1 INFO-VAX	Mon, 31 Mar 2003	Volume 2003 : Issue 177       Contents: Re: 7.3-1 and EXBYTLIM; Re: Apache %SYSTEM-F-NOPRIV on RIGHTSLIST.DAT after upgrade  Re: COV Sponsors Re: COV Sponsors Re: COV Sponsors Re: COV Sponsors Re: COV Sponsors# Earth to Andrew, do you get it now? ' Re: Earth to Andrew, do you get it now? ' Re: Earth to Andrew, do you get it now?  Re: Fortran Guru requested Gotta have those ol' time docs" Re: Gotta have those ol' time docs How to get timing  OPCOM debugging  Re: OPCOM debugging  Re: OPCOM debugging  Re: OPCOM debugging P RE: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retai) Removing files marked for delete problem. - Re: Removing files marked for delete problem. - Re: Removing files marked for delete problem. - Re: Removing files marked for delete problem. * Re: Structure of  the TCPIP Services group* Re: Structure of  the TCPIP Services group* Re: Structure of  the TCPIP Services group* Re: Structure of  the TCPIP Services groupD Re: SYSTARTUP_VMS (was: Three questions re: the VMS startup process)D Re: SYSTARTUP_VMS (was: Three questions re: the VMS startup process)D Re: SYSTARTUP_VMS (was: Three questions re: the VMS startup process)D Re: SYSTARTUP_VMS (was: Three questions re: the VMS startup process)& Re: TCPIP Services and IP masquerading& Re: TCPIP Services and IP masquerading& Re: TCPIP Services and IP masquerading  Re: VAX XDM, help needed please!7 Re: VMS Software Product & Online Documentation Library 7 RE: VMS Software Product & Online Documentation Library 7 Re: VMS Software Product & Online Documentation Library 7 Re: VMS Software Product & Online Documentation Library  RE: VMS Upgrade Questions  Re: VMS Upgrade Questions 9 Re: Which PAK for DECNET Phase IV under Hobbyist Program? 1 Re: [MOZILLA V1.3] Compatibility with JAVA V1.4 ?   F ----------------------------------------------------------------------  % Date: Sun, 30 Mar 2003 20:34:50 -0500   From: John Santos <JOHN@egh.com> Subject: Re: 7.3-1 and EXBYTLIM 4 Message-ID: <1030330203045.416D-100000@Ives.egh.com>  & On Sun, 30 Mar 2003, David Webb wrote:  Y > In article <1030329002915.416F-100000@Ives.egh.com>, John Santos <JOHN@egh.com> writes: + > >On Fri, 28 Mar 2003, Paul Webster wrote:  > > J > >> I have an application (but no source code) that works on VMS 7.1 (and) > >> I think 7.2) but is failing on 7.3-1  > >>  @ > >> The failure is %SYSTEM-F-EXBYTLM, exceeded byte count quota > >>   > H > >Completely different thing that has worked fine for years, but is now* > >dying on VMS 7.3-1 with the same error. > > B > >TCPWare IP-over-DECnet line to a remote site.  The local end isG > >an AlphaServer 1200 running VMS V7.3-1 (upgraded from V7.2-1 Tuesday C > >night), DECnet-Plus, TCPware 5.6-2.  Remote end is VAX VMS V7.1, # > >DECnet-Plus V7.1, TCPware 5.3-3.  > >  > O > I thought TCPware's Decnet over IP  only worked with DECNET Phase IV not with  > DECNet-Plus / PHASE V.K > I tried to load it on an Alpha recently with PHASE V on it and got a load @ > of errors about not being able to create the line and circuit.  > You may be correct, but this is irrelevent.  I'm talking about# IP-over-DECnet, not DECnet-over-IP.   @ We are now using Ph V everywhere, and use its IP transport layer9 (TPCONS) to do DECnet to sites with only IP connectivity.    >  > David Webb > VMS and Unix team leader > CCSS > Middlesex University   --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Mon, 31 Mar 2003 07:36:18 +0200  From: Dirk Munk <munk@home.nl>D Subject: Re: Apache %SYSTEM-F-NOPRIV on RIGHTSLIST.DAT after upgrade1 Message-ID: <b68k5c$cb$1@news4.tilbu1.nb.home.nl>   O Did you run the config procedure after the upgrade? After each upgrade of CSWS  T you MUST run the config procedure, otherwise you get all kind of Privilege problems.   ------------------------------    Date: 30 Mar 2003 16:01:42 -0800( From: qwqwqwqw70@hotmail.com (Jon Power) Subject: Re: COV Sponsors = Message-ID: <71367ac8.0303301601.44f3f14a@posting.google.com>   D Well David, My name is Jon Power. The reason for the way too obviousF jab at the free vms group was the fact that we were banned for writingF apparantly what the people at the 'old free vms mailing' didn not likeC (so much for free speech). Im very heartened to hear that it is now E that it is 'old' As for the other posting regarding Sector7 appearing F in the google spashes, IBM had nothing to do with this. IBM do not ownE any part of Sector7. We happen to be allied to them because IBM truly B understand the scope and complexity of migration services and theyD have a business model that can ease the transition for OpenVMS usersD to IBM hardware. IBM really do have their act together, dont confuseB the IBM of 20 years ago with the IBM of today. If anyone out thereE still thinks of IBVM in terms of batch processing mainframes, you owe C it to yourselves to take an new look. As for VMS migrations we have B completed more that any other organization in the world. From realD time C&C to banking, stock exchanges, manufacturing, communications.F We have migrated systems 'natively' and using our subsystem layer thatE makes UNIX & NT appear to the applications as 'OpenVMS'. We have also @ migrated many OpenVMS applications to LINUX and will be happy toE answer any technical questions that anyone has free of charge. We now > have methodologies for extracting the business logic from OVMSF applications are recreating the logic in new development environments.C So, if anyone has a questions on the following topics, feel free to E email me and I'll do my best to ensure that we answer your questions. D Important questions anyone consider migrating from OpenVMS to UNIX /A NT should ask: How do I replace the complex '$' API on the target ? system ? What do I do with VMS extensions to the base languages A (COBOL, PL/I, BASIC, FORTRAN,C,C++ etc etc) What do I do with the E Database (RMS, DBMS, Rdb (RDML & SQL) What are the steps in a complex B VMS migration ? How do I control the divergence of the source baseC (development continues while the migration project is in progress - D how to limit the re-do) ? How do I manager the transition from Async< IO (QIO) to a thread model How do I create a project plan toE accurately estimate the time and cost of the migration ? What are the @ specific metrics that get captured by the plan How do I create aE migration test plan (different from a development test plan) What are D the 'gotcha's. To sum up, the technical process and programming partD of the OVMS migration project is really only a part of ther over allC project. The effort put int the technical transition plan should be B about 25% the remaing 75% is the surrounding processes required toD make a migration project sucessfull. What we mean by 'sucessfull' isE accurately costed, timed, planned and managed to that plan to hit all @ targets. When we finish a project plan for an OpenVMS to UNIX/NTE project we have a project plan with between 3000 and 10000 individual F line items. Once you have the plan, you resource it and then level theF plan for the available resources and skill levels. Add the hourly costE ofthe resources and you have the time and cost of the project. So, if D anyone woudl like to contact me with specific questions on migratingD OpenVMS applications to UNIX(incl LINUX) or NT feel free to do so atF jon.power(at)us.sector7.com We do have a comercial business to run, so> dont expect too rapid response but we will try to get round to? answering your questions as quickly as we can. If we get a vast ? response to this posting then we will create a web site for the D questions and answers. So, from 'gib.senip' -- which considering theC amount of spam I get on this subject - I was way before my time and D could have made millions - I urge you to contact us (s7) if you have< any problems or questions. Remember that sucessfully OpenVMSB migrations are all about complex project plans, line items (tasks)= which have some basis in reality ( we have been doing OpenVMS B migration since 1988 so we have a few hundred projects to draw ourF metrics from). If anyone is interested our DCL was put on the originalE OSF/1 CDROM and Digital used to have a stand a DECUS (when it used to D be fun --- remember those days) where you got a t-shirt if you couldD guess if it was Sector7's DCL or Digitals DCL. Also, we helped writeC the book 'how to migrate VMS application to Digital UNIX' - you can E still down load it from the HP web site. Digital wrote a flyer to all F of the OpenVMS users saying that one of the 3 reasons to buy OSF/1 wasA Sector7's tool set. We also trained 40 DEC engineers in India and F created a migration factory for VMS to DEC OSF/1. If there are any DEC= people (and I differentiate between DEC, Compaq, and HP) with F migration skills anywhere in the world I would like to hear from them.F We are being inundated with OpenVMS to UNIX migrations from every partB of the globe. Thanks for taking the time to read this posting. Jon Power    ------------------------------  % Date: Sun, 30 Mar 2003 22:06:05 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>  Subject: Re: COV Sponsors ' Message-ID: <3E87BEAD.50AE3F7F@fsi.net>    Simon Clubley wrote: > ] > In article <3E863FA1.C8D3348B@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:  > > 8 > > Yes. Sector7 ads appear frequently on Google Groups. > > J > > Jon whats-'is-name at one time came around to the old Free-VMS mailingE > > list, even going so far as forging his mail headers and using the I > > pseudonym "gib.senip@reknaw.com". Class guy, huh? (Apologies, Didier. K > > Sorry you had to hear it from me.) I can probably unearth a post or two ( > > and excerpt it, if anyone doubts me. > >  > " > What kind of things did he say ?  = See Jon's response to my post, and draw your own conclusions.    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Sun, 30 Mar 2003 22:09:08 -0600 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>  Subject: Re: COV Sponsors & Message-ID: <3E87BF64.920CAF6@fsi.net>   Jon Power wrote: > 3 > Well David, My name is Jon Power. [Mucho deletia]   ! Thanx for the reminder (I think).   H Sue Skonetski - meet Jon Power, apparently #1 guy at Sector/7. Jon, meetH Sue Skonetski: all-around sweetheart and head of the OpenVMS Ambassadors program.  A In case you hadn't noticed, Jon, while you are in the business of G serving refugees of VMS's rapidly dwindling customer base, the majority E of this group seeks to sustain VMS, unlike (apparently) Sue's bosses.   G So, while you may gain some referrals from surfacing at this time, much F of the response will likely be similar to the group's general response to Andrew Harrison of Sun fame.   D Just wanted to prepare you for what may come. Otherwise, my comments stand as posted.  : FYI: There is a new FreeVMS project around. Search Google.  D From a business perspective, you might want to consider looking intoG Charon-VAX, a commercial, supported VAX emulator for Windows, Linux and H OpenVMS-Alpha. At least one Charon-VAX distributor frequents this group.H Charon-VAX allows OpenVMS-VAX and OpenVMS-VAX applications (binaries) toF run on foreign platforms with no recompiling. I hear an Alpha emulator= may be in the works for IA64, but that's purely rumor, AFAIK.    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Mon, 31 Mar 2003 00:26:12 -0500 * From: "Stanley F. Quayle" <stan@stanq.com> Subject: Re: COV Sponsors , Message-ID: <3E878B24.8861.7016A8@localhost>  1 On 30 Mar 2003 at 22:09, David J. Dachtera wrote: G > >From a business perspective, you might want to consider looking into I > Charon-VAX, a commercial, supported VAX emulator for Windows, Linux and J > OpenVMS-Alpha. At least one Charon-VAX distributor frequents this group.J > Charon-VAX allows OpenVMS-VAX and OpenVMS-VAX applications (binaries) toH > run on foreign platforms with no recompiling. I hear an Alpha emulator? > may be in the works for IA64, but that's purely rumor, AFAIK.   ? Thanks for mentioning me (sort of)...  Yes, I'm the CHARON-VAX  ? distributor who lurks here.  I'd be glad to discuss your needs.   D As far as CHARON-Alpha, there doesn't seem to be a business need -- A you can still get brand-new Alphas.  When they're unavailable ... 
 --Stan Quayle  Quayle Consulting Inc.  
 ----------C Stanley F. Quayle, P.E. N8SQ  +1 614-868-1363  Fax: +1 614 868-1671 1 8572 North Spring Ct. NW, Pickerington, OH  43147 = Preferred address:  stan@stanq.com       http://www.stanq.com    ------------------------------  % Date: Sun, 30 Mar 2003 23:55:47 -0500 ( From: David Froble <davef@tsoft-inc.com> Subject: Re: COV Sponsors , Message-ID: <3E87CA53.4030008@tsoft-inc.com>   Jon Power wrote:  # > Well David, My name is Jon Power.      <snip>  9 > When we finish a project plan for an OpenVMS to UNIX/NT G > project we have a project plan with between 3000 and 10000 individual 
 > line items.     P Gee, sounds complicated, and expensive.  In addition, many here would call it a P 'downward' migration.  Other than DEC's stupidity in the past, why would anyone # want to engage in such an activity?    > ( we have been doing OpenVMSD > migration since 1988 so we have a few hundred projects to draw our > metrics from).    I Same question, what reasons did people give for this expensive, complex,   downward migration?     H > We are being inundated with OpenVMS to UNIX migrations from every partD > of the globe. Thanks for taking the time to read this posting. Jon > Power   $ Same people that are buying bridges?    K Dave, who is doing a VMS to windoz project, and knows for certain that the  P customer has no idea of what they are doing, but is willing to spend big bucks, R and is NOT willing to listen.  Give you best opinion, then take the money and run!     --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com6 T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486   ------------------------------    Date: 30 Mar 2003 18:07:54 -0800( From: bob@instantwhip.com (Bob Ceculski), Subject: Earth to Andrew, do you get it now?= Message-ID: <d7791aa1.0303301807.7412168f@posting.google.com>    Andrew writes:   >>> 9 > There is no OpenVMS magic bullet get used to it, I have 6 > provided you with enough examples over the years and7 > to be frank the argument is getting pretty pointless.  <<<   K One of the good things in OpenVMS are the Descriptors. With this OpenVMS is J protected against "buffer overflow" Also there was a better security modelH (e.g. xhost + versus security tripple transport, user, host). But in theK newer times there are a lot of tools and programs without the philosophy of D OpenVMS. The most one came from UNIX. So a lot of buggy tools become standard in OpenVMS.K The other good thing is, that memory must be allocated to the user. If not, J you will get ACCVIO. Under UNIX, user have the right to read every part ofJ memory without any limit. Also user can get rights (roots), which he never5 should get. Normaly this is unpossible under OpenVMS.    Best regards Rudolf Wingert    ------------------------------  # Date: Mon, 31 Mar 2003 02:49:59 GMT ( From: "Mark E. Levy" <mlevy70@attbi.com>0 Subject: Re: Earth to Andrew, do you get it now?- Message-ID: <r1Oha.268108$L1.76571@sccrnsc02>   5 "Bob Ceculski" <bob@instantwhip.com> wrote in message 7 news:d7791aa1.0303301807.7412168f@posting.google.com...   	 [deletia]   7 > should get. Normaly this is unpossible under OpenVMS.   
 "unpossible"?    ------------------------------  % Date: Sun, 30 Mar 2003 23:22:19 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>0 Subject: Re: Earth to Andrew, do you get it now?/ Message-ID: <3E87B442.5A6B0773@vl.videotron.ca>    Bob Ceculski wrote: M > One of the good things in OpenVMS are the Descriptors. With this OpenVMS is % > protected against "buffer overflow"     M No you are not. You can build a descriptor that tells the system you have 15k J byte buffer when in fact the allocated space is only 15 bytes. And you get buffer overflow.  N And most of the unix-like io routines that write to a buffer have as argument,9 a maximum length and provide an actual length many times.   K One has to be careful with strcpy. But strcpy is not OS related. And a good * programmed may use strncpy or even memcpy.   ------------------------------  % Date: Sun, 30 Mar 2003 20:03:42 -0500   From: John Santos <JOHN@egh.com># Subject: Re: Fortran Guru requested 4 Message-ID: <1030330193941.416C-100000@Ives.egh.com>  ) On Sun, 30 Mar 2003, Paddy O'Brien wrote:    >  >  > John Santos wrote:% > > On Sat, 29 Mar 2003 PRSTSC wrote:  > >  > > T > >>I encounter some trouble porting a fortran/vax program to Alpha (no warnings at / > >>compile time, either with or without /F77):  > >> > >>$ run spcinv	 > >>../..  > >>T > >>%SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=000000000000* > >>0030, PC=FFFFFFFF808E5380, PS=0000001B3 > >>%TRACE-F-TRACEBACK, symbolic stack dump follows O > >>   image    module    routine             line      rel PC           abs PC T > >>                                             0 0000000000000000 FFFFFFFF808E5380T > >>  SPCINV  INVADERS  SLEEP                  141 0000000000000ADC 0000000000030ADCT > >>  SPCINV  INVADERS  SPECIAL_GRAPHIC        579 0000000000002BB8 0000000000032BB8T > >>  SPCINV  INVADERS  MOVE                    42 00000000000003D0 00000000000303D0T > >>  SPCINV  INVADERS  INVADERS                27 0000000000000000 0000000000000000T > >>                                             0 FFFFFFFF802513D4 FFFFFFFF802513D4 > >>% > >>132         SUBROUTINE SLEEP(TAG) P > >>133 C       Set timer and wait for TAG seconds.  TAG is a CHARACTER*5 stringS > >>134 C       of the form ##.##, which is the seconds, and hundredths of seconds. B > >>135 C       To delay game for 5 seconds, TAG would be '05.00'.( > >>136         IMPLICIT INTEGER*4 (A-Z)# > >>137         EXTERNAL SS$_NORMAL % > >>138         DOUBLE PRECISION TIME , > >>139         CHARACTER DELTA_TIME*9,TAG*5 > >>140 E > >>141         DELTA_TIME='0 ::'//TAG          !Concatenate strings. N > >>142         RET_STAT=SYS$BINTIM(DELTA_TIME,TIME)    !ASCII to binary time.D > >>143         IF(RET_STAT.NE.%LOC(SS$_NORMAL))STOP'BINTIM failed.'Q > >>144         RET_STAT=SYS$SETIMR(,TIME,,)            !Set timer, event flag 0. D > >>145         IF(RET_STAT.NE.%LOC(SS$_NORMAL))STOP'SETIMR failed.'S > >>146         RET_STAT=SYS$WAITFR(%VAL(0))            !Wait for flag 0 to be set. D > >>147         IF(RET_STAT.NE.%LOC(SS$_NORMAL))STOP'WAITFR failed.' > >>148         RETURN > >>149         END  > >> > >>Thanks,  > >> > >>D. > >  > > G > > I don't think I'm a Fortran guru, not having programmed in it in 28 H > > years, but I think the problem is that the subroutine is lying aboutH > > TIME being a double-precision float.  It is really a 64-bit integer. > > H > > On a VAX all bit patterns are legal floats (or at least the hardwareD > > doesn't complain about them when you MOVD them), but not true onC > > Alphas.  It "fixes" them.  Since TIME is really an int, any bit C > > pattern is legal.  Some of the other followups imply there is a B > > floating point exception occuring somewhere in the background. > >  > John,  > I > It is a plausible explanation, but I'm not sure whether it works quite   > like that. > J > Time is 64 bits long, but no "use" is made of it in the routine.  It is A > passed to sys$bintim where it is written to and then passed to  B > sys$setimr where it is read.  There is no use which changes its E > representattion as a bit pattern.  No matter how it is declared in  G > SLEEP, as long as it is 64 bits, the interaction of the two routines   > should work correctly.  ? I know there is nothing explicitly done to TIME in the routine, B except to use it as a place to store the 64-bit value.  What I wasD concerned about is that the compiler might be doing something behindA the program's back to manipulate it (e.g. move it to a register). = My reason for suspecting that was that someone in a follow up = mentioned an FP exception, and TIME is the only float.  Maybe  /NOOPT would cure that?   H > There should be no moving of the data: it is written to TIME and then  > the same bit pattern is read.   = But on Alpha, the 1st few args are passed in registers, maybe D the compiler is doing a LDD (with implicit or explicit Double->GiantC conversion) when calling $SETIMR.  (The arg is a pointer to a Quad, > but maybe it decides to pass a copy of the quad or something.)  B Another thought; the prototype for SETIMR (and BINTIM) say the argC is a pointer to Quad int, but the storage is declared as an Double; B could it be doing an implicit type conversion from a temp variable> into and out of TIME before/after calling the system services?  F > I think what you are referring to with the MOVD instruction is that I > G-floating and D-floating need a shift to add one exponent bit between  J > the two.  D- is default on VAX, but G- is default on Alpha.  If that is H > what you are referring to, it should not be a problem here, as per my  > comment in my last paragraph.   E Also Alpha's can (tend to? perfer to? I don't do much FP programming)  use IEEE floating as well.   > E > > I don't know how or if you can use 64-bit (Quad) ints in Fortran, A > > but it seems very likely to me.  If not, maybe you can create A > > a 2-element 32-bit (Long) int array in a common, and point to 0 > > the 1st element instead of using the DOUBLE? > > J > Yes, on Alpha (at least, I don't think it is available on VAX) you have J > INTEGER*8, but for the same use of TIME, you can have INTEGER*4 TIME(2) K > -- *4 being the default.  No need to be in COMMON, a local variable will  
 > suffice.  = What tangled webs we weave when first we practice to deceive!   B Sounds to me like INTEGER*8 is the right way to do it on an Alpha,9 (and IA64?) but INTEGER*4 TIME(2) may be portable to VAX.    > > P.S. Great project!  > > C > > But I found that since losing hardwired terminals, (LAT instead B > > of DZ's), the response of interactive games became too erraticB > > to make playing pleasureable.  I think a DECterm would be evenB > > worse.  In space invaders, I would fire a couple of test shotsC > > to guage the relative speed of the target and the missile, then C > > fire again when the target got to the right point (I would hide G > > behind a shield, pop out, fire and pop back.)  On LAT, the response F > > was always a little jerky, so I would start missing a lot.  In theH > > early rounds, I could move and fire again, but in later rounds, when3 > > things were moving faster, this was impossible.  > > E > > I could be mis-remembering my strategy- I don't think I've played A > > since our 11/780 went away, ten years ago almost to the day -  > > 13-MAR-1993.  :-(  > > J > I tried compiling, linking and running.  The first two were successful, J > but the program did not like that I am on a DECterm.  It claimed access J > to only a VT100 or VT52.  Since it was only a spurious exercise, I have " > not bothered to change the code. >  > Regards, Paddy >  >  > I > ***********************************************************************  > E > "This electronic message and any attachments may contain privileged @ > and confidential information intended only for the use of the D > addressees named above.  If you are not the intended recipient of E > this email, please delete the message and any attachment and advise D > the sender.  You are hereby notified that any use, dissemination, 9 > distribution, reproduction of this email is prohibited.  > C > If you have received the email in error, please notify TransGrid  C > immediately.  Any views expressed in this email are those of the  ? > individual sender except where the sender expressly and with  E > authority states them to be the views of TransGrid.  TransGrid uses @ > virus scanning software but excludes any liability for viruses > contained in any attachment. > > > Please note the email address for TransGrid personnel is now& > firstname.lastname@transgrid.com.au" > I > ***********************************************************************  >  >  >    --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Sun, 30 Mar 2003 17:54:38 -0500 * From: Jerry Leichter <leichter@smarts.com>' Subject: Gotta have those ol' time docs & Message-ID: <3E8775AE.79DA@smarts.com>  G [Boy, it's interesting to look at comp.os.vms again.  It's been years.  E Nice to see many of the same names - and, in a way, nicer to see some 
 new ones!]  D Are you, or do you know, a packrat who keeps around old VMS-related E documentation?  I need stuff specifically related to DECnet, and even F more specifically to DECnet management.  I'd especially love to get myG hands on the old POLYCENTER docs.  The older the better; anything after B about 1995 is of no interest.  CD or hardcopy is fine.  Obviously,E on-line is easiest - but not necessary.  I am in a position to pay to % either buy or just borrow this stuff.   H I know I used to lambaste people for asking for personal replies but ...F I very much doubt there's broad interest in this issue - and it's niceE to see that comp.os.vms still gets quite a bit of traffic anyway.  So 3 ... if you could drop me a line, I'd appreciate it.    Thanks!  							-- Jerry    ------------------------------  % Date: Sun, 30 Mar 2003 20:31:15 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>+ Subject: Re: Gotta have those ol' time docs / Message-ID: <3E878C36.1D69F69A@vl.videotron.ca>    Jerry Leichter wrote: E > Are you, or do you know, a packrat who keeps around old VMS-related G > documentation?  I need stuff specifically related to DECnet, and even * > more specifically to DECnet management.   G > on-line is easiest - but not necessary.  I am in a position to pay to ' > either buy or just borrow this stuff.   : For $1000, I'll tell you you can get the documentation at:( http://h71000.www7.hp.com/doc/index.html  B Payment can be made in chocolate bars or in real money :-) :-) ;-)   ------------------------------    Date: 30 Mar 2003 20:03:36 -08002 From: texascheryl@hotmail.com (Cheryl Hoefelmeyer) Subject: How to get timing< Message-ID: <7112dd15.0303302003.e3abe18@posting.google.com>  F In a previous post, to demonstrate the speed of a spawn, W.F.J Mueller wrote:   >In one session execute. > < > $                         write sys$output "1 - "+f$time()< > $ spawn/nolog/nosym/nokey write sys$output "2 - "+f$time()< > $                         write sys$output "3 - "+f$time()< > $ spawn                   write sys$output "4 - "+f$time() > @ >In the other execute the certainly CPU bound program, e.g. this Fortran code >  >         program allcpu > 100     continue >         goto 100
 >         end  > F >Now measure the time of a SPAWN with and without a competing process. On a; >DEC 3000 - M300LX and with a heavy symbol and logical name  environment A >(about 140 process logicals and 700 symbols) I got the following  timing:  > ' >   Node idle CPU bound process running  >  > SPAWN  0.720 sec  146 sec !!! # > SPAWN/NO.. 0.270 sec   19 sec !!!  >   C What command(s) is/are used to get the timing? I'd like to try this  experiment myself.   Thanks,  Cheryl   ------------------------------  % Date: Sun, 30 Mar 2003 17:18:46 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca> Subject: OPCOM debugging/ Message-ID: <3E875F25.F8DB3778@vl.videotron.ca>   I For a while now, OPCOM on my all mighty microvax has decided it no longer @ wants to broadcats its message to OPA0: (or any other terminal).  H REPLY/STATUS yields nothing, just goes back to Dollar prompt. (it shouldN normally display the OPCOM classes that have been enabled on that terminal, as" well as any outstanding requests).  K Perhaps this bug occured when the second node in the cluster crashed due to  TCPIP some days ago.  O OPCOM is still writing to OPERATOR.LOG messages from both nodes in the cluster.   G OK, i know I can kill OPCOM and then use the STARTUP.COM to restart it.   K However, before doing this, is there anything I could do to try to find out E what had actually happened ? Are there any debgging tools for OPCOM ?    (This is VAX VMS 7.2)   K I find that hearing the bell from OPA0: in the next room  helps me a lot to N debug problems, and to find out if someone has succesfully connected , or whenL someone tries to constently access my system. And the lack of those bells is driving me nuts :-) :-)    ------------------------------  # Date: Mon, 31 Mar 2003 00:09:23 GMT 6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) Subject: Re: OPCOM debugging3 Message-ID: <TGLha.62784$Xa3.586362@news.chello.at>   b In article <3E875F25.F8DB3778@vl.videotron.ca>, JF Mezei <jfmezei.spamnot@vl.videotron.ca> writes:J >For a while now, OPCOM on my all mighty microvax has decided it no longerA >wants to broadcats its message to OPA0: (or any other terminal).  > I >REPLY/STATUS yields nothing, just goes back to Dollar prompt. (it should O >normally display the OPCOM classes that have been enabled on that terminal, as # >well as any outstanding requests).  > L >Perhaps this bug occured when the second node in the cluster crashed due to >TCPIP some days ago.  > P >OPCOM is still writing to OPERATOR.LOG messages from both nodes in the cluster. > H >OK, i know I can kill OPCOM and then use the STARTUP.COM to restart it. > L >However, before doing this, is there anything I could do to try to find outF >what had actually happened ? Are there any debgging tools for OPCOM ?  / How about $ SEARCH SYS$SYSTEM:OPCOM.EXE "DEBUG" 7 which shows "OPC$DEBUG_FILE:" and "OPC$DEBUG_TERMINAL:"  Maybe it helps...    --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  # Date: Mon, 31 Mar 2003 02:52:51 GMT ( From: "Mark E. Levy" <mlevy70@attbi.com> Subject: Re: OPCOM debugging- Message-ID: <74Oha.268208$F1.47715@sccrnsc04>   = "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in message ) news:3E875F25.F8DB3778@vl.videotron.ca... K > For a while now, OPCOM on my all mighty microvax has decided it no longer B > wants to broadcats its message to OPA0: (or any other terminal).  - Are your terminals being set to /NOBROADCAST?    ML   ------------------------------  % Date: Mon, 31 Mar 2003 00:23:18 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca> Subject: Re: OPCOM debugging. Message-ID: <3E87C289.89A1146@vl.videotron.ca>   "Mark E. Levy" wrote: / > Are your terminals being set to /NOBROADCAST?     F Tail between the legs, foot deeply into mouth, blushing out of extreme embarassement !!!    Thanks. That was it.  ' Somehow, OPA0: was set to /NOBROADCAST.   I Suggestion to the VMS engineers: When one does a REPLY/STATUS, perhaps it / could warn if the terminal is set /NOBROADCAST.   L I have to assume that perhaps I was in some editor or other itility that was< abruptly cancelled and a temporary setting became permanent.   ------------------------------  % Date: Sun, 30 Mar 2003 19:39:23 -0500 ' From: "Main, Kerry" <Kerry.Main@hp.com> Y Subject: RE: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retai T Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF402660EC1@kaoexc01.americas.cpqcorp.net>   Andrew, Andrew ..   ? <<< But Oracle applications performance wasn't one of them. >>>   H So, are you saying that these GS Series Customers with Oracle Financials- never considered performance as a criteria???   ? Come on, you can do better fud than that .. Lets not get silly.   , While a tad dated now, here are some quotes:; http://h18002.www1.hp.com/alphaserver/gs/quotes/oracle.html    Regards     
 Kerry Main Senior Consultant  Hewlett-Packard (Canada) Co.! Consulting & Integration Services  Voice: 613-592-4660  Fax   : 613-591-4477 Email: kerryDOTmain@hpDOTcom-     (remove the DOT's and replace with "."'s)  OpenVMS DCL - the original .COM    -----Original Message-----' From: Andrew Harrison SUNUK Consultancy71 [mailto:Andrew_No.Harrison_No@nospamn.sun.com]=20c Sent: March 28, 2003 11:24 AMS To: Info-VAX@Mvb.Saic.ComOC Subject: Re: OpenVMS.org: Marvel article and HP's press release for H Marveland Alpha Retain Trust Alpha Retain Trust Alpha Retain Trust AlphaE Retain Trust Alpha Retain Trust Alpha Retain Trust Alpha Retain Trusta% Alpha Retain Trust Alpha Retain Trustt         jlsue wrote:J > On Thu, 27 Mar 2003 10:30:31 +0000, Andrew Harrison SUNUK Consultancy=200 > <Andrew_No.Harrison_No@nospamn.sun.com> wrote: >=20 >=20 >> >>Fred Kleinsorge wrote: >>J >>>Can we just wrap this all up in a shorthand format, or do we have to=20 >>>go through the ritual.r >>>oE >>>Sun:   Benchmarks don't mean anything, especially the ones we look  bad on.fH >>>Guy:  Well, how about actual performance stories from real customers. >> >>Sun: Well where are they thencF >>Guy: Ahh well thats a bit tricky how about Bank of Austria instead ? >>Sun: Sorry bzzt. >> >=20 >=20I > Yeah.  You're sorry because you can't contradict that story with any=20M
 > real facts.u >=20I > The article stands on it's own as proof that there are customers who=20 ; > use the GS160 with Oracle Financials - your proverbial=20iI > transaction-oriented workloads - even though you claim that it can't=20e
 > be true.  3 Sure there are but Bank of Austria isn't one, theirq3 reference is for Wall Street not Oracle Financials.o   Did you actually read it ???  7 And I am sure that there are Oracle Financials users ont3 GS servers but they didn't select the boxes becausee of their performance did they.  + Could have been one of a number of reasons.    Perfered supplier. Trashed margins. Right colour box.c  . But Oracle applications performance wasn't one of them.   regardsC Andrew Harrison    ------------------------------  # Date: Sun, 30 Mar 2003 21:12:11 GMT-! From: rob.buxton@wcc.spam.govt.nz02 Subject: Removing files marked for delete problem.& Message-ID: <3e875bf8.2059943832@news>   Hi All,3  F On a cluster running 7.2 (Vax) and 7.2-2 (Alpha) I've got a shadow set mounted on all nodes.0  F DFU and anal /disk /norepair show a couple of files marked for delete.  B If I run anal/disk/repair it still doesn't clean up the files, theE first run through showed 3 files markde for delete. It did clean up 1k? of the files. I've run the anal/disk/repair on all nodes in theE cluster. each report runs through like:B Analyze/Disk_Structure/Repair for _DSA1300: started on 31-MAR-2003 08:55:20.82w  . %ANALDISK-I-OPENQUOTA, error opening QUOTA.SYS" -SYSTEM-W-NOSUCHFILE, no such file: %ANALDISK-W-DELHEADER, file (653,68,1) AR_CUSTOMER.DAT;854         marked for deleterA %ANALDISK-W-DELHEADER, file (709,9,1) AR_CUSTOMER_ADDRESS.DAT;851e         marked for deleteC  , Any reasons why it cannot clean up the file?F Any methods of cleaning up these files other than by anal/disk/repair?   TIA as always,   Rob.   ------------------------------  % Date: Sun, 30 Mar 2003 16:47:24 -0400l0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>6 Subject: Re: Removing files marked for delete problem./ Message-ID: <3E8757CD.CC0842FA@vl.videotron.ca>o  " rob.buxton@wcc.spam.govt.nz wrote:< > %ANALDISK-W-DELHEADER, file (653,68,1) AR_CUSTOMER.DAT;854 >         marked for deleteIC > %ANALDISK-W-DELHEADER, file (709,9,1) AR_CUSTOMER_ADDRESS.DAT;851C >         marked for delete  > . > Any reasons why it cannot clean up the file?  
 $MC SYSMAN SYSMAN>SET ENV/CLUSTERH SYSMAN>DO PIPE SHOW DEV/FILES <disk name> | search sys$input AR_CUSTOMER  I This will show if any process is still using those versions of the files.u  N SHOW DEV/FILES on one node will not show the files opened by another node. ButM when you try to Delete, it will see that the file is locked by someone on thec& cluster and you get the above message.  F If you have one or more processes in your cluster that access/update aN different version of a database than the processes on the other nodes, you mayM have an even bigger problem in your hands and in fact, you may no longer want I to delete those files if you come to the conclusion that you will have totV merge any transaction written to the old files after the new version had been created.   ------------------------------  % Date: Sun, 30 Mar 2003 23:37:58 +0200.& From: Andreas Fassl <afassl@progis.de>6 Subject: Re: Removing files marked for delete problem./ Message-ID: <b67o05$1sl$02$1@news.t-online.com>d   Hi,g   have you tried $ set volume/rebuild ? What results do you get from a $ sho dev dsa3100:/windows ?    Regardst   Andreasd uxton@wcc.spam.govt.nz wrote:e   >Hi All, >eG >On a cluster running 7.2 (Vax) and 7.2-2 (Alpha) I've got a shadow seth >mounted on all nodes. > G >DFU and anal /disk /norepair show a couple of files marked for delete.  >aC >If I run anal/disk/repair it still doesn't clean up the files, thesF >first run through showed 3 files markde for delete. It did clean up 1@ >of the files. I've run the anal/disk/repair on all nodes in the	 >cluster.l >each report runs through like:,C >Analyze/Disk_Structure/Repair for _DSA1300: started on 31-MAR-2003s >08:55:20.82 >n/ >%ANALDISK-I-OPENQUOTA, error opening QUOTA.SYS # >-SYSTEM-W-NOSUCHFILE, no such fileS; >%ANALDISK-W-DELHEADER, file (653,68,1) AR_CUSTOMER.DAT;854  >        marked for deleteB >%ANALDISK-W-DELHEADER, file (709,9,1) AR_CUSTOMER_ADDRESS.DAT;851 >        marked for delete > - >Any reasons why it cannot clean up the file?rG >Any methods of cleaning up these files other than by anal/disk/repair?  >s >TIA as always,o >g >Rob., >e >  e >    ------------------------------  # Date: Mon, 31 Mar 2003 00:47:27 GMT ! From: rob.buxton@wcc.spam.govt.nz 6 Subject: Re: Removing files marked for delete problem.& Message-ID: <3e878fda.2073226992@news>  , On Sun, 30 Mar 2003 16:47:24 -0400, JF Mezei( <jfmezei.spamnot@vl.videotron.ca> wrote:  # >rob.buxton@wcc.spam.govt.nz wrote:w= >> %ANALDISK-W-DELHEADER, file (653,68,1) AR_CUSTOMER.DAT;854e >>         marked for deleteD >> %ANALDISK-W-DELHEADER, file (709,9,1) AR_CUSTOMER_ADDRESS.DAT;851 >>         marked for delete >> t/ >> Any reasons why it cannot clean up the file?  >I >$MC SYSMANt >SYSMAN>SET ENV/CLUSTER(I >SYSMAN>DO PIPE SHOW DEV/FILES <disk name> | search sys$input AR_CUSTOMERt >hJ >This will show if any process is still using those versions of the files. >tO >SHOW DEV/FILES on one node will not show the files opened by another node. ButmN >when you try to Delete, it will see that the file is locked by someone on the' >cluster and you get the above message.j > G >If you have one or more processes in your cluster that access/update aeO >different version of a database than the processes on the other nodes, you maypN >have an even bigger problem in your hands and in fact, you may no longer wantJ >to delete those files if you come to the conclusion that you will have toW >merge any transaction written to the old files after the new version had been created.   F Thanks JF, seems as though there was a process that hadn't been closedB down by the usual mechanisms and it did indeed still hold locks on+ thiose files. Now removed it and it's fine.h   ------------------------------  # Date: Sun, 30 Mar 2003 21:05:40 GMT.; From: "John Gemignani, Jr." <jon-nope@thiswontworkossc.net>i3 Subject: Re: Structure of  the TCPIP Services groupI= Message-ID: <E_Iha.23802$TW2.3790660@news1.news.adelphia.net>e  = "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in messagee) news:3E87260A.77ECE888@vl.videotron.ca...w > "John Gemignani, Jr." wrote:L > > I absolutely enjoy being able to work while tanning down by the pool ... in > > the summer, of course. >w >.E > How is the group that maintains/develop the  TCPIP Services product  structured ????w >wB > Seems that one person is located on a beach on the Gold Coast of
 Australia,G > while the other is located next to his pool in a place where there isS winter.o >aL > What is the history behind the group ? Was it formed/merged from differentI > groups that were in different locations ? In how many locations is that  groupdH > split/distributed ? Do specific locations have specific tasks (coding,H > testing, documentation) or do specific locations have specific product, > responsabilities (kernel, SMTP, FTP etc ?) >aI > I know that the former ALL-IN-1 group has donated and maintains some ofD thelL > mail applications such as the IMAP server. And they're in Reading England.H > (they are offloading actual ALL-IN-1 support/maintenance to Bangalore, India).r > G > Here is my concern: the structuring of the whole product, notably thew locationJ > of files, could do with a major cleanup. I don't like having hundreds ofG > thousands of files in SYS$MANAGER for instance. Is there some centrali locationG > that orchestrates all of the product, or is there a certain degree ofr > independance ? >aI > For instance, the DHCP GUI for TCPIP 5.3 has a broken help facility: itv hasaL > all the help topics, but each one tells you to go lookup www.join.com, butL > that site no longer exists.  What is the future of that application if the2 > company that wrote it has gone out of business ?  L You're asking for a LOT of information, and I'll try to answer some of them.  F Since we are a DEC/Compaq/HP networking product, we're more integratedG into the system than, let's say, TCPware.  Third party products tend to  setuptD all of their parts to be in directories under their own control with logicals pointinge to them.  J I agree with your cleanup point.  The product has grown incredibly since I joinedJ the startup (UCX V1.0) group in 1986, when we were one supervisor and fiveK engineers in the Ultrix group.  We are now a much larger group (although wex haveI lost a number of people to downsizing) and are part of the OpenVMS group.e One'L of the side effects is the level of support for older releases that we often get I driticized for.  Since we move with the operating system we adapt with ito andgL it becomes increasingly difficult to qualify with more and more versions, so we have to draw a line somewhere.  J Also, understand that there is a long span of time from when we start work on aL release to the time that it actually comes out -- usually 1-2 years.  During thatL time frame things outside of the group often change (i.e. www.join.com couldJ be one of them).  We are about to go to field test for V5.4 for submission sometime in the fall.a  J Regarding the DHCP future, I will ask the engineer who works on it to find> this topic and provide a reply, as I don't know the specifics.  I Concering the group: we have several formal projects, such as the kernel,sA NFS, security, management, mail, performance, maintenance, QA anda documentation.F Other pieces are distributed among the group members either staticallyF (same responsibilities release to release) or dynamically (as it needs attention).   C For myself, I have been involved with the group since 1986.  I haveeG worked on many, many pieces in the product.  I typically work from homeaH in the mornings and head into the office in the early afternoon and stay intoL the (sometimes late, late) evening/night.  During the summer I am by my poolC until about 11 AM with the laptop, then head into the office in thew
 afternoon.  B We have group members in Colorado, Connecticut, Massachusetts, New
 Hampshire, England, France and Australia.  E I have been working mainly on features and performance in the network  kernel, NFScH server, INETDRIVER (Alpha/IPF), TNDRIVER and TCPIP$SDA for this release.@ I have worked on other pieces in the product but only as needed.  < If you have other specific questions, feel free to email me.   ------------------------------  % Date: Sun, 30 Mar 2003 20:26:52 -0400s0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>3 Subject: Re: Structure of  the TCPIP Services group . Message-ID: <3E878B30.95B787C@vl.videotron.ca>   "John Gemignani, Jr." wrote:N > the (sometimes late, late) evening/night.  During the summer I am by my poolE > until about 11 AM with the laptop, then head into the office in thea > afternoon.  L If I were you, I'd fear for my life. Don't your coworkers have a high degree- of jealousy at your lifestyle ? (Just joking)   F Is there a history for the various groups in each of the locations youM mentioned ? For instance, how come you have surfer types on the gold coast ofe: australia involved in TCPIP ? How come they got involved ?   ------------------------------  # Date: Mon, 31 Mar 2003 03:47:52 GMTn; From: "John Gemignani, Jr." <jon-nope@thiswontworkossc.net> 3 Subject: Re: Structure of  the TCPIP Services groupl= Message-ID: <ITOha.23931$TW2.3907909@news1.news.adelphia.net>h  = "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in messagen( news:3E878B30.95B787C@vl.videotron.ca... > "John Gemignani, Jr." wrote:K > > the (sometimes late, late) evening/night.  During the summer I am by my. poolG > > until about 11 AM with the laptop, then head into the office in the  > > afternoon. > G > If I were you, I'd fear for my life. Don't your coworkers have a highp degree/ > of jealousy at your lifestyle ? (Just joking)   1 (I know that you're joking, and I'm not offended) H Well, most of them are gone when I'm in there working late, so I suppose we'rehK even.  Also consider that I'm working here at home, always doing something,e onH hardware that I bought and built myself from ebay.  Believe me, I.am not
 surroundedJ by women fanning me and feeding me grapes.  When you look at the amount ofJ work that I get done, you'd be surprised that only one person is doing it. There isL a lot of stuff that has gone into the product that people don't even realize because it's2 kernel or exec code and you don't use it directly.   >hH > Is there a history for the various groups in each of the locations youL > mentioned ? For instance, how come you have surfer types on the gold coast of< > australia involved in TCPIP ? How come they got involved ?  K The person in Australia worked in the group on an international relocation.i HeL returned home quite a while ago and continues to work on performance and newF features back in his home country.  He has been doing some intense and valuableL performance work.  The upcoming failSAFE work is also his.  (Matt-- stand up and take a bow.)  H We don't have any "groups" of people in any one location except the main grouprI in Littleton, MA.  The rest of us are working from either an HP office int
 their areaF or from home.  There is a weekly meeting and most of us conference in.
 Aside fromI that we do a lot of work via email and tunnel in.  I myself carry Storagey Works L drives with builds back and forth.  Since I updated the build procedures for V5.0,tF I also put together procedures to create build environments on CDs and
 removeableH drives so that we can take them with us.  I did not do the latest set of
 changes toI support the IPF builds (Lee -- stand and take a bow), but we are actuallye buildingH images for IPF right now.  We're waiting on the OS to see if our efforts	 have beennH successful.  I will be involved with the debugging and providing help to
 other membersiJ of the group as we start to get into it.  Needless to say, I won't have an IPF at home,D so I may not be able to spend too much time by the pool this summer.   ------------------------------  + Date: Mon, 31 Mar 2003 04:39:28 +0000 (UTC)n) From: Dan Foster <dsf@globalcrossing.net>d3 Subject: Re: Structure of  the TCPIP Services groupe2 Message-ID: <slrnb8fhj9.qg.dsf@gaia.roc2.gblx.net>  & If I might interject a comment here...  J My team is not too unlike how the UCX/TCPIP team is structured, as members are all over the country.t  H Much of us works from home, also, but we find it to be no different from> working in an office except for fewer face-to-face interactionK opportunities. We still work regular hours, and work in rooms dedicated for J offices without any interruptions -- for instance, one of the requirements@ in the paperwork that we have to sign is that we cannot have anyG unsupervised children at home while we are working - they have to be at 4 daycare, school, watched by the spouse, or whatever.  G This also permits us to work much more than the standard 8 hours/day ifaG needed (and we usually donate a lot of time/effort on our own, anyway).y  D We still do go in to the office if there's one nearby, for meetings,? special presentations, meeting with vendors, etc. Works out ok.w  K Interestingly enough, this is sometimes misinterpreted by people in *other*tK teams as being somehow luxurious or allowing us to work differently than ifaJ we were at an office. Not so, but _they_ don't always understand that! So,H yes, we do see some misunderstood jealousy in some other groups. I thinkH that if they were in our shoes for a week or a month, they would quickly sing a very different tune!.  F If anything, it's actually _easier_ to keep 8x5 hours in an office. :)E Because *THAT* usually means no obligation to respond at other times. , (At least, for here, that's often the case.)  G With practice, time, and experience, working from home is no different,/K with a strict discipline. For instance, I very rarely schedule appointmentssI during work day, and that is equally as true whether or not I work in theFF office on any given day. If I do, I follow normal procedure by gettingG approval from manager, notifying the team about unavailability, etc. No ! different no matter where I work.   F Doesn't really matter where I work because both the office at home andF office at the data center/centre are equally as nice and appropriatelyG equipped so I can do anything. Just that I can respond faster when I am E at home especially during off-hours as it saves me an one hour drive.0   -Dan   ------------------------------  % Date: Sun, 30 Mar 2003 21:40:51 -060001 From: "David J. Dachtera" <djesys.nospam@fsi.net> M Subject: Re: SYSTARTUP_VMS (was: Three questions re: the VMS startup process)3' Message-ID: <3E87B8C3.770832E3@fsi.net>    brandon@dalsemi.com wrote: > ? > In regards to the problems of LAT not starting up properly...r > K > I have made our startup routines modular and submit them to a batch queuel* > during the normal VMS startup procedure. > M > In SYLOGICALS I mount needed disk drives and execute all command files that N > have a file naming convention of *$LOGICALS.COM;  This allows me to maintainM > related logicals (i.e., LAT, TCPIP, applicaiton, etc.) in one command file.nN > So rather than modify a SYLOGICALS.COM procedure that has become 3,000 linesL > long, I can easily update a small command file.  It is much easier to workN > with 10 or 20 lines rather than the combined 3,000 lines of code.  I use the3 > f$search to locate and execute the command files.P > N > In SYSTARTUP_VMS I delete & create a startup batch queue (job limit = 1) andL > submit in order each startup process (mount disks, DECnet, TCPIP, PW, LAT,P > applications, etc) with path directed /log qualifier.  This allows me to loginQ > to our server during the boot process (obviously after SYLOGICALS has executed)nP > and subsequently monitor the startup process.  I also allow provissions in theN > startup process to allow DECnet only to start (what I call a service level). > K > So what does it have to do with LAT?  This method allows me to review the Q > output of the LAT startup in the log file generated during the startup process.f > ) > It also allows for a faster boot time. i  ? *** WARNING *** WARNING *** WARNING *** WARNING *** WARNING ***i  F As you have undoubtedly noticed, shortly after SYSTARTUP_VMS finishes,) you see a message like so on the console:l  H %SET-I-INTSET, login interactive limit = 64, current interactive value = 2   E A little known secret is that until this message appears, user loginsdH via the usual procedures (Job Controller) are disabled COMPELETELY. ThisG DOES NOT mean that logins are set zero, for indeed they are set at bootsF time to the CURRENT value of IJOBLIM. (Go into SYSGEN, USE CURRENT andG SHOW IJOBLIM. Be careful to distinguish between CURRENT and ACTIVE.) ItcC means that the Job Controller will not create interactive processeseC until at least one SET LOGINS/INTER=value command is executed afters	 boot-up.    @ One known exception to this is Multinet, where TELNET inbound isH possible anytime after the Multinet Master server completes its startup,C so long as inbound Telnet is enabled in the Multinet configuration.e   What does this mean to you?a  E Simply, it means that if SYSTARTUP_VMS finishes before your stream ofeG batch jobs, users with sufficient privilege by default can log into the A system before all of the disks are mounted, services are started,h( applications/databases are started, etc.  D What if you don't start TCP/IP until very late in the sequence? SomeH services/applications/databases may fail to start properly unless TCP/IP is up and running.  F PROCEED WITH __V_E_R_Y__ __S_E_V_E_R_E_L_Y__ __E_X_T_R_E_M_E__ CAUTION !!!!!!   --   David J. Dachterad dba DJE Systemsb http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho//   ------------------------------  % Date: Sun, 30 Mar 2003 22:05:26 -0600i From: brandon@dalsemi.comtM Subject: Re: SYSTARTUP_VMS (was: Three questions re: the VMS startup process) 1 Message-ID: <03033022052629@dscis6-0.dalsemi.com>l  B >  *** WARNING *** WARNING *** WARNING *** WARNING *** WARNING ***  : No Will Robinson!  Too much Lost In Space... I think!  ;-)    I >  As you have undoubtedly noticed, shortly after SYSTARTUP_VMS finishes, , >  you see a message like so on the console: >  oK >  %SET-I-INTSET, login interactive limit = 64, current interactive value =y >  2    Sure have... a solution below...  H >  A little known secret is that until this message appears, user loginsK >  via the usual procedures (Job Controller) are disabled COMPELETELY. ThisnJ >  DOES NOT mean that logins are set zero, for indeed they are set at bootI >  time to the CURRENT value of IJOBLIM. (Go into SYSGEN, USE CURRENT andnJ >  SHOW IJOBLIM. Be careful to distinguish between CURRENT and ACTIVE.) ItF >  means that the Job Controller will not create interactive processesF >  until at least one SET LOGINS/INTER=value command is executed after >  boot-up.   ) Add this to your SYSTARTUP_VMS procedure:   * $!---------------------------------------- $! disable users from login F $! the set logins actually takes place in vms$lpbegin-050_startup.com;E $! set the startup$interactive_logins symbol to disable logins thoughoE $! users with /defpriv=oper will be able to gain access to the systema $!! $ startup$interactive_logins == 0    Self-explanitory.   E Then the last command procedure that you execute sets the logins to ae@ sufficient value - I use a logical set in SYLOGICALS to do this.  O Note that /defpriv=OPER?  This will allow you into the system regardless of the  set logins=0    C >  One known exception to this is Multinet, where TELNET inbound isfK >  possible anytime after the Multinet Master server completes its startup,yF >  so long as inbound Telnet is enabled in the Multinet configuration. >  l >  What does this mean to you? >   H >  Simply, it means that if SYSTARTUP_VMS finishes before your stream ofJ >  batch jobs, users with sufficient privilege by default can log into theD >  system before all of the disks are mounted, services are started,+ >  applications/databases are started, etc.t  N Users with sufficient privileges are allowed to access the system.  Period.  AM priv user (at least in my line of work) has the gray-matter to understand thet
 implications.n  O Since users access our server using TCP/IP, I can limit access two ways; (1) by * only starting DECnet and/or (2) set log/=0    G >  What if you don't start TCP/IP until very late in the sequence? SomeaK >  services/applications/databases may fail to start properly unless TCP/IPe >  is up and running.i >  eI >  PROCEED WITH __V_E_R_Y__ __S_E_V_E_R_E_L_Y__ __E_X_T_R_E_M_E__ CAUTIONa	 >  !!!!!!,  G My comment here - If you are starting TCP/IP very late in the sequence,iM regardless of SUBMIT or @ then you are shooting yourself in the foot.  As Red $ would say... (too much 70's for ya!)  M My startup submit's in order those procedures that need to start first and inaM ORDER.  This is command logica regardless.  There is absolutely no difference 8 in a submit vs. an @, you need to keep the orer or else.       John Brandon VMS Systems Administrator* Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wk  972.371.4003 fxn   ------------------------------  % Date: Sun, 30 Mar 2003 22:29:29 -0600-1 From: "David J. Dachtera" <djesys.nospam@fsi.net>3M Subject: Re: SYSTARTUP_VMS (was: Three questions re: the VMS startup process) ' Message-ID: <3E87C429.93322DC2@fsi.net>A   brandon@dalsemi.com wrote: > [snip]+ > Add this to your SYSTARTUP_VMS procedure:. > , > $!---------------------------------------- > $! disable users from loginnH > $! the set logins actually takes place in vms$lpbegin-050_startup.com;G > $! set the startup$interactive_logins symbol to disable logins though G > $! users with /defpriv=oper will be able to gain access to the system  > $!# > $ startup$interactive_logins == 0r >  > Self-explanitory.e > G > Then the last command procedure that you execute sets the logins to anB > sufficient value - I use a logical set in SYLOGICALS to do this. > Q > Note that /defpriv=OPER?  This will allow you into the system regardless of the  > set logins=0  C Exactly my point. (Far too) Many applications require users to holdiE outlandish privileges just to run, although this is fading out as thew ISV base shrinks.u  : Speaking from experience here - the danger is *VERY* real!   -- r David J. Dachtera- dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/R   ------------------------------  % Date: Mon, 31 Mar 2003 00:33:39 -0400t0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>M Subject: Re: SYSTARTUP_VMS (was: Three questions re: the VMS startup process)e/ Message-ID: <3E87C4F6.CE1E8687@vl.videotron.ca>    brandon@dalsemi.com wrote:I > My comment here - If you are starting TCP/IP very late in the sequence,tH > regardless of SUBMIT or @ then you are shooting yourself in the foot.   
 Why is that ?n  O > My startup submit's in order those procedures that need to start first and in O > ORDER.  This is command logica regardless.  There is absolutely no difference-: > in a submit vs. an @, you need to keep the orer or else.  K I have a batch queue with a single initiator (sorry for IBM/MVS terminogy).1M During startup, i submit a bunch of jobs in the order which they must executea to that batch queue.  N The problem with the TCPIP services is that the startup should really be splitD in 2 parts: the core startup and the "layered applications" startup.  M What I do in SYSTARTUP_VMS is that when it gets to the end, it then waits for+D the device BG: to become available, at which point it ends, allowing DECwindows to start.  N I think that one could use this to start other products in parralel as soon as BG becomes available.   L It may be good to get BG: up ASAP, but one may not necessarily setup all the, other TCPIP services right away during boot.  H On the other hand, with 5.3, TCPIP ahs a big all-mighty monkey hidden inH INET$ACP that starts services defined in the database (and automaticallyI restarts if they fail). So it may no longer be possible to start the core45 without having it start all the layered applications.    ------------------------------  # Date: Mon, 31 Mar 2003 03:27:39 GMTe; From: "John Gemignani, Jr." <jon-nope@thiswontworkossc.net>-/ Subject: Re: TCPIP Services and IP masquerading = Message-ID: <LAOha.23922$TW2.3901582@news1.news.adelphia.net>M  < "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message! news:3E87B40A.C7034F0F@fsi.net...g > "John Gemignani, Jr." wrote: > > @ > > "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message% > > news:3E864617.DD1DF0E4@fsi.net...o > > > Patrick Young wrote: > > > > E > > > > The quick answer I guess is yes/no (gateway yes, Masq - no) ?n > > > >e > > > > However... > > > >sK > > > > At home, I have a cable modem connected to a hub which is connectedeK > > > > to my OpenVMS Alpha system (OpenVMS 7.3-1/TCPIP Services 5.3) and a H > > > > Linux laptop (yes, I know I said I would not buy Intel, uh hum - stuff,L > > > > but it is running Redhat 8 and Window(tm) is nowhere near it's hdd - Il= > > > > need to get to my OpenVMS boxen while on the move :-)e > > > >gJ > > > > Telstra (ISP) only allow one computer at a time to authenticate to > > > > their gateway. [snip]m > > > L > > > Then a broadband router/firewall would be ideal for you. One device  -F > > > the router - authenticates to your ISP (probably PPPoE) and thenL > > > provides NAT to the rest of your home LAN. Smaller, cheaper and easier? > > > to run than another entire computer just for routing/NAT.e > > >eJ > > > I use the dial-up equivalent here, a 3com 56K LANmodem and have fiveJ > > > computers on-line, six when you include the laptop my group just got > > > from work. > >eI > > The 3COM LANmodem has one nice feature that I wish the LinkSys had --y itsoI > > own DNS.  The 3COM is a "dial on demand" router and, with its own DNSa builtrK > > in, DHCPs leases are recorded as well as the ability to define a seriesp ofK > > static addresses for non-DHCP devices as well.  I don't believe that itt > > caches lookups, though.o >tJ > I have a Netgear FR314 that I've never been able to use because we can'tI > get broadband here. I'd hoped a downtown Chicago provider would be ablenD > to service us, but they screwed up and failed to design sufficientF > capacity into their network. So, I still have the router I bought in > anticipation.d >g4 > Likewise, the Netgear doesn't seem to provide DNS. >w. > Here's a thought I've been mulling, however: >tF > I use static IP addresses on all my local machines. So, their setupsI > point to the LANmodem as the "next guy upstream to answer DNS queries". C > The thought occurs to me whether the LANmodem can be told to seeklI > someone else upstream via the LAN as *HIS* "next guy upstream to answer E > DNS queries". Dunno - don't think so, but may be worth exploring...e >o@ > ...or perhaps play some games with sub-zone delegation and the
 > LANmodem...R >O
 > Dunno... >  > -- > David J. Dachtera1 > dba DJE Systems  > http://www.djesys.com/ >@* > Unofficial Affordable OpenVMS Home Page:! > http://www.djesys.com/vms/soho/   I When the LAN modem connects, the provider has the ability to identify DNS K information for it when it's assigned its address.  I know that when I usedrG to use NT to do my DHCP I had the ability to provide a lot of different-; information to the client, such as default routes and such.i  J I don't have the LAN modem here any longer (I have it set up at the parishI office for their internet access), but I recall that's how it all worked.r   ------------------------------  % Date: Sun, 30 Mar 2003 21:20:42 -0600i1 From: "David J. Dachtera" <djesys.nospam@fsi.net>!/ Subject: Re: TCPIP Services and IP masqueradings' Message-ID: <3E87B40A.C7034F0F@fsi.net>    "John Gemignani, Jr." wrote: > > > "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message# > news:3E864617.DD1DF0E4@fsi.net...p > > Patrick Young wrote: > > >tC > > > The quick answer I guess is yes/no (gateway yes, Masq - no) ?m > > >  > > > However... > > >-I > > > At home, I have a cable modem connected to a hub which is connectedyI > > > to my OpenVMS Alpha system (OpenVMS 7.3-1/TCPIP Services 5.3) and aeM > > > Linux laptop (yes, I know I said I would not buy Intel, uh hum - stuff,6L > > > but it is running Redhat 8 and Window(tm) is nowhere near it's hdd - I; > > > need to get to my OpenVMS boxen while on the move :-)o > > >gH > > > Telstra (ISP) only allow one computer at a time to authenticate to > > > their gateway. [snip]  > >eJ > > Then a broadband router/firewall would be ideal for you. One device  -D > > the router - authenticates to your ISP (probably PPPoE) and thenJ > > provides NAT to the rest of your home LAN. Smaller, cheaper and easier= > > to run than another entire computer just for routing/NAT.  > >:H > > I use the dial-up equivalent here, a 3com 56K LANmodem and have fiveH > > computers on-line, six when you include the laptop my group just got > > from work. > K > The 3COM LANmodem has one nice feature that I wish the LinkSys had -- itssM > own DNS.  The 3COM is a "dial on demand" router and, with its own DNS builtiL > in, DHCPs leases are recorded as well as the ability to define a series ofI > static addresses for non-DHCP devices as well.  I don't believe that ite > caches lookups, though.s  H I have a Netgear FR314 that I've never been able to use because we can'tG get broadband here. I'd hoped a downtown Chicago provider would be ableHB to service us, but they screwed up and failed to design sufficientD capacity into their network. So, I still have the router I bought in
 anticipation.l  2 Likewise, the Netgear doesn't seem to provide DNS.  , Here's a thought I've been mulling, however:  D I use static IP addresses on all my local machines. So, their setupsG point to the LANmodem as the "next guy upstream to answer DNS queries".:A The thought occurs to me whether the LANmodem can be told to seekaG someone else upstream via the LAN as *HIS* "next guy upstream to answerhC DNS queries". Dunno - don't think so, but may be worth exploring...   > ...or perhaps play some games with sub-zone delegation and the LANmodem...c   Dunno...   --   David J. DachteraO dba DJE Systems- http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/e   ------------------------------  # Date: Mon, 31 Mar 2003 04:02:49 GMT ( From: "Mark E. Levy" <mlevy70@attbi.com>/ Subject: Re: TCPIP Services and IP masquerading - Message-ID: <J5Pha.269477$F1.47963@sccrnsc04>s  F "John Gemignani, Jr." <jon-nope@thiswontworkossc.net> wrote in message7 news:LAOha.23922$TW2.3901582@news1.news.adelphia.net...   K > When the LAN modem connects, the provider has the ability to identify DNSiH > information for it when it's assigned its address.  I know that when I usedI > to use NT to do my DHCP I had the ability to provide a lot of differentt= > information to the client, such as default routes and such.  >bL > I don't have the LAN modem here any longer (I have it set up at the parishK > office for their internet access), but I recall that's how it all worked.   H But the typical ISP dial-up doesn't use DHCP to assign the address. It'sF done as part of the PPP negotiation. You are correct, though, that DNS- information can be passed on at connect time.    ML   ------------------------------  + Date: Sun, 30 Mar 2003 23:30:13 +0000 (UTC)l, From: lewis@probe.mitre.org (Keith A. Lewis)) Subject: Re: VAX XDM, help needed please!e. Message-ID: <b67um5$m6b$1@newslocal.mitre.org>   "Diego CLAEYS" <diego.claeys@pandora.be> writes in article <CTeha.11575$t_2.1348@afrodite.telenet-ops.be> dated Sat, 29 Mar 2003 10:50:10 GMT:+ >Here is a question from a newbie hobbyist,9I >I have a VAX 4000-300 without display card, I have installed OpenVMS 7.3 / >with DECWindows 1.2.6 (Motif) and TCPWARE 5.6.cE >I would like to know if it is possible to have XDM running on my vax,E >(without display) and have a graphical login on my PC (OpenBSD) with: >something like X -query???<   Yes, you should be able to.>  < >Connectivity is OK = I can ping www.google.com from my VAX.H >DECWindows transport is OK = TCPIP = I can export the display with "SET( >DISPLAY/CREATE/NODE=.../TRANSPORT=TCPIP  L This command is roughly equivalent to a "setenv DISPLAY ..." on a unix box. I The test to see if you did it correctly comes with the command to run a X  application.  For examplec       $ run sys$system:decw$clockM  K If that works, you have exported the display correctly.  If it doesn't, yousH might need to do an "xhost +" command on your PC, as somebody mentioned.  9 >%%%%%%%%%%%  OPCOM  29-MAR-2003 11:45:38.11  %%%%%%%%%%%o  >Message from user SYSTEM on VAX9 >%TCPware_XDM-I-DEBUG, computed display name: localhost:0l  I It shouldn't be "localhost", it should be your PC's name or IP address.  yJ Does your PC have a name, according to the host table on the Vax?  If not,J consider adding it.  TCPware should work with just an IP, but I have never! used it so I don't know for sure.d  H If you can't get XDM to work, you can do a rough equivalent by exporting@ your display and then running the Decwindows startlogin utility.  1     $ set display/create/node=your_pc/trans=tcpipr     $ mc decw$startlogin  7 This will attempt to put a Vax login screen on your_pc.r  + --Keith Lewis              klewis$mitre.orgR> The above may not (yet) represent the opinions of my employer.   ------------------------------    Date: 30 Mar 2003 15:35:05 -0800* From: leichter@smarts.com (Jerry Leichter)@ Subject: Re: VMS Software Product & Online Documentation Library= Message-ID: <d1775491.0303301535.2cdb6a3a@posting.google.com>    Wow!  Talk about good timing!a  A After several years of not visiting comp.os.vms, I came back hereiF today to ask about ... getting hold of old VMS docs!  I'm specificallyD interested in stuff related to DECnet, and even more specifically toB DECnet management.  I'd especially love to get my hands on the oldD POLYCENTER docs.  The older the better; anything after about 1995 is? of no interest.  CD or hardcopy is fine.  Obviously, on-line is0 easiest - but not necessary.  B I am in a position to pay to either buy or just borrow this stuff.  ; If you have this material around ... please drop me a note.(   Thanks!R@                                                         -- Jerry   ------------------------------  % Date: Sun, 30 Mar 2003 21:42:24 -0500e' From: "Main, Kerry" <Kerry.Main@hp.com> @ Subject: RE: VMS Software Product & Online Documentation LibraryT Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF402660EC2@kaoexc01.americas.cpqcorp.net>   Re: old doc's ..  
 As a fyi -+ http://h71000.www7.hp.com/doc/archived.htmla   Regards   
 Kerry Main Senior Consultantd Hewlett-Packard (Canada) Co.! Consulting & Integration Servicesn Voice: 613-592-46604 Fax   : 613-591-4477 Email: kerryDOTmain@hpDOTcom-     (remove the DOT's and replace with "."'s)e OpenVMS DCL - the original .COMe   -----Original Message-----4 From: Jerry Leichter [mailto:leichter@smarts.com]=20 Sent: March 30, 2003 6:35 PM To: Info-VAX@Mvb.Saic.Com @ Subject: Re: VMS Software Product & Online Documentation Library     Wow!  Talk about good timing!A  G After several years of not visiting comp.os.vms, I came back here today$@ to ask about ... getting hold of old VMS docs!  I'm specificallyD interested in stuff related to DECnet, and even more specifically toB DECnet management.  I'd especially love to get my hands on the oldG POLYCENTER docs.  The older the better; anything after about 1995 is ofdF no interest.  CD or hardcopy is fine.  Obviously, on-line is easiest - but not necessary.  B I am in a position to pay to either buy or just borrow this stuff.  ; If you have this material around ... please drop me a note.r   Thanks!i@                                                         -- Jerry   ------------------------------  % Date: Sun, 30 Mar 2003 23:18:10 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>@ Subject: Re: VMS Software Product & Online Documentation Library/ Message-ID: <3E87B34A.4E307510@vl.videotron.ca>    "Main, Kerry" wrote:I > After several years of not visiting comp.os.vms, I came back here today-1 > to ask about ... getting hold of old VMS docs!    C Mr Main, you are lying. I saw you on comp.os.vms last week. :-) :-)R  = In other words, learn to configure your software to abide by:8  - ftp://ftp.rfc-editor.org/in-notes/rfc1855.txte  J (Nettiquette, in particular the use of > to quote text, quoted text at theP top, and NEVER NEVER send a post with more quoted text than what you add to it.)   ------------------------------  % Date: Mon, 31 Mar 2003 07:59:36 +0200c From: Dirk Munk <munk@home.nl>@ Subject: Re: VMS Software Product & Online Documentation Library2 Message-ID: <b68lgv$ntb$1@news4.tilbu1.nb.home.nl>  P Slightly of topic maybe, but a nice tip if you want to have all docs on line on & a VMS system with X-Windows interface.  K Create 3 LD devices (LD is on the freeware CD), and use backup to copy the tJ contents of the docs CD's to these devices. Mount the LD devices with the P correct label/logical (AXPDOC etc.), and run the setup utility from the docs CD.  = And don't forget to install Netscape, or better Mozilla/CSWB.c  = Result: all docs on line, and much faster then with CD-Rom's.n  P And of course you can have more docs sets on line too, for instance Vax and AXP 1 manuals. You only have to create more LD devices.    ------------------------------  % Date: Sun, 30 Mar 2003 15:55:09 -0700e& From: "rlfitch" <rlfitch@peakpeak.com>" Subject: RE: VMS Upgrade Questions/ Message-ID: <003601c2f70f$6d0c2dc0$0100a8c0@pj>   G This is very timely for me as I would like to upgrade from (Alpha) v7.2 H to v7.2-2 (or 7.3) but I'm confused as to which patch(es) to apply.  AnyE help is appreciated in deciphering the filenames of the patches. Plus H how to determine exactly what version the system is at and which patches: have been installed. An upgrade 'path' would help greatly.   thanks,  Ransom Fitch     -----Original Message-----( From: John Santos [mailto:JOHN@egh.com] & Sent: Saturday, March 29, 2003 7:42 AM To: Info-VAX@Mvb.Saic.Comu" Subject: Re: VMS Upgrade Questions    ( On Sat, 29 Mar 2003, Gary L. Ross wrote:  @ > We are upgrading from VMS 6.2-1H3 on our AlphaServer 4100 and H > AlphaStation 255/233 to VMS 7.2-1.  I have 2 questions regarding this 
 > upgrade. > ; > 1.)  Is VMS 7.2-1 supported on the AlphaStation 255/233 ?  > D > 2.)  While locating ECO's for the new OS version, I found a ratherF >       odd situation on which ECO's should be applied first.  Here is >       what I'm finding:w > & > Install DEC-AXPVMS-VMS721_PCSI-V0100' > and    DEC-AXPVMS-VMS721_UPDATE-V0300g >  > No problems up to this point.n > ? > Here is where I get confused on what should be installed nextn >   > DEC-AXPVMS-VMS721_AUDSRV-V02007 > * Requires SYS-V1000 but I can only locate SYS-V1200)z >   H No, you don't need to install SYS-V1000 first.  You just need to install& it as well to get all the corrections.     > DEC-AXPVMS-VMS721_SYS-V1200 1 > * Requires FIBRE_SCSI-V0500, MOUNT96-V0300, andt4 >    SYSLOA-V0200.  Located FIBRE_SCSI-V0600 though. >    Same...     $ > DEC-AXPVMS-VMS721_FIBRE_SCSI-V0600 > * Requires SYS-V1200 >    Same...e    ! > DEC-AXPVMS-VMS721_MOUNT96-V0300n- > * Requires SYS-V1100, FIBRE_SCSI-V0500, andm >    CLIUTIL-V0200     Same...e   > ! > DEC-AXPVMS-VMS721_CLIUTIL-V0200h > * Requires AUDSRV-V0200H >    Same...l    8 > It like one vicious cycle.  Am I missing something ??? >   H Yes.  :-)  The fine print says "The following remedial kit(s), or later,  F must be  installed BEFORE installation of this, or any required kit:",H which is not the same thing as: "In order to receive all the correctionsD listed  in  this kit,  the following remedial kits, or later, should also be installed:"s  C Only PCSI-V0100 and UPDATE-V0300 need to be installed before any of = these other patches.  The rest can be installed in any order.a  H Also, where patches say they require a reboot...  I've *never* needed toC reboot for each one.  Just install the lot of them and reboot once.t  F BTW, 7.2-1 is no longer officially supported (or support is going away/ RSN.)  I would definitely go to 7.2-2 or 7.3-1.f   > Gary L. Ross > Park Nicollet Health Servicesn >  >  >  >    --   John Santos, Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  # Date: Mon, 31 Mar 2003 00:45:50 GMTe! From: rob.buxton@wcc.spam.govt.nzh" Subject: Re: VMS Upgrade Questions& Message-ID: <3e878ea7.2072919190@news>  0 On Sat, 29 Mar 2003 00:42:17 -0600, Gary L. Ross  <rossgl@parknicollet.com> wrote:  > >We are upgrading from VMS 6.2-1H3 on our AlphaServer 4100 andA >AlphaStation 255/233 to VMS 7.2-1.  I have 2 questions regardingo >this upgrade. >R< >1.)  Is VMS 7.2-1 supported on the AlphaStation 255/233 ?   >.= I would strongly suggest going to 7.2-2 as this is the latest > "staging" version and will be supported for longer than 7.2-1.  C >2.)  While locating ECO's for the new OS version, I found a rathereE >      odd situation on which ECO's should be applied first.  Here isu >      what I'm finding: >e% >Install DEC-AXPVMS-VMS721_PCSI-V0100 & >and    DEC-AXPVMS-VMS721_UPDATE-V0300 >o >No problems up to this point. >y> >Here is where I get confused on what should be installed next >  >DEC-AXPVMS-VMS721_AUDSRV-V0200d6 >* Requires SYS-V1000 but I can only locate SYS-V1200) >-F Others have clarrified the issues,but also note that these updates areA cumulative. So, sys-v1200 will contain all that sys-v1000 did andeF more. So, ifsys-v1000 is a minimum, then sys-v1200 will more than meet the requirements.t  C And the best site for ECOs and which ones to pick is (and I hate toe4 say this living is New Zealand 8-) ) an Aussie site:@ http://ftp.support.compaq.com.au/pub/ecoinfo/ecoinfo/version.htm       >DEC-AXPVMS-VMS721_SYS-V12000 >* Requires FIBRE_SCSI-V0500, MOUNT96-V0300, and3 >   SYSLOA-V0200.  Located FIBRE_SCSI-V0600 though.r >o# >DEC-AXPVMS-VMS721_FIBRE_SCSI-V0600o >* Requires SYS-V1200e >e  >DEC-AXPVMS-VMS721_MOUNT96-V0300, >* Requires SYS-V1100, FIBRE_SCSI-V0500, and >   CLIUTIL-V0200d >   >DEC-AXPVMS-VMS721_CLIUTIL-V0200 >* Requires AUDSRV-V0200 >o7 >It like one vicious cycle.  Am I missing something ???  >e
 >Gary L. Rosss >Park Nicollet Health Services >a >    ------------------------------  % Date: Sun, 30 Mar 2003 22:14:04 -0600m1 From: "David J. Dachtera" <djesys.nospam@fsi.net> B Subject: Re: Which PAK for DECNET Phase IV under Hobbyist Program?' Message-ID: <3E87C08C.ED747592@fsi.net>    Paul Sture wrote:e > ] > In article <3E864163.6162CC81@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:s > > Paul Sture wrote:  > >>c > >> In article <slrnb8b820.qg.dsf@gaia.roc2.gblx.net>, Dan Foster <dsf@globalcrossing.net> writes:eK > >> > Thanks to all who followed up. I registered and loaded DVNETEND, and R > >> > then installed DECNET Phase IV, started it, added proxy entries. EverythingO > >> > appears to be working great. This is a single system, no other VMS boxesrD > >> > or routing needed so it's no great loss not to have DVNETRTG. > >> >N > >> > Just curious about something...the Hobbyist program also provides a PAK, > >> > called DVNETEXT. What's that one for? > >> > > >>, > >> I believe that is for DECnet-Plus. From< > >> http://h71000.www7.hp.com/doc/72final/6496/6496pro.HTML > >>Q > >> "The specific license required on your system is determined by the functionsg > >> you want to use:n > >>K > >> o - Basic function license (DVNETEND) --- provides end system support.r > >>P > >> o - Extended function license (DVNETEXT) for Alpha systems --- provides endK > >>     system support, DECdts server, cluster alias, and OSI applications  > >>     gateways. > >>N > >> o - Extended function license (DVNETRTG) for VAX systems --- provides endM > >>     system support, DECdns server, DECdts server, cluster alias, and OSI  > >>     applications gateways.w > >'L > > AFAIK, (I'll see if I can unearth the DECnet-IV doc.'s), DVNETEXT serves > > these purposes:b > >gJ > > On VAX, it permits a VAX executor to function as router within an areaD > > (synchronous links to remote LANs, etc.) A VAX area routing node > > requires the DVNETRTG PAK. > >oJ > > On Alpha, DVNETEXT permits mutliple DECnet lines/circuits to be active; > > simultaneously (DECnet Alpha does not provide routing).t > >u7 > > ...again AFAIK. I am open to correction, of course.g > > D > I dug out the above quote by plugging DVNETEXT into the search boxE > at www.openvms.compaq.com. The results included quite a few ask thetC > Wizard topics, one of which pointed at the SPDs as the definitive  > place to look. > E > I seem to remember the last time I went hunting for SPDs I ended upeB > going in circles, so I'll admit to chickening out at that point.  @ The SPD (if nothing else) might be found in the [.DOCUMENTATION]F directory under DECnet-IV of a current OpenVMS distro. I looked at theA V7.2-2 CD and it has the SPD (only) as .PS and .TXT in that path.l   -- e David J. Dachterae dba DJE Systemsr http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/n   ------------------------------    Date: 30 Mar 2003 21:14:38 -0800& From: jordan@ccs4vms.com (Rich Jordan): Subject: Re: [MOZILLA V1.3] Compatibility with JAVA V1.4 ?= Message-ID: <cc5619f2.0303302114.5436d906@posting.google.com>   [ Colin Blake <colin@theblakes.com> wrote in message news:<3E78F0A4.7080504@theblakes.com>...sI > Please upgrade from M1.2-1, A1.3 and B1.3 to M1.3 (1.3 final). Believe  I > me, there are problems with Java support which are not fixed until 1.3 e< > final. If you haven't seen them yet, you've been lucky :-)  D I just completed a series of upgrades to my PWS600au in order to getB Mozilla 1.3 (final) running.  VMS 7.2-2 (from 7.2-1), Motif 1.2-6,E TCPIP V5.1 eco 4 (was at eco 3). Also installed the Java SDK requirediF patches.  TCPIP V5.3 is waiting on media, but V5.1 should work per the release notes.   VMS722_ACRTL-V0200 VMS722_RMS-V0400/ VMS722_UPDATE-V0100   <--- for VMS722_SYS-V0100p  
 (plus LAN V3)e  @ Mozilla 1.3 seemed to run with the 1.3.1 plugin, though I reallyD didn't try anything specific to test, so I got the 1.4.0 release andD installed it in another directory tree (ODS-5 disk).  I have updated@ the logins, and copied the new plugin as required in the install? instructions.  When I go to a site that tries to activate Java,u? Mozilla now crashes with the following messages on the DECterm:a  > INTERNAL ERROR on Browser End: Could not read ack from browser' System error?:: interrupted system calla  D The user account has its UAF settings at appropriate levels, and theC SDK release notes were checked for new requirements (which were all3E met...).  Normal browsing and mail/news works fine as long as I don'tu go to a site running java.  B Don't have time to fall back to 1.3.1 tonight, though I'll do thatC just to be sure since I was hitting a couple of sites that actuallyt/ make java work hard (or try to).  Any thoughts?r   Rich Jordanm   ------------------------------   End of INFO-VAX 2003.177 ************************