E. Feldman) wrote in message news:<b096a4ee.0210171722.53f859f6@posting.google.com>... n > > whohe@whoever.com (DL Phillips) wrote in message news:<af0dc2ea.0210170727.4d03db0b@posting.google.com>... > > > > 9 > > > > $ BACKUP DISK1:[BLAH],DISK2:[BLAH],DISK3:[BLAH] - $ > > > >   MKA0:SAVE-SET-NAME.BCK !.  > > > G > > > Ouch! This should not be one save_set! There is no way to easily  K > > > restore these files. This should be three backups to three save_sets.  > > I > > But VMS BACKUP allows users to do this, and in some cases it may well I > > be an okay thing to do. Maybe the user is transferring the files to a F > > much bigger disk on a remote system. It would make sense for that. > > G > > But that doesn't really matter, because the fact that it is allowed 5 > > means that it cannot be ignored as a possibility.  > > 9 > > > >     $ BACKUP SEARCH_LIST: MKA0:SAVE-SET-NAME.BCK   > > > > L > > > > where SEARCH_LIST is a logical name search list. Then there's no wayC > > > > during the restore to know whether SEARCH_LIST had multiple I > > > > equivalence names or not during the save operation; hence, BACKUP K > > > > would not have enough information to tell whether to read the whole 9 > > > > save set or not. So in general, it can't be done.  > > > M > > > So, backup parses the input spec(s) and translates all logicals anyway, B > > > doesn't it? How else would it "know" which files to process? > > @ > > Yes it does. But during the restore, all it has access to isF > > SEARCH_LIST. And from that alone, there is no way to tell what itsH > > value was during the save operation. So BACKUP would have to analyzeG > > the input parameter during the save *and* record the result of that I > > analysis in the save set. But BACKUP doesn't do that, so it has to go H > > through the whole tape to be sure there is not another occurrence ofG > > the same directory-filename combination. Perhaps BACKUP could abort I > > after the selected file is found during a restore of an image backup. & > > I don't see any problem with that. > D > Well, since it's Friday, I guess I'll take the time to ponder thisE > a bit more. Google won't post it for hours, anyway and I'll be long  > gone to the old pub by then. > I > That SEARCH_LIST is a logical name and isn't meaningful to /SELECT list J > during a restore (unless the logical is translated to a symbol and used I > as /select='search_list'). A /select list cannot contain a device name, 	 > though.   ? [Obviously my poor writing skills have confused some readers. I A apologize for that and will now attempt to be more clear about my 
 argument.]  E I never said to use SEARCH_LIST as the argument for /SELECT. And yes, C I know, you can't include a device name in that argument either. In - fact, I assumed that in my previous argument.   I > If there are multiple directories with the same name within a save_set, I > then the backup procedure is wrongly designed whether backup allows it  I > to happen or not. I can't think of any reasonable case where a restore  H > would yield a consistantly desirable result if the directories containH > files with the same name. You would need to use either /new_version orH > /replace and I'm afraid my pondering doesn't compute that one today:-) [...]   C You're right. However, it can be done and people will make mistakes B (even system managers (!)) and it might happen that files with theF same name and directory will end up in the same save set. I introduced  the idea of commands of the form  9 $ BACKUP DISK1:[BLAH],DISK2:[BLAH] tape:save-set-name.bck   ? simply to illustrate that. I did not suggest that such commands C consititued a good procedure, except for copying files from several E small disks to one large disk. And yes, you are right that that could F cause trouble, but in most cases, it would probably not be too hard toD straighten out if it is your files that are being moved and you knowE that the chances of dir-file matches from different disks are low and D you are in a hurry and other things that may come to mind if I had a/ lot more time to think about it were also true.    In fact, you could also have  C $ BACKUP SEARCH_LIST1:,LOGICAL_NAME1:,SEARCH_LIST2:,SEARCH_LIST3: -  tape:blah.bck   F And it might happen that the logical names overlap unexpectedly. Yeah,A somewhat unusual, but still possible, and you'd again end up with C multiple files with the same name in the same directory in the same A save set which, in fact, could be different if they were modified  during the save operation.  C Now, given the possiblity of such commands, I argued about what you F would have to have BACKUP do to determine whether it needs to read the> entire save set or not and not have any possibity of missing a; matching file (one that matches the argument of the /SELECT  qualifier).   D So, I argued that DURING THE RESTORE OPERATION, BACKUP would have toC check THE ORIGINAL BACKUP COMMMAND, THE ONE USED TO CREATE THE SAVE C SET IN THE FIRST PLACE, and look there to see if the ORIGINAL input E specifier was a single file-spec, or, instead, a comma-separated list F of file-specs. THIS HAS NOTHING TO DO WITH THE ARGUMENT OF THE /SELECTC QUALIFIER EXCEPT THE FACT THAT SUCH ARGUMENT (the /SELECT qualifier D argument) BE COMPLETE IN DIRECTORY, NAME, TYPE, AND POSSIBLY VERSION NUMBER. OK?   D Now though it seems like a rather weird thing to do, BACKUP could beE rewritten so as to have it analyze THE ORIGINAL BACKUP COMMAND, which D IS stored IN THE SAVE SET. By examining THE ORIGINAL BACKUP COMMAND,D it could determine whether the ORIGINAL input-specifier was a singleC file-spec, or, instead, a comma-separated list of file-specs. If it B found a comma-separated list, then BACKUP would definitely have toF read the entire save set to be sure it didn't miss any matching files.E So you would then think that if it found a single file-spec, it could @ just abort once it found and completely restored a match. Right?  E NO! You'd be wrong. This is where I introduced the possibility of the D input specifier being a search list. I AM TALKING ABOUT THE ORIGINALE INPUT SPECIFIER, THE ONE USED TO ***CREATE*** THE SAVE SET, ***NOT***  RESTORE IT.   > So, if DURING THE RESTORE, BACKUP analyzed the ORIGINAL backupD command, and it found a single file-spec, say, SEARCH_LIST, WHICH IS? IN NO WAY SPECIFIED IN THE BACKUP COMMAND USED TO CARRY OUT THE E RESTORE OPERATION, there is always the possiblity that said file-spec F is a logical name, and could in fact, in some instances, actually be aD search list. I AM TALKING ABOUT THE ORIGINAL INPUT SPECIFIER USED IN= THE BACKUP COMMAND THAT CREATED THE SAVE SET. THAT COMMAND IS A AVAILABLE DURING THE RESTORE OPERATION BECAUSE ONE CAN DISPLAY IT F USING BACKUP/LIST. THEREFORE, SUCH INFORMATION IS READILY AVAILABLE TOB BACKUP, EVEN THOUGH IT MAY SEEM A BIT OF A WEIRD THING TO ACTUALLY READ IT AND MAKE USE OF IT.   E Now, since the original input specifer could have been a search list, C BACKUP is going to have to read the entire save set anyway. And the B only exception I can think of is doing a selective restore from anB image save set, in which, I would think, there is no possiblity ofF duplicate dir-file combinations. I was trying to concoct a scenario inE which there would be as a consequence of users renaming things during E the SAVE operation, but unless I have overlooked something, and I may C well have, it should be safe to rewrite BACKUP to have it abort the = restore once the first matching file was found and completely > restored. (Actually, I never tried something like BACKUP/IMAGEC DISK_SEARCH_LIST ..., but I'd be really surprised if such a command  would work.)  B But then I thought of another complication. When BACKUP stores theE ORIGINAL COMMAND in the save set, it removes the quotation marks from = the /COMMENT="All-exciting comment", which would make it more 5 difficult to parse and/or analyze the STORED COMMAND.   B My point is that what at first seems like a simple matter actuallyB turns out not to be. And the more you think about it, the more the gotchas keep piling on.   F And for a program whose primary goal is data integrity, I think it may@ be better to not add this feature so as not to add an additional@ potential source of missing data due to unforseen circumstances.   OK?   F > So, anyhow, this is all just a pointless ponder because backup worksE > the way is works, and aside from the lack of /statistics, it really B > does work just fine. I'd just have done it a little differently.  D OK. But be sure not to leave any possibility of missing any matching files!   Disclaimer: JMHO Alan E. Feldman    ------------------------------  % Date: Sat, 19 Oct 2002 02:19:16 -0400 * From: "Bill Todd" <billtodd@metrocast.net>$ Subject: Re: Dell retakes #1 PC spot2 Message-ID: <pDmdnRP-WeTJZC2gXTWcqg@metrocast.net>  0 "Arne Vajhj" <arne@vajhoej.dk> wrote in message# news:3DB054FF.6050102@vajhoej.dk...  > JF Mezei wrote:  > E > > Not quite VMS related, but does outline how merger promises don't 	 generally  > > pan out... >  > F > > HOUSTON -- Dell Computer Corp. (DELL) regained its standing as the world's L > > largest personal-computer maker just two quarters after losing the crown toE > > Hewlett-Packard Co. according to a newly released market-research  report.  > > F > > In the quarter ended Sept. 30, Dell, of Austin, Texas, shipped 5.2 million PCs D > > to H-P's 5.0 million PCs to retake the crown, according to a new
 estimate from / > > International Data Corp., Framingham, Mass.  > 5 > You can conclude that HP may not have got what they  > thougth they bougth. > 6 > But you could also interpret as if Compaq would have. > been in big big problems without the merger.  G Or you could interpret it as the logical result of merging two troubled H companies both run by incompetents.  About the best argument against theG merger was that without it said incompetents would have gotten the boot F (yielding at least the hope of improved leadership) instead of gettingI another couple of years to continuing destroying the combined enterprise.    - bill   ------------------------------  % Date: Sat, 19 Oct 2002 02:24:13 -0400 * From: "Bill Todd" <billtodd@metrocast.net>$ Subject: Re: Dell retakes #1 PC spot2 Message-ID: <6Sednd2Tv5HgZy2gXTWcqg@metrocast.net>  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3DB0573E.240C9371@videotron.ca... > John Smith wrote: ; > > Did you ever have any doubt it would turn out this way?  > E > I didn't. But Carly's promises seemed to have some credibility with  analystsL > and reporters who believed that HP could truly become #1 in the PC market. > L > In the end, HP will have paid a hell of a lot just to get iPAQ, Tandem and VMS ' > in terms of technologies it may keep.   F Well, it did get quite a bit more than that:  some really good storageG technology, a stadium-full of service people, a large existing customer J base, ...  The problem is that it doesn't understand any better what to doL with what it got than it understood what to do with what it already had:  ifL its leadership were competent, it *would* be #1 in PCs, and have a good shotC at being #1 in the enterprise, storage, and service spaces as well.    - bill   ------------------------------  % Date: Sat, 19 Oct 2002 02:36:22 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca> $ Subject: Re: Dell retakes #1 PC spot, Message-ID: <3DB0FD57.6B6948B5@videotron.ca>   Bill Todd wrote:J > companies both run by incompetents.  About the best argument against theI > merger was that without it said incompetents would have gotten the boot 5 > (yielding at least the hope of improved leadership)   M On the other hand, the major reduction in the combined wintel sales will turn = out to be good since HP will lose less money on wintel crap.    L I personally think it was wrong of HP to continue to show the "Compaq" brandE on PCs and advertising after the merger. Folks are less likely to buy % mechandise from a dying/dead company.   C As far as leadership is concerned, I have absolutely no hope that a J replacement would have been any better than Capellas with regards to VMS. K Pfeiffer was a strong man who probably had a clash with the board. CapellassL was probably a very weak "yes sir" type of guy who bent to the wishes of theL board, knowing he doesn't have any credibility nor strenght to take them on.L As a result, I would say that Capellas' policies had the full support of hisM board, and as a result, that board would have selected a clone of Capellas toa, replace him and continued the same policies.  J With people like Ken Lay on Compaq's board, the board probably prefered toR focus on financial deals with Intel rather than strategic technological decisions.   ------------------------------  % Date: Sat, 19 Oct 2002 03:06:39 -0400o- From: JF Mezei <jfmezei.spamnot@videotron.ca>k$ Subject: Re: Dell retakes #1 PC spot, Message-ID: <3DB1046F.EF6639FE@videotron.ca>   Bill Todd wrote:H > Well, it did get quite a bit more than that:  some really good storageI > technology, a stadium-full of service people, a large existing customers
 > base, ..  K Point taken for the storage stuff.  But as far as the service folks, aren'ttD those in excess sice there would be a lot of duplication getwene theJ ex-digital and ex-HP folks ? (tandem has a small more focused force not so spread out around the world).h  N But in terms of the customer base, look at what happend to their wintel sales.L  Consider how much they pissed the Tru64 customers totally, and rteally madeL VMS customers ponder their future due to the forced migration to an unwanted- platform. How much will they really retain ? n  K The Tande, folks are not in a dead end situation. MIPS hasn't been declaredfN dead from what I heard. So if IA64 doesn't pan out and Intel abandons it, they could stay on MIPS.   N By the time they will have ported today's Trucluster on HP-UX, the competitors7 will had had 4 years to move their clustering forwards.D  L When you're in a marathon race, if you really need to change your shoes, youK wait until the new shoes are in your hands before you even think of killingpH your old shoes. If you prematurely remove your shoes before the new onesI arrive, then you're stuck during a long time period while the competitorse! continue to run full speed ahead.    ------------------------------  % Date: Sat, 19 Oct 2002 04:37:37 -04000* From: "Bill Todd" <billtodd@metrocast.net>$ Subject: Re: Dell retakes #1 PC spot2 Message-ID: <lW-dnQLBeftchCygXTWc3w@metrocast.net>  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3DB1046F.EF6639FE@videotron.ca... > Bill Todd wrote:J > > Well, it did get quite a bit more than that:  some really good storageK > > technology, a stadium-full of service people, a large existing customerh > > base, .. >hF > Point taken for the storage stuff.  But as far as the service folks, aren'tF > those in excess sice there would be a lot of duplication getwene theL > ex-digital and ex-HP folks ? (tandem has a small more focused force not so > spread out around the world).o >eI > But in terms of the customer base, look at what happend to their wintel  sales.I >  Consider how much they pissed the Tru64 customers totally, and rteallyh madeE > VMS customers ponder their future due to the forced migration to ann unwanted. > platform. How much will they really retain ?  E You seem to have lost track of the statement that I was responding tovI (perhaps you shouldn't have snipped it, or the rest of my response).  TherK fact that HP does not have the foggiest notion of how to make effective usemI of what it bought does not mean that the value of what it bought was nil,wH only that HP is wasting that value (especially the customer base and the0 service organization), just as it's wasting VMS.  E If HP had had competent leadership that had given Curly a good enoughhK personal deal to shut him up, it could have started leveraging (rather thannD destroying) what it was buying even before the sale went through.  IJ sincerely doubt that Curly would have been inclined to tell Carly that sheL couldn't handle things as she saw fit (or that he would have gotten anywhereI even if he had tried).  And for that matter, she probably could have keptr0 him from killing Alpha had she had the sense to.  K Hell, if Carly had been competent, Walter Hewlett probably would have stooda up and applauded the merger.  J Compaq was a disaster, but mostly because of the way it was run:  it stillI had assets of great value (possibly even as much as HP paid for them) hadnK they been put to the right use.  Unfortunately, HP is a similar disaster in A quality, and with the acquisition an even larger one in quantity.t   - bill   ------------------------------  % Date: Sat, 19 Oct 2002 04:52:57 -0400c* From: "Bill Todd" <billtodd@metrocast.net>$ Subject: Re: Dell retakes #1 PC spot2 Message-ID: <LricnYoaX7fEgCygXTWcpA@metrocast.net>  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3DB0FD57.6B6948B5@videotron.ca... > Bill Todd wrote:L > > companies both run by incompetents.  About the best argument against theK > > merger was that without it said incompetents would have gotten the boot 7 > > (yielding at least the hope of improved leadership)a >SJ > On the other hand, the major reduction in the combined wintel sales will turn> > out to be good since HP will lose less money on wintel crap.  D When sales go down, unless the problem was that suppliers were beingI over-paid (rather unlikely in this case) losses go up - until capacity is F slashed and the company has recovered from the resulting restructuringJ costs.  If Compaq had gone down that toilet it would have happened quicklyK enough that there would have been something worthwhile left to salvage, but J by the time HP has (because of the year-plus accountability void caused by, the merger confusion) there may well not be.   >pH > I personally think it was wrong of HP to continue to show the "Compaq" brand G > on PCs and advertising after the merger. Folks are less likely to buyd' > mechandise from a dying/dead company.h  H The suggestion that any significant portion of HP's problems is based onJ poor brand name selection (even if I agreed with your premise that it *is* poor) is ludicrous.e   >eE > As far as leadership is concerned, I have absolutely no hope that a K > replacement would have been any better than Capellas with regards to VMS.eD > Pfeiffer was a strong man who probably had a clash with the board. CapellasJ > was probably a very weak "yes sir" type of guy who bent to the wishes of thenJ > board, knowing he doesn't have any credibility nor strenght to take them on.eJ > As a result, I would say that Capellas' policies had the full support of his L > board, and as a result, that board would have selected a clone of Capellas to. > replace him and continued the same policies.  J That would depend on how and which leadership was replaced, I'd think.  IfF the move had been initiated by the stockholders (as well it might haveJ been), it could have put the fear of God into the BoD (or what was left ofK it) and resulted in a major course change.  If the change had resulted fromlF a take-over by someone competent, the BoD would have been a non-issue.   - bill   ------------------------------  % Date: Sat, 19 Oct 2002 13:00:31 +02000 From: Dirk Munk <munk@home.nl>- Subject: Fun: First patch for VMS 7.5 sighted5& Message-ID: <3DB13B4F.6060800@home.nl>  D I was checking the list of patches for AXP VMS, and found the first D patch for OpenVMS 7.5 ! It is a Fortran patch, so if you're running ' Fortran on OpenVMS 7.5 you are warned !f   ------------------------------  # Date: Sat, 19 Oct 2002 15:32:35 GMTt. From: peter@langstoeger.at (Peter LANGSTOEGER)1 Subject: Re: Fun: First patch for VMS 7.5 sightede4 Message-ID: <nWes9.79415$N_6.1120604@news.chello.at>  G In article <3DB13B4F.6060800@home.nl>, Dirk Munk <munk@home.nl> writes:-E >I was checking the list of patches for AXP VMS, and found the first rE >patch for OpenVMS 7.5 ! It is a Fortran patch, so if you're running e( >Fortran on OpenVMS 7.5 you are warned !  J Now that explains why I didn't find the FORTRAN ECOs the list announced...   -- r Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  # Date: Sat, 19 Oct 2002 17:19:52 GMTe= From: peter.watkinson1@<nospam>ntlworld.com (Peter Watkinson)i" Subject: Re: graphics card problem1 Message-ID: <3db1942c.7390140@news.cable.ntl.com>h   Hi dave,  D It's an S3 trio it also says S3 Trio644 V2/DX on the chip. It's made by a firm called Sparkle?e  E I've also now got a Diamond fire GL 1000 pro Permedia 2 but somehow IP! need to ftp the drivers for this.   E My plan is to get the S3 working get Dec Windows up and then copy thec# drivers for the Permedia card over.?  % Thanlks for any help you can give me!e   cheers,      Peter Watkinsony             Is it a TRIO or a Virge ???v     -- David B Turner	 Sales Dptv Island Computers US Corporation? 2700 Gregory Street 	 Suite 180e Savannah GA 31404e Tel: 912 447 6622n Fax: 912 201 0096h sales@islandco.com www.islandco.com' http://www.islandco.com/legal-email.htm_   We sell Alpha Systems !p* All emails are checked for Virus and WormsD "peter.watkinson1@ ntlworld.com (Peter Watkinson)" <nospam> wrote in messages, news:3da563bb.10446343@news.cable.ntl.com...3 > On Wed, 9 Oct 2002 20:00:39 -0400, "David Turner"   > <dbturner@islandco.com> wrote: > % > >go to www.islandco.com/support.htmw > >t > >j >  >n > Hi,E > G > I've got a S3 VGA card plus Hobbyist Open VMS 7.2 and I'm also getingOH > a no video device error. Is there an S3 kit on the CD that needs to be > installed also?w > 	 > cheers,y >l >o > Peter WatkinsonV' > peter.watkinson1<nospam>@ntlworld.como > remove <nospam> to reply :-) >e        ! On Fri, 11 Oct 2002 10:25:46 GMT, > peter.watkinson1@<nospam>ntlworld.com (Peter Watkinson) wrote:  2 >On Wed, 9 Oct 2002 20:00:39 -0400, "David Turner" ><dbturner@islandco.com> wrote:. > $ >>go to www.islandco.com/support.htm >> >> >o >n >Hi, >eF >I've got a S3 VGA card plus Hobbyist Open VMS 7.2 and I'm also getingG >a no video device error. Is there an S3 kit on the CD that needs to beh >installed also? >E >cheers, >l >n >Peter Watkinson& >peter.watkinson1<nospam>@ntlworld.com >remove <nospam> to reply :-)v   Peter Watkinson % peter.watkinson1<nospam>@ntlworld.comA remove <nospam> to reply :-)   ------------------------------   Date: 19 Oct 02 11:00:45 +0200) From: p_sture@elias.decus.ch (Paul Sture)t@ Subject: Re: HP tells VMS users "Relax, you are sitting pretty"!) Message-ID: <HpkaULpxSrjM@elias.decus.ch>h  ` In article <8o6qqu8le1qba09vrls6belptqq45lme7i@4ax.com>, Alan Greig <a.greig@virgin.net> writes:5 > On Tue, 15 Oct 2002 17:38:50 GMT, "Fred Kleinsorge".& > <kleinsorge@star.zko.dec.com> wrote: >  >>  >>Fred is a moron:    YES  /  NO >>Andy is a moron:    YES /  NO/ > > > Gordon is a moron. Probably only UK readers will get this :)  
 Ah, memories.a  O What amused me about that song was a previous boss called Gordon who was of thepI "scream and shout at them" management school. He made my life hell to thesF point where I had to quit before I thumped him. Fortunately one of theL data prep girls found me another job, which I got, which was my introduction	 to  PDPs.a  H I bought that girl the single as a thank you, and her reply was that theK whole of the accounting and IT departments already had personal copies that.. they used to play as soon as they got home :-)   -- e
 Paul Sture Switzerland    ------------------------------   Date: 19 Oct 02 11:12:27 +0200) From: p_sture@elias.decus.ch (Paul Sture)r@ Subject: Re: HP tells VMS users "Relax, you are sitting pretty"!) Message-ID: <84mgi+c4IsUZ@elias.decus.ch>   a In article <mf20rukipcma9qk5ch26mu9j2rpqo8fhov@4ax.com>, David M Smith <dsmit115@csc.com> writes: A > On 17 Oct 2002 17:43:04 -0700, susan_skonetski@hotmail.com (Sue4 > Skonetski) wrote:g > " >>At the risk of getting blasted   >>G >>Neither Fred nor Andy is a moran.  I personally know Fred and he is af2 >>wonderful person and most definatly not a moran. >>D >>I have not had the pleasure of meeting Andy, but Andy provides all7 >>kinds of services for this newsgroup and even to VMS.u >>F >>Both Fred and Andy have view points which is a good thing, the worldF >>is full of dull gray people with no opinions, no color, no emotion. ? >>Sometimes we may not like a certain color but all are useful.m >>H >>Andy does have some good points and he certainly keeps us on our toes.G >> He also lets us know that SUN is watching what we are doing and they E >>would only do that if they felt we were compitition.  Also think ofc? >>all the interesting reading both Andy and Fred provide to theoA >>newsgroup.  Just look at the amount of response to this string.t >> >>just my opinion. >>sue  > F > Thanks for posting a reasonable response to this thread, rather thenD > continuing the name calling. It's interesting (to me) that a woman8 > took this reasonable view rather than a boy...er, man!  F LOL. I was going to say something about school playground behaviour...   -- B
 Paul Sture Switzerland    ------------------------------  % Date: Sat, 19 Oct 2002 10:50:06 +0200v From: Dirk Munk <munk@home.nl>9 Subject: Re: installing the Adobe Acrobat Viewer for Javag& Message-ID: <3DB11CBE.4050203@home.nl>  G Don't try to use it. It will not work with any Java version newer then nD 1.1.8 (even if Adobe claims that it will). Use XPDF, that will work.  
 DTL wrote:  4 > I copied the "Adobe Acrobat Viewer for Java" from 7 > http://www.adobe.com/products/acrviewer/acrvdnld.html  > and installed it.  >  > Did not find the howto.o >i > So, I did: >e > DTL02> java LAUNIXGC.SH  >l > amd I got: >a- > java.lang.NoClassDefFoundError: launixgc/sh  >o > Can anyone help, please? >t	 > Thanks,p >  > D. >s   ------------------------------  % Date: Sat, 19 Oct 2002 13:11:28 +0200 6 From: Arne =?ISO-8859-1?Q?Vajh=F8j?= <arne@vajhoej.dk>) Subject: Re: java on VMS - help requested ) Message-ID: <3DB13DE0.8030808@vajhoej.dk>s   John McLean wrote:  E > I'm having problems with trying to work with Java servlets on a webc > server running on VMS.     Which servlet container ?W  F >                       I just don't think that java is up to the taskE > of providing anything that is even vaguely approaching an efficientl& > method.  Can someone please advise ?    9 If used correctly and with sufficient ressources Java cano	 run fast.a  H > I would like to create a web application that is run from a web server	 > on VMS.u    > When you say "web application" do you specifically mean a J2EE< web-application as in a war-file or do you use the term more
 generally.  I >           On the VMS machine I want to be able to use logical names forh$ > all the directories and filenames,    1 Logical names for files should work fine in Java.g  J >                                    use a large number of global symbols,    C I do not really think symbols make much sense in a servlet context. B It is not DCL. And it is not obvious where the symbol table should
 come from.  = > access RMS indexed files (some of which have multiple keys).    ; You will need to use JNI to do that. Java as being platformc: independent does not know about VMS index-sequential file.  6 Use JNI to call C code. Or use JNI to call C code that# call your code in another language.d  7 I have previously posted alittle example with access toa' index-sequential files Java->C->Pascal.5  H >                                                             and I wantD > to be able to call command procedures which call other procedures.    = Have you looked at the Bridgeworks product. It is free. It iss; from Compaq (now HP). And it makes accessing DCL procedures" from Java a piece of cake.  E > The whole thing runs happily under DCL (lots of it) and is fast andfF > smooth, but it is only available on VMS terminals or workstations. I@ > would like to extend its flexibility by getting it on the web.    # That is a common requirement today.E  B > I may be wrong but as I understand it, java cannot retain globalJ > symbols, cannot retain logical names (unless I define a new name table),J > cannot efficiently call a series of command procedures and cannot accessJ > RMS indexed files without major effort. (I have seen methods of enablingJ > java to access RMS indexed files but this creates non-standard java whenI > one basic aim of java is to be a standard interpreted language - "writeo > anywhere, run anywhere".)e    & I would stay away from global symbols.    ? Java do use logicals for file names etc.. To translate logicals F you can either use System.getenv or use JNI. To set logicals you would need to use JNI.  > Accessing command procedures is easy with Bridgeworks. Without it wil be a bigger task.  : Accessing index-sequential files is not that difficult via@ JNI. Yes - that is not portable. But neither is index-sequential= files ! If you want portability use a relational database ande JDBC.9    2G > Much of the problem seems to be that java must spawn a new process inrG > order to do anything which wants to use the capabilities that are notcH > available in C on Unix.  When there is a sequence of interactions that@ > need to take place, the overheads of this approach become very+ > significant and performance will be poor.i    	 Not true.0  ; Java calls C with JNI. C can call any native language using2 VMS calling standard.o  9 Yes - you will need a process to run DCL. That is more ore9 less unavoidable (well the OSU HTTP server has managed toE8 implement reusage but that was builtin from start in the architecture).    G > Am I wrong in my assumptions ?  Is there any reasonable way to create E > the software that I need or should I resort to CGI scripts, perhapsTJ > using a mailbox to communicate with a VMS process, one that can keep itsF > environment of logical names and global symbols, that can access RMSI > indexed files and can call other procedures in a sensible and efficienta	 > manner?l    . You are rigth about much and wrong about some.  # I think you should study JNI a bit.b  - And I definattly think you should try out the - Bridgeworks product. It really makes a lot ofp4 Java to VMS stuff much easier (it also uses JNI, but the code is generated for you).T   > Some general comments: > I > Am I wrong or is java just an extension of C on unix ?  In other words,sI > is it a continuation of the concepts that unix and C people consider isrF > industry-standard, while users of proprietary operating systems fromI > DEC/CPQ/HP and IBM consider these to be "lowest common denominator" andD; > woefully inefficient for a lot of commercial processing ?m    . The syntax of Java is a simplified C++ syntax.  ; The JVM concept goes back to stuff like USCD P-code Pascal.l  5 The platform independence obviously has made it a bitt1 "lowest common denominator". That is unavoidable.   7 Compaq has not done much to provide Java tools for VMS.d9 The major exception is the great but very unknown producth Bridgeworks.  ; Java is definatetly capable of commercial processing. It is 9 "THE" choice for enterprise computing today. Small thingsI8 get developed for NT using MS technology. Big things get5 developed with Java and J2EE. Financial, OLTP, e-bizza7 etc. stuff then my guess is that >80% of completely newh3 stuff is being done in Java and J2EE. Java and J2EEf  business is way bigger than VMS.    F > I can understand that a java applet should be in a standard languageI > because it is handled by the browser on any platform, but I fail to seerI > any real need to create servlets in a standard language.  Each platform.H > has its strengths and weaknesses so it seems reasonable to extend java > to suit the platform.      Hmmm.h  A If you use J2EE technology to access OS specific technology, thens. the portability is ofcourse of little benefit.  9 But for a new application that uses relational databases,E! then portability is a good thing.-  < I have personally tried taking a complete ear-file developed= and tested on Windows and deployed on VMS and it just worked.P  ? You do not seem to have much benefiot from the portability, but 
 other may.   You wil only benefit from:8    - the availability of feature-rich products available      (many of them free !)?    - the availbility of books, training courses, web ressources.5    - the availability of programmers with Java skills @    - the availability of software packages (try compare what you@      can find for XML in Cobol/Fortran/Pascal with what you find      for Java)  I > An application running via a web server under one operating system willoD > almost inevitably need changes if it is moved to another operatingG > system.  Attempting to use "standard" java servlets - ie. with no o/suB > specific enhancements - seems a very high price to pay for beingJ > "standard" because sites rarely change web servers onto other platforms.    / True if you web-enable an existing application.a  8 Not true of you architect a new application from scratch with portability in mind.i    oC > Why not extend the java software to suit the capabilities of each < > operating system in the most useful and efficient manner ?    
 Not allowed !   ' That was what SUN kicked Microsoft for.T  7 What HP should do was really push integration products.h  ; Libraries that accesses VMS functions like index-sequentiald files.  - Make people aware of the Bridgeworks product.o  	 Etc.etc..   A > If java is not up to the task then maybe should just move on too > something better.c    5 You can critisize Java for many things. I got severalp1 things about Java and J2EE that really annoys me.o  6 But the fact is that there are a huge industry support	 for Java.   3 It has been estimated that there are 3 million Javat programmers.  8 Java is moving. You can argue whether it is good or bad. But it is moving.u   Arne   ------------------------------   Date: 19 Oct 02 17:31:32 +0200) From: p_sture@elias.decus.ch (Paul Sture)l) Subject: Re: java on VMS - help requestedu) Message-ID: <lnAoY6D3xWfk@elias.decus.ch>   i In article <3DB050AA.14DD012B@swissonline.delete.ch>, John McLean <mcleanj@swissonline.delete.ch> writes:b <massive snip>   > C > Why not extend the java software to suit the capabilities of each < > operating system in the most useful and efficient manner ?  > John, you have just reminded me about compatibility mode COBOLA more than 20 years ago. Because it was "industry standard", therel@ was no such thing as global variables, nor passing parameters byA the VAX Calling Standard. The latter could be done to call systema4 services, but it was easier to write a bit of macro.  C It was crap to write/maintain, and didn't perform very well either.a   Just my 5 rappen (cents).    -- s
 Paul Sture Switzerland    ------------------------------  # Date: Sat, 19 Oct 2002 12:54:14 GMTo" From: Guy Peleg <guy.peleg@hp.com>" Subject: Re: Linker tip of the day& Message-ID: <3DB155E8.5FB949A3@hp.com>   Brian,  M No problem backporting to V7.2-2 and even for a VAX. As we in engineering arei being measuredO by cases, please log a service call with your local support center and ask thems to escalte the callu to engineering.o   Thanks,e   Guye   Brian Tillman wrote:  B > >We have lately released two linker kits. VMS73_LINKER-V0300 andJ > >VMS731_LINKER-V0100. The kits increase the linker's internal I/O buffer0 > >resulting in significant reduce in Link time. >iN > Surely that change should be easily incorporated into V7.2's linker as well,K > for those people on prior-version support.  And of course you've made thel# > same fixes on OpenVMS VAX, right?r >!: > In fact, wouldn't the V7.3 linker run just fine on V7.2? > --C > Brian Tillman                   Internet: tillman_brian at si.comsC > Smiths Aerospace                          tillman at swdev.si.com.? > 3290 Patterson Ave. SE, MS      Addresses modified to preventm> > Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@": >        This opinion doesn't represent that of my company   ------------------------------   Date: 19 Oct 2002 17:53:30 GMT( From: bill@cs.uofs.edu (Bill Gunshannon) Subject: Re: Multia help needede5 Message-ID: <aos66p$p79bc$1@ID-135708.news.dfncis.de>t  ' In article <3DB0B971.930E0C6A@fsi.net>, 4 	"David J. Dachtera" <djesys.nospam@fsi.net> writes: > Bill Gunshannon wrote: >> iH >> Well, I'm begging again.  Is there anychance anyone here has a coupleF >> of SIMMs suitable for use in a Multia that they would be willing toH >> part with??  I have been unable to find any anywhere.  I would reallyE >> like to find out if the Multia I have works, but alas, not without  >> memory.  :-(t > F > Aren't those 72-pin parity SIMMs? They're (almost) a dime a dozen on > e-Bay! >   F Well, although I neither like nor trust Ebay I went for a look anyway.  F I found three entries.  One is a mixture of parity and non-parity withD the actual mix not specified.  It's currently a $.01 but consideringE they don't say what your bidding on....  And you also don't know what F the reserve is.  Hmmm.  Unknown reserve, unknown item. Yeah, I can see bidding on that.  F Then there's another.  50 16Meg SIMMS that puts a max of 32 in the boxF and 48 to sit ont he shelf cause nothing else uses them.  I don't knowH if 32M is even enough to run VMS on an Aplpha.  I have more than that inI some of my VAX.  And, the bid is already up to $50.00.  A lot to pay justG to find out if the box works.   I One more.  Two 32's.  Bid not to high.  I'll try it, but somehow I expectLJ that there will be a last minute bid $.50 over whatever I put in that will take it.  I So, I'm still asking.  Anybody got some old 72M Parity SIMMs they want totI get rid of?  Surely there isn't much of anything that still uses them.  I 5 don't even know if they are still being manufactured.i   bill   -- OJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   i   ------------------------------   Date: 19 Oct 02 17:02:10 +0200) From: p_sture@elias.decus.ch (Paul Sture)l4 Subject: RE: Originator of detached process from DCL) Message-ID: <vgLRmNaoP7Xq@elias.decus.ch>o  W In article <01C2768F.D1F4D6C0@sulfer.icius.com>, Shane Smith <ssmith@icius.com> writes:E1 > RSTS/E had that. Also see "HELP ATTACH" on VMS.T >   K As also, did CTS 300 on RT-11. Extremely handy - start a job on a terminal,pL input a few parameters, then detach to free up the terminal for another job.  K If the job went "dead", typically an error condition or it wanted more diskt8 space, you could simply do an attach to see the message.  c > Shanea >  > -----Original Message-----8 > From: David J. Dachtera [mailto:djesys.nospam@fsi.net]( > Sent: Friday, October 18, 2002 9:31 AM > To: Info-VAX@Mvb.Saic.ComS6 > Subject: Re: Originator of detached process from DCL >  >  > Paul Repacholi wrote:  >>  + >> Joshua Cope <Joshua.Cope@hp.com> writes:E >>  G >> > On Alpha V7.3-1, there's a new GETJPI code for the parent process:C >> Y0 >> >   $ write sys$output f$getjpi("","CREATOR") >> >      20602D44 >> u0 >> But are we not talking of DETACHED processes? >> o2 >> And why, pray tell, do we call them `detached'? > C > From memory, there once was a way (in other operating systems) totG > "detach" a process from it's controlling terminal; that is, convert ac3 > "foreground" process into a "background" process.h >  > -- o > David J. Dachterac > dba DJE Systems  > http://www.djesys.com/ > * > Unofficial Affordable OpenVMS Home Page:! > http://www.djesys.com/vms/soho/b -- e
 Paul Sture Switzerlande   ------------------------------   Date: 19 Oct 02 16:57:04 +0200) From: p_sture@elias.decus.ch (Paul Sture)i* Subject: OT - we're back on old cars again) Message-ID: <HtCNPkb6v8RI@elias.decus.ch>c  ` In article <aopf6o$os57g$1@ID-135708.news.dfncis.de>, bill@cs.uofs.edu (Bill Gunshannon) writes:2 > In article <01C27688.3E519900@sulfer.icius.com>,) > 	Shane Smith <ssmith@icius.com> writes:gI >> I had a FIAT 126; a 2 cylinder, 650cc shoebox sized pile of ****. When G >> the starter motor died I push started it - alone. Run along with it,t >> jump in, crash it into gear.. >> tH >> I had a couple of less polite alternative acronym expansions for that	 >> thing.a >> s > I > I had an 850 when I lived in Europe back in 71-72.  I mentioned here intK > another thread carrying a spare fanbelt around in Corvairs.  For this carnH > you carried a spare head gasket.  I got real good at emergency engine K > rebuilds on the side of the road.  The engine was an engineering marvel. fN > Engineers world- wide marveled at the fact that it actually ran at all.  :-) > A Sorry, cannot resist this one. Back in the '70s I read about some F guy who upgraded his Fiat 500 to an 850. I might have the models wrong@ there, but the outcome was that one engine worked clockwise, theB other anti-clockwise, so he ended up with a car that was very fast in reverse.e  G My experience with such? When I first passed my driving test, my motherhG had a 2 cylinder, rear engined 600 cc NSU Prinz, which I got to drive. i  C It could beat an 850cc Mini from traffic lights, but was absolutelyoH lousy on roadholding. The back end went sideways with little provocationL and in winter a couple of  hundredweights of sand in the luggage compartment% at the front helped with the braking.   D Then of course, there was the time when loaded up with camping gear,E I was trying to stop toward the bottom of a steep hill, and literally,G pulled myself off the seat in order to put more pressure on the brakes.i  A Ah, the joys of motoring, in the days when one could fix anythingnC with the odd spanner or two, a pair of pliers, and a screwdriver...a     -- n
 Paul Sture Switzerland    ------------------------------  % Date: Sat, 19 Oct 2002 14:16:24 +0200 6 From: Arne =?ISO-8859-1?Q?Vajh=F8j?= <arne@vajhoej.dk>2 Subject: Re: OT: Oracle's take on Microsoft's .NET) Message-ID: <3DB14D18.6040409@vajhoej.dk>    Shane Smith wrote:  * > http://www.oracle.com/ip/net/claim1.html >  > I don't think they like it.h   True.n  * Not the first time, that Oracle is talking against Microsoft.   Arne   ------------------------------   Date: 19 Oct 02 17:50:35 +0200) From: p_sture@elias.decus.ch (Paul Sture)- Subject: Re: TLZ09 on Alpha?) Message-ID: <CfD2C4Okv77N@elias.decus.ch>o  j In article <5.1.0.14.2.20021018135711.00ad4fc0@raptor.psccos.com>, Dan O'Reilly <dano@process.com> writes:. > At 12:10 PM 10/18/2002, Arne Vajh=F8j wrote: >>Dan O'Reilly wrote:  >>I >>>Is this supported, or am I going to have to use this only on my VAXen?h5 >>>This is the tabletop model, by the way (TLZ09-DB).n >> >>Why should it not work ? >>, >>SCSI devices tend to work as SCSI devices. > G > Yes and no.  I tried a TK50 on my AS200 and it would get all sorts ofn/ > errors.  Works fine on my Vaxstation 4000/60.t > C Was the Tk50 even a true SCSI device? (whatever true SCSI means...)    -- t
 Paul Sture Switzerland    ------------------------------  + Date: Sat, 19 Oct 2002 11:11:04 -0500 (CDT)e From: sms@antinode.org Subject: Re: TLZ09 on Alpha?) Message-ID: <02101911110438@antinode.org>t  ) From: p_sture@elias.decus.ch (Paul Sture)hE > Was the Tk50 even a true SCSI device? (whatever true SCSI means...)i  D    The TK50Z-FA was used with the VAXstation/MicroVAX 2000.  I don'tF wish to restart the SCSI-yes-or-no argument about that one.  (However,F long ago I did try a -FA on a VAX 4000 model 300 with a KZQSA, and had no trouble.)  E    The TK50Z-GA was used with the VAXstation 3100 (among others), andB? I've heard no arguments about that not having a SCSI interface.h  D    I have a (little-used) -GA on a KZPBA-CX adapter in my AlpSta 200> 4/233 (VMS V7.2-1), and it also seems to work just fine there.  E    My long-standing problem with the built-in SCSI bus on that systemnH was finally traced to the lack of termination at the CD-ROM drive end of$ the bus, not that anyone else cares.  H ------------------------------------------------------------------------  C    Steven M. Schweda               (+1) 651-699-9818  (voice, home)lC    382 South Warwick Street        (+1) 763-781-0308  (voice, work)aG    Saint Paul  MN  55105-2547      (+1) 763-781-0309  (facsimile, work) 9    sms@antinode.org                sms@provis.com  (work)S   ------------------------------   Date: 19 Oct 02 11:23:49 +0200) From: p_sture@elias.decus.ch (Paul Sture) 7 Subject: Re: VMS tabletop without Office(tm): no chanceo) Message-ID: <cVu4bCl9Eijx@elias.decus.ch>r  w In article <01KNTFS7XV4Y9OF8NK@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:t< >> > Who will start porting MS Office(tm) to VMS/DECwindows? >> iG >> No one, I hope.  The last thing VMS needs is apps that provide theirs >> own security holes. >  > Right. > + >> How about a port of OpenOffice, instead?  >  > Better, but still not good.  > F > If you like Microsoft products, use them.  If you don't, using some D > "replacement" project which still has to dance to the tune of the K > moving-target proprietary Microsoft specifications is just going part of t
 > the way. > H > Why not go ALL the way and say that the only sensible way to exchange K > information between various computing platforms is to use a format which  I > is not proprietary to, or for the most part determined by, just one of t > the players. >i  4 IIRC StarOffice now uses XML with that goal in mind.  tC > Open SOURCE with respect to the source code for applications is apC > different matter entirely.  On the other hand, there should be NOoJ > substitute for open standards when the transfer of files between various > platforms is concerned.  > I > Impossible?  Maybe.  I don't think so.  But a self-fulfilling prophecy oK > if people don't think about a completely different paradigm just because lE > they think that Microsoft is too powerful and it will never happen o	 > anyway.a  D The only reason I bought Office 97 was that my (mainly bean counter,B non-technical) customers kept sending me Excel spreadsheets, which' if couldn't read with my existing apps.a > E > What really gets me are folks who post here and when there is some fJ > problem with their email, their news reader or whatever, complain about G > the Microsoft software they are using.  Folks, you can be completely a@ > Microsoft-free TODAY, certainly for your private stuff.  Your F > corporation might be another matter, but in principle that's just a  > matter of time.  >-  D Even those who are stuck with MS Windows as their (cough) OS, do notF _have_ to use Outlook. There are plenty of cheap or free alternatives.   B > I don't know if Patrick Daly lurks here.  Like many LaTeX gurus D > (including the developer, Lamport, who work(ed?) for the research G > division at DEC), he's a VMS man.  Perhaps he could relate the story tG > about how he got all the secretaries at his institute to use LaTeX.  lH > There are many communities in which all processing is done in LaTeX.  K > For those who complain that it is not WYSIWYG (though one should keep in sG > mind that, often, what you see is all you've got), there are WYSIWYG g > interfaces available for it. >a  E Agreed. I had a team of French secretaries quite happily using RUNOFF M 20 years ago. They claimed they had little or no English, but they managed to.: master it in preference to their golfballs at the time :-)   I > Powerful.  Portable (runs on all platforms---source code also publicly iD > available).  Can do all you need to do.  THE replacement for WORD.   Better quality too :-) -- c
 Paul Sture Switzerland    ------------------------------  % Date: Sat, 19 Oct 2002 12:32:14 +0200m6 From: Arne =?ISO-8859-1?Q?Vajh=F8j?= <arne@vajhoej.dk>7 Subject: Re: VMS tabletop without Office(tm): no chanceh) Message-ID: <3DB134AE.9060704@vajhoej.dk>o   Ed Wensell III wrote:    > Didier Morandi wrote:s9 >>Who will start porting MS Office(tm) to VMS/DECwindows?j >> >>I ask the question.  > 9 > Maybe the question should be who's going to take on thee > OpenOffice/StarOffice VMS?  ! That at least would be possible !i  4 I have serious doubts how much it would be used, but2 that is one of the good thing about free software:2 1000 users paying 0 $ is the same as 1000000 users paying 0 $ !   Arne   ------------------------------   Date: 19 Oct 02 17:15:49 +0200) From: p_sture@elias.decus.ch (Paul Sture)o7 Subject: Re: VMS tabletop without Office(tm): no chanced) Message-ID: <thmlH5hvDZqT@elias.decus.ch>o  c In article <$Owna8veIj07@eisner.encompasserve.org>, koehler@encompasserve.org (Bob Koehler) writes:ey > In article <01KNTFS7XV4Y9OF8NK@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:m >> nI >> Why not go ALL the way and say that the only sensible way to exchange iL >> information between various computing platforms is to use a format which J >> is not proprietary to, or for the most part determined by, just one of  >> the players.o > I >    Because we have a customer which has declared that Microsoft Word is 0 >    thier standard document interchange format. >   H But increasingly over the last year or less, I am finding that .PDFs are+ being used, both on the web and internally.    -- n
 Paul Sture Switzerlandt   ------------------------------   Date: 19 Oct 02 17:47:21 +0200) From: p_sture@elias.decus.ch (Paul Sture) 7 Subject: Re: VMS tabletop without Office(tm): no chance ) Message-ID: <w9ATE$qPOEa2@elias.decus.ch>   c In article <LB9sQu3NwWUC@eisner.encompasserve.org>, Kilgallen@SpamCop.net (Larry Kilgallen) writes: e > In article <$Owna8veIj07@eisner.encompasserve.org>, koehler@encompasserve.org (Bob Koehler) writes:sz >> In article <01KNTFS7XV4Y9OF8NK@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes: >>> J >>> Why not go ALL the way and say that the only sensible way to exchange M >>> information between various computing platforms is to use a format which sK >>> is not proprietary to, or for the most part determined by, just one of n >>> the players. >> .J >>    Because we have a customer which has declared that Microsoft Word is1 >>    thier standard document interchange format.r > J > After trying for the past 3 days to get such a document from a customer,J > I would say that "Microsoft Word Format" is an under-specification.  OneK > must determine the version and platform.  Even their RTF is not portable.r  C Before your comment, I would have called RTF "the _most_ portable".o   -- n
 Paul Sture Switzerlandm   ------------------------------   Date: 19 Oct 02 18:01:34 +0200) From: p_sture@elias.decus.ch (Paul Sture)d7 Subject: Re: VMS tabletop without Office(tm): no chance ) Message-ID: <TX5wa2jB5G4j@elias.decus.ch>w  W In article <3DB06CE3.EED109C1@Free.fr>, Didier Morandi <Didier.Morandi@Free.fr> writes:  > Bob Koehler wrote: >> dz >> In article <01KNTFS7XV4Y9OF8NK@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes: >> >J >> > Why not go ALL the way and say that the only sensible way to exchangeM >> > information between various computing platforms is to use a format which K >> > is not proprietary to, or for the most part determined by, just one oft >> > the players.t >>  J >>    Because we have a customer which has declared that Microsoft Word is1 >>    their standard document interchange format.c > L > Tell the Customer that there ARE standards. And the standard today is pdf. > N > Microsoft will understand that one day when all FrontPage sites will be justQ > unreadable by any other navigator than ie and when the Field starts to be upsetnQ > about it. There IS a W3C standard and an HTML 4 standard and and and. MicrosoftoO > has changed their default font characters to make text unreadable by 51 years7Q > old eyes using Netscape. This is dishonest. The antislash for subdirectories is:N > dishonest, all these new accentuated characters coding witch^H^H^H :-) which> > break Netscape 4.x is dishonest (does this word exist, btw?) > M > Do you know that Netscape HAD to include non standard coding to allow their.+ > users to be able to read FrontPage sites?c > 7 > In French, we say "the world is walking on its head".5 >  Bien-sur  C Let just try chucking www.microsoft.com at http://validator.w3.org/n   Total, abject failure.  G Firstly, "Fatal Error: no document type declaration; will parse withoutL validation"E  K So we then choose a manual override of HTML 4.01 transitional, and it still0 barfs with multiple errors.   @ Some statement about "the world's leading software company", eh? -- a
 Paul Sture Switzerland9   ------------------------------   End of INFO-VAX 2002.578 ************************