1 INFO-VAX	Thu, 06 Dec 2001	Volume 2001 : Issue 677       Contents:? 1248           Your Vacation Winning !                  1715619  DCPS 1.1 for serial ports  Re: DCPS problem Urgent $ Re: DCPS slow by printing LN06-Files' Re: Example of server code using thread & Re: File cache performance for dummies' Re: Is it a DEC C problem, or is it me?  Re: KFPSA compatibility  Re: KFPSA compatibility  KZPSC and SWXCR questions " Memo:  Re: HP OpenView and OpenVMS+ Need to breakout TAR file on OpenVMS system / Re: Need to breakout TAR file on OpenVMS system / Re: Need to breakout TAR file on OpenVMS system  Re: Oracle problems. Re: Oracle problems. Re: Oracle problems.! Re: Oracle RDB announcment/letter  Re: PC connection # Re: PCSI problem with release notes # Re: PCSI problem with release notes  Problem with the Info-VAX list? # RE: Problem with the Info-VAX list? P Re: Q: Tool or script to remove nonprinting characters from a SET HOST log? log? Re: RECALL does not work Re: RECALL does not work RE: RECALL does not work Re: RECALL does not work" Testing software for OpenVMS/Alpha& Re: Testing software for OpenVMS/Alpha  RE: tubes (was: RE: DEC is DEAD)  Re: tubes (was: RE: DEC is DEAD) RE: Veritas on VMS?  Re: Veritas on VMS?  Re: Veritas on VMS?  Re: Veritas on VMS?  Re: Veritas on VMS?  Re: Veritas on VMS? F Re: VMS Effort (Was: Microsoft Pyramid Collapses Enron and Hewlett PacF Re: VMS Effort (Was: Microsoft Pyramid Collapses Enron and Hewlett PacJ Re: VMS Effort (Was: Re: Microsoft Pyramid Collapses Enron and Hewlett PacJ Re: VMS Effort (Was: Re: Microsoft Pyramid Collapses Enron and Hewlett Pac) Re: VMS marketing (advertising) sighting. ) Re: VMS marketing (advertising) sighting. ) Re: VMS marketing (advertising) sighting. ) Re: VMS marketing (advertising) sighting. 4 Re: Was the Alpha to Intel agreement ever approved ?L RE: Why does file access preformance over WAN depend greatly on the command?P Re: Why does file access preformance over WAN depend greatly on the command? com/ Re: XML Test Problems & Java Performance Query. / Re: XML Test Problems & Java Performance Query.   F ----------------------------------------------------------------------   Date: Thu, 6 Dec 2001 03:19:17% From: 1917156travelincentives@aol.com H Subject: 1248           Your Vacation Winning !                  17156199 Message-ID: <iss.433e.3c0e79f3.8f2c0.1@mx2.east.saic.com>   > You have been specially selected to qualify for the following:  0 Premium Vacation Package and Pentium PC Giveaway6 To review the details of the please click on the link # with the confirmation number below:    http://vacation.81832.com    Confirmation Number#Lh340 J Please confirm your entry within 24 hours of receipt of this confirmation.  " Wishing you a fun filled vacation!J If you should have any additional questions or cann't connect to the site % do not hesitate to contact me direct: , mailto:vacation@btamail.net.cn?subject=Help!   ------------------------------  % Date: Wed, 05 Dec 2001 21:31:11 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca> " Subject: DCPS 1.1 for serial ports+ Message-ID: <3C0ED84C.490918D@videotron.ca>   P I have DCPS V1.1 on VAX-VMS driving a single postscript printer via serial port.  K I remembered some sort of text about having hardware flow control to enable N binary postscript. I finally got around to making the wire between the vax andN the printer full modem control. But low and behold, I find that DCPS no longer0 seesm to have any options for binary postscript.  H Looking in the DCPS$EXECUTIONS_QUEUE.COM it seems that the /HOSTSYNC andL /TTSYNC are hardcoded in there. So I removed them and put /COMMSYNC and thatQ *seems* to work. (although I have yet to test some really large postscript file).   5 Has anyone else dealt with this issue ? Any gotchas ?   M Also, I traced the stuff going to the printer and found out that DCPS sends a J whole lot of crud prior to sending my postscript code. Is there some magicK incantation or logical that would result in DCPS just testing to see if the J printer is ready and if so, then send my code to it without surrounding it; with all the crud that gets sent before and after the job ?   J In particular, if I print 5 pages on 5 pieces of paper, the job never endsM until I put a 6th piece of paper so that the printer will accept all the crud N that DCPS adds after my job (thinking it will be another page). Has anyone had that problem too ?  F I am curious as to the implications of my print file having the properM Postscript Document Structure Convention commentsat the top and end, but DCPS P adding lots of crud before and after my data. Doesn't that confuse the printer ?  J Finally, I finally found out that the large delay before the job prints isK spent entirely waiting for the DCPS crud before my job. There seems to be a K fair amount of time between each of the modules in DCPS$LIBRARY get sent to A the printer prior to my job. Is there a way to make this faster ?    ------------------------------  % Date: Wed, 05 Dec 2001 15:21:34 -0500 0 From: Paul Anderson <paul.r.anderson@compaq.com>  Subject: Re: DCPS problem Urgent; Message-ID: <051220011521347205%paul.r.anderson@compaq.com>   @ In article <3C0CFEB7.DBCDF6E1@medecine.unige.ch>, Alain Chappuis) <alain.chappuis@medecine.unige.ch> wrote:   	 > I used:  >  > stop/queue/next printer  > delete/queue printer > @sys$startup:dcps$startup.com  >  > after them the queue is ready  > 1 > but nothing to do and allways the same message!   D Do you have another printer you can use?  If so, do you have similar problems on that printer?   D You could collect a DCPS "diag" file that might give me a clue as to what is happening.  H ------------------------------------------------------------------------G A DCPS "diag" file helps diagnose the communication between the OpenVMS  system and the printer.   5 To create a DCPS diag file, define this logical name:   0     $ DEFINE /EXECUTIVE_MODE /SYSTEM DCPS$TEST 1     * and then stop and start the DCPS queue(s).  ! Print with the following command:   E     $ PRINT /NOTIFY /PARAMETERS=(your_parameters,DIAG=FILE=BOTH) file   D Diag files contain all data sent between the host and the printer in
 files called:   *     SYS$LOGIN:DCPS$JOB_entry_DIAG_FROM.LOG(     SYS$LOGIN:DCPS$JOB_entry_DIAG_TO.LOG     D Diag files contain all print data information, so be cautious if the5 files being printed contain confidential information. H ------------------------------------------------------------------------   Paul   --    Paul Anderson   OpenVMS Engineering    Compaq Computer Corporation    ------------------------------  % Date: Wed, 05 Dec 2001 15:45:12 -0500 0 From: Paul Anderson <paul.r.anderson@compaq.com>- Subject: Re: DCPS slow by printing LN06-Files ; Message-ID: <051220011545122290%paul.r.anderson@compaq.com>   < In article <3c0d3ac1$1@news.kapsch.co.at>, Peter LANGSTOEGER <eplan@kapsch.net> wrote:   K > In article <261120011043145585%paul.r.anderson@compaq.com>, Paul Anderson & > <paul.r.anderson@compaq.com> writes: > K > > You're welcome, but I don't understand why defining the NO_SYNC logical K > > name would solve your problem.  If it did, I'm glad, but the only thing J > > the logical name does is to bypass the status check DCPS issues when aF > > queue first starts.  Yes, your Lexmark printer probably needs thisI > > logical name defined (my Optra T616 certainly does) but it would only K > > cause the queue to not start, not affect the timing of jobs printing on  > > an already-running queue.  > > > Are you about that (means, Paul, I assume you are sure ;-) ? > 7 > I had different observations the last time I checked. M > Status is not checked at queue start time (you even can start a queue while I > the printer is switched off) but on job start time. And then the timing G > of every job (but not on every file in the same job) _is_ effected...   F You're correct.  I think I've inherited brain cramps from having kids.  D Nothing is sent by DCPS when the queue starts.  It's the query aboutF printer model that's sent only at the beginning of the first job after@ the queue starts.  The status request (control-T) is sent at the beginning of every job.   G Usually, the status request is either responded to by the printer or is D ignored.  In other words, DCPS will get an answer back right away orF will wait forever, causing the job to remain "Starting" forever.  It'sB possible that the printer could wait some period of time to answer@ (maybe it needs time to switch between language-sensing mode and8 PostScript, for example) and that would cause the delay.   So please excuse my faux-pas.    Paul   --    Paul Anderson   OpenVMS Engineering    Compaq Computer Corporation    ------------------------------   Date: 5 Dec 2001 15:37:09 -0800 1 From: jason_odonnell@erinet.com (Jason O'Donnell) 0 Subject: Re: Example of server code using thread= Message-ID: <5c8ffd05.0112051537.67000b84@posting.google.com>   O Binh Nguyen <binh@egh.com> wrote in message news:<3C0E0D2E.75D04927@egh.com>... 
 > Rushlan, > Q > I need to write a server that will listen on port 443 (secure port) for client. M > When a client wants to talk, I'll spun a thread to listen to the port while Q > the original thread process the request.  The number of clients could be in the = > thousands!  None of the threads need to talk to each other.  >  > Binh >    Binh,   ? On an OpenVMS platform one of the performance considerations is E process creation and deletion.  I would recommend creating a listener F process and have a set of worker processes with multiple threads (less= than 32) for each process.  All of these should be created at F application startup.  So, if you need 1000 connections maximum and youE had 20 threads per worker process, you would need one listener and 50  worker processes.   B If you search this site, you should find some discussion on how toB hand sockets off between processes.  We have implemented a process such as this on our own system.    Food for thought.    JMOD   ------------------------------   Date: 5 Dec 2001 11:00:42 -0800 + From: stephane_paquin@hotmail.com (SPaquin) / Subject: Re: File cache performance for dummies < Message-ID: <fdd7874.0112051100.21a04d3e@posting.google.com>  / Just to let you know I just found the answer in U http://www.compaq.com/support/asktima/operating_systems/00917ED0-DD2916C0-1C0069.html    In one line only.   ? Our problem in reason number 2. It seems I have found a support 	 source...    Thanks to all who helped.    Stphane  o stephane_paquin@hotmail.com (SPaquin) wrote in message news:<fdd7874.0112040716.2c9402a9@posting.google.com>... A > I am new to performance analysis so forgive my errors. I have 3 G > identical VAX VMS 3100-98 512Mb having the same software on it. VAX A F > is backup, has 54 processes(11 users). VAX B(104 proc, 53 users) andD > C(80 proc,31 users) are in production. The same compilation job isD > slower on VAX A than on VAX C. Since VAX A has almost no workload,B > this is puzzling. I have read the OpenVMS Performance ManagementE > manual and searched my systems for the IO cache performance. I used D > MON FILE and MON IO for 24h to get a sense of the hit % and other. > Here is some data: > % >                 VAX A      B      C & > DIR FCB           63%      71     99& > DIR DATA          16       22     99& > FILE HDR          35       52     95 > & > DIRECT IO         54       60     19& > BUFFER IO         33       99     56' > PAGE READ IO RATE 2.1      3.2    2.6  > G >  I noticed a very bad hit rate for DIR DATA on VAX A and B, excellent G > on C. How is that possible? I found out that the DIR DATA hit rate is H > dependant on ACP_DIRCACHE parameter. I have searched the current value > for this parameter.  > % >                 VAX A      B      C : > ACP_DIRCACHE    65535*   65535* 16384   (*=maximum size) > H > The fact that the current value has grown to the maximum size on A andB > B is puzzling. And C is getting great performance from a smallerC > cache. The command sh mem/cach/full gives a VIOC Read Hit Rate of B > 94-96% range for all vax with sizes of 11394, 61053 and 14223. AH > defrag runs on all VAXes every day. Last item, the direct IO count forG > the same compilation job on the VAXes is 2500 on VAX A and B, 1000 on  > VAX C. > F > I think the file system cache is wrong. How can I determine if it is9 > working properly ? What other investigations can I do ?  >  >  Suggestions are welcome.  >  Thanks to VMS gurus > 
 >  SPaquin   ------------------------------  % Date: Wed, 05 Dec 2001 20:20:57 +0100 & From: Michael Joosten <joost@c-lab.de>0 Subject: Re: Is it a DEC C problem, or is it me?$ Message-ID: <3C0E7399.41C6@c-lab.de>   Martin Vorlaender wrote: > . > John Eisenschmidt (jeisensc@aaas.org) wrote:2 > > "John Eisenschmidt" <jeisensc@AAAS.ORG> wrote: > C > > So I figured I'd be brave and see how incompatible the code is.  > > - > > $ CC /DEBUG/INCLUDE=TCPWARE_INCLUDE foo.c  > > 9 > >     if (getpeername(sockp, &clit_addr, &addrlen) < 0)  > > ..........................^ L > > %CC-W-PTRMISMATCH, In this statement, the referenced type of the pointerL > > value "&clit_addr" is "struct sockaddr_in", which is not compatible with > > "struct sockaddr".= > > at line number 256 in file DSA100:[QUXX.R71A.SRC]foo.c;59  > = > I don't know enough about network programming to rate this.  >    The code should have read B if (getpeername(sockp, (struct sockaddr*)&clit_addr, &addrlen) <0)  G as the prototypes only takes sockaddr, whereas the actual arguments are = always sockaddr_in or sockaddr_un (err, or sockaddr_dent ??))  Sort of 'polymorphism' in C...   But not critical...   9 > >     if (getpeername(sockp, &clit_addr, &addrlen) < 0) + > > ......................................^ L > > %CC-W-PTRMISMATCH, In this statement, the referenced type of the pointerK > > value "&addrlen" is "int", which is not compatible with "unsigned int". = > > at line number 256 in file DSA100:[QUXX.R71A.SRC]foo.c;59  > L > Ignore this one. If you change addrlen's declaration to "unsigned int", it6 > will probably generate a warning at some other line. >   G Yes, old code. Nowadays, addrlen is supposed to be a of 'size_t', which E is normally an 'unsigned int'. You might try to change the definition C (not declaration!) of addrlen into 'size_t addrlen;', provided that 5 'stdlib.h' is also included (which seems to be case).          > M > They're just warnings, not errors. Like the "int" vs. "unsigned int" issue, N > they can be harmless. But the other type conflict - char[] vs. char - really > looks like an error. >   / Yes, you should inspect this one more closely.       --  * Michael Joosten, SBS C-LAB, joost@c-lab.de* Fuerstenallee 11, 33094 Paderborn, Germany, Phone: +49 5251 606127, Fax: +49 5251 6060658 C-LAB is a cooperation of University Paderborn & SIEMENS   ------------------------------   Date: 5 Dec 2001 15:03:11 -0600 + From: kuhrt@encompasserve.org (Marty Kuhrt)   Subject: Re: KFPSA compatibility3 Message-ID: <df2qY6VvZhw8@eisner.encompasserve.org>   C I'd boot the system, try it and find out for sure.  I went a couple B of rounds with a vendor over a couple of KZPSAs (or was it KZPBAs)C that wouldn't show at the console, but worked fine.  You might also ( want to see if the firmware is upto rev.  U In article <uIzO7.386$lI5.814392@paloalto-snr2.gtei.net>, dittman@dittman.net writes: @ > Is the KFPSA compatible with the PWS 500au?  My system doesn't? > see the board, which makes me suspect the board is bad, but I 1 > thought I'd check before contacting the seller.  > --   > Eric Dittman > dittman@dittman.net ? > Check out the DEC Enthusiasts Club at http://www.dittman.net/    ------------------------------  # Date: Wed, 05 Dec 2001 22:37:40 GMT  From: dittman@dittman.net   Subject: Re: KFPSA compatibility: Message-ID: <UkxP7.1359$292.274406@paloalto-snr1.gtei.net>  E : I'd boot the system, try it and find out for sure.  I went a couple D : of rounds with a vendor over a couple of KZPSAs (or was it KZPBAs)E : that wouldn't show at the console, but worked fine.  You might also * : want to see if the firmware is upto rev.  D When the KFPSA is installed the SCSI drives attached to the built-inE ISP1040 controller and KZPBA (or whatever the Symbios C810 controllernC is named) do not show up, so I can't boot.  I've got the latest PWS  firmware loaded. --   Eric Dittman dittman@dittman.netl= Check out the DEC Enthusiasts Club at http://www.dittman.net/t   ------------------------------  % Date: Wed, 05 Dec 2001 23:53:12 -0600 / From: Chris Scheers <chris@applied-synergy.com>d" Subject: KZPSC and SWXCR questions3 Message-ID: <3C0F07C8.C8DB69F8@applied-synergy.com>l  D I've been setting up a KZPSC (1 channel) configuration and have some
 questions.  D Using RA200RCU 4.78-08, printing the configuration shows the option:  # 	Auto Rebuild Management : Disabled1  G I haven't found any documentation for this (or even how to change it). : What is this option?    C I have SWXCR V2.0 loaded on AlphaVMS V7.1.  Where can I get a latergF version of SWXCR?  Is it better to switch to SWCC?  Does SWCC have anyD more capabilities than SWXCR (for use with a KZPSC)?  Is there a VMS= based version of SWCC?  (The display utility, not the agent.)i   Thanx!  G -----------------------------------------------------------------------r$ Chris Scheers, Applied Synergy, Inc.  C Voice: 817-237-3360            Internet: chris@applied-synergy.com a   Fax: 817-237-3074    ------------------------------  % Date: Wed, 05 Dec 2001 17:22:38 +0000s From: paul.beaudoin@hsbc.com+ Subject: Memo:  Re: HP OpenView and OpenVMSaE Message-ID: <OF86FCC827.F2463C30-ON80256B19.005F271A@systems.uk.hsbc>r   <snip>  7 >  $   snmp_trapsnd == "$SYS$SYSTEM:TCPIP$SNMP_TRAPSND"d	 >usage iscA >snmp_trapsnd enterprise agent-address generic specific timeticksaG >[-v version] [-c community] [-h host] [-p port] [-tcp] {variable [typek >value]}  H very interesting.... Is there anywhere I can get doc on how to use this?   Many thanksf   Paul        ' ** HSBC's website is at www.hsbc.com **a  D ********************************************************************B  This message and any attachments are confidential to the ordinaryB  user of the e-mail address to which it was addressed and may also>  be privileged. If you are not the addressee you may not copy,8  forward, disclose or use any part of the message or itsC  attachments and if you have received this message in error, pleaseEB  notify the sender immediately by return e-mail and delete it from
  your system.t  =  Internet communications cannot be guaranteed to be secure oreA  error-free as information could be intercepted, corrupted, lost,n>  arrive late or contain viruses. The sender therefore does not?  accept liability for any errors or omissions in the context ofv?  this message which arise as a result of Internet transmission.o  tD  Any opinions contained in this message are those of the author and ?  are not given or endorsed by the HSBC Group company or office p=  through which this message is sent unless otherwise clearly sA  indicated in this message and the authority of the author to so e3  bind the HSBC entity referred to is duly verified.s  D ********************************************************************   ------------------------------  $ Date: Wed, 5 Dec 2001 11:35:03 -05003 From: "Pohl, Kathy" <Kathy.Pohl@itec.mail.suny.edu> 4 Subject: Need to breakout TAR file on OpenVMS systemI Message-ID: <0B41ACBECF83734D98ABC83DD8E097E15DE7C8@cipher.itec.suny.edu>6  J This message is in MIME format. Since your mail reader does not understand< this format, some or all of this message may not be legible.  ' ------_=_NextPart_001_01C17DAA.CA38B1B0s Content-Type: text/plain;  	charset="iso-8859-1"a     	Hi,  E 	Is there a way to breakout a TAR file on an Alpha OpenVMS system.  I . need to open a TAR file that contains all the D 	license paks for a TRU64 UNIX server, but I do not have access to a
 UNIX system. h   	Thanks, 	Kathy      ' ------_=_NextPart_001_01C17DAA.CA38B1B0d Content-Type: text/html; 	charset="iso-8859-1"a+ Content-Transfer-Encoding: quoted-printablee  1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">r <HTML> <HEAD>9 <META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =a charset=3Diso-8859-1">@ <META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
 5.5.2653.12">n: <TITLE>Need to breakout TAR file on OpenVMS system</TITLE> </HEAD>  <BODY> <BR>  > <P>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <FONT SIZE=3D2 = FACE=3D"Arial">Hi,</FONT>S </P>  > <P>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <FONT SIZE=3D2 =B FACE=3D"Arial">Is there a way to breakout a TAR file on an Alpha =G OpenVMS system.&nbsp; I need to open a TAR file that contains all the =l </FONT></P>m  > <P>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <FONT SIZE=3D2 =H FACE=3D"Arial">license paks for a TRU64 UNIX server, but I do not have =  access to a UNIX system. </FONT> </P>  > <P>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <FONT SIZE=3D2 = FACE=3D"Arial">Thanks,</FONT>=? <BR>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <FONT SIZE=3D2 =t FACE=3D"Arial">Kathy</FONT>. </P> <BR>   </BODY>. </HTML>C) ------_=_NextPart_001_01C17DAA.CA38B1B0--e   ------------------------------  % Date: Wed, 05 Dec 2001 20:53:52 -0500 - From: JF Mezei <jfmezei.spamnot@videotron.ca>d8 Subject: Re: Need to breakout TAR file on OpenVMS system, Message-ID: <3C0ECF90.38415170@videotron.ca>   > "Pohl, Kathy" wrote: > 
 >         Hi,o > N >         Is there a way to breakout a TAR file on an Alpha OpenVMS system.  I/ > need to open a TAR file that contains all thee > M >         license paks for a TRU64 UNIX server, but I do not have access to a_ > UNIX system.  6 Look at the freeware CD, available online somewhere atC http://www.openvms.compaq.com  there should be a tar utility there..   ------------------------------  # Date: Thu, 06 Dec 2001 02:53:30 GMT 2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)8 Subject: Re: Need to breakout TAR file on OpenVMS system1 Message-ID: <K4BP7.190$BK1.3576@news.cpqcorp.net>A   In article <0B41ACBECF83734D98ABC83DD8E097E15DE7C8@cipher.itec.suny.edu>, "Pohl, Kathy" <Kathy.Pohl@itec.mail.suny.edu> writes:g  ?   Please turn off MIME when sending email to INFO-VAX.  Thanks.pA   Your PC tool likely has an option that permits you to use plainuA   text when sending email -- please use plain text messages with  @   the INFO-VAX list and with the comp.os.vms newsgroup.  Thanks!;   (The INFO-VAX list has a bi-directional gateway with the i   comp.os.vms newsgroup.)i  C :	Is there a way to breakout a TAR file on an Alpha OpenVMS system.y  F   Yes.  There exist tar utilities for OpenVMS, with versions available   on the OpenVMS Freeware.    G   The OpenVMS Frequently Asked Questions (FAQ) has references to tools tF   such as tar, and also has pointers to the Freeware website.  The FAQ*   is available at various sites including:  -       http://www.openvms.compaq.com/freeware/-  7   The Freeware is available at various sites including:n  -       http://www.openvms.compaq.com/freeware/A  D   If you have an OpenVMS V7.3 distribution kit, there is a tar tool E   included in the [000tools] area present on each of the two volumes f   of the Freeware CD-ROM disks.l  C   Please remember to check the FAQ, if for no other reason than in wD   kindness to the folks in the newsgroup that have answered certain B   questions, um, a few times. :-)  But seriously, checking the FAQ>   first is of interest to you, too -- you can search the (textA   version of the) FAQ to get answers to the common questions far hA   more quickly than posting and then waiting for a response.  :-)e    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  5 Date: Wed, 5 Dec 2001 19:53 +0000 (GMT Standard Time)n- From: dmacdonald@cix.co.uk (Duncan Macdonald)  Subject: Re: Oracle problems.sC Message-ID: <memo.20011205195301.10979A@dmacdonald.compulink.co.uk>   0 In article <9uh0g8$7kk$1@newsg1.svr.pol.co.uk>, : LGBowden@bowdenfamily.fsnet.co.uk (Leigh G. Bowden) wrote:  ; > We have an AlphaServer 2100 4/200 with two CPU's and 1GB  
 of memory andY= > plenty of disks - all shadowed. As far as I am aware it is s using Oracle > 7.1.5 on OpenVMS 6.2 AXP.  > < > It appears to be unreliable. After about three weeks this  system will lock: > up and the console will report "Insufficient memory for  operation" type = > errors and will allow nobody to logon even at the console.   The only way to $ > get out of this to power cycle it. > 2 > Has anybody experienced this before and got any  suggestions for a fix??s >  >  > < Some versions of Oracle (7.1.5 was one!!) have a bug in the ; Oracle listener that causes a user mode logical name table B< to be created for each SQL*Net connection. These tables are 3 not removed when the SQL*Net connection terminates. 9 Each of these tables eats a small amount of paged pool - f7 when the system cannot expand paged pool further , the  ; system gets to the state where it is no longer possible to k log in or run images.o/ To see if this is the problem enter the command ( SH LOG */TABLE=LNM$SYSTEM_DIRECTORY/full9 If you see hundreds (or thousands!!) of Oracle user mode  = tables then you need to write a small command file to be run V  at intervals to delete them all.   ------------------------------  $ Date: Wed, 5 Dec 2001 21:11:58 +0100- From: "Martin Hoogenboom" <martinh@xs4all.nl>C Subject: Re: Oracle problems. 3 Message-ID: <006e01c17dc9$1be6e5e0$8149130a@WWS006>n   Hi,w  8 Sounds like a problem we still have using some oracle 7.F The Remote oracle engineers use a program ORA_AGENTWORKS to moitor the	 databasesnF This program has a memory leak and keeps using virtual memory until it reaches it quota.g  G For some reason at the end it fauls up the system. Usually we crash the J program and everything returns to normal. The oracle engineers restart the4 program and in a few weeks it starts all over again.  L AFAIK Oracle is investigating the problem but for now we've simply increasedE the pagefile quota (still got plenty left) This will probably help usQ through for a few months.m  L If you need it i mixed a little dcl program that monitors a process and puts6 requested info in a CSV file. That's how we found out.   Regards7 Martin Hoogenboom@   ----- Original Message -----; From: "Leigh G. Bowden" <LGBowden@bowdenfamily.fsnet.co.uk>D To: <Info-VAX@Mvb.Saic.Com>m) Sent: Tuesday, December 04, 2001 12:07 AMl Subject: Oracle problems.h    H > We have an AlphaServer 2100 4/200 with two CPU's and 1GB of memory andI > plenty of disks - all shadowed. As far as I am aware it is using Oraclel > 7.1.5 on OpenVMS 6.2 AXP.. > L > It appears to be unreliable. After about three weeks this system will lockI > up and the console will report "Insufficient memory for operation" typesL > errors and will allow nobody to logon even at the console. The only way to$ > get out of this to power cycle it. >tI > Has anybody experienced this before and got any suggestions for a fix??    ------------------------------  % Date: Wed, 05 Dec 2001 22:02:58 +0100.( From: Paul Sture <paul.sture@bluewin.ch> Subject: Re: Oracle problems.i- Message-ID: <VA.000004d8.cb2232ec@bluewin.ch>v  G In article <Tj6P7.69396$Z2.1019213@nnrp1.uunet.ca>, Peter Weaver wrote:oH > "Leigh G. Bowden" <LGBowden@bowdenfamily.fsnet.co.uk> wrote in message+ > news:9uh0g8$7kk$1@newsg1.svr.pol.co.uk...nJ > > We have an AlphaServer 2100 4/200 with two CPU's and 1GB of memory andK > > plenty of disks - all shadowed. As far as I am aware it is using Oracle  > > 7.1.5 on OpenVMS 6.2 AXP.M > >eN > > It appears to be unreliable. After about three weeks this system will lockK > > up and the console will report "Insufficient memory for operation" typerN > > errors and will allow nobody to logon even at the console. The only way to& > > get out of this to power cycle it. > > K > > Has anybody experienced this before and got any suggestions for a fix??i > M > I do not know if this will help or not, but it sounds like a problem we hadtL > with an old version or Oracle (I don't think it was 7.1.5, but it may haveL > been). Make sure you take a look at what the procedure is doing and run itK > in test mode (i.e.  comment out the actual deassign and double check that N > the table the .COM wants to deassign is actually no longer in use before youM > let the job run completely) since I do not know what versions of VMS/Oracleg > we had had this problem on.m > N That was my diagnosis too. When the logical name tables are "full", you can't  even perform a new logon.  ___n
 Paul Sture Switzerlandn   ------------------------------  # Date: Wed, 05 Dec 2001 23:25:17 GMTh6 From: "Andy Bustamante" <a_c_bustamante@earthlink.net>* Subject: Re: Oracle RDB announcment/letterF Message-ID: <x1yP7.31438$WC1.3202892@newsread2.prod.itd.earthlink.net>  D Thanks Sue,  I'm on the road next week and this will good news to my" customers (and senior management).   -- Andy Bustamante  Remove the ASCII 95s to reply-      = "Sue Skonetski" <susan.skonetski@compaq.com> wrote in messager- news:qtqN7.2097$RL6.63071@news.cpqcorp.net...  > got this last night. >r > suee >g6 > ____________________________________________________ >n > Dear Rdb Customer: >EA > On June 29, shortly after Compaq announced plans to consolidatee >>G > Compaq's entire 64-bit server family on the Itanium architecture fromi >rE > Intel, I wrote to you regarding Oracle Rdb, Oracle CODASYL DBMS and  > B > the Itanium Processor Family. At that time, I said we would work >tB > closely with Compaq and our customers to determine the best path >p< > forward for Rdb. I am writing to you now with an update on >m; > developments regarding a port of Rdb and DBMS to Itanium.- >-H > From our discussions with you, our customers, over the past months, it >sE > is clear that many of you are very interested in seeing Oracle Rdb,o >eC > Oracle CODASYL DBMS and Oracle9i available on OpenVMS on Itanium.C >VG > As a result of this interest, Juan Jones, Oracle's Vice President fors > B > the Systems Platforms Division, recently, released the following >= > statement: >D9 > Oracle and Compaq have a long and successful history oft >n> > delivering enterprise solutions to our OpenVMS customers. In >J: > July, Oracle released Rdb 7.1 for OpenVMS. In September, >G: > Oracle9i for OpenVMS was released. Given Compaq's recent >u? > announcement to consolidate its 64-bit servers on the Itaniumn > > > Processor Family (tm), Oracle's current plan is to team with >-> > Compaq and work toward a delivery of Oracle Rdb and Oracle9i >p< > for OpenVMS on IPF based upon Compaq's current engineering >e
 > roadmap. >0H > Consistent with this direction we have developed a preliminary project >:E > plan for our porting activities for both Rdb and CODASYL DBMS. This% >lC > preliminary plan was developed with the assistance of the OpenVMS, >t@ > Engineering Group at Compaq. This plan is based on the current >0* > delivery dates provided to us by Compaq. >r! > Highlights of the plan include:> >iF > * By early next year, Compaq will provide us with cross compilers to >N> > allow us to begin initial testing of our code in the Itanium >, > environment. > B > * In 2003, Compaq will deliver to us and their other partners an >hE > early version of OpenVMS for Itanium that will allow us to complete  >s< > the port and begin testing of Rdb in this new environment. >w@ > * In 2004, Compaq will deliver the first production release of >f > OpenVMS on Itanium.  >aB > Oracle normally ships supported versions of Rdb and CODASYL DBMS > F > within 90 days of Compaq's production release date for new operating >kC > systems and processors. Assuming Compaq delivers Itanium software  >uG > according to their announced schedule and that this software performsn >eF > as expected, we will deliver the first production release of Rdb and >tH > CODASYL DBMS for Itanium within 90 days of Compaq's production release >d > date for OpenVMS on Itanium. >tG > These are still early days and Compaq's plans and road maps are still  > F > evolving. We expect to learn more as we and Compaq progress with our >eA > respective porting efforts. In the meantime we will continue tot >o2 > enhance and improve Rdb on the OpenVMS platform. >rE > We will provide an updated report on our progress on Itanium at the  >pG > Rdb Forum at Oracle Headquarters near San Francisco on December 1 ande >0G > 2 as well as at other Rdb Forums to be scheduled throughout the worldj >:C > in 2002. I hope to see you at one of these events. (You can stilln >yF > register for the Forum at our web site, http://www.oracle.com/rdb/.) >o > Best regards,  >n
 > Kevin Duffy- > ! > Oracle Rdb Development Directore >tH > ---------------------------------------------------------------------- >tH > I sent you this mail because of your business relationship with Oracle >SC > Rdb. If you no longer wish to receive mailings from us about Rdb,  >SD > please reply to this message with the word "Remove" in the subject >tE > line of your message. We will immediately remove your name from ourf >C > mailing list.u >i >  >  >  >n >i   ------------------------------  # Date: Wed, 05 Dec 2001 21:17:59 GMT-2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: PC connection1 Message-ID: <bawP7.185$BK1.3508@news.cpqcorp.net>8  T In article <9ul11k$pb6$1@news1.kornet.net>, "David Lee" <phongle@kornet.net> writes:I :I went through the OpenVMS FAQ (part 2/5), hoping that I would find someaM :articles that would help in configuring my PC and an Alpha ES-40 running VMSiM :7.2 so it can see each other.  I want to be able to ftp the files from AlphamM :to the PC running NT 4.0.   I know it's not just plug the RJ-45 cable to the-M :FastEthernet Hub.  Something needed to be done on these machines, just don'tr :know what needed to be done  H   Please go ask the Microsoft Windows folks for details on the MicrosoftI   operating system and networking.  AFAIK, most Microsoft Windows systems*G   are configured and operate as clients and this provide an FTP client;nF   the systems do not operate as servers -- there are very likely some E   configurations around that likely provide server services, however.nC   That said, I would encourage you to continue this portion of the aD   question and the discussion in a Windows-oriented newsgroup -- theF   folks in the Microsoft Windows newsgroups will undounbtedly prove a /   better resource for (most) Windows questions.a  G   As for OpenVMS, you need to install and configure TCP/IP Services or cE   an equivilent IP stack on OpenVMS -- if you choose TCP/IP Services, B   V5.1 with ECO is current.  If you install and boot V5.1, it willB   attempt to start as a DHCP client, and off you (mostly :-) go...   C :If anyone have any ideas, I would appreciate some helps or advices   C   The OpenVMS documentation is referenced in the OpenVMS Frequently E   Asked Questions (FAQ).  Also, this topic arises fairly regularly intB   the Ask The Wizard area -- pointers from those questions to the F   OpenVMS documentation (and specifically the TCP/IP Services manuals)   are the usual fare.*  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Wed, 05 Dec 2001 20:36:29 +0100a, From: Didier Morandi <Didier.Morandi@gmx.ch>, Subject: Re: PCSI problem with release notes& Message-ID: <3C0E773D.533943B5@gmx.ch>   Martin Vorlaender wrote: > 	 > Hi all,n > G > in the development of the next (hopefully final) release of ht://Dig,lI > I'm putting together a PCSI kit, but hit a showstopper with the releases > notes: > ' > The statements in the PCSI$DESC file:_ > + >   file [HTDIG]HTDIG0301-5R5.RELEASE_NOTESa >     generation 03010505- >     release noteso2 >     source [000000]HTDIG0301-5R5.RELEASE_NOTES ;. >   file [HTDIG]GNU_GENERAL_PUBLIC_LICENSE.TXT >     generation 03010500=5 >     source [000000]GNU_GENERAL_PUBLIC_LICENSE.TXT ;" >  > The command used:E >  > $ PRODUCT PACKAGE htdig -3 >     /BASE=axpvms - >     /PRODUCER=gnu -p# >     /SOURCE=HTDIG_VMS_BUILDDIR: ->( >     /DESTINATION=HTDIG_TOP_BUILDDIR: -O >     /MATERIAL=AXPMV$DKA300:[USER.MV.PROJEKTE.HTDIG.3_1_5.REL5.install_vms.] -p >     /FORMAT=SEQUENTIAL - >     /LOG - >     /TRACE >  > withO > HTDIG_VMS_BUILDDIR=AXPMV$DKA300:[USER.MV.PROJEKTE.HTDIG.3_1_5.REL5.BUILD_VMS] E > HTDIG_TOP_BUILDDIR=AXPMV$DKA300:[USER.MV.PROJEKTE.HTDIG.3_1_5.REL5]p >  > results in the messages: > P > %PCSI-E-OPENIN, error opening AXPMV$DKA300:[USER.MV.PROJEKTE.HTDIG.3_1_5.REL5.; > INSTALL_VMS.][HTDIG]HTDIG0301-5R5.RELEASE_NOTES; as inputpA > %PCSI-I-PKGFIL, packaged [000000]GNU_GENERAL_PUBLIC_LICENSE.TXTa > M > So other files in AXPMV$DKA300:[...INSTALL_VMS.][000000] get packaged okay,EJ > but it looks for the release notes in the material root mangled with the > *destination* directory.  < I do not see any mangling here. It looks for the material in= AXPMV$DKA300:[USER.MV.PROJEKTE.HTDIG.3_1_5.REL5.INSTALL_VMS.]jK as you told PCSI to go in the package command. There it expects to find the8G relnotes file in subdirectory [.HTDIG], which gives a full file spec of B AXPMV$DKA300:[USER.MV.PROJEKTE.HTDIG.3_1_5.REL5.INSTALL_VMS.HTDIG] HTDIG0301-5R5.RELEASE_NOTESR   Is the file really in there?B Or maybe 'it' does not like the hyphen. Try an underscore instead.   My two centsP (I didn't know we could use concealed directories in the /material qualifier...)   D. -- mG   ----------------------------------------------------------------------E MORANDI Consulting.  WEB: http://Didier.Morandi.Free.fr/index_us.htmloE Pflanzschulstrasse 53, 8004 Zurich, Switzerland. GSM: +41 79 705 4670p/ 19, chemin de la Butte, 31400 Toulouse, France.s  H Disaster Recovery Plans, Computer Security Audits, DEC OpenVMS ExpertiseH On parle franais, Man spricht Deutsch, Habla Castellano, English spoken   ------------------------------  % Date: Thu, 06 Dec 2001 06:25:29 +0100 2 From: martin@radiogaga.harz.de (Martin Vorlaender), Subject: Re: PCSI problem with release notes; Message-ID: <3c0f0149.524144494f47414741@radiogaga.harz.de>c  - Didier Morandi (Didier.Morandi@gmx.ch) wrote:i > Martin Vorlaender wrote:- > >   file [HTDIG]HTDIG0301-5R5.RELEASE_NOTES	 > >     generation 03010505T > >     release notesn4 > >     source [000000]HTDIG0301-5R5.RELEASE_NOTES ; ...MM > > %PCSI-E-OPENIN, error opening AXPMV$DKA300:[USER.MV.PROJEKTE.HTDIG.3_1_5.eB > > REL5.INSTALL_VMS.][HTDIG]HTDIG0301-5R5.RELEASE_NOTES; as input ...aL > > but it looks for the release notes in the material root mangled with the > > *destination* directory. >w> > I do not see any mangling here. It looks for the material in? > AXPMV$DKA300:[USER.MV.PROJEKTE.HTDIG.3_1_5.REL5.INSTALL_VMS.]oI > as you told PCSI to go in the package command. There it expects to findtJ > the relnotes file in subdirectory [.HTDIG], which gives a full file specG > of AXPMV$DKA300:[USER.MV.PROJEKTE.HTDIG.3_1_5.REL5.INSTALL_VMS.HTDIG]F > HTDIG0301-5R5.RELEASE_NOTES   ? You didn't take into account the "source" clause. From the PCSIe Developer's Guide:  
   file "name"    ...t   source "source"g  J   Specifies the name of the file to package that supplies the contents forI   the file specified in the "name" parameter of the file statement. [...]kE   Use this option when the input file for the package operation has a-E   different relative file specification than the output file your kit     installs on the user's system.  ' So while it should get *installed* into-3 PCSI$DESTINATION:[HTDIG]HTDIG0301-5R5.RELEASE_NOTES-* the file's contents should be fetched fromC AXPMV$DKA300:[USER...INSTALL_VMS.000000]HTDIG0301-5R5.RELEASE_NOTES-5 Which works just fine for other files in the package.0  ) > Or maybe 'it' does not like the hyphen.c  . Thought about that, too. But I didn't test it.   > Try an underscore instead.  0 Hrmpf... But the product's version is 3.1-5R5...F I wouldn't have a problem with that if this weren't the release notes.  D > (I didn't know we could use concealed directories in the /material > qualifier...)d  # An extension to your PCSI_MENU? ;-)k  F A Compaq employee has passed this problem into PCSI development. SeemsC like the combination of "release notes" and "source" triggers this.e@ For now, I think I'll just leave out the "release notes" clause.  A Granted I could set up the source to look just like the installedo@ package (to avoid using the "source" clause), but this has grown= historically (<big g> when I think that I started the port inbD September). And everything else aside, it should work as documented.   cu,a   Martin -- oD                        |  Martin Vorlaender  |  VMS & WNT programmer1   OpenVMS: When you    |  work: mv@pdv-systeme.denE   KNOW where you want  |     http://www.pdv-systeme.de/users/martinv/m8   to go today.         |  home: martin@radiogaga.harz.de   ------------------------------  $ Date: Wed, 5 Dec 2001 10:44:30 -0600+ From: Christopher Smith <csmith@amdocs.com>s( Subject: Problem with the Info-VAX list?L Message-ID: <3B55D7F383B0D31197D9009027541CBF1170DF3D@cmiexch1.cmi.itds.com>  H I haven't been getting Info-VAX stuff for days.  Anyone else having this problem?   Regards,   Chriso  ! Christopher Smith, Perl Developert Amdocs - Champaign, IL   /usr/bin/perl -e '? print((~"\x95\xc4\xe3"^"Just Another Perl Hacker.")."\x08!\n");g 'o      ------------------------------  $ Date: Wed, 5 Dec 2001 17:15:10 -0500* From: WILLIAM WEBB <WWEBB1@email.usps.gov>, Subject: RE: Problem with the Info-VAX list?- Message-ID: <0033000043841036000002L062*@MHS>r  5 =0AConsidering the temperament implied by some of ther+ posts of late, consider yourself fortunate.m   WWWebb   -----Original Message-----/ From: Info-VAX-Request@Mvb.Saic.Com at INTERNETU* Sent: Wednesday, December 05, 2001 4:48 PMB To: Webb, William W Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNET( Subject: Problem with the Info-VAX list?    H I haven't been getting Info-VAX stuff for days.  Anyone else having thi= s  problem?   Regards,   Chrisy  ! Christopher Smith, Perl Developerh Amdocs - Champaign, IL   /usr/bin/perl -e '? print((~"\x95\xc4\xe3"^"Just Another Perl Hacker.")."\x08!\n");s '=   ------------------------------  # Date: Thu, 06 Dec 2001 03:24:20 GMTo1 From: "David J. Dachtera" <djesys.nospam@fsi.net> Y Subject: Re: Q: Tool or script to remove nonprinting characters from a SET HOST log? log?D' Message-ID: <3C0EE505.68849D52@fsi.net>l  * Alan Winston - SSRL Admin Cmptg Mgr wrote: > ] > In article <3C0C4D7D.FF47DEC0@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:e > >Nic Clews wrote:  > >>/ > >> Alan Winston - SSRL Admin Cmptg Mgr wrote:r > >> > > >> > Comp.os.vmsers -- > >> >O > >> > I remember vaguely reading something about a tool or script specifically"D > >> > designed to clean up the log files you get from SET HOST/LOG. > >>I > >> Umph, a quick and dirty way is to use EDT and create a listing file.e > >> > >> From command line mode: > >> > >> SET NONUMBERS > >> PRINT filespec.LIS  > > , > >Yeah - that's what I was gonna suggest... > >< > K > I think I was inadequately clear.  I wanted the nonprint characters to goc9 > away rather than be replaced with symbolic equivalents.n   Well, you might try this, then:o  C Use a terminal program on a PC. Set it to log the session to a diskaB file. Then, TYPE out your SETHOST.LOG. Stop the terminal program'sG session logging. You should now have the text of the log in a disk fileiC on your PC, sans escape sequences and control codes (other than the.G <CR><LF> pair at the end of each record). Transfer that (as ASCII) backt@ to VMS by whatever means is available (FTP, Kermit, WRQ Protocol ("VAXLINK2"), etc.).   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  # Date: Wed, 05 Dec 2001 19:06:47 GMT 3 From: sy18889@rabbit.fmr.com (Bradford J. Hamilton) ! Subject: Re: RECALL does not workt. Message-ID: <bfuP7.11$M3.70@news-srv1.fmr.com>   Hello,  / Slightly OT, but inquring minds want to know...   P Do the references below to V7.3-1 mean that it has been announced (other than a . hazy "2H2002?)?  If so, when may we expect it?   Thanks,C Brad  R In article <3C0DC494.30F6F79@compaq.com>, Guy Peleg <guy.peleg@compaq.com> writes:E >The recall command implemented in DCL. It has nothing to do with theb >terminal driver.w- >DCL algorithem is similar to the following :e >r >Check mode : >if mode eql interactive then add command to recall buffer >nK >I have no idea why we limit RECALL to interactive mode only. I will try to ( >look around for reasonable explanation,B >If I couldn't find one, expect RECALL to support batch mode soon. >xI >Just FYI, the recall command was modified in V7.3-1 to support selectivem >search of the buffer. >t3 >$ RECALL DIR - returns the last directory command.e >tI >In V7.3-1 $ RECALL/ALL DIR will return a list of all directory commands.e >i
 >Guy Peleg >OpenVMS Engineering >i >e >g   Bradford J. Hamilton  bradhamilton@mediaone.net	(home) brad.hamilton@fmr.com		(work)g  ; "All opinions that I express are my own, not my employer's"    ------------------------------   Date: 5 Dec 2001 14:31:50 -0600p- From: Kilgallen@SpamCop.net (Larry Kilgallen)b! Subject: Re: RECALL does not workh3 Message-ID: <QJjLEb8N5imP@eisner.encompasserve.org>   d In article <bfuP7.11$M3.70@news-srv1.fmr.com>, sy18889@rabbit.fmr.com (Bradford J. Hamilton) writes:  R > Do the references below to V7.3-1 mean that it has been announced (other than a 0 > hazy "2H2002?)?  If so, when may we expect it?  C I don't know whether it has been announced for any particular date,-D but I feel that I have enough experience to say that _Guy's_remarks_6 do not mean it is announced (which is what you asked).  @ Once some featurette has been checked in to the VMS master packs= (32-bit and 64-bit) for a particular release, it is much moren% trouble to remove it than to ship it.d  @ I am sure we all understood Guy's statement about the feature toA include an implicit "unless we find something horribly wrong with > the feature that erases your disks and causes your building to burn down" :-)   ------------------------------  $ Date: Wed, 5 Dec 2001 15:46:09 -05001 From: "Farrell, Michael" <MFarrell@voltdelta.com>n! Subject: RE: RECALL does not worktO Message-ID: <025766C9BBC5D511A4ED00B0D0F08C23163090@ny_exchange1.maintech1.com>O  K This is wonderful.  I've often wondered why it hadn't been there all along.f( It'll make a big difference around here.  K While we are on the subject of V7.3-1 goodies, is there any chance that theeK SEARCH command could report out record numbers larger than 6 digits.  RightrI now, we get "******" when the record found in a particular file is beyondnK the 999,999th position.  I think this was available in a prior release (6.2 K or 7.1) but  it disappeared.  I know that we complained about it to the CSCiJ and were given a patched version of SEARCH that did this properly and toldE that it would be factored into the next VMS release.  Didn't happen.!o  K Our files routinely have record counts ranging from 1 million to 12 million A and the search command is the only practical way to find specificn information in them.  K (I'll leave my other request for an enhancement in SEARCH for another time.nE Has to do with wildcard characters being allowed in a search string.)d   Thanks,f   Mike Farrell  G P.S.:	One more thing to love about VMS.  The ability to load huge filese< (100 megabytes+) into the text editor.  Don't think I could  	do that in MS Word!     > -----Original Message-----< > From:	sy18889@rabbit.fmr.com [SMTP:sy18889@rabbit.fmr.com], > Sent:	Wednesday, December 05, 2001 2:07 PM > To:	Info-VAX@Mvb.Saic.Comr# > Subject:	Re: RECALL does not workl >  > Hello, > 1 > Slightly OT, but inquring minds want to know...I > J > Do the references below to V7.3-1 mean that it has been announced (other	 > than a t0 > hazy "2H2002?)?  If so, when may we expect it? > 	 > Thanks,r > Brad > L > In article <3C0DC494.30F6F79@compaq.com>, Guy Peleg <guy.peleg@compaq.com>	 > writes:eG > >The recall command implemented in DCL. It has nothing to do with the  > >terminal driver.e/ > >DCL algorithem is similar to the following :r > > 
 > >Check mode < > >if mode eql interactive then add command to recall buffer > > J > >I have no idea why we limit RECALL to interactive mode only. I will try > to* > >look around for reasonable explanation,D > >If I couldn't find one, expect RECALL to support batch mode soon. > >tK > >Just FYI, the recall command was modified in V7.3-1 to support selective  > >search of the buffer. > >e5 > >$ RECALL DIR - returns the last directory command.- > >-K > >In V7.3-1 $ RECALL/ALL DIR will return a list of all directory commands.e > >o > >Guy Peleg > >OpenVMS Engineering > >  > >I > >  >  > Bradford J. Hamilton" > bradhamilton@mediaone.net	(home) > brad.hamilton@fmr.com		(work)h > = > "All opinions that I express are my own, not my employer's"    ------------------------------  % Date: Thu, 06 Dec 2001 07:08:53 +0100 : From: Karl Rohwedder <extern.karl.rohwedder@volkswagen.de>! Subject: Re: RECALL does not worka, Message-ID: <3C0F0B75.3030201@volkswagen.de>  ) So I can purge my own command definition:B  A   FRC == "Pipe If F$Search("SYS$SCRATCH:FRC.OUT").nes. "" Then $ eC delete/Nolog/NoConfirm SYS$SCRATCH:FRC.OUT.* ; RECALL /ALL > SYS$SCn7 RATCH:FRC.OUT ; Search SYS$SCRATCH:FRC.OUT /Match=And "c   John Santos wrote:  & > On Wed, 5 Dec 2001, Guy Peleg wrote: >  > F >>The recall command implemented in DCL. It has nothing to do with the >>terminal driver.. >>DCL algorithem is similar to the following : >> >>Check mode; >>if mode eql interactive then add command to recall bufferm >>L >>I have no idea why we limit RECALL to interactive mode only. I will try to) >>look around for reasonable explanation,tC >>If I couldn't find one, expect RECALL to support batch mode soon.y >>J >>Just FYI, the recall command was modified in V7.3-1 to support selective >>search of the buffer.e >>4 >>$ RECALL DIR - returns the last directory command. >>J >>In V7.3-1 $ RECALL/ALL DIR will return a list of all directory commands. >> > 
 > Huzzah!! >  >  >>Guy Peleg  >>OpenVMS Engineeringd >> >      -- n  - mit freundlichen Gruessen | with best regardst   Karl RohwedderB iT-Ingenieurteam     | Ellernbruch 11       | D-38112 BraunschweigA Telefon: 0531/515521 | Telefax: 0531/515531 | Mobil: 0172/5434843 E   E-Mail: rohwedder@decus.decus.de           | iT-IngTeam@t-online.dei,           karl.rohwedder@it-ingenieurteam.de DATEX-P: 4505018005::ROHWEDDER   ------------------------------  $ Date: Wed, 5 Dec 2001 21:27:08 -0600' From: "Oscar Lerma" <olerma@sc2000.net>d+ Subject: Testing software for OpenVMS/Alpha " Message-ID: <3c0ee6ac$1@ultranews>  J Does anyone know of any software to test applications on the OpenVMS/Alpha	 platform?a  
 Thank You, Oscar    ------------------------------  % Date: Thu, 06 Dec 2001 14:39:58 +0010 ' From: <paddy.o'brien@zzz.tg.nsw.gov.au>P/ Subject: Re: Testing software for OpenVMS/Alpha-5 Message-ID: <01KBJV2KBZTE001608@tgmail.tg.nsw.gov.au>-   Oscar Lerma wrote:  > >Does anyone know of any software to test applications on the  >OpenVMS/Alpha
 >platform?  @ You're probably after DTM (Digital Test Manager) for regression ! testing.  This is part of DECset.i   Regards, Paddy   ------------------------------  $ Date: Wed, 5 Dec 2001 09:09:15 -0500* From: WILLIAM WEBB <WWEBB1@email.usps.gov>) Subject: RE: tubes (was: RE: DEC is DEAD)r- Message-ID: <0033000043763311000002L012*@MHS>.   =0A****Like Milli Vanilli?   -----Original Message-----/ From: Info-VAX-Request@Mvb.Saic.Com at INTERNET * Sent: Wednesday, December 05, 2001 7:11 AMB To: Webb, William W Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNET) Subject: RE: tubes (was: RE: DEC is DEAD)i    H > Actually these days there are a lot of tube amp simulators, DSP based=  D > digital devices that sound like tube amps. Not a valve to be seen.   Kind of like Charon-VAX?  :-)A  H I believe one of the first was Tom Scholz's Rockman.  "Get rid of those=  A towers of 19-inch racks and get the same sound from a small box".C  F Somehow, I think the groupies are more impressed by the stacks of Real Tube Amplifiers.  : ****It wouldn't surprise me of some heavy-metal gods theseA days just have them for show, with the sound actually coming from. somewhere else.****=   ------------------------------  # Date: Thu, 06 Dec 2001 03:25:58 GMTA1 From: "David J. Dachtera" <djesys.nospam@fsi.net>D) Subject: Re: tubes (was: RE: DEC is DEAD)U' Message-ID: <3C0EE568.9FC22DD7@fsi.net>    Art Rice wrote:a >  > David J. Dachtera wrote: >  > > Phillip Helbig wrote:. > >>L > >> > Most (probably more than 99%) of televisions still have at least one.J > >> > As do a very high percentage of computers for the very same reason.
 > >> > :-) > >>I > >> One area where tubes are still preferred to transistors is in guitaroL > >> amplifiers.  The idea here is NOT high-fidelity; the tubes and speakersJ > >> contribute substantially to the sound.  (Specifically, the non-linearL > >> effects of the tubes when overdriven leads to a "soft" distorted sound,K > >> and the typical speakers act as low-pass filters, which is why even inlG > >> the studio the sound going into the mixing desk often comes from a0K > >> microphone placed in front of a speaker, rather then directly from the_J > >> amplifier or even the guitar itself (perhaps after a pre-amp), thoughH > >> this is not uncommon for bass guitars, where most amps and speakers+ > >> contribute less to the sound quality.)i > >.E > > There's a device called a "tube pre-amp" which is exactly that: a-G > > "valve" circuit (as the Brits would say) which imparts its inherent/H > > "warmth" to the tone of the sound. Goes after your digital tape deckH > > and mixing board or CD player and before your power amp for a really > > sweet playback!u > >e6 > > Makes for some rather tasty music, actually, IMHO. > >sJ > > I don't get to play much anymore, and when I do play, it's mostly bassK > > these days. Sometimes, I still pick up the 12-string guit-box and strum./ > > an old familar tune just to hear the music.  > >r > > *HEARTFELT SIGH* > >3 > M > Antonio had no idea what he was starting.  You can't beat DEC and you can'tx0 > beat the warm glow of a Fender Twin Reverb :>)   AMEN to THAT, bro!   -- S David J. Dachterae dba DJE SystemsT http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/N   ------------------------------  $ Date: Wed, 5 Dec 2001 14:01:59 -0500> From: "Koska, John C. (LNG-MBC)" <John.C.Koska@lexisnexis.com> Subject: RE: Veritas on VMS?M Message-ID: <3D35AD137AAAD411A6BA0008C7B1B12D016025E2@MBCALBEXC03.BENDER.COM>w  A Not sure what you mean by "rumors that VMS engineering is workingXD on ways to address the time crunch.", but certainly volume shadowing@ with new minicopy has done wonders at my site in addressing the > shrinking time windows.  Namely, our shadowsets are made whole& extremely fast now with write bitmaps.  @ If you have info on "rumors", which are not the above, I am sure' many of are interested in hearing them.t   :) jck   > -----Original Message-----8 > From: Dave Gudewicz [mailto:david.gudewicz@abbott.com], > Sent: Wednesday, December 05, 2001 1:17 PM > To: Info-VAX@Mvb.Saic.Coms > Subject: Re: Veritas on VMS? >  >  > Interesting idea.  But...e > ; > That's sounds like doing a backup of a backup to me, a 2   > rather than 1 step? > process.  And let's not forget the other way.  Restore would  
 > require the = > same journey.  Then there's the time element of creating a   > saveset on a diskI; > then taking that file and creating a .tar, which Veritas   > uses.  We are faced3@ > with ever shrinking time windows for getting our backups done. > @ > And I've heard rumors that VMS engineering is working on ways  > to address the > time crunch. >  > < > "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message( > news:3C0E579D.1A75CDFD@videotron.ca... > > Dave Gudewicz wrote:@ > > > BTW, none of the VMSers here are thrilled about NOT using 
 > VMS backup.  > > ? > > How about using BACKUP to create a saveset on a spare disk n > drive, and thenP> > > using that Veritas software to backup a single large file  > (the backup save  > > set) from that spare drive ? >  >    ------------------------------  $ Date: Wed, 5 Dec 2001 13:33:06 -06001 From: "Dave Gudewicz" <david.gudewicz@abbott.com>t Subject: Re: Veritas on VMS?8 Message-ID: <9ulssk$igq$1@fizban.fizban.pprd.abbott.com>  J The rumor I heard had "something" to do with physical backups Vs. image orK inc.  That and a few other things being looked at by VMS engineering wrt to 9 backup.  Maybe plans have changed since.  Donno for sure.r  I "Koska, John C. (LNG-MBC)" <John.C.Koska@lexisnexis.com> wrote in messageeG news:3D35AD137AAAD411A6BA0008C7B1B12D016025E2@MBCALBEXC03.BENDER.COM...iC > Not sure what you mean by "rumors that VMS engineering is working-F > on ways to address the time crunch.", but certainly volume shadowingA > with new minicopy has done wonders at my site in addressing the @ > shrinking time windows.  Namely, our shadowsets are made whole( > extremely fast now with write bitmaps. >gB > If you have info on "rumors", which are not the above, I am sure) > many of are interested in hearing them.P >  > :) jck >m > > -----Original Message-----: > > From: Dave Gudewicz [mailto:david.gudewicz@abbott.com]. > > Sent: Wednesday, December 05, 2001 1:17 PM > > To: Info-VAX@Mvb.Saic.Com2  > > Subject: Re: Veritas on VMS? > >< > >1 > > Interesting idea.  But...P > >m< > > That's sounds like doing a backup of a backup to me, a 2 > > rather than 1 step@ > > process.  And let's not forget the other way.  Restore would > > require theu> > > same journey.  Then there's the time element of creating a > > saveset on a diskI< > > then taking that file and creating a .tar, which Veritas > > uses.  We are facedIB > > with ever shrinking time windows for getting our backups done. > >?A > > And I've heard rumors that VMS engineering is working on waysh > > to address the > > time crunch. > >i > > > > > "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message* > > news:3C0E579D.1A75CDFD@videotron.ca... > > > Dave Gudewicz wrote:A > > > > BTW, none of the VMSers here are thrilled about NOT usinge > > VMS backup.M > > >m@ > > > How about using BACKUP to create a saveset on a spare disk > > drive, and thene? > > > using that Veritas software to backup a single large file3 > > (the backup save" > > > set) from that spare drive ? > >g > >t   ------------------------------  % Date: Wed, 05 Dec 2001 13:06:45 -0700y+ From: Linda Luik <p14175@email.sps.mot.com>e Subject: Re: Veritas on VMS?1 Message-ID: <3C0E7E55.2BF52950@email.sps.mot.com>s  J I got the following email from Peter Weaver last January regarding Veritas and VMS backups. Linda    ========================- From: "Peter Weaver" <peter.weaver@stelco.ca>  To: Info-VAX@Mvb.Saic.Comw  @ NetBackup works fine to backup and restore individual files. ForD example if you are using it to backup the disks where the programersB write code and a programmer purged away something they need to getC back. BUT IT WILL NOT WORK IN A DISASTER RECOVERY SITUATION. Do not @ depend on it to backup your system drive. Do not depend on it toD restore an entire disk. The manual that comes with the software even> tells you that it can not be used to recover the system drive.  E Management was sold on the idea that we did not need Operation peopleeF to go around and swap tapes on the individual machines, but in the endE we ended up still doing image backups to tape and doing the NetBackup)= to a tape robot. If a person needs one file restored then thevA NetBackup people can do it from their desk. If more than one filea> needs to be restored we go to the VMS backup savesets on tape.  ; ===========================================================m     Dave Gudewicz wrote:  9 > Anyone using Veritas for their backup solution for VMS?m >yF > We're kinda being pushed that way, but have discovered some problems > withJ > system disk backups.  Seems like some of the directories are not gettingH > backed up on a standalone test box.  This would imply similar problems > with) > a cluster, which we haven't tested yet." >0G > BTW, none of the VMSers here are thrilled about NOT using VMS backup.D >4 > -- > Dave Gudewiczu   --
 Linda Luik Motorola GIS 2200 W. Broadway Rd.	 AZ09-M555m Mesa, Arizona 85202n Phone: 480-655-4432b FAX:   480-655-3659  Pager  1-888-772-5230l linda.luik@motorola.comr   ------------------------------  % Date: Wed, 05 Dec 2001 10:38:44 -0800 % From: Dean Woodward <deanw@rdrop.com>  Subject: Re: Veritas on VMS?) Message-ID: <3C0E69B4.E0F93E04@rdrop.com>l   Jan-Erik Svderholm wrote:  >  > Hi. = > Maybe not related, but is't quite common to run Rdb backupso? > with RMU/BACKUP first to disk, then backup the RBF files withe9 > VMS BACKUP to tape using the regular backup routine(s).m) > Also a "backup of a backup" in a way...l  E Exactly, and it's proven damn handy a time or two to have the currentn RBF on a disk.   ------------------------------  % Date: Wed, 05 Dec 2001 21:12:03 -0500@- From: JF Mezei <jfmezei.spamnot@videotron.ca>s Subject: Re: Veritas on VMS?, Message-ID: <3C0ED3D1.1893A62C@videotron.ca>  M Anyone who has worked on a system other than VMS and had problems restoring acB backup will want to insist on using VMS BACKUP to do your backups.  N What would be nice though is if VMS BACKUP had various options/filters for itsH save sets. For instance, a filter to create/read ZIP files or tar files.  L So you could backup your system disk with VMS BACKUP, but the save-set wouldJ be a .ZIP file that could be used from other systems with file header infoK discarded. Or you could use BACKUP to read .TAR files and restore them to ae VMS system.t  L Heck, VMS BACKUP could have an "open" input/output filter architecture whereK you could write your own input/output filters. The filter could be handed 2lF files for every VMS file. The first one contains the VMS specific file> structure information and the second contains the actual data.   ------------------------------  # Date: Thu, 06 Dec 2001 04:05:27 GMTr1 From: "David J. Dachtera" <djesys.nospam@fsi.net>o Subject: Re: Veritas on VMS?& Message-ID: <3C0EEEA8.7918AA8@fsi.net>   Dave Gudewicz wrote: > 9 > Anyone using Veritas for their backup solution for VMS?R > K > We're kinda being pushed that way, but have discovered some problems withaJ > system disk backups.  Seems like some of the directories are not gettingM > backed up on a standalone test box.  This would imply similar problems witha) > a cluster, which we haven't tested yet.r > G > BTW, none of the VMSers here are thrilled about NOT using VMS backup.   F I'm sure the management will be less than thrilled with the results of your next DR test, also.  F Oh, yeah - if Abbott carries disaster insurance, make sure the carrierF doesn't get wind of this, or the premiums may suddenly increase by the% square of the square of the exponent.i   -- E David J. Dachterap dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/c   ------------------------------  % Date: Wed, 05 Dec 2001 22:02:58 +0100 ( From: Paul Sture <paul.sture@bluewin.ch>O Subject: Re: VMS Effort (Was: Microsoft Pyramid Collapses Enron and Hewlett Pac - Message-ID: <VA.000004d7.cb2232ce@bluewin.ch>r  J In article <VgqP7.519$Xx2.89415@nwrddc02.gnilink.net>, Jeff Killeen wrote:7 > "Bill Todd" <billtodd@metrocast.net> wrote in messageh> > news:B%cP7.245757$dk.16561028@bin1.nnrp.aus1.giganews.com... > >lC > > "Hoff Hoffman" <hoffman@xdelta.zko.dec.nospam> wrote in messagey/ > > news:ombP7.125$BK1.2931@news.cpqcorp.net...] > >e > > ...s > >DJ > > >   There are specific individuals that are (often repeatedly) posting > veryN > > >   specific and strongly-held opinions to this and to other newsgroup andM > > >   to various websites, and I foresee little or nothing that anyone here  > at/ > > >   Compaq can do to change these opinions.t > > N > > Then your perception is remarkably limited.  Or perhaps you meant that youM > > foresee little or nothing that anyone at Compaq *will* do to change those 4 > > opinions, which is an entirely different matter. > N > I believe 99% of the IT professionals, who are actual OVMS customer for whomL > this more that a debating exercise, will make rational business decisions.N > If Compaq delivers a viable port to the ISV's very early in 2003 Compaq willC > find those rational business decisions break in Compaq's favor...u > K I'm beginning to anticipate timescale problems, depending on the lead time 3 introduced by the ISVs.   O According to slide 5 of Gaitan D'Antoni's Porting presentation, EV79 EOL is at dN the end of 2005. We end customers will have to wait for the ISVs (Oracle etc) J to do _their_ port first before we can even start on our own effort. It's  starting to look tight.r ___i
 Paul Sture Switzerland    ------------------------------   Date: 5 Dec 2001 16:05:48 -0600a- From: Kilgallen@SpamCop.net (Larry Kilgallen)-O Subject: Re: VMS Effort (Was: Microsoft Pyramid Collapses Enron and Hewlett Pac<3 Message-ID: <uVYVp5Pj0Gfe@eisner.encompasserve.org>C  X In article <VA.000004d7.cb2232ce@bluewin.ch>, Paul Sture <paul.sture@bluewin.ch> writes:  Q > According to slide 5 of Gaitan D'Antoni's Porting presentation, EV79 EOL is at $P > the end of 2005. We end customers will have to wait for the ISVs (Oracle etc) L > to do _their_ port first before we can even start on our own effort. It's  > starting to look tight.    But Oracle gets a head start.a   ------------------------------   Date: 5 Dec 2001 14:38:13 -0600e+ From: young_r@encompasserve.org (Rob Young)iS Subject: Re: VMS Effort (Was: Re: Microsoft Pyramid Collapses Enron and Hewlett Pac 3 Message-ID: <Wq4zbR+FTaZX@eisner.encompasserve.org>v  Y In article <1011205105407.39582A-100000@Ives.egh.com>, John Santos <JOHN@egh.com> writes:o    3 >> 	Mark Gorham's letter made it to that same site:t > ! > Which site is "the same site"? n  N Click on the referenced URLs.  I don't see anything messed up yet.  The threadF only has a handful of contributors.  If you can't sort it out, you may0 wish to visit www.deja.com and look at it there.   				Rob    ------------------------------  $ Date: Wed, 5 Dec 2001 19:34:23 -0500  From: John Santos <JOHN@egh.com>S Subject: Re: VMS Effort (Was: Re: Microsoft Pyramid Collapses Enron and Hewlett Pace6 Message-ID: <1011205193032.16710A-100000@Ives.egh.com>   On 5 Dec 2001, Rob Young wrote:@  [ > In article <1011205105407.39582A-100000@Ives.egh.com>, John Santos <JOHN@egh.com> writes:e >  > 5 > >> 	Mark Gorham's letter made it to that same site:- > > # > > Which site is "the same site"?   > P > Click on the referenced URLs.  I don't see anything messed up yet.  The threadH > only has a handful of contributors.  If you can't sort it out, you may2 > wish to visit www.deja.com and look at it there. > 	 > 				Rob   ? I was confused by Hoff's post, which followed up your change of1C reference to the inquirer articles, but which I took to be refering:D to the billparish/netslaves web sites that Fred dismissed (two postsA earlier in the thread.)  So I'm still not sure if Hoff was sayingfC it wasn't worth replying to the Inquirer articles or to the earlier?
 site or both.a   -- e John Santosl Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Wed, 05 Dec 2001 21:43:34 +0000 4 From: Andrew Swallow <andrew.swallow@baesystems.com>2 Subject: Re: VMS marketing (advertising) sighting.. Message-ID: <3C0E9506.C6A7BF9D@baesystems.com>   Dave Gudewicz wrote: >  [snip] > H > I didn't think that marketing included **creating** the right product.4 > Creating a market, yes.  Creating the product.  ?? > 6 Marketing includes writing a letter to the head of R&D1 telling him that the customers want the following 7 features...   This may involve producing say 3 models -u a.  every feature imaginable1 b.  a cheap (and reliable) one that only supports "     the facilities used every day.5 c.  the pretty one that comes in 20 different covers.o  1 Marketing will also have to play customer to R&D. 1 Housewives like to treat machines as black boxes,o5 anything inside is the repairman's job.   Translatings- their needs into something technical that R&Du can understand is difficult. -- t7 _______________________________________________________l Andrew Swallow   ------------------------------   Date: 5 Dec 2001 16:32:20 -08000# From: jgessling@yahoo.com (jojimbo)d2 Subject: Re: VMS marketing (advertising) sighting.= Message-ID: <dc2d8031.0112051632.6046a4b2@posting.google.com>f  O "Dave Gudewicz" <david.gudewicz@abbott.com> wrote in message news:<9ulcmb$ga7> gH > I didn't think that marketing included **creating** the right product.4 > Creating a market, yes.  Creating the product.  ?? > 7 > Otherwise, thanks for the good news.  Its refreshing.  > 	 > Dave...r  G Well, I'm way out on a limb here, but in the interest of getting people $ to think deeper about marketing...    C Suppose there are people in a hot country getting thirsty.  That's cE a market (demand).  Perhaps it's satisfied by a guy with a bucket anddL a dipper.  Now along comes someone that organizes an effort to get bottles, G label them with a catchy name and put water in them.  That's marketing,mD (and merchandising to put them on the shelf in the right place) and D if they put a sign up "cool water here" that's the promotion part of
 marketing.  G So yes, I think the main result of marketing is creating a product. By sB product, I don't just mean the pyhsical thing siting on the shelf,D I also include getting into "design wins" in the case of a chip.  Or9 getting in shelf space, in the case of a new potato chip.a  B Think about this.  If you are a system architect are you going to B specify alpha or itanium?  The decision has a lot more to do with ? than what's better (technically).  It's more similar to what's  $ available on the shelf in the store.   Later,     Jim.   ------------------------------  # Date: Thu, 06 Dec 2001 03:41:49 GMTs1 From: "David J. Dachtera" <djesys.nospam@fsi.net>p2 Subject: Re: VMS marketing (advertising) sighting.' Message-ID: <3C0EE91D.D0954C26@fsi.net>    jojimbo wrote: > [snip]B > Back to my first sentence, marketing includes creating the right > product,     Well, Digital got that right.T   > pricing it correctly,   5 Neither Digital nor Compaq ever could relate to that.t  % > positioning it in the marketplace, a  E Why not let the market position the product? Who knows the customers'-) needs better: the customer or the vendor?4   > as > well as whatB > I've heard many of you call marketing which is exactly promotion > (advertising).  H Well, sort of. "Promotion" and "marketing" are sort of like "parenthood"E and "having children". Dysfunction results from trying to divorce the A two. Marketing, by definition, includes not only advertising, butg@ passing out trinkets at trade shows, sending out things like theB "OpenVMS umbrellas", buying ad space/time in the media, mentioningF products in presentations, ... the list goes on, but I think the point is covered.h  4 > Please try to understand the difference, what mostD > of you are asking for is promotion, perhaps a valid complaint, but8 > please be aware that this is only a part of marketing.  ? Understood. Understand also that advertising is a key marketing:G strategy, the one most frequently - and usually most easily - employed,dF as well as being one of the most effective, if not THE most effective.H Hence, the push for advertising as a primary marketing strategy: how canC you expect the market to be interested in a product it doesn't know  about?   -- d David J. Dachtera  dba DJE Systemsa http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/o   ------------------------------  # Date: Thu, 06 Dec 2001 04:01:20 GMTk1 From: "David J. Dachtera" <djesys.nospam@fsi.net> 2 Subject: Re: VMS marketing (advertising) sighting.' Message-ID: <3C0EEDB1.6923B222@fsi.net>d   jojimbo wrote: > [snip]I > Well, I'm way out on a limb here, but in the interest of getting people $ > to think deeper about marketing... > [snip]  H If you can't get anyone to understand the foundation, do the intricacies$ of the superstructure really matter?  E If the train never leaves the station, does it matter how fast it can C go? How long it is? What it is carrying? Whether or not it's on/off6	 schedule?D  B Let's get past the main road block first: "putting the word in theE streets". There'll be plenty of time to bicker about nits and detailso later.  D Oh, and here's a nit about your example that you forgot to consider:  H Suppose the guy with the bucket and dipper also happens to have a ratherE deep and exceptionally pure well, and offers his dippers of water forGH $10,000 each. Suppose the guys selling the bottled water get it from theG municipal supply of a town some distance away, but offer theirs for theo usual $1 each.  ? Who's gonna sell more water? (Hint: Pricing is a key element ofr
 "marketing".)t   -- o David J. Dachterai dba DJE Systemsa http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/a   ------------------------------   Date: 5 Dec 2001 13:06:20 -0600 - From: Kilgallen@SpamCop.net (Larry Kilgallen) = Subject: Re: Was the Alpha to Intel agreement ever approved ? 3 Message-ID: <WNNF$13xli4B@eisner.encompasserve.org>d  V In article <3C0E65D0.4090409@compaq.com>, John Reagan <john.reagan@compaq.com> writes:  J > Not speaking for Compaq nor I have seen the contract or anything, but I H > wasn't aware that Compaq needed regulator approval for the Intel deal. > K > They didn't need it when selling Rdb to Oracle, selling various products g! > to CA, etc.  Is this different?   B When they announced the Intel deal, Compaq said that is was not an: exclusive license so there would be no regulatory problem.   ------------------------------  $ Date: Wed, 5 Dec 2001 12:48:21 -08001 From: "Farrell, Michael" <MFarrell@voltdelta.com>nU Subject: RE: Why does file access preformance over WAN depend greatly on the command?dO Message-ID: <025766C9BBC5D511A4ED00B0D0F08C23163091@ny_exchange1.maintech1.com>.  H Why is this message coming back incessantly?  I have received over 30 of these in the last two days.  n (About 10 of them today alone.)e   Mike Farrell   > -----Original Message-----( > From:	Ian Parker [SMTP:parker@gol.com]) > Sent:	Sunday, December 02, 2001 5:39 PM  > To:	Info-VAX@Mvb.Saic.ComwG > Subject:	Re: Why does file access preformance over WAN depend greatly  > on the command?  > > > In article <3c026f81.29477957@news.demon.co.uk>, Jim Johnson6 > <Jim.Johnson@software-exploration.nospam.com> writes > >Alan, > >cC > >You've just encountered the SQO optimization for DAP/FAL in RMS.c > > H > >SQO (Sequential-only) is a bit set in the FAB on open when the callerG > >knows that it will never randomly read a record.  TYPE knows that isaH > >the case, and '@' knows it isn't (e.g. handling of GOSUBs and GOTOs). > >iG > >For the network access, SQO is used to initiate a bulk data transferiI > >protocol that doesn't send back any ACK messages until the eof is hit.gB > >This means that the server can pack in as many data messages as9 > >possible, and just keep sending them until it is done.p > > G > >If SQO isn't used, then a much slower protocol that is essentially at, > >request-response pair per record is used. > > ' > >That's the difference you're seeing.v > >  > >Fwiw, > >Jim.  > >e > > A > >On 26 Nov 2001 08:17:17 -0800, SPAMSINK2001@YAHOO.COM (Alan E.  > >Feldman) wrote: > >s
 > >>Hello, > >>D > >>I have trouble running a command file across the ocean. The file= > >>resides on node 1. When I run it on node 1, it runs fine.v > >>? > >>Node 2 is on the same LAN as node 1. When I run the commandeC > >>@NODE1::TO.COM on node 2, it takes longer, but not much longer.= > >>J > >>Node 3 is across an ocean. When I log in to node 3 and run the commandF > >>@NODE1::TO.COM, it takes much, much longer. But when I TYPE it, it > >>only takes a little longer.t > >>I > >>This is summarized in the following table (times given in hh:mm:ss.ccf > >>format): > >> > >>-- > >>> > >>Command            | Node 1 | Node 2 |  Node 3 | Node 4  |> > >>                   | (same) | (LAN)  |  (WAN)  | (LAN)   |> > >>                   |        |        |IP tunnel|IP tunnel|> > >>-------------------+--------------------------------------: > >>TYPE NODE1::TO.COM |   0.22 |   0.54 |    4.61 |  1.64: > >>   @ NODE1::TO.COM |   0.24 |   2.34 | 2:37.97 |  5.42 > >>  > >>(Node 3 is across an ocean.) > >>H > >>Now, my question is, if it takes only 4.61 seconds to type it acrossJ > >>the WAN, why does it take more than 157 seconds to execute it when the3 > >>normal execution time is only about .2 seconds?w > >>I > >>The node 4 column represents two nodes that are in the same room, butoG > >>on two different IP subnets, connected via a DECnet over IP tunnel. J > >>Apparently, the problem is associated with going overseas via the WAN. > >>I > >>We are using default DECnet settings. The WAN consists of a shared T1 C > >>line and a timeplex at each end. We are using TCPWARE V5.3 witheE > >>DECnet-over-IP tunnels. We are running VMS 6.1 and 6.2 and DECneteF > >>phase IV, but version mismatches between any two test nodes do notA > >>affect this problem. Can anyone explain this and/or make somea$ > >>recommendations for improvement? > >>G > >>We have a similar problem with one of our applications. It needs toeI > >>open and read an overseas file and it takes much longer than the timerG > >>it would take to simply copy it over with the DCL COPY command. (ItdI > >>needs to open for synchronization purposes.) Some of this may have to H > >>do with the application updating other machines, but it appears thatJ > >>most of it is due to a problem similar to, or the same as, the problem/ > >>that makes @ take much longer over the WAN.n > >>% > >>Thanks in advance for your help. l > >> > >>Disclaimer: JMHO > >>Alan E. Feldman  > >>afeldman&gfigroup.como > >r > >Jim Johnson > >Software Exploration, Ltd.i, > >(remove '.nospam' from the reply address) > J > And the solution that I use, if circumstances permit, is for the commandI > procedure to copy itself down to the local node and then run this localuG > copy.  It can make a big performance difference for a small amount ofe > extra coding.  > 	 > Regardsi >  > -- f > Ian Parker   ------------------------------  # Date: Wed, 05 Dec 2001 21:54:37 GMTi1 From: "Mark D. Jilson" <jilly@clarityconnect.com>.Y Subject: Re: Why does file access preformance over WAN depend greatly on the command? com 2 Message-ID: <3C0E975C.68E4CFB7@clarityconnect.com>  A Since I didn't get them I'd be asking your news server admin thisa	 question.    "Farrell, Michael" wrote:  > J > Why is this message coming back incessantly?  I have received over 30 of > these in the last two days.-! > (About 10 of them today alone.)c >  > Mike Farrell >  > > -----Original Message-----* > > From: Ian Parker [SMTP:parker@gol.com]+ > > Sent: Sunday, December 02, 2001 5:39 PM  <snipped the rest>   --  D Jilly	- Working from Home in the Chemung River Valley - Lockwood, NY0 	- jilly@clarityconnect.com			- Brett Bodine fan. 	- Mark.Jilson@Compaq.com			- since 1975 or so, 	- http://www.jilly.baka.com               -   ------------------------------   Date: 5 Dec 2001 14:55:27 -0600p- From: koehler@encompasserve.org (Bob Koehler)08 Subject: Re: XML Test Problems & Java Performance Query.3 Message-ID: <eq9oAGf+Pw9R@eisner.encompasserve.org>l  i In article <3C0E31F4.153B07B6@gtech.com>, Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com> writes:o > Bob Koehler wrote:H >>    The JRE wants MyClass to be in MyClass.java, and the Java compilerD >>    wants MyClass to be in MyClass.java.  On case insensitive fileE >>    systems you can have MyClass in MYCLASS.JAVA and myclass.class, D >>    but you still have to have MyClass.java passed to the compiler >>    and MyClass to the JRE.b >  > ???? > G > JRE wants MyClass to be in MyClass.class - JRE is runtime environmentc > noth4 > compiler and it does not need MyClass.java at all.  H OK, in painfull detail:  when you invoke the JRE via the java command itH wants the byte code for MyClass to be in MyClass.class.  When you invokeD the compiler via the javac command it sometimes wants the source for* MyClass to be in MyClass.java (see below).  > > You can pass anything to javac (compiler), but java (runtime > interpreter)' > requires it to look as MyClass.class.S > 
 > Example: >  > $ javac foobar.javat > $ javac FOOBAR.JAVAd  H   I'm not sure that worked with all versions of the JDK for VMS, and I'mA   fairly sure it doesn't work for all versions of the JDK for all I   case-insensitive file systems.  I have seen it work (I know it works onLI   my old JDK on WNT), but I'm fairly sure I've also seen it fail.  Never (F   recorded the details, but I have reason to associate the failure it F   with VMS because I had to work around MMS file name upper casing to     get Java code built under MMS.   ------------------------------  $ Date: Wed, 5 Dec 2001 19:08:23 -0500  From: John Santos <JOHN@egh.com>8 Subject: Re: XML Test Problems & Java Performance Query.6 Message-ID: <1011205184357.16710B-100000@Ives.egh.com>  ! On 5 Dec 2001, Bob Koehler wrote:b  k > In article <3C0E31F4.153B07B6@gtech.com>, Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com> writes:r > > Bob Koehler wrote:J > >>    The JRE wants MyClass to be in MyClass.java, and the Java compiler3                                                ^^^^eG Bob, I think Arne is refering to the probable typo in the above line...c  F > >>    wants MyClass to be in MyClass.java.  On case insensitive fileG > >>    systems you can have MyClass in MYCLASS.JAVA and myclass.class,:F > >>    but you still have to have MyClass.java passed to the compiler > >>    and MyClass to the JRE.i > >  > > ???? > > I > > JRE wants MyClass to be in MyClass.class - JRE is runtime environmentO > > notm6 > > compiler and it does not need MyClass.java at all. > J > OK, in painfull detail:  when you invoke the JRE via the java command itJ > wants the byte code for MyClass to be in MyClass.class.  When you invokeF > the compiler via the javac command it sometimes wants the source for, > MyClass to be in MyClass.java (see below). > @ > > You can pass anything to javac (compiler), but java (runtime > > interpreter)) > > requires it to look as MyClass.class.s > >  > > Example: > >  > > $ javac foobar.java  > > $ javac FOOBAR.JAVA5 > J >   I'm not sure that worked with all versions of the JDK for VMS, and I'mC >   fairly sure it doesn't work for all versions of the JDK for alleK >   case-insensitive file systems.  I have seen it work (I know it works oneK >   my old JDK on WNT), but I'm fairly sure I've also seen it fail.  Never  H >   recorded the details, but I have reason to associate the failure it H >   with VMS because I had to work around MMS file name upper casing to " >   get Java code built under MMS.  H I've only used java a little bit on VMS (mostly on Solaris, and not thatA much there, so I'm not an expert), but I think Solaris absolutelyf? requires the case to match for both javac and java, but on VMS, G javac is *not* case-sensitive about file names.  (ODS-2, I've not triedyG it on ODS-5.)  IIRC from my limited experience, java is case-sensitive,?F and so you have to play games with quoting, and fake Unix-style paths.  E This is what I think is going on: the file name of the class file is  C constructed from the class name on the command line, so the command.I line has to match what is in the class file, but VMS (at least for ODS-2)9E is case-blind when it opens the file.  For example, java sees MyClass E in the command line, with its case carefully preserved by quoting, so C it asks VMS to open "MyClass.class", and F11BXQP, being case-blind,nE opens "MYCLASS.CLASS".  Java reads the file, looking for "MyClass" aseE a class name in it (case-sensitive search), finds it, so it loads andmE executes the java byte code for it.  If you don't specify the correcttE case for the class name, java will open the right file, but the case-o5 sensitive search for the class in the file will fail.l   -- a John Santoss Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------   End of INFO-VAX 2001.677 ************************