1 INFO-VAX	Fri, 17 Mar 2006	Volume 2006 : Issue 151       Contents:C Re: Can I use DCPS LPD temporary files as a print to file solution? C Re: Can I use DCPS LPD temporary files as a print to file solution? C Re: Can I use DCPS LPD temporary files as a print to file solution? 0 Re: Device Naming: (was OPA0 Console connection)1 DS15, Radeon 7500, OpenVMS 7.3-2 DECWindows Crash  Re: GLIB and GTK for OpenVMS ? Re: GLIB and GTK for OpenVMS ? Re: GLIB and GTK for OpenVMS ?1 Re: Image Backup and Restore of Bound Volume sets / Re: Intalation of OpenVMS Alpha V7.3-1 question * Re: Is setting NIC duplex a dynamic thing? Re: LSI Logic 6200U3LP-S% Mail Utility - Logical Names It Needs ) Re: Mail Utility - Logical Names It Needs ) Re: Mail Utility - Logical Names It Needs ) Re: Mail Utility - Logical Names It Needs  Re: NIS under VMS ( OpenVMS, Itanium, and Video Card Support Test% Re: [F$GETQUI] Please explain RAD ;-)   F ----------------------------------------------------------------------  # Date: Thu, 16 Mar 2006 18:58:07 GMT  From: d_gillbilly@hotmail.com L Subject: Re: Can I use DCPS LPD temporary files as a print to file solution?8 Message-ID: <glaj121siau2m44v8obk45lkit1qnghgs1@4ax.com>   Thanks Paul for this info.  F On Thu, 16 Mar 2006 18:06:21 GMT, Paul Anderson <paul.anderson@hp.com> wrote:  9 >In article <8k3j12hosd4magp6bk1tgrb0bkfccpi8dq@4ax.com>, ! ><d_gillbilly@hotmail.com> wrote:  > E >> Reading the DCPS system manglers guide v2.5 (Nov 2005), I find out H >> that a temporary file can get created when spooling to a LPD queue. IG >> suspect that the temporary file would be postscript that I could use  >> (in the short term).  >  >...snip...  > E >> I'm thinking that I could create a stopped LPD que (that points at 9 >> nothing?) and dig my file out of DCPS$SPOOL_DIRECTORY.  > G >A stopped LPD queue won't work because DCPS won't create the temporary   >file until the job is printing. > D >You might get closer to what you want if you define the unsupportedH >logical name DCPS$SPOOL_KEEP.  This will keep the temporary file in theG >DCPS spool directory after the job is done printing.  Now if you don't H >want the file to print, specify a remote queue instead of a printer and >don't start the remote queue. >   E I'm glad I didn't start on the LPD side first. I may try this anyway. C Even if it's only to look at the generated PS. Then I could compare $ the size of the resulting PDF files.    J >> Is it possible to set up a DCPS queue so that the output goes to a disk' >> file rather than a physical printer?  >>   >> The Answer is : >>  K >>   This print-to-file capability is being considered for a future release 
 >>   of DCPS.  > D >This is still true.  Print-to-file is being considered for a future@ >release.  It's closer than it used to be since we now, with LPDH >spooling, have the mechanism to make such a file.  The question now is,  >how should a user request this    Ick. I never thought that far.B If it needs to be an execution que, use P2? (raw_post, raw_pcl...)2 If it needs to be a generic que, P3? DATA=RAW_POST; Or maybe a brand new entity that just looks like a printer?   @ Unless the PS generated is specific for a printer, I don't thinkE a printer is even required. But if it wasn't a printer, would DCPS be B the appropriate solution? (only because they seem to have the only native VMS PS converter :-(   $ >and what should the file be called? >   F My first thought would be x:file_name.PS or x:file_name.PCL, but I canF see this causing problems. Printing x:file_name.ps to a raw_post queue' would cause me to lose my source file.    : Maybe a logical DCPS$RAW_PS_FILE? (As long as it's not x:)D We use a construct like this so that the print to file option of ourE print routine has a scratch work area. I can then intercept this file G and do some post select, pre print processing. But ultimately, the user   specifies the output file name.   G Needing to specify the input and the output name would require a change 8 to our print routine (but that's what I'm doing anyway).  - Seemingly simple sometimes isn't always so...   0    Thanks for helping me see the bigger picture,       Duane    >Paul    ------------------------------  % Date: Thu, 16 Mar 2006 17:48:57 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> L Subject: Re: Can I use DCPS LPD temporary files as a print to file solution?, Message-ID: <4419EB4A.42B0404D@teksavvy.com>   d_gillbilly@hotmail.com wrote:B > Unless the PS generated is specific for a printer, I don't thinkG > a printer is even required. But if it wasn't a printer, would DCPS be D > the appropriate solution? (only because they seem to have the only > native VMS PS converter :-(     & If you have fairly simple ascii files:  ? CONVERT/DOC myfile.txt/format=TEXT  myfile.ps/format=POSTSCRIPT   & (not sure if it is "TEXT" or "ASCII").  0 But I don't think it does ANSI escape sequences.  F BTW, when using the term "form" you need to qualify this. VMS uses theE term "form" to specify a group of print parameters applied to a job (  SHOW QUEUE/FORM/FULL    G And in Postscript a "form" is  Postscript level 2 feature that lets you F define a postscript function to draw a form, and it is only rasterizedG once, and subsequent calls to that form are much faster because it uses  an in-memory image of the form.     G And in the "real" world, a form refers to a page with pretty logos  and  boxes with data in them :-)    ------------------------------  # Date: Fri, 17 Mar 2006 01:48:04 GMT ; From: "Jeffrey H. Coffield" <jeffrey@digitalsynergyinc.com> L Subject: Re: Can I use DCPS LPD temporary files as a print to file solution?> Message-ID: <ozoSf.54598$H71.42385@newssvr13.news.prodigy.com>   d_gillbilly@hotmail.com wrote: > Hello fellow VMSer's,  > ? >    I am currently developing a proof of concept for customer. A > Instead of printing a document, they want to email it. By using E > a little DCL, lib$spawn, ghostscript, mime and sff, I can send out, C > as PDF's, a very high percentage of my output. I can even extract B > setup files from POST.TLB, marry it with an ANSI data stream andC > email single pages. But I can't seem to turn the page (it creates A > corrupted PDF's). It could be a problem with ghostscript or the @ > coding of the postscript in the library (but I think I'm doing> > something to corrupt a dictionary stack when I marry the twoE > data streams). Then it occurred to me that I'm trying to duplicate  D > the functionality of DCPS, so instead of re-inventing yet another J > wheel, I decided to see if I could capture the output from a DCPS queue. >   G I do a lot of this sort of thing for my customers using Ghostscript to  G convert ps to pdf. The earlier versions of Ghostscript have significant C problems creating pdf files. I use version 8.51 (which I have also  I ported to Itanium). If you send me a sample ps file and it's setup file,  ; I will take a look at it to see if the problems are simple.   
 Jeff Coffield    ------------------------------  # Date: Thu, 16 Mar 2006 22:27:44 GMT   From: John Santos <john@egh.com>9 Subject: Re: Device Naming: (was OPA0 Console connection) * Message-ID: <ADlSf.8670$Jz4.2534@trnddc08>   JF Mezei wrote:  > David J Dachtera wrote:  > C >>Hhmmm... which do we ASSIGN as the equivalence string of the $CD, + >>$CD_ROM, etc.logical name? ...$DVD, etc.?  >  > % > Each site should do this manually.   > G > It is much easier than playing with some "plug and play" utility that  > just doesn't cooperate.  >  > I > What could be done is to procide with a prepupulated but system manager F > updateble RMS file that maps various device names with a class name.P > (DISK, CD, REMOVABLE, SCANNER etc). So "SEAGATE ST31055N" would map to "DISK". > I > You plug in a new obscure device, you can add that device's name to the C > table and the system would then recognize it as the proper class.  >  > H > You could then build your own logicals by scanning the devices and theB > first "CD" enountered gets "CD1", the second one gets "CD2" etc.  C My systems (work, personal, customer systems we manage) use logical G names for all disks and tape drives, usually things like "CD0:", "CD1:" H for the CD drives, etc., defined with /tran=concealed at system startup,C so the users don't see the hardware names, and if a system reconfig H changes them we just need to change one command file.  This works prettyH well, except in a few cases where the hardware names intrude and confuseG people.  This is in places like ANA/ERROR, DIAGNOSE, SHOW DEVICE, etc., 5 and in OPCOM messages (Shadowing, mount-verify, etc.)   C I think a useful and fairly easy way to deal with this is to have a D special logical name table (e.g. the "canonical device name" table),H with reverse translations of the device logical names.  This table wouldB *NOT* be part of the normal logical name translation process, justE sitting out there to be looked at by interested parties.  It would be H populated by the system manager, site-specifically with what ever namingF conventions the system manager likes.  Things that know about hardwareG names (ANALYZE/ERROR, DIAGNOSE, SHOW DEVICE, F$GETDVI, F$DEVICE, OPCOM, H etc. could be modified to translate the hardware name (e.g. $47$DKA300:)C to "CD1:" by looking in this table.  (This could be controlled by a D qualifier, for example "SHOW DEVICE/NAME=CANONICAL DK" would displayE the canonical translation if it exists instead of the hardware name.)   B To make this as easy as possible (and minimize typos), a qualifierB could be added to ASSIGN and/or DEFINE to populate both the normal! table and the reverse name table.   5 $ ASSIGN/SYSTEM/TRAN=CONC TAPE1: $2$MKA400:/CANONICAL   > would define the logical name TAPE1 in the system table (as it? does now), but would also define the logical name $2$MKA400: in + the canonical name table pointing to TAPE1.   A It would be up to the system manager to ensure there was a 1 to 1 > mapping of the canonical names to real device names.  It could= error out if $2$MKA400: was already defined, or it could just   replace with the new definition.     --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------    Date: 16 Mar 2006 16:04:08 -0800' From: "mjjerabek" <mjjerabek@gmail.com> : Subject: DS15, Radeon 7500, OpenVMS 7.3-2 DECWindows CrashC Message-ID: <1142553848.594771.107430@z34g2000cwc.googlegroups.com>   > I have an OpenVMS V7.3-2 system running on a DS15 running with6 DECwindows 1.3-1 with the following patches installed:   Graphics 400
 Update 500 DWMotif ECO2  F We run a custom HMI application that when poked at properly will crashC the X-server with a single mouse click. I can also achieve the same E crash using the same VMS system, and a Radeon card running in a DS10. G With the DS10, if I replace the Radeon card with a different video card ? (PowerStorm 4D10T for example), the problem disappears. The HMI % application is a 2D only application.   B In the past year I have worked with the DECwindows folks to detectA these types of crashes which have resulted in fixes placed in the & GRAPHICS 300 and GRAPHICS 400 patches.  G This new crash most likely related to a failure doing memory allocation G within the Radeon driver with an improper error response (just like the A last problem I had). I have several questions for this newsgroup:   E 1) Anyone else seeing problems like this, especially after installing * the latest patches for VMS and DECwindows?  D 2) I have been looking for a consistent list of sysgen and authorizeG quotas and parameters. After searching the online documentation, I have B found that nothing has changed in the DECwindows release notes andF installation documentation for these quotas for 5-10 years. Any of youG DECwindows/VMS engineers listening to this group care to supply updates ? for this information, or an updated DECwindows tuning document?   F 3) I have been told that VMS 8.2+ and DECwindows 1.5 do not have theseF same problem with the Radeon card. We intend to move up to VMS 8.2 andE DECwindows 1.5 when we do our port to Itanium. My fear is that I will D have similar nightmares with DECwindows and the Radeon cards in this$ new environment. Are my fears valid?  B I am now in the clutches of DECwindows and VMS engineering with myG Radeon problem. Any and all help from newsgroup members will be greatly  appreciated.  
 M. J. Jerabek  HSQ Technology   ------------------------------    Date: 16 Mar 2006 22:21:05 +01006 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)' Subject: Re: GLIB and GTK for OpenVMS ? , Message-ID: <4419e4d1$1@news.langstoeger.at>  ` In article <44184663$0$8735$9b622d9e@news.freenet.de>, "Klaus-D. Bohn" <info@it-bcsb.de> writes:J >does everybody tried to move these libraries (www.gtk.org) to the OpenVMS
 >systems ?+ >I mean the new version (2.x), not the old.   + Have you tried contacting VMS engineering ? * They did provide GTK V1 (for/via MOZILLA).> Maybe there is a GTK V2 underway and we don't know it (yet)...   just my 0.02   --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  # Date: Thu, 16 Mar 2006 23:37:37 GMT 7 From: John Malmberg <malmberg@dskwld.zko.hp.compaq.dec> ' Subject: Re: GLIB and GTK for OpenVMS ? 2 Message-ID: <5FmSf.4822$RL6.1262@news.cpqcorp.net>   davidc@montagar.com wrote:C > Actually, I have some software written for the 1.2 GTK libraries. F > Also, there are some really nice desktop applications (like GnuCash)0 > that use GTK, that would make a nice VMS port. > I > Since there has been talk about desktop apps on OpenVMS, this is a good F > port for OpenVMS.  GTK is used for many of the Gnome applications onI > Linux, and of course, the GIMP.  It's an excellent toolkit for building I > X-Window apps (or even WIN32 apps if you use the GTK Windows port, I've I > done it).  Glade is a nice tool for doing the layout and generating the G > baseline code generation.  It's not a full IDE, but it's very easy to J > build a multiple dialog GUI application framework in pretty short order.  F I was looking for the names of specific applications that the lack of H GTK+ 2.x is the only thing preventing them from being ported to OpenVMS ) and what their commercial value would be.    -John ! malmberg@dskwld.zko.hp.compaq.dec  Personal Opinion Only    ------------------------------  # Date: Fri, 17 Mar 2006 00:15:04 GMT 7 From: John Malmberg <malmberg@dskwld.zko.hp.compaq.dec> ' Subject: Re: GLIB and GTK for OpenVMS ? 2 Message-ID: <ccnSf.4825$dQ6.4196@news.cpqcorp.net>   JOUKJ wrote: > Klaus-D. Bohn wrote: > A >> Yes, i know. But i found only the old and not a newer version.  >>K > But it is there see : http://nchrem.tnw.tudelft.nl/openvms/software1.html  > and look for the gtk2+ item.  D I would not consider Linker warnings to be something that should be F ignored.  It usually is an indication that something serious is wrong  with the build procedure.   K > Note that it is realy "alpha" and not finished. I got some tests running  G > correctly but only on screens which ghave the Xrender Extenstion. So  4 > hardly anything works yet on a native VMS display.  F So far I have not found anything that requires the Xrender Extension, G but I have not yet tried building anything that requires glib 2, since  2 my build of glib 2.x is not yet somewhat "stable".  ? There are 59 tests in glib 2.8.6, and I do not see any of them   attempting to do X-11 output.   I Of the 59 tests, 4 will not run on OpenVMS 8.2 for the following reasons:   G Test child-test fails because it needs a UNIX style fork() and can not  " be quickly changed to use vfork().  G Test convert-test fails because neither OpenVMS or the I18N for it kit  I supply converters for the CP1255 character set, and I do not know of any  9 alias names for it that might be present in the I18N kit.   C Test file-test fails because it is testing for symbolic links, and  G unless you are on the 8.3 Eft or using the RMS SYMLINKS SDK, those are  
 not there.  D Test unicode-encoding fails because neither OpenVMS or the I18N kit E supply converters for the UTF-16 character set.  And again, I do not  7 know of any aliases for that set that might be present.   C I still have 2 tests failing, and 2 tests that are generating some  F diagnostics that I have not yet fully investigated.  The rest are now  passing for me.    In cairo 1.0.2:   E Test path-data is checks for the behavior of passing an out of range  A error status to the error handler and expecting the results of a  B previous version of cairo, and is not compatible with the current + version, which exits with an assert() call.   I Test ft-font-create-for-ft-face is broken in that it is selecting a face  H which does not exist and does not check the error, and starts passing a C truncated structure to other routines.  Eventually this causes bad   things to happen.   B I am seeing some tests in cairo failing from what looks like some / rounding issues on floating point calculations.   I So far none of these look like OpenVMS specific issues.  A google search  E is showing that LINUX on x86 does floating point precision different   than other platforms.    In gettext 0.14.5:  H The msgexec program is trying to use fork() in a way that will not work I on OpenVMS, and the msgfilter program is trying to set non-blocking I/O.  D   I have not looked at converting them to be able to run on OpenVMS.  I The full make check step takes about 36 hours on a 466Mhz DS10 with fast   wide SCSI disks.  F I am still working on resolving the differences that glib and gettext D have revealed between gnu libiconv and the iconv + I18N kit that is G supplied by HP.  Currently I have 14 tests failing out of the 227 test  H that were run, mostly because of missing character sets.  22 tests were F not run because the test procedure determined that the character sets # required for them were not present.   B > Also be prepared to install a lot of packages and do some (yet) I > undocumented configuration. If you have questions ask them me off line.   I The gnome prefix has not been registered with HP, so "gnome$" should not  G be used in logical names, symbols, or filenames.  Use "gnome_" instead.   F The use of $ in public program symbols, filenames, logical names, and E DCL symbols is reserved to HP or the registered owners of facilities. B See the OpenVMS FAQ for information on how to register a facility.  G And please who ever may be doing porting and getting your changes back  G into the main distribution sites, do not name MMS description files or  I DCL command files makefile.vms.  This breaks building the products under  E GNV as the GNV gnu make utility will try use makefile.vms instead of   makefile, and then fail.  . > All help making this port better is welcome.  H As near as I can tell, there appears to be at least 5 or more different H people or groups now independently working on porting the components of ? GTK+ 2.x or other required libraries to OpenVMS with almost no   coordination between them.  H Discussions of such activity would be on topic for the vmsnet.vms-posix F newsgroup, and announcements of milestones would be on topic for both  there and this forum.    -John ! malmberg@dskwld.zko.hp.compaq.dec  Personal Opinion Only    ------------------------------  + Date: Fri, 17 Mar 2006 00:26:53 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply): Subject: Re: Image Backup and Restore of Bound Volume sets$ Message-ID: <dvcvod$6et$1@online.de>  E In article <dv80ct$pmc$2@newslocal.mitre.org>, klewis@OMEGA.MITRE.ORG  (Keith A. Lewis) writes:    N > I never figured out how to do an image restore on a bound set.  Luckily, you > can do a non-image restore.     H You could do a BACKUP/PHYSICAL of each of the individual disks.  Should E be even more like the original than /IMAGE (which could be better or  ' worse, but in most cases won't matter).    ------------------------------    Date: 16 Mar 2006 22:16:15 +01006 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)8 Subject: Re: Intalation of OpenVMS Alpha V7.3-1 question* Message-ID: <4419e3af@news.langstoeger.at>  o In article <dvacc3$kou$1@nsnmrro2-gest.nuria.telefonica-data.net>, Paco Alguacil <paco.linux@gmail.com> writes: M >I have a question. I have installed VMS on an AlphaServer DS20 in a  9Gb HD, D >I have made the instalation default, and I added decnet phase V, C,H >fortran, ada and mozilla and I noticed that the disk is almost full. IsD >this normal? I think the OS and utilities must fit in 2Gb. Im doing( >something wrong (problably)? Any ideas?  O I've a 18GB disk and have only 11GB free (ok, I have APACHE/JAVA/MOZILLA/PERL). @ 2GB enough ? Maybe only for VMS. But not for other useful stuff.= But I admit, you should have at least 1-2GB more freespace...   G VMS is distributed in (at least partially) compressed kits. And running H a system means you have them uncompressed (for performance) and also youK have files on your disk which you don't have on your CD (like PAGEFILE.SYS, I SWAPFILE.SYS, SYSDUMP.DMP, SYS$ERRLOG.DMP and so on). And these files can I be big nowadays with todays memory consumption and memory in the systems.   L Consider also, you probably have multiple versions of these (and some other)I files because you surely didn't add a PAGEFILE=0 and SWAPFILE=0 into your J SYS$SYSTEM:MODPARAMS.DAT (and so the AUTOGEN resized them - built a newer,8 smaller one, without deleting/purging the previous one).  I Consider purging (start with a $ PURGE/CONFIG SYS$SYSDEVICE:[000000...]), N a bigger system disk (recommended), less layered products installed (shudder),J secondary (instead of primary) page/swap files, compressed system dump and' or dump file of system disk (DOSD), ...   C Check also, that you don't have SET PROCESS/DUMP entered somewhere, B or every failing image leaves a process dump file around (but this- is nowhere default and thus not very likely).    Good luck (and come again)   --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  # Date: Thu, 16 Mar 2006 20:08:26 GMT % From: Rick Jones <rick.jones2@hp.com> 3 Subject: Re: Is setting NIC duplex a dynamic thing? 2 Message-ID: <_AjSf.4799$OH6.4283@news.cpqcorp.net>  - Carl Friedberg <frida.fried@gmail.com> wrote: @ > When the interface is set to auto-configure, the specification@ > requires that, if there is a fixed speed at the other end, the1 > interface must confugure itself to half-duplex.   F > This means you will get a lot of errors, since one end is using fullE > duplex, but the other end is using half duplex. The only way to fix D > this is to set the NIC to full duplex, 100 mbps, and for that, you > need to use LANCP.    & Or change the other end to be autoneg.  E Don't forget that when/if you upgrade to Gigabit, all that hardcoding " has to be tossed-out the window...  $ The boiler-plate I use on the topic:    How Autoneg is supposed to work:  E When both sides of the link are set to autoneg, they will "negotiate" > the duplex setting and select full duplex if both sides can do full-duplex.  C If one side is hardcoded and not using autoneg, the autoneg process E will "fail" and the side trying to autoneg is required by spec to use  half-duplex mode.   E If one side is using half-duplex, and the other is using full-duplex, # sorrow and woe is the usual result.   E So, the following table shows what will happen given various settings 
 on each side:   +                  Auto       Half       Full   ,    Auto        Happiness   Lucky      Sorrow  ,    Half        Lucky       Happiness  Sorrow  /    Full        Sorrow      Sorrow     Happiness   C Happiness means that there is a good shot of everything going well. D Lucky means that things will likely go well, but not because you did@ anything correctly :) Sorrow means that there _will_ be a duplex
 mis-match.  D When there is a duplex mismatch, on the side running half-duplex youD will see various errors and probably a number of late collisions. OnA the side running full-duplex you will see things like FCS errors. F Note that those errors are not necessarily conclusive, they are simply
 indicators.     D Further, it is important to keep in mind that a "clean" ping (or theA like - eg "linkloop") test result is inconclusive here - a duplex F mismatch causes lost traffic _only_ when both sides of the link try toF speak at the same time. A typical ping test, being synchronous, one atF a time request/response, never tries to have both sides talking at the
 same time.  
 rick jones --  G oxymoron n, Hummer H2 with California Save Our Coasts and Oceans plates F these opinions are mine, all mine; HP might not want them anyway... :)D feel free to post, OR email to rick.jones2 in hp.com but NOT BOTH...   ------------------------------    Date: 16 Mar 2006 17:22:42 -06004 From: kuhrt.nospammy@encompasserve.org (Marty Kuhrt)! Subject: Re: LSI Logic 6200U3LP-S 3 Message-ID: <cl7aRLYnLbCE@eisner.encompasserve.org>   h In article <20060316142333.41b6271d.fishtank.spam@web.de>, Marc Schlensog <fishtank.spam@web.de> writes: > On 15 Mar 2006 15:07:42 -0600 7 > kuhrt.nospammy@encompasserve.org (Marty Kuhrt) wrote:  > @ >> I found a LSI Logic PCI U3 SCSI card, part number 6200U3LP-S.? >> Any chance of this working in an Alpha system like a DS10 or 
 >> a 500a? > ? > I don't think so. That -S is a hint that it's the version for I > Sparc/Solaris, which has, IIRC, a firmware different from the x86/Alpha  > model. >   ? Just to follow up.  I put this card in a 500a and the power on  ' console output looked something like...   B "What ever you put in slot four is not going to work.  Stop that!"  , That would seem to confirm Marc's statement.   ------------------------------    Date: 16 Mar 2006 14:06:50 -0800( From: "Craig Dedo" <Dedo.Craig@mayo.edu>. Subject: Mail Utility - Logical Names It NeedsB Message-ID: <1142546810.865857.68290@i40g2000cwc.googlegroups.com>  F What logical names need to be defined so that the OpenVMS Mail Utility works properly?   C This afternoon our OpenVMS Mail Utility stopped working.  If anyone . tries to send any mail, they get a message of:( %SYSTEM-F-IVLOGNAM, invalid logical name  C I tried looking through the OpenVMS technical documentation without F success.  There is very little information on the technical aspects of7 configuring the Mail Utility so that it works properly.    Craig T. Dedo, VMS Consultant $ Mayo Central Lab for Clinical Trials& Superior Drive Support Center 2-110-26 3050 Superior Drive NW Rochester, MN   55905-1700 Voice Phone:  (507) 538-4636 Fax Phone:    (507) 284-0615 Mobile Phone:  (414) 412-5869  E-mail:  <Dedo.Craig@mayo.edu>   ------------------------------  # Date: Thu, 16 Mar 2006 23:24:50 GMT # From: hoff@hp.nospam (Hoff Hoffman) 2 Subject: Re: Mail Utility - Logical Names It Needs2 Message-ID: <6tmSf.4819$oP6.2192@news.cpqcorp.net>  m In article <1142546810.865857.68290@i40g2000cwc.googlegroups.com>, "Craig Dedo" <Dedo.Craig@mayo.edu> writes: G :What logical names need to be defined so that the OpenVMS Mail Utility  :works properly?  C   That's kind of an open-ended question, unfortunately -- there are C   all sorts of system logicals that can cause problems with MAIL...     D :This afternoon our OpenVMS Mail Utility stopped working.  If anyone/ :tries to send any mail, they get a message of: ) :%SYSTEM-F-IVLOGNAM, invalid logical name   @   What is the sequence of DCL and/or MAIL commands leading up to
   this error?   >   What changes were made to the system proximate to the error?  0   What is the OpenVMS version, platform and ECO?  B   If this is a SEND to an SMTP address, I'd look at whether or notA   the TCP/IP Services environment (or the third-party IP stack in D   use) has been configured and started.   On OpenVMS V6.2 and later,@   this process is largely automatic when a compliant IP stack isC   installed -- on yet older releases, there were a few set-up steps    that were required.   C   The core logical name for configuration is MAIL$SYSTEM_FLAGS, but '   that should not cause this behaviour.   B   And given your location in Minnesota, you do need to ensure thatD   the logical names don't simply freeze up and crack -- we have that7   same sort of problem here in New Hampshire, too.  :-)     N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------G        Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[\0100]hp.com    ------------------------------  % Date: Thu, 16 Mar 2006 17:44:16 -0500 % From: BRAD <bradhamilton@comcast.net> 2 Subject: Re: Mail Utility - Logical Names It Needs* Message-ID: <4419EA40.6030201@comcast.net>   Craig Dedo wrote: H > What logical names need to be defined so that the OpenVMS Mail Utility > works properly?  > E > This afternoon our OpenVMS Mail Utility stopped working.  If anyone 0 > tries to send any mail, they get a message of:* > %SYSTEM-F-IVLOGNAM, invalid logical name   IVLOGNAM,  invalid logical name   % Facility:     SYSTEM, System Services   I Explanation:  A name string exceeds the maximum length permitted, or has  F a length of 0, or is longer than the specified maximum length logical F name string, table name string, or equivalence string.  Logical names D are limited to a length of 255 characters; global section names are H limited to 43 characters; common event flag, process names, and cluster I names are limited to 15 characters. An error can be returned when a name  F table is defined and the name for the new name table contains invalid  characters.   H User Action:  If the error occurs during command processing, verify the G command. If the error occurs during execution of a program, check that  G the character string descriptors pointing to name strings indicate the   correct lengths.  G (This is for V7.3-2; you didn't indicate which version you are using).  H It looks as though you're going to have to look at logical name lengths.   ------------------------------    Date: 16 Mar 2006 16:34:13 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) 2 Subject: Re: Mail Utility - Logical Names It Needs3 Message-ID: <V5GByj8qIO7O@eisner.encompasserve.org>   m In article <1142546810.865857.68290@i40g2000cwc.googlegroups.com>, "Craig Dedo" <Dedo.Craig@mayo.edu> writes: H > What logical names need to be defined so that the OpenVMS Mail Utility > works properly?  > E > This afternoon our OpenVMS Mail Utility stopped working.  If anyone 0 > tries to send any mail, they get a message of:* > %SYSTEM-F-IVLOGNAM, invalid logical name  F In my opinion, your problem is more likely to be an extraneous logical* name than a missing one.  Try the command:   	HELP/MESSAGE IVLOGNAM  E > I tried looking through the OpenVMS technical documentation without H > success.  There is very little information on the technical aspects of9 > configuring the Mail Utility so that it works properly.   D I doubt that you are running into an explicit configuration problem.C It is more likely an inadvertent logical name.  If you have another G system, compare contents of the system logical name tables.  (I presume D you have verified that this happens to all users and is not an issue! with a group logical name table.)    ------------------------------  # Date: Fri, 17 Mar 2006 00:41:58 GMT + From: Jack Patteeuw <jjpatteeuw@nospam.net>  Subject: Re: NIS under VMS< Message-ID: <qBnSf.57099$Jd.8875@newssvr25.news.prodigy.net>   tomarsin2015@comcast.net wrote: + > Is NIS avaiable for VMS either Alpha/VAX.  > thanks > phil >     C What problem are you trying to solve that makes you think NIS is a   possible solution ?   F Living in "both worlds" the only NIS "feature" that I wish existed in F VMS was netgroups for TCPIP communication proxies (so I don't have to A specify 35 hosts for each and every proxy; and don't suggest * !)    ------------------------------    Date: 16 Mar 2006 16:47:49 -0800' From: "mjjerabek" <mjjerabek@gmail.com> 1 Subject: OpenVMS, Itanium, and Video Card Support C Message-ID: <1142556469.947699.288160@z34g2000cwc.googlegroups.com>   E With last orders for Alpha based systems deadline approaching, we are B about to dive, head first into a port of a large VAX/Alpha OpenVMS= software system. From scanning this news group, and published E literature, it does not look to be a real big deal at all, but I have 
 some worries.   F During the past year I have had a hellish time working around problemsF running DECwindows using the currently supported Radeon 7500 card. OurB HMI application is only a 2D application, but I seem to be able to@ crash DECwindows at will doing simple, perfectly legal x-windows operations.   A With this port, I will have to deal with a new version of OpenVMS A (8.2+), DECwindows (1.5+), and a new set of video cards and their  drivers. My question is,  ? What video cards will be supported in the newer VMS and Itanium G environment. From the literature I believe that the current Radeon 7500 ? card, plus a Radeon 7000-like card are supported. Is this true?   1 What other cards will be supported in the future?   D Will any of these cards be using "current" technology instead of theA current trend of being 3-4 years behind in video card technology?   F In addition to the above questions, will I likely have the same driver? problems/crashes as I have had in my current Alpha environment?   > Any and all help from newsgroup members (Or OpenVMS/DECwindows' engineers) will be greatly appreciated.    ------------------------------  # Date: Thu, 16 Mar 2006 21:15:52 GMT ' From: Bob Edwards <Edwards@comcast.net> 
 Subject: Test < Message-ID: <Edwards-80929A.16154816032006@news.verizon.net>   Test   --   Test   ------------------------------    Date: 16 Mar 2006 22:36:39 +01006 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER). Subject: Re: [F$GETQUI] Please explain RAD ;-), Message-ID: <4419e877$1@news.langstoeger.at>  ` In article <4418DE60.9B7BE1BA@comcast.net>, David J Dachtera <djesys.nospam@comcast.net> writes:E >Based on what F$GETQUI() returns, I'm gonna have to go with the idea F >that RAD=0 is the default if not set explicitly and -1 = /NORAD until/ >someone pipes up with something authoritative.   H And I came to the conclusion, that F$GETQUI RAD doesn't provide the sameG information as SHOW QUEUE does. (I wonder what SYS$GETQUI will tell...)   9 I have a lot of queues which all provide 0 instead of -1. D And all of them are old queues. That means they haven't been createdD in the last couple of years. Every queue I so far created since thenG shows -1. And every queue I did a SET QUEUE/NORAD (which per definition G does nothing as the queue already is /NORAD) now also shows -1 for RAD.   L So, I assume, this is a field somewhere in the SYS$QUEUE_MANAGER.QMAN$QUEUESH file, which didn't get updated by the VMS upgrade which brought us /RAD.J And also F$GETQUI doesn't check the other field (which got upgraded) whichM SHOW QUEUE uses. Sounds bogus, but for me sounds more logical than your idea.   0 So, Guy/Hoff, please, your word is required. TIA   --   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   ------------------------------   End of INFO-VAX 2006.151 ************************