1 INFO-VAX	Sun, 01 Jun 2003	Volume 2003 : Issue 301       Contents:& Re: Burning CDs in ISO/Files-11 format& Re: Burning CDs in ISO/Files-11 format& Re: Burning CDs in ISO/Files-11 format% Re: Controller for VAXStation 4000 90  Re: Install Directory  Re: Install Directory 	 My AS2100  Re: Need OpenVMS (any version) Re: Need OpenVMS (any version)$ Re: Needed, history of startup files Re: Overclocking VAXen Re: Portents of VMS death  Re: Portents of VMS death  Re: Portents of VMS death E Re: Separating Development from Production data (was: Portents of...)   Re: StorageTek l700 on VMS 7.3-1, Re: SYS$QIOW - Wtih more Details - Final Try8 Re: Your Own Installation Tool? (was: Install Directory) [OT] best HTML code cleaner?  Re: [OT] best HTML code cleaner?  F ----------------------------------------------------------------------  % Date: Sat, 31 May 2003 23:47:13 +0100 ) From: Antonio Carlini <arcarlini@iee.org> / Subject: Re: Burning CDs in ISO/Files-11 format & Message-ID: <3ED930F1.5030304@iee.org>   VAXman- wrote:: > The instructions are still there:  www.tmesis.com/CDrom/ > I > These instructions do require duplication of the data on the CD.  Thus, J > only about 300MB of data can be placed on the CD.  The instructions makeH > used of freely available software tools too.  I toyed with the idea ofG > mapping the data to save space.  Others have done this (David North). , > I don't have a URL handy for David's work.  1 The early version of David's ODSISO tool is here: .    http://support.tditx.com/~odsiso/index.html  7 There is a later version (v1.1 IIRC) that he pointed me 7 at when I mailed him about this - this was at least two  years ago, maybe longer.  4 It does exactly what it says on the tin (i.e. if you/ want 700MB visible as ODS-2 and ISO9660, that's 0 what you get - assuming the files are suitable).  6 With some manual mangling you can also have files that4 are visible only under ODS-2 and only under ISO9660. Bootable CDs worked for me too.   7 The only confusing bit initially was that "Alpha" meant  "post-VAX" and not "pre-beta".  : As for internal DEC tools that did this - if they exsited,7 noone was willing to admit to it when I asked. I didn't ; see much possibility for unreliability with ODSISO (barring 5 program bugs, of course) since you basically build an 8 ODS-2 disk, then find where all the files live in blocks5 and build a set of ISO9660 structures that map to the  required files.    Antonio    --   --------------- - Antonio Carlini             arcarlini@iee.org    ------------------------------  # Date: Sat, 31 May 2003 23:33:00 GMT # From: hoff@hp.nospam (Hoff Hoffman) / Subject: Re: Burning CDs in ISO/Files-11 format 2 Message-ID: <MYaCa.1696$Qx4.1170@news.cpqcorp.net>  R In article <3ED930F1.5030304@iee.org>, Antonio Carlini <arcarlini@iee.org> writes:  ; :As for internal DEC tools that did this - if they exsited, 8 :noone was willing to admit to it when I asked. I didn't< :see much possibility for unreliability with ODSISO (barring6 :program bugs, of course) since you basically build an9 :ODS-2 disk, then find where all the files live in blocks 6 :and build a set of ISO9660 structures that map to the :required files.    D   I'm presently expecting to be working in this area Real Soon Now, B   for reasons not particularly germane to the current discussion.   C   No commitments nor announcements nor technical details here, etc. 
   (Sorry.)  N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com    ------------------------------  # Date: Sun, 01 Jun 2003 03:52:38 GMT 8 From: "Jerome H. Fine" <jhfineb9rv@b9rvnospamcompsys.to>/ Subject: Re: Burning CDs in ISO/Files-11 format 4 Message-ID: <3ED97874.A88DB9C2@b9rvnospamcompsys.to>   >Larry Kilgallen wrote:   d > > In article <BAF9515C.8DE1%JCam90502@jcameron.com>, Jeff Cameron <JCam90502@jcameron.com> writes: > K > > Is it possible to burn CD's in a dual ISO/Files-11 format so that files J > > retain their VMS record attributes when mounted on a VMS system and beP > > readable by PCs and UNIX systems? (MACintosh would also be desiable, but not > > a requirement). H > I believe Brian Schenkenberger has posted instructions on his web siteJ > for doing that but not sharing data.  That is, there are separate copies. > of the data for ODS and ISO directory trees.   Jerome Fine replies:  < I don't know if this information will help since my goal was% dual ISO9660 / RT-11 file structures.   7 In addition, I also managed to be able to NOT duplicate ? the actual files, ONLY the 2nd RT-11 file structure was needed.   : (a)  I first burned 4 sample files under Windows 98 SE and;       Nero Burning to a CD-RW (I did not want to waste a CD 7       so I used a CD-RW - this was only the first step)   A (b) I copied the CD image to a file on the hard drive.  Note that E       while Nero Burning is able to copy the contents of a file which A       is a CD image to a CD, the opposite is not possible.  In my ?       case, I mounted the CD under the RT-11 emulator Ersatz-11 E       along with an empty file on the hard drive that was the correct ;       size and used the DUP program in RT-11 to copy the CD #       image to the hard drive file.   B (c)  By searching through the ISO9660 file structure, I found thatF       the first file started at block 100 (sector 25) on the CD image.D       All files started on a sector boundary (multiple of 4 blocks).=       The IS09660 file structure occupied blocks 64 to 99 (or "       sectors 16 to 24 on the CD).  A (d)  Since the first file was V5.03 of RT-11, I copied that RT-11 H       file structure to block 0 through block 11 on the hard drive file.C       In addition, I changed the boot device from an RL02 using the >       device driver DL.SYS with the monitor file RT11AI.SYS to:       the MSCP device driver DUX.SYS with the monitor file<       RT11XM.SYS so that it would be more convenient to boot>       the CD on a real DEC  PDP-11 hardware with a SCSI  CDROM       drive.  ? (e)  I also added a 4th RT-11 directory segment in blocks 12,13 F       on the hard drive file which specified block 100 as the starting;       block number and did an RT-11 CREATE command for each 9       of the 4 RT-11 files in the ISO9660 file structure.   ? Note that (b), (c), (d) and (e) were done using RT-11.  The net D result was that the first file (the V5.03 distribution of RT-11) was on the CD in 3 ways:C (1)  As individual files which could be booted directly under RT-11 / (2)  As a DSK file which was an RL02 file image 3 (3)  As a single file in the ISO9660 file structure   C The other 3 files under the ISO9660 file structure were also in the > RT-11 file structure as DSK files which were RL02 file images.  = NOTE that the only addition to the CD image was an RT-11 file = directory and boot blocks in block zero to block 13 which the @ ISO9660 file structure does not use.  NONE of the 4 actual files1 under the ISO9660 file structure were duplicated.   D There are a few small details that only RT-11 users will appreciate. Ask me if you need them!  @ One problem that I have is that the above plan ONLY works easilyF with files that are in RT-11 partition zero.  This is because no RT-11B device can have more than 65536 blocks.  As a result, RT-11 splits< any device with more than 65536 blocks into RT-11 partitions? that each have exactly 65536 blocks.  This is the same solution ) used by DOS V3.3 if I remember correctly.   J If the total size of all of the files is less than 300 MBytes or less thanD 10 RT-11 partitions, then the problem with non-zero RT-11 partitionsI is most easily solved by duplicating the files since the capacity of a CD J is 20 RT-11 partitions.  There really is no point in a fancy solution.  InC addition, since a CD-R is so inexpensive these days, having a fancy J solution when there is insufficient capacity to duplicate all the files isC not worth the cost.  BUT, the challenge is so satisfying if a fancy 5 solution can be found that the effort is worth while.   B The fancy solution is to some how keep the first 64 blocks of eachA RT-11 partition (each group of 65536 blocks) empty.  This is easy 9 for RT-11 partition zero since that is how ISO9660 works.   A So in addition to mentioning that the above is possible for RT-11 C partition zero, I am requesting help for non-zero RT-11 partitions. ? What would do the job is a "mkiso.exe" program which leaves the > first 64 blocks out of groups of 65536 blocks empty.  The rest@ could be done in RT-11 to set up RT-11 file directories for each RT-11 partition.   Sincerely yours,   Jerome Fine  --4 To obtain the original e-mail address, please remove5 the ten characters which immediately follow the 'at'. 8 If you attempted to send a reply and the original e-mail7 address has been discontinued due a high volume of junk 5 e-mail, then the semi-permanent e-mail address can be 7 obtained by replacing the four characters preceding the . 'at' with the four digits of the current year.   ------------------------------  % Date: Sat, 31 May 2003 23:56:13 +0100 ) From: Antonio Carlini <arcarlini@iee.org> . Subject: Re: Controller for VAXStation 4000 90& Message-ID: <3ED9330D.1060702@iee.org>   Justin McMichael wrote: J > I think I found out what it is.  It's a DSW21-AA synchronous controller.N > The problem is, noone has it.  Nothing on eBay, nothing (of value) on Google > or any other search engine.   ? In all of my years of supporting the WAN Device Drivers product > (which includes a driver for the DSW21/41/42/43 cards) I never@ had a call logged that related to the DSW21. I assume that means? that not many of them were ever sold (not surprising since they = were specifically for workstations and most of those would be + on someone's desk, not hooked up to a WAN).   A Unless your people have a specific need for a VAXstation 4000-90, 8 you might consider a MicroVAX 3100-90 (or -95 or -96) or= a VAX 4000-100 (or any -10x except the -108A). Then you could ; use a DSW41 (single line) or DSW42 (dual line) which may be < easier to find. The MicroVAX 3100-98 (and maybe -88) and the9 VAX 4000-108A would need a DSW43 (and will happily fry if 8 you put a PCI card into the slot that looks like PCI but really isn't ...).  @ Finally, you originally said "asynchronous" - the DSW series areB synchronous. If you really do want asynch, you need something like< a DHW42 (or maybe do asynch via LAT or reverse-telnet though a terminal server etc.).   Antonio    --   --------------- - Antonio Carlini             arcarlini@iee.org    ------------------------------    Date: 31 May 2003 16:04:27 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)  Subject: Re: Install Directory3 Message-ID: <2j9zJjxhzJoH@eisner.encompasserve.org>   W In article <6S8Ca.1693$Nu4.424@news.cpqcorp.net>, hoff@hp.nospam (Hoff Hoffman) writes: g > In article <57ee6a43.0305290638.25812eb1@posting.google.com>, wsmith@smausa.com (Wendy Smith) writes: E > :What, if any, is the convention regarding the directory into which N > :third-party software should be loaded in preparation for an installation ofJ > :the software? We are currently installing in [000000], and a client hasL > :indicated it is "more appropriate" to install in SYS$MANAGER. I am new toH > :the Alpha VMS world, and thus have no fact-based opinion of my own... > G >   You will want to read the PCSI (PRODUCT INSTALL) and the VMSINSTAL  F >   utility documentation and you will want to read the OpenVMS User'sD >   Guide, as a start.  (I will here assume you are not writing yourH >   own installation tools, as that is far more involved and difficult.)  E Steve, I think you really ought to look at existing discussion before F posting.  As it turns out, Wendy _is_ writing (or maintaining) her ownE installation tools.  The pros and cons of doing that have been widely   discussed in this topic already.   ------------------------------  # Date: Sat, 31 May 2003 21:09:22 GMT # From: hoff@hp.nospam (Hoff Hoffman)  Subject: Re: Install Directory1 Message-ID: <6S8Ca.1693$Nu4.424@news.cpqcorp.net>   e In article <57ee6a43.0305290638.25812eb1@posting.google.com>, wsmith@smausa.com (Wendy Smith) writes: C :What, if any, is the convention regarding the directory into which L :third-party software should be loaded in preparation for an installation ofH :the software? We are currently installing in [000000], and a client hasJ :indicated it is "more appropriate" to install in SYS$MANAGER. I am new toF :the Alpha VMS world, and thus have no fact-based opinion of my own...  E   You will want to read the PCSI (PRODUCT INSTALL) and the VMSINSTAL  D   utility documentation and you will want to read the OpenVMS User'sB   Guide, as a start.  (I will here assume you are not writing yourF   own installation tools, as that is far more involved and difficult.)  C   [000000] is the root directory of a disk volume or the root of a  E   concealed logical, neither of which are usually appropriate for the C   typical layered product installation.  The disk root directory is E   usually left for user and application directories, and a few system $   files specific to the disk volume.  G   The usual installation target for layered product startup procedures  I   is in SYS$STARTUP:, and when operating with your own tools in a cluster E   environment, you will usually want to specify this directory as the H   cluster-common directory SYS$COMMON:[SYS$STARTUP].  For management andE   related tasks and for startups on older products, use SYS$MANAGER:  E   (SYS$COMMON:[SYSMGR]).  As you are new to OpenVMS, you will want to F   familiarize yourself with the cluster-common system disk format and G   with the system logical names -- one set of files for multiple nodes. F   (This is something that applications ported from other platforms mayF   not expect or assume -- the per-node configuration is certainly alsoH   possible, but most OpenVMS applications play nicely within a cluster.)  E   A few executable files (with no set number and no set limit) can be E   loaded into the SYS$SYSTEM: (SYS$COMMON:[SYSEXE]) system directory, H   but applications requiring large numbers of application-specific filesJ   should have an application-specific directory or directory tree created.  D   Shareable images belong in SYS$SHARE: (SYS$COMMON:[SYSLIB]) or canD   be accessed elsewhere by definition of logical name.  (See the FAQD   for a pointer to the shareable image cookbook, particularly if you   have common code.)  F   PCSI is the prefered software installation tool, while VMSINSTAL is B   the older tool.  If you choose to use VMSINSTAL, please keep theB   file installation tasks entirely separate from the configuration7   tasks, as this makes an eventual move to PCSI easier.   D   I personally prefer not to link on-site, as this makes maintenanceD   and support rather more difficult -- shipping pre-built executableB   images allows me to maintain source listings (with machine code)F   and link maps, all of which can be used to diagnose run-time errors.A   For information on writing log files from applications, see the F   FAQ; search for the shared access SHR stuff and the SS$_DEBUG stuff.  D   All files stored in any shared areas (eg: SYS$STARTUP, SYS$SYSTEM)D   should use an application-specific facility prefix -- on files, onB   all logical names, etc.  Anything that can collide with another H   application.  For products, facilities can be registered by contactingH   the (IIRC) product[-at-]hylndr.sqp.zko.dec.com address and requesting,G   completing, and submitting a product facility registration form.  The F   facility prefix keeps you from colliding with other packages or with   OpenVMS itself.   E   In addition to the PCSI and VMSINSTAL documentation and the OpenVMS C   User's Guide and the System Manager's documentation, the Guide to C   Modular Programming and the calling standard can be useful reads. G   All are part of the OpenVMS documentation.  (Again, it is best if you F   build some familiarity with this material, as it helps integrate theI   product, to economically support it, and to discuss it -- the "install  G   directory", for instance, isn't terminology commonly used on OpenVMS, K   and the INSTALL utility is largely unrelated to product installation. :-)   N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com    ------------------------------  # Date: Sun, 01 Jun 2003 03:02:09 GMT 1 From: Michael Austin <maustin@firstdbasource.com>  Subject: My AS21001 Message-ID: <3ED967F5.44F39A2@firstdbasource.com>   C Well it finally happened.  I was making a backup of my system after F noticing one of my 7-member raid set drives had died and was using theC only 3 good 4MM DAT tapes I had left -- which contained my previous  known good backup.    ( And can you guess what happened next????  F You probably guessed it.  A second member of the raid-set died.  AfterF all they are about 8 or 9 year-old 2.1GB drives that have run non-stopH (expcept for the 5 days it took me to move from NC to Kansas City) since they were initially installed.  E So, When it comes back up, you can expect to see a whole new look and D feel to my web sites.  I guess I should break down and buy some moreD replacement tapes -- and look to see if I can find some new 4 or 9GB* drives (green bricks) that are real cheap.   --   Regards,  6 Michael Austin            OpenVMS User since June 19847 First DBA Source, Inc.    Registered Linux User #261163 7 Sr. Consultant            http://www.firstdbasource.com    ------------------------------    Date: 31 May 2003 20:03:39 -00004 From: Doc.Cypher <Use-Author-Address-Header@[127.1]>' Subject: Re: Need OpenVMS (any version) 6 Message-ID: <20030531200339.21791.qmail@gacracker.org>  L On 31 May 2003, holitska_a@cut-it-outludens.elte.hu (Holi - Holitska Andrs) wrote:  E >  Currently you can't put even hobbyist CDs out to the net, not even H >  for your domain/friends/etc., but of course I'm sure there are peopleF >  who do sharing none the less (((spelled correctl?))), maybe on some? >  popular free file sharing network like KaZaA or eDonkey2000.   I That's the _last_ thing the Hobbyist program needs - distributions flying I about file sharing networks.  It would also be very stupid to pick up any : operating system distribution from such a suspect source.   D >  Of course you can ( = it is possible) share a CD image illegally,G >  but it would be much better if HP/Compaq would make OpenVMS Hobbyist E >  really OPEN to the community (at least in binary form... full free F >  source code will remain a dream for a long long time I fear... even >  for the VAX version).  I It is open _within_ a community.  That community being Encompass, or your J local user group.  There is nothing stopping you from borrowing media fromJ someone within that community.  In fact, I've seen it recommended numerousI times on this newsgroup.  With some of the layered products that Hobbyist E licenses can be obtained for, this is the only way you can get them.    H >  If it would be legal, I would put it on our site too. 'OpenVMS servedA >  to the community by an OpenVMS system.' Sounds really nice. :]   M I wouldn't, really.  If Microsoft's legal team chases people who host old DOS K versions, I think HP's legal team would have to take action against someone N offering VMS for download by anyone.  Even if HP were to sanction distributionI across the Internet they would require that those offering it verify that 3 anyone downloading it is a DECUS/Encompass member.       Doc. --  K OpenVMS.  Eight out of ten hackers                   https://vmsbox.cjb.net K           prefer *other* operating systems.        http://althacker.cjb.net    ------------------------------    Date: 31 May 2003 22:16:49 +0200B From: holitska_a@cut-it-outludens.elte.hu (Holi - Holitska Andrs)' Subject: Re: Need OpenVMS (any version) ! Message-ID: <cCOVhUQTD5ll@ludens>   m In article <20030531200339.21791.qmail@gacracker.org>, Doc.Cypher <Use-Author-Address-Header@[127.1]> writes: N > On 31 May 2003, holitska_a@cut-it-outludens.elte.hu (Holi - Holitska Andrs) > wrote: > F >>  Currently you can't put even hobbyist CDs out to the net, not evenI >>  for your domain/friends/etc., but of course I'm sure there are people G >>  who do sharing none the less (((spelled correctl?))), maybe on some @ >>  popular free file sharing network like KaZaA or eDonkey2000. > K > That's the _last_ thing the Hobbyist program needs - distributions flying K > about file sharing networks.  It would also be very stupid to pick up any < > operating system distribution from such a suspect source.   H   You'r right about that, I wouldn't do it either. I'm not saying peopleC   should download it from there. (If it looked that way, sorry, and #   DON'T download from those sites!)   E >>  Of course you can ( = it is possible) share a CD image illegally, H >>  but it would be much better if HP/Compaq would make OpenVMS HobbyistF >>  really OPEN to the community (at least in binary form... full freeG >>  source code will remain a dream for a long long time I fear... even  >>  for the VAX version).  > K > It is open _within_ a community.  That community being Encompass, or your L > local user group.  There is nothing stopping you from borrowing media fromL > someone within that community.  In fact, I've seen it recommended numerousK > times on this newsgroup.  With some of the layered products that Hobbyist G > licenses can be obtained for, this is the only way you can get them.    H   (I don't know if borrowing/giving away a media is allowed, but I guess   you'r right about it.)  B   I don't have anything against the Hobbyist program, thank God itB   exists, and if you know someone from the community, then you canE   get the CDs some way, but if your just someone, who is just curious C   what VMS is, and you don't know anything about the community, you E   will have to pay $s so you can try it, and this fact scares peoples D   away I think. You can't try it for free if you don't know anybody.  E   If your an adventurer that sais 'Well, I've tried Windows and Linux C   on my new Itanium, now, are there any other OS-es out there?' and J   you find OpenVMS you'll see, that even a tryout version costs something.F   Some people will say 'Oh, i'll have to pay? No thanks.' others will,F   'Hmm, maybe there are a few OpenVMS users in my neighborhood...' andA   find the community. We are loosing a few people here and there.   I >>  If it would be legal, I would put it on our site too. 'OpenVMS served B >>  to the community by an OpenVMS system.' Sounds really nice. :] > O > I wouldn't, really.  If Microsoft's legal team chases people who host old DOS   H   I said (wrote, whatever), if it would be legal, meaning if it would be   legal to put it up.   M > versions, I think HP's legal team would have to take action against someone P > offering VMS for download by anyone.  Even if HP were to sanction distributionK > across the Internet they would require that those offering it verify that 5 > anyone downloading it is a DECUS/Encompass member.    J   If it would be a requirement, I would host it that way ( = ensuring thatF   the downloader is a DECUS/Encompass member if it's possible, if not,   then I wouldn't host it).   F   It's is no problem, that registration/membership is required. I justH   don't like the fact that _if_ you don't know anybody, you can't try itI   for free. OpenVMS doesn't have a free media. The internet would be one. J   It would allow VMS to be more widely spread if VMS did have a free media   my oppinion.  H   BUT the Hobbyist program is good this way too, waaay better than if itI   wouldn't exist. I'd like to thank you all, who made/make this possible.    Thank you!    C   Perhaps we're back at 'Not enough advertisement for OpenVMS _and_ D   local DECUS group.', maybe this is the core problem. If anyone whoC   hears about VMS would also immediately hear about the local DECUS '   group, that would erase all problems.     D   Btw., do any of you know about libraries where OpenVMS CD could beC   lent/lend (((I allways confuse these two...)))? Would it be legal C   if a library would offer such a service (for Hobbyist CD I mean)?   A   If someone in charge could answer this, perhaps I could talk to A   the informatics library on our university to offer Hobbyist CDs =   to students who'd like to try it out. That would be a major    step forward.   @   Could someone please clarify if this would be legal? (Could be5   done at other universities/libraries too if it is.)    > Doc. > --  M > OpenVMS.  Eight out of ten hackers                   https://vmsbox.cjb.net M >           prefer *other* operating systems.        http://althacker.cjb.net    Bye: <Holi>    PS: Good sig :]    ------------------------------   Date: 1 Jun 03 07:09:45 +0200 ) From: p_sture@elias.decus.ch (Paul Sture) - Subject: Re: Needed, history of startup files ) Message-ID: <WHhG4rzNlTjN@elias.decus.ch>   c In article <bbabbc$75dh2$1@ID-143435.news.dfncis.de>, "Hans Vlems" <hvlems.nieuw@zonnet.nl> writes: F > Others have posted answers; this is what my memory thinks it was :-) > K > The startup files named SY*.COM are called by SYS$SYSTEM:STARTUP.COM. DEC I > recommended that you edit these files, e.g. to mount disks. I never did K > that, for the simple reason that you can stop SYSTARTUP*.COM from running N > thru SET STARTUP_P1 "MIN". AFAIK you have no control over the SYLOGICALS.COM > file or its nephews. >   F Thanks, you have reminded me of a long forgotten piece of information.  A You do have some control over the execution of SYLOGICALS.COM, in : that it is executed with STARTUP_Pn (n=1-8) as parameters.  J BTW, STARTUP_P1 can also have the value "MINI", so I would recommend doing something like :  K $ if f$extract(0,3,p1) .eqs. "MIN" then exit ! or whatever suits your needs     --  
 Paul Sture   ------------------------------  % Date: Sat, 31 May 2003 20:05:42 -0400 - From: "Richard Tomkins" <tomkinsr@@istop.com>  Subject: Re: Overclocking VAXen 2 Message-ID: <WrbCa.266$RT6.284577912@news.nnrp.ca>  G OK, the VAX 8550 upgrade was a jumper on the backplane to get the speed I increase. The 8500 was too fast when it was finished so the microcode was I filled up with NOP's and a jumper put in place to make them run. when you K got an 8550 upgrade for $20,000, you got a new backplane with the jumper in K place and new microcode load. Customers figured they were getting something I special with a new backplane exchange. The 8530 figures in here somewhere  too.   rtt    ------------------------------  % Date: Sat, 31 May 2003 22:03:45 -0400 ( From: David Froble <davef@tsoft-inc.com>" Subject: Re: Portents of VMS death, Message-ID: <3ED95F01.5010700@tsoft-inc.com>   Bradford J. Hamilton wrote:   q > In article <a44Ca.307042$M81.29385@news02.bloor.is.net.cable.rogers.com>, "John Smith" <a@nonymous.com> writes:  > 5 >>"jlsue" <jefflsxxxz@sbcglobal.net> wrote in message 4 >>news:jvffdvgav6k8rb4o6jtgehcq699q5c70fv@4ax.com... >> > <snip> > C >>>And it's not that is "isn't what VMS is for", it's that it is an E >>>unsupported configuration (potentially).  I still say that putting  >>> 
 >>Dev/test >>C >>>in a cluster (security domain) with production would get you hit  >>> 
 >>big-time >>E >>>with the auditors I've known.  Especially in the financial sector.     O Most auditors don't know shit!  It's not an issue of the same cluster, it's an  P issue of security.  Give the developers privs, and the auditors would be right. M   Even if the dev people are trusted, what about their actions taking down a  Q system or the whole cluster?  If the developers and testers cannot in any manner  I harm operations or data, being on the same system/cluster isn't an issue.   M Do believe that the most trusted person, with privs, can inadvertantly cause  N problems.  I'm proof of this!  My own opinion is that development and testing P should be on seperate systems/clusters/whatever, not to protect operations, but < to protect the developers from their sooner or later "oops!"   Dave     --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com6 T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486   ------------------------------   Date: 1 Jun 03 07:30:27 +0200 ) From: p_sture@elias.decus.ch (Paul Sture) " Subject: Re: Portents of VMS death) Message-ID: <DX2qv+hC$P3k@elias.decus.ch>   o In article <4l4Ca.307109$M81.52589@news02.bloor.is.net.cable.rogers.com>, "John Smith" <a@nonymous.com> writes:  >    [snip]  :  And besides, when was the last time HP spent any money onC > advertising/marketing VMS so I can look good when I present to my 8 > bosses?" That or they are just swinging with the wind. >   E That just about sums up the problem for me. I was talking to a couple I the other day, coincidentally both working in IT. He had heard of Alphas, A she hadn't, but neither had heard of this peculiar OS called VMS.   J The talk I heard from golf club types a few years ago was all about movingE to Windows. Herd mentality is what I thought, but those were the guys * with working chequebooks. Simple isn't it?  C Can you drop me a line please? Please use firstname at lastname dot & homeip.net, so we can keep it private.   --  
 Paul Sture   ------------------------------   Date: 1 Jun 03 07:33:09 +0200 ) From: p_sture@elias.decus.ch (Paul Sture) " Subject: Re: Portents of VMS death) Message-ID: <xIaVEMjj5qv6@elias.decus.ch>   o In article <a44Ca.307042$M81.29385@news02.bloor.is.net.cable.rogers.com>, "John Smith" <a@nonymous.com> writes:  > 5 > "jlsue" <jefflsxxxz@sbcglobal.net> wrote in message 4 > news:jvffdvgav6k8rb4o6jtgehcq699q5c70fv@4ax.com.../ >> On Thu, 29 May 2003 16:36:16 -0400, JF Mezei  > <jfmezei.spamnot@istop.com> 	 >> wrote:  >> >> >jlsue wrote:? >> >> How do you get from a discussion about a specific cluster  > configuration 1 >> >> into a generic, VMS marketplace discussion?  >> >C >> >Because not everyone has an application that justifies buying a  > dozen nodes toF >> >make a full fledged cluster where any node can go down without any	 > impact. A >> >There are many who would like to choose VMS for a small scale 
 > system that D >> >provides production on one node, test on another with production > mirrored onto G >> >the test system so that the test system can be used when production  > fails. >> >G >> >If you tell them that this isn't what VMS is for, then they will go 	 > to Unix D >> >and Windows. Is the goal to only keep the "crme de la crme" of > customers, or D >> >to expand the VMS marketplace and capture the HUGE potential for > small $ >> >clusters in smaller businesses ? >>F >> If I stated anywhere that only configurations with a dozen nodes is > our > >> target, please show me where.  We are still talking about a > specific,   >> mis-configured imho, cluster. >>C >> And it's not that is "isn't what VMS is for", it's that it is an E >> unsupported configuration (potentially).  I still say that putting 
 > Dev/testC >> in a cluster (security domain) with production would get you hit 
 > big-timeE >> with the auditors I've known.  Especially in the financial sector.  >  > G > Maybe in the banking/insurance sector. In the brokerage sector things H > are often played faster and looser - even at some of the biggest names< > in the world - especially at the operating division level. > F And then try the automotive or warehousing sectors, where they look to& shave every last penny from the costs.     --  
 Paul Sture   ------------------------------    Date: 31 May 2003 23:06:54 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen) N Subject: Re: Separating Development from Production data (was: Portents of...)3 Message-ID: <YCU$ddfwFA+g@eisner.encompasserve.org>   W In article <3ED95F01.5010700@tsoft-inc.com>, David Froble <davef@tsoft-inc.com> writes:   C > It's not an issue of the same cluster, it's an issue of security. = > Give the developers privs, and the auditors would be right. L > Even if the dev people are trusted, what about their actions taking down aK > system or the whole cluster?  If the developers and testers cannot in any1 > mannerK > harm operations or data, being on the same system/cluster isn't an issue.   G So long as one is running only VMS and not SEVMS there are no mandatoryVF access controls and giving developers access to the network containingA production data presents a security risk, even without privilege.   B Developers can change the software so that in production it altersJ the protection of files and makes production data available "for testing".  E It is astounding the security breaches that can be committed "for thenJ good of the project", and even get the approval of first level management.   ------------------------------  % Date: Sat, 31 May 2003 21:02:02 +0200-/ From: "Marc Van Dyck" <marc.vandyck@brutele.be>a) Subject: Re: StorageTek l700 on VMS 7.3-1n* Message-ID: <bbau6v$q7b$1@news.brutele.be>   David,  6 If you look for SCSI-FC compatibility, you should look6 at the products made by TD Systems, a company located,6 I believe, somewhere in New England. They have a range9 of products, named Omniserve, which is basically a switcht: with SCSI and/or FC ports, and lots of goodies (tape drive> sharing/locking, tape to tape copy, lun remapping, and so on).6 They are very responsive: when we report a problem, we< usually get a microcode fix a day or two later. We use those; boxes to share all our tape drives (STK 9840 SCSI, 9840 FC, > 9940 FC, ...) with all our hosts, across two sites a few miles8 apart. In the OpenVMS world, it interfaces well with SLS: and DCSC. With Unix and NT, we use Veritas. Worth a check.> Contact me off line if you want more info. Disclaimer : I have8 no personal interest with the TD Systems company, just a happy customer.a   Marc.h  < "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message! news:3ED6B364.176A3DE7@fsi.net...u > "John N." wrote: > > I > > ok,  forget the dumb question in my last post.  I did run the commandl IOK > > FIND_WWID, (followed by IO AUTO) and it did find FIVE of the EIGHT tape  > > drives.  That is progress. >e > O.k. Now here's a big rub... >tF > On the NSRs (FC-SCSI bridges, sometimes called "Crossroads" switchesJ > since they seem to be the only company in the country (world?) who makesJ > them, everyone else -including HP - just rebadges them, and maybe tweaksI > the firmware a bit), if you're remapping SCSI id's, make sure that each-I > MAP starts with LUN 0. Apparently, VMS can't handle it if, for example,-E > the first NSR presents LUNs 0-4 and the second presents 5-whatever.h >oH > I kinda wish you had elected to hold SYSMAN IO AUTOCON until after youI > had resolved the visibility problem. The only way to clear the VMS data3J > structures it (and FIND_WWID) populates is to reboot the entire cluster. >DH > Note that if you want, before your next reboot, you can change the MGAJ > device units numbers to any you wish - you're not stuck with what SYSMANI > arbitrated. However, if you've already configured other software to use2@ > the existing device units, well, you know how that stuff kinda > snowballs... >u > -- > David J. Dachteraq > dba DJE Systems  > http://www.djesys.com/ > * > Unofficial Affordable OpenVMS Home Page:! > http://www.djesys.com/vms/soho/    ------------------------------  # Date: Sat, 31 May 2003 21:27:25 GMTc# From: hoff@hp.nospam (Hoff Hoffman) 5 Subject: Re: SYS$QIOW - Wtih more Details - Final Try-2 Message-ID: <179Ca.1694$Nu4.1560@news.cpqcorp.net>  P In article <vdh8p3ctb9mh84@corp.supernews.com>, "Vivek" <visoni@bmc.com> writes:  L :...I create a mailbox with CREMBX call and set max msg size to 1024 and the :buffer lenght 2048.  E   This means you can transfer a single I/O of slightly less than 10240D   bytes (due to mailbox overhead) and the mailbox can store slightlyG   less than 2048 bytes of unread data.  (These I/O settings are further.G   limited by the system MAXBUF parameter setting and your BYTLM processcG   quota and a few other similar quotas, but values of 1024 and 2048 areoF   usually well within the typical and well within the default limits.)  I :Now I can assign this mailbox to another process using the ASSIGN or the 
 :CREMBX call.  :  :I used CREMBX call.  B   You will want to be aware of the nature of the logical names andC   the mailboxes, as associated with the UIC groups of the processesoB   involved.  Outside of the current UIC group, the mailbox logical/   names are not generally visible (by default).   J :Can the max msg size and the buffer size be different from the ones which :actually created this mailbox.s  D   If the application is not creating the mailbox, the values are notD   used -- the $crembx call assigns a channel to an existing mailbox.  J :For msg size less than 1024 my read process aborts. But if more than 1024 :what will happen.  E   Um, I'd encourage you to try it -- OpenVMS is usually quite good atRD   not being particularly nasty when you are experimenting, providingE   status codes and errors and other useful information.  But you will    get an I/O error.a     Also see:   F   http://h71000.www7.hp.com/freeware/freeware40/srh_examples/mbxdemo.c  D   For details on debugging and on common coding mistakes, please seeD   Ask The Wizard topics (1661) and (7552), and the other topics thatE   referenced from those two.  These will discuss the expected status r8   checks, event flags, and IOSBs and debugging and such.  %     http://h71000.www7.hp.com/wizard/s  2     http://h71000.www7.hp.com/wizard/wiz_1661.html2     http://h71000.www7.hp.com/wizard/wiz_7552.html     :Any performace affects.  J   Mailboxes have some definite limits, particularly for remote operations.F   My preferences would be for the intra-cluster communicatons ($ICC*) G   system services (limited to communications within an OpenVMS Cluster)dF   or for the use of a network transport such as DECnet or (more widelyJ   available) IP, personally -- these are more flexible over the long haul.    N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comb   ------------------------------  # Date: Sat, 31 May 2003 22:49:13 GMTa# From: hoff@hp.nospam (Hoff Hoffman)lA Subject: Re: Your Own Installation Tool? (was: Install Directory)e2 Message-ID: <JjaCa.1695$pc4.1676@news.cpqcorp.net>  i In article <20030529152226.15080.00000767@mb-m14.news.cs.com>, ka2doug@cs.commoc.sc (DL Phillips) writes:u :Larry Kilgallen wrote:e? :>In article <57ee6a43.0305290638.25812eb1@posting.google.com>,A) :>wsmith@smausa.com (Wendy Smith) writes:rF :>> What, if any, is the convention regarding the directory into whichO :>> third-party software should be loaded in preparation for an installation ofhK :>> the software? We are currently installing in [000000], and a client has M :>> indicated it is "more appropriate" to install in SYS$MANAGER. I am new toEI :>> the Alpha VMS world, and thus have no fact-based opinion of my own...  :>C :>Temporary directory storage during installation should be managedsC :>by either VMSINSTAL or PRODUCT INSTALL, the two supported methodsl6 :>of installing software, under control of the vendor. :nO : Yes, this is the digi-paqard way, but during the  2+ decades that I've been aeM :TP software supplier of VMS applications I've found it unwise to rely on thes :whims of digi-paqard.  G   HP and third-party vendors have a monstrously large base of installedcH   software, and no real desire to re-write large tracts of it.   BecauseE   of this, upward-compatibility is a central goal for the HP OpenVMS 1   Engineering organization.s  J   I can foresee cases where the existing software and existing tools were,H   um, used creatively -- what can be done in a KITINSTAL and what shouldG   be done in a  KITINSTAL are completely different, for instance.  I've I   encountered KITINSTAL code from authors that clearly never read through J   the VMSINSTAL manuals, for instance.  And there have been the occasionalI   bugs.  And there have been a few areas where there have been deliberatefG   and incompatible changes, with the contractually-required removal of yK   Adobe Display Postscript being a salient example.  But again, maintaining I   upward compatibility is a Big Deal around folks in OpenVMS Engineering.t  N :  It's been easier in our experience to use our own installation method that  :we control ...x  F   This is a trade-off between maintaining and upgrading your own toolsH   and leveraging the expectations of the installers.  Some folks do readG   the installation directions and -- from experience -- many do not.  IoH   have certainly worked with my share of home-grown software installers,L   too.  (Some of these tools work very nicely, and some are, well, hideous.)  G :and not spend wasted man-days learning some new and obscure "langauge"l. :that is liable to change in the next release.  G   Citations, please?  If you have specific complants or specific areas oI   where syntax was changed incompatably, I'm interested in knowing about aG   them.  Syntax and related error checking has improved and there have  G   been various PCSI enhancements made, of course.  (I assume PCSI here,0,   as VMSINSTAL is DCL-based in the extreme.)  F   And again, this area is replete with trade-offs, particularly aroundF   customer expectations and such.  (I can remember having very similarD   discussions over the years with folks that were writing their own D   communications protocols, rather than using IP or DECnet or KermitG   or other mechanism -- some of the factory floor folks seemingly loved2H   to create their own very unique communications schemes and protocols.)     L : I'd say that the PRODUCT INSTALL replacement of VMSINSTAL makes for a goodD :example. The "breakage" of NCP (and DECnet) and the ERRLOG.SYS fileL :(analyse/error) are two other what-the-????'s that happen to come to mind.   D   I will assume this was the NCP emulator provided with DECnet-Plus,F   and fallout from the DECnet-Plus upgrade -- given DECnet Phase IV isF   still around and given the native NCP still works (AFAIK) just fine.  F   If you are referencing this in the context of software installation,F   it would appear you are performing tasks involved with configurationF   -- this is certainly common in some product installations, and it isD   certainly an area which is certainly rather more difficult to move   into PCSI.  C   Details of the ERRLOG.SYS area would be of interest -- there have F   been changes made as OpenVMS has been upgraded, and as new tools areG   made available.  I am not particularly aware of a tie to the area of iE   any of the software installers here -- and the changes made to this B   area of OpenVMS would certainly also affect third-party softwareF   installation tools, as well.  But again, very few installations I amF   aware of -- PCSI, VMSINSTAL, or otherwise -- have ties to this area.  B   As for the availability of the tools, VMSINSTAL continues to be E   available, and even the VMSUPDATE tool is around for the oldest VAXd   kits.d    N  ---------------------------- #include <rtfaq.h> -----------------------------J       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comD   ------------------------------  % Date: Sun, 01 Jun 2003 01:52:21 +0100n" From: Didier Morandi <no@spam.com>% Subject: [OT] best HTML code cleaner?o' Message-ID: <3ED94E45.4070100@spam.com>i  J I'm working on the WEB site of a school, which has been "produced" with W= ORD.  J What is the best code cleaner to remove the (amazing) noise from the html=
  pages,=205 please? Even FrontPage :-( cannot open the sources...o   Thanks,o   D. --=20h- Didier Morandi sarl au capital de 8 000 eurosd-   19 chemin de la Butte 31400 Toulouse Francea/   T=E9l: 33(0)5 6120 1964 Fax: 33(0)5 6154 1928m&           http://www.didiermorandi.com                    RC en cours   ------------------------------  % Date: Sun, 01 Jun 2003 04:22:03 +0200s2 From: martin@radiogaga.harz.de (Martin Vorlaender)) Subject: Re: [OT] best HTML code cleaner?g; Message-ID: <3ed9634b.524144494f47414741@radiogaga.harz.de>d  # Didier Morandi (no@spam.com) wrote:lI > I'm working on the WEB site of a school, which has been "produced" withe > WORD.i >cK > What is the best code cleaner to remove the (amazing) noise from the htmlp> > pages, please? Even FrontPage :-( cannot open the sources...  ? This runs by the name of "demoronizing" in the perl newsgroups.T   Hope it helps,   Martin  	 [--cut--]l- From: Tom Christiansen <tchrist@mox.perl.com>  Newsgroups: comp.lang.perl.miscr Subject: Re: MS-HTML must die!& Message-ID: <37393556@cs.colorado.edu>  Date: 12 May 1999 02:01:26 -0700  $ lr@hpl.hp.com (Larry Rosler) writes:K : I promised to redo the demoroniser (British spelling) in one page.  It's tK : actually somewhat more (88 lines including comments).  It would probably iG : run at least an order of magnitude faster than the original; perhaps   : two.  B Your code has the Sed Essence, but was not so phrased.  Now it is.E Also, I fixed an $ARGV bug you had, and added a multiple break munge.e   --tomt   #!/bin/perl -0777pwl  9 #           De-moron-ise Text from Microsoft Applicationsd # 2 #                   by John Walker -- January 1998/ #                      http://www.fourmilab.ch/k7 #                   revised by Larry Rosler -- May 1999 C #                      http://www.hpl.hp.com/personal/Larry_Rosler/ 6 #	            sedition by Tom Christiansen -- May 1999 #k5 #               This program is in the public domain.t   use strict;u my %tr;d  A #   Eliminate idiot MS-DOS carriage returns from line terminator.r s/\s+$/\n/g;  8 #   Supply missing semicolon at end of numeric entity if #   Billy's bozos left it out. #   Map characters in entities.n s/&#(\d+);?/C      $tr{chr $1} || ($1 < 0x80 || $1 > 0x9F ? "&#$1;" : chr $1)/ge;b  8 #   Now check for any remaining untranslated characters. my $iline = 1;% $1 eq "\n" ? ++$iline : printf STDERReK    "$0: warning -- untranslated character 0x%.2X in line $iline of `%s'\n",t3 	 ord $1, ($ARGV eq '-' ? 'standard input' : $ARGV)b1 	     while /([\x00-\x08\n\x10-\x1F\x80-\x9F])/g;   ; #   Fix unquoted non-alphanumeric characters in table tags.F5 s@(<T(?:ABLE|D|H)\s.*?WIDTH\s*=\s*)(\d+%)@$1"$2"@gis;r  + #   Correct PowerPoint mis-nesting of tags.I; s@(<FONT\s[^>]*>\s*<STRONG>.*?) (</FONT>) (\s*) (</STRONG>)i  @$1$4$2$3@gisx;  ; #   Translate bonehead PowerPoint misuse of <UL> to achieves #   paragraph breaks.y s@<(?:P|/UL)>\s*<UL>@<P>@gi; s@</UL>\s*</P>@@gi;m  8 #   Repair PowerPoint depredations in "text-only slides" s@<P></P>@@gi; s@ (<TD HEIGHT=100)@ <TR>$1@gi;e s@<LI>(?=<H2>)@@gi;r    #   Repair idiotic double breaks s@<BR>(?:\s*<BR>)+@<P>@gi;     BEGIN {a        %tr = (F          #   Fix dimbulb obscure numeric rendering of &amp; &lt; &gt;.          chr 38  => '&amp;',          chr 60  => '&lt;',l          chr 62  => '&gt;',   E          #   Map strategically incompatible non-ISO characters in the @          #   range 0x82 -- 0x9F into plausible substitutes where          #   possible.          "\x82"  => ',',!          "\x83"  => '<em>f</em>',i          "\x84"  => ',,',h          "\x85"  => '...',            "\x88"  => '^',          "\x89"  => '/',r               "\x8B"  => '<',          "\x8C"  => 'Oe',d    m          "\x91"  => '`',          "\x92"  => "'",          "\x93"  => '"',          "\x94"  => '"',          "\x95"  => '*',          "\x96"  => '-',          "\x97"  => '--',t#          "\x98"  => '<sup>~</sup>',n$          "\x99"  => '<sup>TM</sup>',               "\x9B"  => '>',          "\x9C"  => 'oe',,      );e   } 	 [--cut--]r   -- nJ One OS to rule them all       | Martin Vorlaender  |  VMS & WNT programmer7 One OS to find them           | work: mv@pdv-systeme.deoJ One OS to bring them all      |   http://www.pdv-systeme.de/users/martinv/> And in the Darkness bind them.| home: martin@radiogaga.harz.de   ------------------------------   End of INFO-VAX 2003.301 ************************