1 INFO-VAX	Mon, 07 Mar 2005	Volume 2005 : Issue 132       Contents:+ Alternatives to media distribution service? 0 Re: Build problem - GD (but we're almost there!)# Re: Changing Tape Device protection & Re: DECC : toupper/tolower performance  Re: DECC floating point question  Re: DECC floating point question  Re: DECC floating point question LDAP module in Apache 2.0 & Re: new Itanium after Tukwila: Poulson Re: New OpenVMS-Integrity Video  Re: New OpenVMS-Integrity Video  Re: New OpenVMS-Integrity Video  Re: New OpenVMS-Integrity Video  Re: New OpenVMS-Integrity Video  Re: New OpenVMS-Integrity Video  Re: New OpenVMS-Integrity Video E Re: PCSI product install failure - DDIS-E-TNF, invalid element syntax E Re: PCSI product install failure - DDIS-E-TNF, invalid element syntax E Re: PCSI product install failure - DDIS-E-TNF, invalid element syntax 4 Re: Problems with two patches on OpenVMS 7.3-2 Alpha4 Re: Problems with two patches on OpenVMS 7.3-2 Alpha4 Re: Problems with two patches on OpenVMS 7.3-2 Alpha4 Re: Problems with two patches on OpenVMS 7.3-2 Alpha Shannon opines on HP Re: Shannon opines on HP Re: Shannon opines on HP Re: Shannon opines on HP Re: Shannon opines on HP Re: Shannon opines on HP( Sony vs HP: a comparison of CEO oustings, Re: Sony vs HP: a comparison of CEO oustings. Starting JAVA after Upgrading to OpenVMS 7.3.2I TCP/IP - Sockets appear to be restricted to maximum 65,535 byte transfers M RE: TCP/IP - Sockets appear to be restricted to maximum 65,535 byte transfers P Re: TCP/IP - Sockets appear to be restricted to maximum 65,535 byte transfers tr tuning NFS for single-thread Re: VMS Alpha 8.2 with CSWS 2.0  VMS startup problem? Re: VMS startup problem? Re: VMS startup problem? Re: VMS startup problem? [OT]: AMD virtualization. [TCPIP, OpenVMS] Problem with scaleable kernel  F ----------------------------------------------------------------------   Date: 7 Mar 2005 07:33:12 -0800  From: tadamsmar@yahoo.com 4 Subject: Alternatives to media distribution service?C Message-ID: <1110209592.544644.259990@f14g2000cwb.googlegroups.com>   E I see that we are paying $217/month for Media and Doc. Dist. Service.   + Yet, I only upgrade the OS every few years.   ; I wonder if you can just order an update when you want one.   7 I did see that a 7.32 security update patch that is not 7 listed on the patch download site came with 8.2.  So it 5 might be that some patches are only available via the 
 distribution.    ------------------------------  % Date: Mon, 07 Mar 2005 15:23:38 +0100 ' From: JOUKJ <joukj@hrem.stm.tudelft.nl> 9 Subject: Re: Build problem - GD (but we're almost there!) 2 Message-ID: <422C63EA.7040100@hrem.stm.tudelft.nl>   mcbill20@yahoo.com wrote: H > Thanks. That worked. Taking out all the "#" lines allowed the build toI > complete. I now just need to get the same thing to work with GDCHART. I D > am having some issues with the GD library software though. The exe > directory now contains:  > 7 > ANNOTATE.EXE;1          2344   4-MAR-2005 22:54:59.72  > CIRCLETEXTTEST.EXE;17 >                         3203   4-MAR-2005 22:55:14.84 7 > DEMOOUT.GIF;2              9   4-MAR-2005 23:15:33.61 7 > DEMOOUT.GIF;1              9   4-MAR-2005 23:09:45.06 7 > DEMOOUT.PNG;2             18   4-MAR-2005 23:15:33.13 7 > DEMOOUT.PNG;1             18   4-MAR-2005 23:09:44.63 7 > DEMOOUTP.PNG;2            11   4-MAR-2005 23:15:33.89 7 > DEMOOUTP.PNG;1            11   4-MAR-2005 23:09:45.30 7 > FONTSIZETEST.EXE;1      2563   4-MAR-2005 22:55:17.33  > FONTSIZETEST1.PNG;2 7 >                            1   4-MAR-2005 23:21:04.21  > FONTSIZETEST1.PNG;1 7 >                            1   4-MAR-2005 23:14:09.28  > FONTSIZETEST2.PNG;2 7 >                            1   4-MAR-2005 23:21:04.70  > FONTSIZETEST2.PNG;1 7 >                            1   4-MAR-2005 23:14:09.83  > FONTSIZETEST3.PNG;2 7 >                            1   4-MAR-2005 23:21:05.17  > FONTSIZETEST3.PNG;1 7 >                            1   4-MAR-2005 23:14:10.29  > FONTSIZETEST4.PNG;2 7 >                            1   4-MAR-2005 23:21:05.65  > FONTSIZETEST4.PNG;1 7 >                            1   4-MAR-2005 23:14:10.75  > FONTWHEELTEST.EXE;1 7 >                         2565   4-MAR-2005 22:55:19.81 7 > GD2COPYPAL.EXE;1         363   4-MAR-2005 22:55:01.89 7 > GD2TIME.EXE;1            363   4-MAR-2005 22:55:21.96 7 > GD2TOPNG.EXE;1          1021   4-MAR-2005 22:55:03.54 7 > GDDEMO.EXE;1            1281   4-MAR-2005 22:55:24.15 7 > GDPARTTOPNG.EXE;1       1021   4-MAR-2005 22:55:05.32 7 > GDTEST.EXE;1            1680   4-MAR-2005 22:55:27.39 7 > GDTESTFT.EXE;1          3229   4-MAR-2005 22:55:29.93 7 > GDTOPNG.EXE;1            991   4-MAR-2005 22:55:07.05 7 > PNGTOGD.EXE;1            991   4-MAR-2005 22:55:08.79 7 > PNGTOGD2.EXE;1          1020   4-MAR-2005 22:55:10.57 7 > TESTAC.EXE;1            1025   4-MAR-2005 22:55:32.51 7 > WEBPNG.EXE;1             989   4-MAR-2005 22:55:12.95  > H > However, when I try tests (using GDDEMO, GDTEST or WEBPNG) I don't getG > valid images. Actually, none of the existing PNG files that came with E > the distribution seem to be valid either. I copied them to a PC and G > tried to view them but I get "drawing failed" with the .JPG conersion H > and "Could not complete this operation. The internal file signature isI > incorrect." with the .PNG files. This seems odd but then again, I don't E > know much about the GD software yet. I'll poke around some more and  > also try building GDCHART. >  > Bill > E Strange. The sizes of the png files are wrong when I run gddemo on my * system I get the following files created :% Directory $DISK4:[JOUKJ.PUBLIC.gd.gd]     DEMOOUT.GIF;1            192/192  DEMOOUT.PNG;1            278/279  DEMOOUTP.PNG;1           156/156  ! Total of 3 files, 626/627 blocks.    All 3 images look fine.   G If you have VMS7.3-2 I can packup my gd/gdchart directory for you. You   can try my binaries.                   Jouk   ------------------------------  $ Date: Mon, 7 Mar 2005 11:10:55 -0500* From: "Marty O'Connor" <moconnor@dvfs.com>, Subject: Re: Changing Tape Device protection, Message-ID: <393co5F5oqrmuU1@individual.net>   : H : That's EXACTLY how this was manifested. The entries in SYS$DEVICES.DATG : did not configure at boot up; however, in attempting to find out what H : happened, I tried IO REPLACE, and guess what? Somehow the WWIDs of allD : my drives got scrambled. To this day, no one knows how that can be* : possible, but I (and my partner) saw it! :  : -- : David J Dachtera : dba DJE Systems  : http://www.djesys.com/  ^ I believe you. I spent the first few month of our new EVA (about 3 years ago) getting the tape] drives set the way I wanted them setup. The worst thing was that we had to shutdown the whole \ cluster so that it would "forget" the current information and only use the carefully updated information in SYS$DEVICES.DAT.    Marty    ------------------------------  % Date: Mon, 07 Mar 2005 10:11:00 -0800  From: Z <Z@no.spam> / Subject: Re: DECC : toupper/tolower performance * Message-ID: <vX0Xd.10209$q01.557@fe07.lga>    martinkirby12@yahoo.co.uk wrote:H > Note that there is no performance gain (in fact a performance hit) for: > using the macro format if the current locale is not "C".  E Can you supply some code that shows a perf hit when using the macros?    ------------------------------  # Date: Mon, 07 Mar 2005 16:30:19 GMT & From: John Reagan <john.reagan@hp.com>) Subject: Re: DECC floating point question 1 Message-ID: <vk%Wd.1122$0t1.502@news.cpqcorp.net>    Dave Froble wrote:   > I > Now that I got in my required language promotion, I'll guess that your  I > problem is using DOUBLE (8 byte floating point) on an Alpha.  There is  G > no such thing.  Alpha converts such to IEEE floating point, does the  I > calculations, and then converts it back.  Guess what, there are 3 less  , > bits of prescision in IEEE floating point. >   H Alpha has hardware support for VAX F, VAX G, IEEE S, and IEEE T formats B as well as a few instructions to help with VAX D floating.  The C G program on Alpha should use VAX F and VAX G out of the box without any   conversions to/from IEEE.    --   John Reagan / HP Pascal/{A|I}MACRO for OpenVMS Project Leader  Hewlett-Packard Company    ------------------------------  % Date: Mon, 07 Mar 2005 11:19:36 -0500 ' From: Dave Froble <davef@tsoft-inc.com> ) Subject: Re: DECC floating point question 0 Message-ID: <112ov83r9b5og1f@corp.supernews.com>   John Reagan wrote: > Dave Froble wrote: >  >>J >> Now that I got in my required language promotion, I'll guess that your J >> problem is using DOUBLE (8 byte floating point) on an Alpha.  There is H >> no such thing.  Alpha converts such to IEEE floating point, does the J >> calculations, and then converts it back.  Guess what, there are 3 less - >> bits of prescision in IEEE floating point.  >> > J > Alpha has hardware support for VAX F, VAX G, IEEE S, and IEEE T formats D > as well as a few instructions to help with VAX D floating.  The C I > program on Alpha should use VAX F and VAX G out of the box without any   > conversions to/from IEEE.  >   H Has my memory failed me again?  I thought that D-float was converted to < IEEE float for any colculations.  Is it G-float that's used?   ------------------------------   Date: 7 Mar 2005 11:47:57 -0600  From: briggs@encompasserve.org) Subject: Re: DECC floating point question 3 Message-ID: <2J1wWQnglWI$@eisner.encompasserve.org>   Z In article <112ov83r9b5og1f@corp.supernews.com>, Dave Froble <davef@tsoft-inc.com> writes: > John Reagan wrote: >> Dave Froble wrote:  >>   >>> K >>> Now that I got in my required language promotion, I'll guess that your  K >>> problem is using DOUBLE (8 byte floating point) on an Alpha.  There is  I >>> no such thing.  Alpha converts such to IEEE floating point, does the  K >>> calculations, and then converts it back.  Guess what, there are 3 less  . >>> bits of prescision in IEEE floating point. >>>  >>  K >> Alpha has hardware support for VAX F, VAX G, IEEE S, and IEEE T formats  E >> as well as a few instructions to help with VAX D floating.  The C  J >> program on Alpha should use VAX F and VAX G out of the box without any  >> conversions to/from IEEE. >>   > J > Has my memory failed me again?  I thought that D-float was converted to > > IEEE float for any colculations.  Is it G-float that's used?  C If my memory serves, D float is indeed converted to IEEE float for  : calculations.  And you lose 3 bits of precision due to the conversion.   C On VAX, you have the choice of running the C compiler in G floating ? mode or in D floating mode.  The default is D floating.  If you ? run in G floating mode, you want to be sure to link against the  G floating run time library.  = On Alpha, you also have a choice.  But, presumably due to the = poor hardware support for D floating, the default on Alpha is @ G floating.  I assume that want to take care to link against theB D floating run time library if you choose to override the default.  C By default, on Alpha, you get to use a fully supported data type --  G floating.    	John Briggs   ------------------------------   Date: 7 Mar 2005 10:09:43 -0800  From: drew.shelton@sematech.org " Subject: LDAP module in Apache 2.0C Message-ID: <1110218983.070535.277670@z14g2000cwz.googlegroups.com>   D I'm trying to load the LDAP module in SWS 2.0.  At this point I have. just the following line in mod_auth_ldap.conf:  9     LoadModule AUTH_LDAP_MODULE modules/mod_auth_ldap.exe   1 The command "httpd -t" gives the following error:   ;     Can't locate API module structure `AUTH_LDAP_MODULE' in D     file /apache$root/000000/modules/mod_auth_ldap.exe: function not     implemented   F The file apache$root:[modules]mod_auth_ldap.exe does exist.  What am I missing?   Thanks,  Drew    L ============================================================================A Drew Shelton                            drew.shelton@sematech.org < VMS Systems Manager                     office: 512-356-7575< Sematech                                fax:    512-356-7600 2706 Montopolis Drive A Austin, TX 78741-6499                I speak for myself only, not 	 Sematech. D     "There are two major products that come out of Berkeley: LSD and UNIX. D     We don't believe this to be a coincidence." - Jeremy S. AndersonL ============================================================================   ------------------------------  % Date: Sun, 06 Mar 2005 23:39:38 +0800  From: prep@prep.synonet.com / Subject: Re: new Itanium after Tukwila: Poulson - Message-ID: <87sm38g4w5.fsf@prep.synonet.com>   * Bill Todd <billtodd@metrocast.net> writes:   > prep@prep.synonet.com wrote: >  > ...  > F >> BTW, 90 and 65nM are the `new' sizes, don't expect to see 45nM till >> around 2010.  > H > I think that Intel is already sampling 65 nm. product for introductionF > about a year from now; IBM and AMD reportedly aren't far behind.  My? > impression is that at least Intel and IBM have already got 45 D > nm. devices running in the lab:  do you have a solid basis for not' > expecting 45 nm. product before 2010?   D No one has production capable steppers as yet, plus the small detail$ of leakage and working high K gates.  E Intel has a second problem in that IBM holds lots of the key patents.    --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Mon, 07 Mar 2005 11:59:39 GMT " From:   VAXman-  @SendSpamHere.ORG( Subject: Re: New OpenVMS-Integrity Video0 Message-ID: <00A4066B.B19A5208@SendSpamHere.ORG>  r In article <1110164660.97d1f060235acede8ac7ee82fd12e755@teranews>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes: >David J Dachtera wrote: > J >> Yes. However, we want that message to be consistent. That is, VMS info.I >> dissemination should not be exclusive of VMS systems (running Mozilla,  >> Opera, Mosaic, ...).   G Or of my Mac without having to pollute its pristine waters with the be- @ fouling stench of effluent from that Redmond cyber-sewage plant.      E >Which si why choosing an open non proprietary standard to dissminate E >information should be done. And that rules out Microsoft proprietary  >video formats.  > I >If the microsoft video software is able to play standard formats such as H >MPEG, then using open standards would only make VMS materials availableF >to a wider audience without preventing anyone from viewing it, as the! >current proprietary format does.   F Exactly.  You and David and a few others seem to be able to grasp this7 but the current faineant owners of VMS remain clueless.       --  K VAXman- A Bored Certified VMS Kernel Mode Hacker   VAXman(at)TMESIS(dot)COM              5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------  % Date: Mon, 07 Mar 2005 07:26:59 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> ( Subject: Re: New OpenVMS-Integrity VideoB Message-ID: <1110197663.e3275a220fed1b5ca518236d3f458c17@teranews>  ! VAXman-, @SendSpamHere.ORG wrote: H > Exactly.  You and David and a few others seem to be able to grasp this9 > but the current faineant owners of VMS remain clueless.   C But you can't blame the likes of Sue and others. They are slaves to G corporate standards. Heck, the VMS webmaster had to put back the stupid C , really stupid, 1 pixel images to set background colour instead of F using the bgcolor= tag that is built into HTML and far more efficient.  G I can imagine that if someone posted non-approved formats on the HP web H site, they might get fired, especially since there are plenty of rumours of downsizing.   ------------------------------   Date: 7 Mar 2005 05:35:02 -0800 * From: "Alan Greig" <greigaln@netscape.net>( Subject: Re: New OpenVMS-Integrity VideoC Message-ID: <1110202502.379893.267700@f14g2000cwb.googlegroups.com>    VAXman-@SendSpamHere.ORG wrote: G > In article <1110164660.97d1f060235acede8ac7ee82fd12e755@teranews>, JF , Mezei <jfmezei.spamnot@teksavvy.com> writes: > >David J Dachtera wrote: > > F > >> Yes. However, we want that message to be consistent. That is, VMS info. B > >> dissemination should not be exclusive of VMS systems (running Mozilla, > >> Opera, Mosaic, ...).  > E > Or of my Mac without having to pollute its pristine waters with the  be- B > fouling stench of effluent from that Redmond cyber-sewage plant. >  >  > G > >Which si why choosing an open non proprietary standard to dissminate G > >information should be done. And that rules out Microsoft proprietary  > >video formats.  > > C > >If the microsoft video software is able to play standard formats  such as @ > >MPEG, then using open standards would only make VMS materials	 available D > >to a wider audience without preventing anyone from viewing it, as the # > >current proprietary format does.  > C > Exactly.  You and David and a few others seem to be able to grasp  this9 > but the current faineant owners of VMS remain clueless.   G Maybe they will believe Microsoft then because even Windows Media 9 can 3 produce  (more or less) standards compliant output.   B "Windows Media Video 9 VCM provides the quality and flexibility ofG Windows Media Video without requiring the Advanced Systems Format (ASF) G container. ASF is the container specification to which the familiar WMA  and WMV files conform."   C But as Windows Media player plays almost any format other than Real A Media there is no need to use WM9 at all as the only distribution  formst for the media.   E Typically movie trailers are released by the big film companies in an B mpeg-4 format which all the major players (including Windows MediaF Player) can play. Very often with QuickTime recommended. VMS marketingB is obviously aimed at a techically less aware audience than "Dude, Where's my Car."  E Which reminds me. the trailers for Star Wars III and the Hitch Hikers D Guide to the Galaxy Movie are worth a look. Oh and the new series ofG Doctor Who starts on the BBC in the UK later this month but I really am  drifting off topic now.  --
 Alan Greig   ------------------------------  % Date: Mon, 07 Mar 2005 11:26:09 -0500 ' From: Dave Froble <davef@tsoft-inc.com> ( Subject: Re: New OpenVMS-Integrity Video0 Message-ID: <112ovkdgibqke96@corp.supernews.com>   JF Mezei wrote: # > VAXman-, @SendSpamHere.ORG wrote:  > H >>Exactly.  You and David and a few others seem to be able to grasp this9 >>but the current faineant owners of VMS remain clueless.  >  > E > But you can't blame the likes of Sue and others. They are slaves to I > corporate standards. Heck, the VMS webmaster had to put back the stupid E > , really stupid, 1 pixel images to set background colour instead of H > using the bgcolor= tag that is built into HTML and far more efficient. > I > I can imagine that if someone posted non-approved formats on the HP web J > site, they might get fired, especially since there are plenty of rumours > of downsizing.  ) http://www.theinquirer.net/?article=21608   F The firm has also imposed rigid consistency where it is needed least. G Where HP should be dynamic, it is static. And vice-versa. Take message  F delivery: HP spends some ~$300 million USD per annum on marketing and I marketing communications. Where these marketing dollars go is known best  I to HP. The firm just sacked an employee who was arguably the firm's most  H effective message delivery system, which probably saved $150K or so per C year. The opportunity cost inherent in lost sales opportunities is  C impossible to quantify, but the price will likely exceed the $150K  G figure by several zeroes. Perhaps the sacking was due to the fact that  A the employee in question failed to adhere to every aspect of the  G 132-page document HP created to instruct presenters on how to create a  G "One Voice" PowerPoint presentation. The employee may have used one of  E the Four Approved Fonts in the wrong place, or ran afoul of the text  B density police. Regardless, the employee was perfectly capable of ? speaking in his or her own voice, and the deals closed by this  > individual offer proof that substance matters more than style.   ------------------------------  # Date: Mon, 07 Mar 2005 17:12:16 GMT # From: hoff@hp.nospam (Hoff Hoffman) ( Subject: Re: New OpenVMS-Integrity Video1 Message-ID: <QX%Wd.1144$lw1.948@news.cpqcorp.net>   S In article <00A404DC.EBAA949C@SendSpamHere.ORG>, VAXman-  @SendSpamHere.ORG writes: } :In article <FD827B33AB0D9C4E92EACEEFEE2BA2FB594C71@tayexc19.americas.cpqcorp.net>, "Main, Kerry" <kerry.main@hp.com> writes:    :...and once again, .WMV : ; :VMS and Integrity... It's for Micro$oft PeeCee users only!     A   FWIW, I have been looking for a current mplayer port as well as C   any mpeg-related player software, and other related packages, for D   inclusion on the next Freeware.  Anyone have (more) current ports?    N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com    ------------------------------  # Date: Mon, 07 Mar 2005 16:58:27 GMT " From:   VAXman-  @SendSpamHere.ORG( Subject: Re: New OpenVMS-Integrity Video0 Message-ID: <00A40695.70037D3A@SendSpamHere.ORG>  r In article <1110197663.e3275a220fed1b5ca518236d3f458c17@teranews>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes:" >VAXman-, @SendSpamHere.ORG wrote:I >> Exactly.  You and David and a few others seem to be able to grasp this : >> but the current faineant owners of VMS remain clueless. > D >But you can't blame the likes of Sue and others. They are slaves to  D WHERE DID I EVER BLAME SUE??????????????????????????????????????????   --  K VAXman- A Bored Certified VMS Kernel Mode Hacker   VAXman(at)TMESIS(dot)COM              5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------  # Date: Mon, 07 Mar 2005 17:00:44 GMT " From:   VAXman-  @SendSpamHere.ORG( Subject: Re: New OpenVMS-Integrity Video0 Message-ID: <00A40695.C1753DB8@SendSpamHere.ORG>  p In article <1110202502.379893.267700@f14g2000cwb.googlegroups.com>, "Alan Greig" <greigaln@netscape.net> writes: >   >VAXman-@SendSpamHere.ORG wrote:H >> In article <1110164660.97d1f060235acede8ac7ee82fd12e755@teranews>, JF- >Mezei <jfmezei.spamnot@teksavvy.com> writes:  >> >David J Dachtera wrote:  >> >G >> >> Yes. However, we want that message to be consistent. That is, VMS  >info.C >> >> dissemination should not be exclusive of VMS systems (running 	 >Mozilla,  >> >> Opera, Mosaic, ...). >>F >> Or of my Mac without having to pollute its pristine waters with the >be-C >> fouling stench of effluent from that Redmond cyber-sewage plant.  >> >> >>H >> >Which si why choosing an open non proprietary standard to dissminateH >> >information should be done. And that rules out Microsoft proprietary >> >video formats. >> >D >> >If the microsoft video software is able to play standard formats >such asA >> >MPEG, then using open standards would only make VMS materials 
 >availableE >> >to a wider audience without preventing anyone from viewing it, as  >the$ >> >current proprietary format does. >>D >> Exactly.  You and David and a few others seem to be able to grasp >this : >> but the current faineant owners of VMS remain clueless. > H >Maybe they will believe Microsoft then because even Windows Media 9 can4 >produce  (more or less) standards compliant output. > C >"Windows Media Video 9 VCM provides the quality and flexibility of H >Windows Media Video without requiring the Advanced Systems Format (ASF)+                                             E Anything that incorporates "Advanced Systems" is a sleazy brainfuck.       --  K VAXman- A Bored Certified VMS Kernel Mode Hacker   VAXman(at)TMESIS(dot)COM              5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------  # Date: Mon, 07 Mar 2005 14:40:24 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond) N Subject: Re: PCSI product install failure - DDIS-E-TNF, invalid element syntax2 Message-ID: <sJZWd.1113$Kd1.1004@news.cpqcorp.net>  R In article <OFB2569A6A.B3DE25D1-ON85256FBA.0081D4B3-85256FBA.008276D8@metso.com>,  norm.raphael@metso.com writes:  J >This raised a question for me.  If he reinstalls all the stuff with /save >and then does /removeI >(and by extention, if those of us who did /save later do /remove) to get  >the disk space back, ; >is he (are we) good or is he (are we) in danger from this?   & Thank you for this excellent question.  9 I have NOT confirmed this, but, Yes, it seems to me that:   C If you PRODUCT INSTALL /SAVE_RECOVERY_DATA of a patch that includes  SYS$BASE_IMAGE.EXE,  AND * if then do a PRODUCT DELETE RECOVERY_DATA, AND D you then (probably by chance) put another file in the location where+ SYS$BASE_IMAGE.EXE, was privioulsy located,  AND ! you do all this BEFORE re-booting   3 Then you could see "corruption" in that other file.   @ My advice is: do NOT delete recovrey data untill the system disk1 as been re-booted by ALL NODES that boot from it.   & I will pass this on to the developers.   --  J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  # Date: Mon, 07 Mar 2005 14:45:50 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond) N Subject: Re: PCSI product install failure - DDIS-E-TNF, invalid element syntax1 Message-ID: <yOZWd.1114$Kd1.560@news.cpqcorp.net>   - In article <hf7Wd.22367$QQ3.10740@trnddc02>,  " John Santos <john@egh.com> writes: >Charlie Hammond wrote:  ..J >>     NOTE: The steps shown above will not create information in the PCSI1 >>     database about the OpenVMS platform.  i.e.  >>  $ >>         DEC AXPVMS OPENVMS V7.3-2 >>  : >>     However, the operating system itself will be there. >>  $ >>         DEC AXPVMS OPENVMS V7.3-2 > - >Should this line be "DEC AXPVMS VMS V7.3-2"?   K Yes.  (Blame it on the fact that I wrote that on Friday afternoon. <smile>)   I >(Both OPENVMS and VMS show up in PRODUCT SHOW PRODUCT *VMS* on my V7.3-2  >Alpha.)  ? This is correct and normal on all recent OpenVMS Alpha systems. - And on all OpenVMS I64 systems you should see             HP I64VMS OPENVMS V8.0  and           HP I64VMS VMS V8.0   G HOWEVER (too repeat) if you follow my instructions to recreate the PCSI 0 database, you will  NOT see the "OPENVMS" entry. --  J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  # Date: Mon, 07 Mar 2005 15:07:42 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond) N Subject: Re: PCSI product install failure - DDIS-E-TNF, invalid element syntax1 Message-ID: <27_Wd.1117$Kd1.967@news.cpqcorp.net>   ` In article <42292458.D45F227F@comcast.net>, David J Dachtera <djesys.nospam@comcast.net> writes: >Charlie Hammond wrote: 	 >> [snip] L >> We are embarrased to have caused this problem, and we offer our appology. > C >What you may want to consider for the next PCSI ECO is supplying a 1 >utility that will recreate the PCSI database ...   - This has been and continues to be considered.   ; However, fixing things that CAUSE problems is the priority.    --  J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------   Date: 7 Mar 2005 06:47:09 -0800  From: tadamsmar@yahoo.com = Subject: Re: Problems with two patches on OpenVMS 7.3-2 Alpha B Message-ID: <1110206829.431545.23470@z14g2000cwz.googlegroups.com>   Jim Duff wrote: F > Community service announcement:  two patches for OpenVMS 7.3-2 AlphaC > have significant problems.  Both patches are still available from  HP'sC > FTP site, with no indication that they contain any type of error.  > D > The first patch is VMS732_SYS-V0600, which has a faulty version of> > IO_ROUTINES.EXE, causing I/O database corruption for certain	 shadowset F > configurations.  This results in the $GETDVI SHDW_NEXT_MBR_NAME item6 > code returning the same physical name over and over. > F > The second patch is VMS732-UPDATE-V0x00.  Yes, all three versions of the A > patch have the problem.  In this case, the corrected version of F > NETDRIVER.EXE is installed in SYS$SYSTEM, not SYS$LOADABLE_IMAGES as itA > should be.  This one can hang your system, as we found out on a  > production cluster :(   4 What are the circumstances of the hangs?  I have not1 had any hangs. I upgraded our systems to 7.32 and 6 applied all these patches in the last couple of weeks.  5 BTW, anyone with the patch installed can confirm that 7 there is a newer version of NETSERVER.EXE in SYS$SYSTEM 2 but all the command procedures reference the older one in SYS$LOADABLE_IMAGES.    >  > Full write-ups at: > : > http://www.eight-cubed.com/blog/archives/000277.html and6 > http://www.eight-cubed.com/blog/archives/000278.html >  > Jim. > --  ! > jim AT eight DASH cubed DOT com    ------------------------------  # Date: Mon, 07 Mar 2005 15:47:44 GMT  From: Jim Duff <jim@127.0.0.1>= Subject: Re: Problems with two patches on OpenVMS 7.3-2 Alpha G Message-ID: <AI_Wd.110950$Th1.69393@bgtnsc04-news.ops.worldnet.att.net>    tadamsmar@yahoo.com wrote: > Jim Duff wrote:  > F >>Community service announcement:  two patches for OpenVMS 7.3-2 AlphaC >>have significant problems.  Both patches are still available from 	  >>[snip]  > 6 > What are the circumstances of the hangs?  I have not3 > had any hangs. I upgraded our systems to 7.32 and 8 > applied all these patches in the last couple of weeks. > 7 > BTW, anyone with the patch installed can confirm that 9 > there is a newer version of NETSERVER.EXE in SYS$SYSTEM 4 > but all the command procedures reference the older > one in SYS$LOADABLE_IMAGES.  >   D The issue occurs when NETDRIVER attempts a double deletion of a NET B device.  The problem in both cases that we saw (the first back on E OpenVMS 7.3, which generated the original NETACP patches and the one  I just recently on 7.3-2 that uncovered this mis-installation problem) was  I that the process that went RWAST was holding some fairly important locks  H associated with an Rdb database, causing us to have to reboot to regain H access to the production application.  So I guess my sentence "this can 2 hang your system" was not a great choice of words.   Regards, Jim    --   jim AT eight DASH cubed DOT com    ------------------------------  * Date: Mon, 7 Mar 2005 16:13:54 +0000 (UTC)6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)= Subject: Re: Problems with two patches on OpenVMS 7.3-2 Alpha 1 Message-ID: <newscache$63pzci$3jd1$1@news.sil.at>   h In article <5kmWd.104344$Th1.60669@bgtnsc04-news.ops.worldnet.att.net>, Jim Duff <jim@127.0.0.1> writes:F >Community service announcement:  two patches for OpenVMS 7.3-2 Alpha H >have significant problems.  Both patches are still available from HP's B >FTP site, with no indication that they contain any type of error.  1 Did you inform George PAGLIARULO directly, also ?   D >The first patch is VMS732_SYS-V0600, which has a faulty version of H >IO_ROUTINES.EXE, causing I/O database corruption for certain shadowset F >configurations.  This results in the $GETDVI SHDW_NEXT_MBR_NAME item 5 >code returning the same physical name over and over.    This sounds dangerous.J But, alas, I can't reproduce it here (though I have a lot of shadow sets).4 Do I need a 3 member shadowset for this to be seen ?  J >The second patch is VMS732-UPDATE-V0x00.  Yes, all three versions of the A >patch have the problem.  In this case, the corrected version of  I >NETDRIVER.EXE is installed in SYS$SYSTEM, not SYS$LOADABLE_IMAGES as it  A >should be.  This one can hang your system, as we found out on a   >production cluster :(  C Indeed, I have a NETDRIVER.EXE in SYSEXE installed by VMS732_UPDATEs8 However, it is not used, neither is one used in SYS$LDR.G I do have a NET$DRIVER.EXE in SYS$LDR. It is used. It came with DNVOSI.i  J So, I assume, NETDRIVER.EXE is part of DECnet Phase IV (which I don't use)G and a new version should therfore come with a DN4 ECO and not a VMS onenC (and should then be of course installed in the correct location) !!u   >Full write-ups at:  >n9 >http://www.eight-cubed.com/blog/archives/000277.html and 5 >http://www.eight-cubed.com/blog/archives/000278.html    I post here.     Please keep us informed    -- n Peter "EPLAN" LANGSTOEGERe% Network and OpenVMS system specialisto E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  % Date: Mon, 07 Mar 2005 08:19:34 -0800, From: JBloggs@acme.com= Subject: Re: Problems with two patches on OpenVMS 7.3-2 Alphab8 Message-ID: <jhvo21hsag5tqkpgp4c5gve06sjl1tv59a@4ax.com>  A On Mon, 07 Mar 2005 15:47:44 GMT, Jim Duff <jim@127.0.0.1> wrote:e   >tadamsmar@yahoo.com wrote:o >> Jim Duff wrote:  E >The issue occurs when NETDRIVER attempts a double deletion of a NET  C >device.  The problem in both cases that we saw (the first back on pF >OpenVMS 7.3, which generated the original NETACP patches and the one J >just recently on 7.3-2 that uncovered this mis-installation problem) was J >that the process that went RWAST was holding some fairly important locks I >associated with an Rdb database, causing us to have to reboot to regain aI >access to the production application.  So I guess my sentence "this can l3 >hang your system" was not a great choice of words.u  9 Is NETDRIVER used by DECnet Phase IV only?  just curious.?   ------------------------------  $ Date: Mon, 7 Mar 2005 06:38:39 -0500# From: "John Smith" <a@nonymous.com>- Subject: Shannon opines on HP-, Message-ID: <Me2dnfyUYrDcoLHfRVn-hQ@igs.net>  ) http://www.theinquirer.net/?article=21608o   --C OpenVMS - The often imitated but never advertised operating system.2   ------------------------------  % Date: Mon, 07 Mar 2005 07:24:00 -0500i- From: JF Mezei <jfmezei.spamnot@teksavvy.com> ! Subject: Re: Shannon opines on HPyB Message-ID: <1110197486.aa4c79f26ad8a7b9280526fa0123908f@teranews>   John Smith wrote:  > + > http://www.theinquirer.net/?article=21608i  D Standard Shannon fodder. He states that Carly had nothing to do withH alpha's cancellation. That is not true. She had to do with the timing of0 the announcement. (before merger was annoucned).  = What shannon doesn't mention is that Curly was shopping for aiC saviour/buyer for some time, and had tried to woo IBM and only when0E Gerstner laughed at the idea after he looked at the investment bankerwG proposals did Compaq then move on to the next possible buyer. He didn'toD want to kill Alpha in case a buyer might actually be interested in aG viable chip. When IBM and Intel were ruled out as potential buyers, andoF HP was positive, then Curly pulled out the "kill alpha" plans and made them public.  F Had Compaq been more succesful and not needing a buyer, I suspect thatC it would have waited for IA64 to be less laughable and less "in the;C future" before announcing the murder of Alpha. But as soon as Carly G showed up with interest, it was clear that Carly had nothing to do with0A Alpha and Curly could then kill it and give its remains to Intel.3  C In both cases, Pfeiffer and Carly had their eyes set on growing theIH company to compete against IBM. In both cases, they failed royally since> they focused on the wintel crap instead of leveraging the real enterprise stuff.     H What I do not know is what happened to the porting of Tru64 to IA64 thatD had been started under Palmer. Did it continue under Compaq ? If so,A when HP inherited Compaq, wouldn't there have been parts of Tru648I already available on IA64 that could have been easily merged with HP-UX ?2   ------------------------------   Date: 7 Mar 2005 05:34:08 -0800M From: bob@instantwhip.comT! Subject: Re: Shannon opines on HP C Message-ID: <1110202448.807119.310330@z14g2000cwz.googlegroups.com>   . Curly Carly Palmer and all the rest are fools,- history shows it, and any one that hires them  is one also ...2   ------------------------------  % Date: Mon, 07 Mar 2005 06:07:20 -0800 # From: "Tom Linden" <tom@kednos.com>R! Subject: Re: Shannon opines on HP1( Message-ID: <opsm9tiiq7zgicya@hyrrokkin>  E On Mon, 7 Mar 2005 06:38:39 -0500, John Smith <a@nonymous.com> wrote:X  + > http://www.theinquirer.net/?article=21608E >E > --E > OpenVMS - The often imitated but never advertised operating system.9 >4 >R@ According to story in latest Business Weel Dick Hackborn appears to be the Gris Eminence.     --  C Using Opera's revolutionary e-mail client: http://www.opera.com/m2/M   ------------------------------  $ Date: Mon, 7 Mar 2005 10:38:40 -0500# From: "John Smith" <a@nonymous.com> ! Subject: Re: Shannon opines on HPe, Message-ID: <6dKdnR21o9ce6LHfRVn-hQ@igs.net>   JF Mezei wrote:t >  >iE > What I do not know is what happened to the porting of Tru64 to IA64hG > that had been started under Palmer. Did it continue under Compaq ? IfeG > so, when HP inherited Compaq, wouldn't there have been parts of Tru64 C > already available on IA64 that could have been easily merged with 	 > HP-UX ?     H You're mixing apples with oranges thinking about this. It's one thing toK port an os to a different architecture, while it's sometheing else to wedgeDJ an elegant set of little-endian features into a big-endian pile of manure.   --- OpenVMS - The classics never go out of style.c   ------------------------------  % Date: Mon, 07 Mar 2005 12:21:28 -0500 ( From: Bill Todd <billtodd@metrocast.net>! Subject: Re: Shannon opines on HP = Message-ID: <JN-dnUmUMdEHELHfRVn-vA@metrocastcablevision.com>    JF Mezei wrote:7   ...c  E > In both cases, Pfeiffer and Carly had their eyes set on growing the J > company to compete against IBM. In both cases, they failed royally since@ > they focused on the wintel crap instead of leveraging the real > enterprise stuff.a   Wrong.  @ Pfeiffer focused significantly, possibly even primarily, on the G enterprise stuff.  It may well have been a large part of the reason he iJ got kicked out (because Ben Rosen favored a far more PC-centric strategy).  ; Or perhaps you meant to say 'Curly' rather than 'Pfeiffer'.a   - bill   ------------------------------  % Date: Mon, 07 Mar 2005 02:58:39 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com>i1 Subject: Sony vs HP: a comparison of CEO oustingsyB Message-ID: <1110181564.d5e691710c4c89601f1a387e3b21f247@teranews>  H Sony just announced it had ousted its CEO and 7 other executives. But asF part of its announcement, it also announced the appointment of the new* CEO (the first foreign born CEO for Sony).  F HP announced the ousting of La Carly with her immediate dismissal, but no replacement in sight.  F Compaq announced the ousting of Pfeiffer with immediate dismissal, butD no replacement in sight, and they evetually failed to find a willing9 replacement and gave the CEO job to the accountant Curly.t  F In the case of IBM, I believe that Gerstner was appointed before Akers9 left on retirement, so again , no gap without leadership.C  G It worries me that HP may in fact end up like Compaq, unable to attract A a new LEADER and forced to appoint some accountant or MBA that is1) without vision, leadership and backbone. o  C One needs backbone in the case of HP to be able to stand up against$P microsoft and intel in order to make one's own more profitable products succeed.  E It has been over a month since La Carly was ousted. (she's only got 2LG months of PC support left !) and we don't hear any serious news about ah
 replacement.    D When you consider that the ousting was not a surprise for the board,G (and not a surprise for anyone for that matter), you'd think that boardtG members would have covertly begun a search a few months before and havedN a CEO replacement process well underway at the time the ousting was announced.  D Is there nobody interested in the job ? Or is it a case of the board2 being lethargic and unable to find a GOOD leader ?  D Note that Gerstner had been contacted long before the process at IBMG became public and was very reluctant because he didn't see how he could9H help IBM out of its morass. But a few key board members became convincedF he was the one they needed and then managed to convince him to take on the challenge.  H Seems to me that HP might become a victim of a lethargic board that justE sends a kind letter of invitation and accepts a "no thank you" answereA without a fight. And the fact that Curly is even being consideredcG (allegedly) is an indication that the Board is really out of touch with B reality and doesn't even realise that Curly is an incompetant twitB capable of only negotiating selling the company to another outfit.  B If they do bring in Curly after a very long wait for him to becomeG available after winding down MCI, it probably means that they expect HPsJ to be bought by some japanese outfit (perhaps Sony might be a good fit ?).   ------------------------------  % Date: Mon, 07 Mar 2005 03:27:02 -0500d- From: JF Mezei <jfmezei.spamnot@teksavvy.com>-5 Subject: Re: Sony vs HP: a comparison of CEO oustings B Message-ID: <1110183264.eaee1b2e07bfd85a994a26d4d340c658@teranews>   Another example:  2 http://news.airwise.com/story/view/1110143136.html  C Rod Eddington is rumoured to be leaving British Airways mid-year to,F return to Australia. BA's board are already busy rounding up potentialE CEO replacements so that BA isn't without CEO for any length of time.eE Note: in BA's case, the Chairman of the Board != CEO, so the chairman * has enough powers to start such a process.  E Was Carly also chairman of the board ? If so, that would have made it E harder from a corporate governamce perspective to have the board takee action against Carly.i   ------------------------------   Date: 7 Mar 2005 10:22:20 -0800r From: gary.lively@gmail.comn7 Subject: Starting JAVA after Upgrading to OpenVMS 7.3.2eC Message-ID: <1110219740.355953.104650@o13g2000cwo.googlegroups.com>F  C After upgrading to OpenVMS 7.3.2, the current version of Java 1.1.8> stopped working.  E I upgraded to version JAVA131 V1.3-15, but after a reboot it will notf start.  7 My question is - how do I restart Java after a reboot ?o   thanks   ------------------------------  % Date: Mon, 07 Mar 2005 15:51:56 +0000s- From: Roy Omond <Roy.Omond@BlueBubble.UK.Com> R Subject: TCP/IP - Sockets appear to be restricted to maximum 65,535 byte transfers, Message-ID: <393bpoF5u1t9oU1@individual.net>   Gentle colleagues,  / this is becoming a bit of a showstopper for me:   < A very simple pair of programs (Server and Client) are setup@ to transfer a quantity of data via a TCP Socket.  The backgroundD to this is actually in trying to get this to work in UCX v3.3 ECO 3,> whereas it worked flawlessly (and as per the documentation) in: vanilla UCX v3.3 (without ECO), running under VAX VMS 6.2.  < The problem is easily demonstrated on VMS 7.3-2 (Alpha) with= TCP/IP Services v5.4 ECO 4, as well as all the versions after:. UCX v3.3 ECO 3 and above (both VAX and Alpha).  F The Server process establishes a socket, listens for an accept on the F socket (which the Client sends), and sends an amount of data down the E socket to the client.  That's it.  It works fine with data up to and p: including 65,535 bytes.  Beyond this, it returns an error . ("%system-f-ivbuflen, invalid buffer length").> I can find no mention in the documentation "TCP/IP Sockets API? and System Services Programming" of any such restriction on thep amount of data to transfer.m   Here are the 2 small programs:   --- cut here ---   $ create client.c  #include <errno.h> #include <stdio.h> #include <sys/types.h> #include <sys/socket.h>= #include <netinet/in.h>  #include "sys/ioctl.h" #include <netdb.h> /* struct hostent *VGetHostByName(char *name)  {           struct hostent *h;5          int i;             if (name == NULL)
          {            return(NULL);
          }          sethostent( 1 );I  +          while ((h = gethostent()) != NULL)T              {3                  if (strcmp(name,h->h_name) == 0) { &                          endhostent();#                          return(h);=                  }                  i = 0;=2                  while (h->h_aliases[i] != NULL) {A                          if (strcmp(name,h->h_aliases[i]) == 0) {y.                                  endhostent();+                                  return(h);p                          }                          i++;:                  }
          }          endhostent();          return( NULL ); }*/i   main (int argc, char ** argv)l {>          int sock;          int byte_count;          int bcount;          char data[1000000];          int sendbuf;M          int len;   /          static struct hostent *mh,master_host;s'          static struct sockaddr_in sin;m  F          if ((sock = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP)) == -1)
          {+                  perror ("Create Socket:"); ?                  printf ("errno in CreateSock is %d\n", errno);r                  return(-1);
          }            if (argc > 1)
          {=                  printf ("Connecting to host %s\n", argv[1]); -                  mh = GetHostByName(argv[1]);                    if (mh == NULL)                  {3                          perror ("GetHostByName:");,F                          printf ("Can't get host name %s\n", argv[1]);%                          close(sock);u%                          return (-1);                   }
          }
          else0
          {/                  printf ("Need a host name\n");                   return (-1);i
          }            master_host = *mh;n  1          sin.sin_family = master_host.h_addrtype;c          sin.sin_port = 50000;I          memcpy(&sin.sin_addr, master_host.h_addr, master_host.h_length);   E          if (connect(sock, &sin, sizeof(struct sockaddr_in))  ==  -1)f
          {%                  perror ("Connect:");                   return (-1);O
          }            len = sizeof(sendbuf);vE          if (getsockopt(sock,SOL_SOCKET,SO_SNDBUF,&sendbuf,&len) < 0)l
          {0                  perror("getsockopt SO_SNDBUF");
          }2          printf("Send buffer %d bytes\n",sendbuf);            while (1)
          {0                  ioctl(sock, FIONREAD, &bcount);                     if (bcount > 0)                  {                   int cnt = 0;H       if ((bcount = read (sock, (char *)&byte_count,sizeof(int))) == -1)                      {2                         perror ("Read data size");$                         return (-1);                      }< printf ("Read %d bytes, count is %d\n", bcount, byte_count);  $                          bcount = 0;1                          while (cnt < byte_count)k                          {C           if ((bcount = read (sock, data, byte_count - cnt)) == -1)"                              {:                                      perror ("Read data");1                                      return (-1);w                              }:                              printf ("Read %d\n", bcount);+                              cnt += bcount;c                          }?                          printf ("Read %d bytes total\n", cnt);t                          break;                   }
          } }S   $ create server.cr #include <errno.h> #include <stdio.h> #include <sys/types.h> #include <sys/socket.h>  #include <netinet/in.h>a #include <ucx$inetdef.h>   inte0 writevv (int fd, struct iovec * iov, int iovcnt) {L'          int total_bytes = 0, bytes, i;   %          for (i = 0; i < iovcnt; i++)t
          {E                  bytes = write (fd, iov[i].iov_base, iov[i].iov_len);r!                  if (bytes == -1)-                  {+                          perror ("write:");A%                          return (-1);j                  }                  elseg                  {.                          total_bytes += bytes;                  }
          }          return (total_bytes); }5   main (int argc, char ** argv)J {             int serv_sock = -1;          int client_sock = -1;          int len;>            int data_size = 1000;  2          if (argc >= 2) data_size = atoi(argv[1]);  K          if ((serv_sock = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP)) == -1)o
          {+                  perror ("Create Socket:");e?                  printf ("errno in CreateSock is %d\n", errno);t                  return(-1);
          }            len=sizeof(data_size);a  I        if (setsockopt(serv_sock,SOL_SOCKET,SO_SNDBUF,&data_size,len) < 0).
          {0                  perror("setsockopt SO_SNDBUF");
          }  I        if (setsockopt(serv_sock,SOL_SOCKET,SO_RCVBUF,&data_size,len) < 0) 
          {0                  perror("setsockopt SO_RCVBUF");
          }<          printf("Send/Receive buffer %d bytes\n",data_size);  
          {/                  static struct sockaddr_in sin;c:                  sin.sin_family          = htons(AF_INET);1                  sin.sin_port            = 50000;r9                  sin.sin_addr.s_addr = htonl(INADDR_ANY);b  @                  if (bind(serv_sock, &sin, sizeof(sin))  ==  -1)                  {>                          printf("cannot bind socket to port");*                          perror ("Bind:");$                          return(-1);                  }
          }  +          if (listen(serv_sock, 20)  ==  -1) 
          {3                  printf("cannot set listen(s,20)");f$                  perror ("Listen:");                  return(-1);
          }  ?          if ((client_sock = accept(serv_sock, NULL, NULL)) > 0)r
          {%                  struct iovec iov[2];r                  int count;r+                  static char data[1000000];   >                  if (data_size > 1000000) data_size = 1000000;7                  iov[0].iov_base = (char *) &data_size;h.                  iov[0].iov_len = sizeof(int);)                  iov[1].iov_base = &data;u,                  iov[1].iov_len = data_size;  6                  if (writevv(client_sock,iov,2) == -1)                  {,                          perror ("writev:");%                          return (-1);-                  }                  sleep(10); 
          } }l $ server:=$u:[xxxx]serveru $ client:=$u:[xxxx]clientg $ spawn/nowait server 65536v $ client my-node-namew      = Any comments ?  Any suggestions ?  I thinks this, at worst, a : bug in the Socket implementation, or, at "best", a serious omission in the documentation.  ; Note, as mentioned above, this worked fine in UCX v3.3, but' not in UCX v3.3 ECO 3 or later.r   Many thanks in advance,o  	 Roy Omonda Blue Bubble Ltd.   ------------------------------  $ Date: Mon, 7 Mar 2005 11:20:39 -0500# From: "Dan Allen" <dallen@nist.gov> V Subject: RE: TCP/IP - Sockets appear to be restricted to maximum 65,535 byte transfers: Message-ID: <JFEPKAPBPMDFDBOIANGDMEGMGCAA.dallen@nist.gov>  J Well - documentation aside - TCP sockets are byte streams. The size of theL actual TCP packets are not necessarily related to the application read/writeN sizes. Just read and write in chunk sizes of your choosing. The TCP layer will packetize it at whim anyway.     HTH,   Danm   > -----Original Message-----6 > From: Roy Omond [mailto:Roy.Omond@BlueBubble.UK.Com]' > Sent: Monday, March 07, 2005 10:52 AM  > To: Info-VAX@Mvb.Saic.ComoJ > Subject: TCP/IP - Sockets appear to be restricted to maximum 65,535 byte > transfersg >e >e > Gentle colleagues, > 1 > this is becoming a bit of a showstopper for me:i >r> > A very simple pair of programs (Server and Client) are setupB > to transfer a quantity of data via a TCP Socket.  The backgroundF > to this is actually in trying to get this to work in UCX v3.3 ECO 3,@ > whereas it worked flawlessly (and as per the documentation) in< > vanilla UCX v3.3 (without ECO), running under VAX VMS 6.2. > > > The problem is easily demonstrated on VMS 7.3-2 (Alpha) with? > TCP/IP Services v5.4 ECO 4, as well as all the versions aftert0 > UCX v3.3 ECO 3 and above (both VAX and Alpha). >eG > The Server process establishes a socket, listens for an accept on thebG > socket (which the Client sends), and sends an amount of data down the F > socket to the client.  That's it.  It works fine with data up to and; > including 65,535 bytes.  Beyond this, it returns an errorV0 > ("%system-f-ivbuflen, invalid buffer length").@ > I can find no mention in the documentation "TCP/IP Sockets APIA > and System Services Programming" of any such restriction on the0 > amount of data to transfer.e >9  > Here are the 2 small programs: >  > --- cut here --- >t > $ create client.ca > #include <errno.h> > #include <stdio.h> > #include <sys/types.h> > #include <sys/socket.h>b > #include <netinet/in.h>p > #include "sys/ioctl.h" > #include <netdb.h> > /* > struct hostent > *VGetHostByName(char *name)o > {t >          struct hostent *h;( >          int i;? >  >          if (name == NULL) >          { >            return(NULL); >          } >          sethostent( 1 );e >e- >          while ((h = gethostent()) != NULL)- >              {5 >                  if (strcmp(name,h->h_name) == 0) { ( >                          endhostent();% >                          return(h);- >                  } >                  i = 0;:4 >                  while (h->h_aliases[i] != NULL) {C >                          if (strcmp(name,h->h_aliases[i]) == 0) {n0 >                                  endhostent();- >                                  return(h);m >                          } >                          i++;  >                  } >          } >          endhostent(); >          return( NULL ); > }*/r >t > main (int argc, char ** argv)u > {. >          int sock; >          int byte_count; >          int bcount; >          char data[1000000]; >          int sendbuf;? >          int len;  >A1 >          static struct hostent *mh,master_host;X) >          static struct sockaddr_in sin;l >yH >          if ((sock = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP)) == -1) >          {- >                  perror ("Create Socket:"); A >                  printf ("errno in CreateSock is %d\n", errno);g >                  return(-1); >          } >a >          if (argc > 1) >          {? >                  printf ("Connecting to host %s\n", argv[1]);e/ >                  mh = GetHostByName(argv[1]);r" >                  if (mh == NULL) >                  {5 >                          perror ("GetHostByName:"); H >                          printf ("Can't get host name %s\n", argv[1]);' >                          close(sock);.' >                          return (-1);s >                  } >          } >          else, >          {1 >                  printf ("Need a host name\n");  >                  return (-1);> >          } >f >          master_host = *mh;e >e3 >          sin.sin_family = master_host.h_addrtype;n  >          sin.sin_port = 50000;K >          memcpy(&sin.sin_addr, master_host.h_addr, master_host.h_length);r >aG >          if (connect(sock, &sin, sizeof(struct sockaddr_in))  ==  -1)  >          {' >                  perror ("Connect:");g >                  return (-1);M >          } >a! >          len = sizeof(sendbuf);hG >          if (getsockopt(sock,SOL_SOCKET,SO_SNDBUF,&sendbuf,&len) < 0)  >          {2 >                  perror("getsockopt SO_SNDBUF"); >          }4 >          printf("Send buffer %d bytes\n",sendbuf); >  >          while (1) >          {2 >                  ioctl(sock, FIONREAD, &bcount); >c" >                  if (bcount > 0) >                  {  >                   int cnt = 0;J >       if ((bcount = read (sock, (char *)&byte_count,sizeof(int))) == -1) >                      {4 >                         perror ("Read data size");& >                         return (-1); >                      }> > printf ("Read %d bytes, count is %d\n", bcount, byte_count); >r& >                          bcount = 0;3 >                          while (cnt < byte_count)t >                          {E >           if ((bcount = read (sock, data, byte_count - cnt)) == -1)k  >                              {< >                                      perror ("Read data");3 >                                      return (-1);.  >                              }< >                              printf ("Read %d\n", bcount);- >                              cnt += bcount;s >                          }A >                          printf ("Read %d bytes total\n", cnt);y! >                          break;/ >                  } >          } > }  >c > $ create server.c  > #include <errno.h> > #include <stdio.h> > #include <sys/types.h> > #include <sys/socket.h>a > #include <netinet/in.h>  > #include <ucx$inetdef.h> >f > intO2 > writevv (int fd, struct iovec * iov, int iovcnt) > { ) >          int total_bytes = 0, bytes, i;P >U' >          for (i = 0; i < iovcnt; i++)( >          {G >                  bytes = write (fd, iov[i].iov_base, iov[i].iov_len);v# >                  if (bytes == -1)i >                  {- >                          perror ("write:"); ' >                          return (-1);o >                  } >                  elseo >                  {0 >                          total_bytes += bytes; >                  } >          }  >          return (total_bytes); > }e >l > main (int argc, char ** argv)t > {. >. >          int serv_sock = -1;  >          int client_sock = -1; >          int len;  >e  >          int data_size = 1000; >-4 >          if (argc >= 2) data_size = atoi(argv[1]); > M >          if ((serv_sock = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP)) == -1)c >          {- >                  perror ("Create Socket:");gA >                  printf ("errno in CreateSock is %d\n", errno);2 >                  return(-1); >          } >.! >          len=sizeof(data_size);e > K >        if (setsockopt(serv_sock,SOL_SOCKET,SO_SNDBUF,&data_size,len) < 0)I >          {2 >                  perror("setsockopt SO_SNDBUF"); >          } >PK >        if (setsockopt(serv_sock,SOL_SOCKET,SO_RCVBUF,&data_size,len) < 0). >          {2 >                  perror("setsockopt SO_RCVBUF"); >          }> >          printf("Send/Receive buffer %d bytes\n",data_size); >n >          {1 >                  static struct sockaddr_in sin; < >                  sin.sin_family          = htons(AF_INET);3 >                  sin.sin_port            = 50000;s; >                  sin.sin_addr.s_addr = htonl(INADDR_ANY);  > B >                  if (bind(serv_sock, &sin, sizeof(sin))  ==  -1) >                  {@ >                          printf("cannot bind socket to port");, >                          perror ("Bind:");& >                          return(-1); >                  } >          } >o- >          if (listen(serv_sock, 20)  ==  -1)p >          {5 >                  printf("cannot set listen(s,20)");y& >                  perror ("Listen:"); >                  return(-1); >          } >rA >          if ((client_sock = accept(serv_sock, NULL, NULL)) > 0)S >          {' >                  struct iovec iov[2];  >                  int count;$- >                  static char data[1000000];4 >D@ >                  if (data_size > 1000000) data_size = 1000000;9 >                  iov[0].iov_base = (char *) &data_size;m0 >                  iov[0].iov_len = sizeof(int);+ >                  iov[1].iov_base = &data;s. >                  iov[1].iov_len = data_size; >y8 >                  if (writevv(client_sock,iov,2) == -1) >                  {. >                          perror ("writev:");' >                          return (-1);  >                  } >                  sleep(10);u >          } > }  > $ server:=$u:[xxxx]server. > $ client:=$u:[xxxx]clientr > $ spawn/nowait server 65536p > $ client my-node-name  >e >r > ? > Any comments ?  Any suggestions ?  I thinks this, at worst, a0< > bug in the Socket implementation, or, at "best", a serious  > omission in the documentation. >.= > Note, as mentioned above, this worked fine in UCX v3.3, buto! > not in UCX v3.3 ECO 3 or later.  >o > Many thanks in advance,t >t > Roy Omondr > Blue Bubble Ltd. >  >e   ------------------------------  % Date: Mon, 07 Mar 2005 16:56:21 +0000 - From: Roy Omond <Roy.Omond@BlueBubble.UK.Com>tY Subject: Re: TCP/IP - Sockets appear to be restricted to maximum 65,535 byte transfers tr , Message-ID: <393fihF5r7hk6U1@individual.net>   Dan Allen wrote:  L > Well - documentation aside - TCP sockets are byte streams. The size of theN > actual TCP packets are not necessarily related to the application read/writeP > sizes. Just read and write in chunk sizes of your choosing. The TCP layer will > packetize it at whim anyway.  8 Exactly, and that is what appears to be *not* happening.  A I try a write of 65536 bytes, and it generates the error message.e   I.e. it's broken.    ------------------------------  * Date: Mon, 7 Mar 2005 14:58:16 +0000 (UTC)- From: klewis@OMEGA.MITRE.ORG (Keith A. Lewis)7% Subject: tuning NFS for single-threadt. Message-ID: <d0hq68$j3b$1@newslocal.mitre.org>  ? I'm getting RMS timeout messages when reading files over NFS.  S   The setup:    ; Alphastation 500/400, built-in 10baseT ethernet, NFS clientn2 PWS 500au, built-in 100baseTX ethernet, NFS client1 PC, Fedora Core 1, 100baseTX ethernet, NFS server0  9 The alphas are running VMS 7.3-2, TCPIP 5.3, clustered.  0  K Only the 500 is having a problem.  It can read MSCP-served disks OK without-E timeouts but not NFS ones.  The 500au can read NFS just fine.  What IuE suspect is happening is that my 10/100 switch is dropping NFS packets.< because the 500's slow, half-duplex interface can't keep up.  K Is there a way to tune NFS so that it won't put out multiple requests, even>H for a large file, and that the expected returning request size fits in a single ethernet packet?   G My app requires a transfer rate of 512 Kbps, only 5% of the theoreticalu max.  D I know a workaround is to have the 500au transfer the files and thenJ MSCP-serve them to the 500, but I'd really like to get both alphas talking to the Linux server.  0 --Keith Lewis              klewis {at} mitre.org> The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Mon, 07 Mar 2005 23:22:33 +1030s* From: Mark Daniel <mark.daniel@vsm.com.au>( Subject: Re: VMS Alpha 8.2 with CSWS 2.0- Message-ID: <422c4ec9@duster.adelaide.on.net>p   Kari Keronen wrote:t3 > "Tom Linden" <tom@kednos.com> kirjoitti viestissc$ > news:opsmww9xn6zgicya@hyrrokkin... > F >>On Mon, 28 Feb 2005 14:56:05 GMT, Nigel Barker <nigel@hp.com> wrote: >> >>K >>>On Mon, 28 Feb 2005 13:58:10 -0000, "Richard Brodie" <R.Brodie@rl.ac.uk>r	 >>>wrote:g >>>  >>>s< >>>>"Kari Keronen" <kari.keronen@digita.fi> wrote in message7 >>>>news:QyEUd.1282$cg3.1241@reader1.news.jippii.net...  >>>> >>>>F >>>>>KAJAVA$DKA0:[SYS0.SYSCOMMON.APACHE.][000000]APACHE$APR_SHRP.EXE;1@ >>>>>-SYSTEM-W-SYSVERDIF, system version mismatch; please relink >>>>>t >>>>>Known problem or what ? >>>>G >>>>Apparently, something in the Apache port required some system level- >>>>hacking, so I guess1G >>>>you're out of luck until there is a new release. Someone else mightw >>>>have the inside track onH >>>>this... but obviously, porting the 1.3 stream was a higher priority. >>> J >>>CSWS 2.0 does not support OpenVMS 8.2. You'll need to wait for CSWS 2.1 >>>if'J >>>you're running OpenVMS 8.2. This should ship or rather be available for >>>downloadG >>>Real Soon Now.r >>! >>Have you tried installing WASD?M >  > K > No, but if someone has tried WASD on OpenVMS 8.2 Alpha/ODS-5, I'd like toe > know the results.i- > I don't currently have time to do the test.   9 Just installed and built it OK after a WASD-clean reboot.e   KLAATU$ cc /ver ' Compaq C V6.5-001 on OpenVMS Alpha V8.2, KLAATU$ show default    DKA100:[HT_ROOT]h LAATU$ show dev dka100/fullo  J Disk $1$DKA100: (KLAATU), device type DEC RZ28D, is online, mounted, file-D      oriented device, shareable, served to cluster via MSCP Server, 
 error logging       is enabled.
 8< snip 8<I    Volume Status:  ODS-5, subject to mount verification, file high-water t marking,"        write-back caching enabled.  2 Will report on I64 V8.2 when it becomes available.  F +--------------------------------------------------------------------+E   Mark Daniel                         http://wasd.vsm.com.au/adelaideVF   mailto:Mark.Daniel@wasd.vsm.com.au (Mark.Daniel@dsto.defence.gov.au);   A pox on the houses of all SPAMers.  Make that two poxes.nF +--------------------------------------------------------------------+   > -Kari- >  >  >>>--s >>>Nigel Barkere" >>>Live from the sunny Cote d'Azur >> >> >> >>-- tE >>Using Opera's revolutionary e-mail client: http://www.opera.com/m2/h >  >  >    ------------------------------  % Date: Mon, 07 Mar 2005 08:08:47 -0700 ' From: John Willis <willisjo@zianet.com>  Subject: VMS startup problem? / Message-ID: <BE51BC8F.11EC%willisjo@zianet.com>R   Hello,  C Having a bit of a problem starting VMS 5.1. It seems that it begins-D execution of the startup procedure, but then it hangs. I'm wonderingC if there's a way to bypass the startup procedures so I can find outg( what's wrong. Here's the messages I get:     KA630-A.V1.3   Performing normal system tests.f     7..6..5..4..3..e   Tests completed.     Loading system software.     2..1..0..s    E    VAX/VMS Version V5.1     Major version id = 1 Minor version id = 0h ..K $!  Copyright (c) 1988 Digital Equipment Corporation.  All rights reserved.e    A The VAX/VMS system is now executing the system startup procedure.   8 %%%%%%%%%%%  OPCOM   7-MAR-2005 08:07:07.89  %%%%%%%%%%%/ Logfile has been initialized by operator _OPA0: / Logfile is SYS$SYSROOT:[SYSMGR]OPERATOR.LOG;769e                    tD %LICENSE-I-LOADED, DEC FORTRAN was successfully loaded with 50 units    E And then nothing. Almost looks like a system ready to be logged into,rC yet pressing Enter does nothing. I know the terminal emulator works C because I am able to halt the system and type commands successfully0 at the >>> prompt.     Thanks in advance! John   ------------------------------   Date: 7 Mar 2005 07:47:12 -0800 $ From: "AEF" <spamsink2001@yahoo.com>! Subject: Re: VMS startup problem?tB Message-ID: <1110210432.891686.55330@g14g2000cwa.googlegroups.com>   John Willis wrote: > Hello, >5E > Having a bit of a problem starting VMS 5.1. It seems that it beginscF > execution of the startup procedure, but then it hangs. I'm wonderingE > if there's a way to bypass the startup procedures so I can find outu* > what's wrong. Here's the messages I get: >d >. > KA630-A.V1.3 >l! > Performing normal system tests.a [...]i > G > And then nothing. Almost looks like a system ready to be logged into,fE > yet pressing Enter does nothing. I know the terminal emulator worksyE > because I am able to halt the system and type commands successfully  > at the >>> prompt.    	 >>>b/r5:1a  C will perform a conversation boot (use lower- or uppercase). It willp0 give you a SYSBOOT prompt. From there you can do   SYSBOOT> SET STARTUP_P1 "MIN"a  G to perform a minumum boot which will skip SYSTARTUP_V5.COM and probablyeF a couple of the other startup command procedures. If that works, checkE the size of your pagefile. Maybe it's too small. (Just a wild stab in / the dark; it could be other things, of course.)   E What is the recent history of this system? Did it suddenly break? Did;9 you buy it and this you have not successfully booted yet?n  C Be sure to set STARTUP_P1 back to "" when you've fixed the problem!g   ------------------------------   Date: 7 Mar 2005 07:51:47 -0800 $ From: "AEF" <spamsink2001@yahoo.com>! Subject: Re: VMS startup problem?wB Message-ID: <1110210707.252629.28730@o13g2000cwo.googlegroups.com>  
 AEF wrote: > John Willis wrote:
 > > Hello, > >aG > > Having a bit of a problem starting VMS 5.1. It seems that it beginsh> > > execution of the startup procedure, but then it hangs. I'm	 wonderinglG > > if there's a way to bypass the startup procedures so I can find outl, > > what's wrong. Here's the messages I get: [...]l   > >>>b/r5:1c >dE > will perform a conversation boot (use lower- or uppercase). It will-2 > give you a SYSBOOT prompt. From there you can do >w > SYSBOOT> SET STARTUP_P1 "MIN"i > @ > to perform a minumum boot which will skip SYSTARTUP_V5.COM and probablyB > a couple of the other startup command procedures. If that works, checknG > the size of your pagefile. Maybe it's too small. (Just a wild stab in-1 > the dark; it could be other things, of course.)T   Oops! Do   SYSBOOT> CONTINUEj  D to continue after running SYSBOOT commands. You can also run HELP at the SYSBOOT prompt.5   >8G > What is the recent history of this system? Did it suddenly break? Didr; > you buy it and this you have not successfully booted yet?d >tE > Be sure to set STARTUP_P1 back to "" when you've fixed the problem!e   ------------------------------  # Date: Mon, 07 Mar 2005 17:40:13 GMT # From: hoff@hp.nospam (Hoff Hoffman)c! Subject: Re: VMS startup problem? 1 Message-ID: <1m0Xd.1149$lw1.231@news.cpqcorp.net>2  i In article <1110210707.252629.28730@o13g2000cwo.googlegroups.com>, "AEF" <spamsink2001@yahoo.com> writes:K :2 :AEF wrote:h :> John Willis wrote:w  0 :> > Having a bit of a problem starting VMS 5.1.  J   V5.1?  Really?  That predates the ANSI X3.27-1987 work -- this was whereH   the ANSI standards were revised in preparation for Y2K -- and there isJ   a Y2K limit within the V5.1 magtape ACP and INITIALIZE, and AFAIK withinK   other releases prior to V5.1-1.  We were compliant with then-current ANSIhJ   standards, and we tracked the 1987 standards update with changes made in,   (as it was then known) VAX/VMS circa 1988.  F   The fix is to upgrade to at least V5.1-1 or (better) to upgrade to aK   release with Prior Version Support (PVS) or with Current Version Support;-H   to OpenVMS VAX V5.5-2 or (better) to OpenVMS VAX V7.3.  (Or to replaceJ   the VAX box with an Integrity server and OpenVMS I64, but I digress. :-)     :> > It seems that it begins? :> > execution of the startup procedure, but then it hangs. I'm  :> > wonderingH :> > if there's a way to bypass the startup procedures so I can find out- :> > what's wrong. Here's the messages I get:h :[...] :a :> >>>b/r5:1 :>F :> will perform a conversation boot (use lower- or uppercase). It will3 :> give you a SYSBOOT prompt. From there you can doe :>  :> SYSBOOT> SET STARTUP_P1 "MIN" :>J :> to perform a minumum boot which will skip SYSTARTUP_V5.COM and probablyI :> a couple of the other startup command procedures. If that works, checkeH :> the size of your pagefile. Maybe it's too small. (Just a wild stab in2 :> the dark; it could be other things, of course.)  F   Pagefile-related problems usually engender squawking at the console,E   FWIW -- an attempt to generate pagefile critical (PAGECRIT) errors,j   etc., is made.  	 :Oops! Do  :  :SYSBOOT> CONTINUE     Oops!  :-)  Do:t  !     SYSBOOT> SET WRITESYSPARAMS 0      SYSBOOT> CONTINUE   E   Or reference the materials in the FAQ, which has the sequence for a E   conversational bootstrap, and a general explanation of license- and -   startup-related troubleshooting processing.m  9   Updates to the FAQ are welcome; I've got it open again.t  E :to continue after running SYSBOOT commands. You can also run HELP atc :the SYSBOOT prompt. :e :>H :> What is the recent history of this system? Did it suddenly break? Did< :> you buy it and this you have not successfully booted yet?  G   The usual cause for the underlying problem in my experience is a nodeaI   name change (SCSNODE system parameter) within recent times; this causesrE   various of the queue manager commands to hang -- something which is K   arguably a bug -- as the connection cannot be established to the previousfJ   name of the node -- which now looks like another node within the cluster"   -- to start the specified queue.  >   To avoid this, I usually add an /ON qualifier onto the queueF   INITIALIZE/QUEUE/START command, and explicitly specify the (current):   node name of the local node as the target for the queue.  H   FWIW and yes, the FAQ has details on changing the node name, too.  :-)  F :> Be sure to set STARTUP_P1 back to "" when you've fixed the problem!  F   If y'all follow the sequence in the FAQ, you get this automatically.E   You won't need the reboot either, as the sequence shown in the FAQ eG   can allow you to get a fully running system with the standard startuptE   -- after making the correction, of course -- without having to make :   the second reboot to get the environment back to normal.    N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faqtN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comc   ------------------------------  $ Date: Mon, 7 Mar 2005 10:56:07 -0500# From: "John Smith" <a@nonymous.com>r! Subject: [OT]: AMD virtualization , Message-ID: <Ko2dnSEEPtwH5LHfRVn-2Q@igs.net>  ; http://news.zdnet.com/2100-3513_22-5600552.html?tag=nl.e589    --  F OpenVMS    - The often imitated but never advertised operating system.  K R.I.P. Alpha - I could have had an EV8 but all I got was this lousy Itanic.m   ------------------------------  * Date: Mon, 7 Mar 2005 14:42:21 +0000 (UTC)6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)7 Subject: [TCPIP, OpenVMS] Problem with scaleable kernelo1 Message-ID: <newscache$lukzci$58b1$1@news.sil.at>9   If I understood correctly, then   4 1) TCPIP V5.4 is not supported on OpenVMS Alpha V8.26 2) TCPIP V5.5 is not supported on OpenVMS Alpha V7.3-2/ 3) TCPIP V5.4 has a scalable kernel in addition N (enabled with $ DEF/SYS TCPIP$STARTUP_CPU_IMAGES "PERF=ALL" in SYLOGICALS.COM). 4) TCPIP V5.5 has only the new scalable kernelF 5) Application xyz unfortunately doesn't work with the scalable kernel  D This means, Upgrade to OpenVMS Alpha V8.2 is not possible. Correct ?< If not correct, eg. how to enable old kernel on TCPIP V5.5 ?   TIAe   -EPLAN  H PS: I do know, that TCPIP V5.5 ECO1 fixes the SSH $ in unsername problem@ (already fixed in TCPIP V5.4 ECO4). When is V5.5 ECO1 expexted ? -- o Peter "EPLAN" LANGSTOEGERa% Network and OpenVMS system specialistg E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------   End of INFO-VAX 2005.132 ************************