1 INFO-VAX	Fri, 03 Aug 2001	Volume 2001 : Issue 428       Contents:= Re: ??== Installed DCPS 2.0: Two observations and a question.  Re: absolute beginner  Re: absolute beginner ( Re: Alpha:  an invitation to communicate' Any way to get a BA350 to work with NT? + RE: Any way to get a BA350 to work with NT? + Re: Any way to get a BA350 to work with NT? + RE: Any way to get a BA350 to work with NT? + Re: Any way to get a BA350 to work with NT? # Re: C v6.2 vs v6.4 volatile keyword B Re: Comments from a VMS customer (and I would like to stay as one)B Re: Comments from a VMS customer (and I would like to stay as one)B Re: Comments from a VMS customer (and I would like to stay as one)B Re: Comments from a VMS customer (and I would like to stay as one)B Re: Comments from a VMS customer (and I would like to stay as one)B Re: Comments from a VMS customer (and I would like to stay as one) Re: DCL... Oops...' DECset enhancements for ODS-5 and Java?  Re: Dismounting a drive  Re: Dismounting a drive  Re: f$parse glitch$ RE: Few People in DEC Understood....$ Re: Few People in DEC Understood....$ Re: Few People in DEC Understood....( Re: How2 DECnet-copy with two passwords?6 Re: IPF Console Bootstrap (was: No chance for OpenVMS)1 Re: Looking for a Pine.exe for TCPIP 5.x services 1 Re: Looking for a Pine.exe for TCPIP 5.x services 1 Re: Looking for a Pine.exe for TCPIP 5.x services = Marketing - any kind !  was  Few People in DEC Understood.... A Re: Marketing - any kind !  was  Few People in DEC Understood.... 1 Re: Media Kits (was Re: IPF Console Bootstrap...) 1 Re: Media Kits (was Re: IPF Console Bootstrap...) 1 RE: Media Kits (was Re: IPF Console Bootstrap...) ( Mystery jumper on MicroVAX 3100 Model 80 openssh  Re: OpenVMS  + Itanium Re: OpenVMS  + Itanium Re: OpenVMS  + Itanium0 Re: Outsourcing of VMS support to Digital India!0 Re: Outsourcing of VMS support to Digital India!0 Re: Outsourcing of VMS support to Digital India!0 Re: Outsourcing of VMS support to Digital India!0 Re: Outsourcing of VMS support to Digital India!0 Re: Outsourcing of VMS support to Digital India!0 RE: Outsourcing of VMS support to Digital India!0 Re: Outsourcing of VMS support to Digital India!< Re: Printer settings was - Re: Print Queue's on IP Addresses3 Re: Problem with floppy drive on an Aplastation 200 3 Re: Problem with floppy drive on an Aplastation 200  Re: Quorum Disk Questions  Re: Quorum Disk Questions  Re: Quorum Disk Questions  Re: Quorum Disk Questions  Re: Quorum Disk Questions  Re: Quorum Disk Questions  Re: Reboot to close a file ????  Re: Reboot to close a file ????  Re: Reboot to close a file ????  Re: Reboot to close a file ????  Re: Reboot to close a file ????  Re: Reboot to close a file ????  Re: Reboot to close a file ????  ssh port forwarding  Re: ssh port forwarding " Re: Sun goes after Alpha user base" Re: Sun goes after Alpha user base Re: SYS$LOGIN:[some.directory] Re: SYS$LOGIN:[some.directory] Re: SYS$LOGIN:[some.directory] Re: SYS$LOGIN:[some.directory] Re: SYS$LOGIN:[some.directory] Re: SYS$LOGIN:[some.directory] tcpip 5.1 dhcp error message Third postcard from Sun  Re: Third postcard from Sun  Re: Third postcard from Sun  Re: Third postcard from Sun  trapping telnet disconnects 	 Re: VAX C 	 Re: VAX C 	 Re: VAX C & Re: Virus Alert at Compaq (2001.08.02) Re: VMS Prompt value Re: VMS Prompt value Re: VMS Prompt value Re: WATCHER compile problems< When/how to send answerback string during logon negotiations@ Re: When/how to send answerback string during logon negotiations7 You Get What You Pay For, a.k.a., There's No Free Lunch ; Re: You Get What You Pay For, a.k.a., There's No Free Lunch < Re: [OT] Market corrections (was Re: Alpha: an invitation to  F ----------------------------------------------------------------------  $ Date: Fri, 3 Aug 2001 09:41:52 +0200, From: "Bart Zorn" <B.Zorn@TrueBit.nospam.nl>F Subject: Re: ??== Installed DCPS 2.0: Two observations and a question.: Message-ID: <3b6a561f$0$2955$e4fe514c@newszilla.xs4all.nl>  L I am aware of the fact that is really no reason that DCPS should not supportJ autostart queues. My point is, I prefer support for the configuration "outI of the box". If I have to configure systems for clients (as opposed to my F "own" system), I prefer to stay as close to the documented features asF possible. So if DCPS$STARTUP could be modified or rewritten to support# autostart, I would appreciate that.    Regards,  	 Bart Zorn     F "Brian Tillman" <tillman_brian@notnoone.notnohow.com> wrote in message news:3b69692e$1@news.si.com...G > >This week I also installed DCPS V2.0. The thing that I missed is the 	 > ability J > >to configure auto start and auto failover queues with DCSP$STARTUP.COM. ItH > >appears that this file has not been modified since version 1.5 or so. > E > The ability is there in the symbiont.  Here is aan excerpt from our  startup: >  > $ > $ CALL DEFINE_QUEUE 1A_HPLJ4000N_1, > """ip_rawtcp/1a_hplj4000n_1.si.com:9100"""# > $ CALL DEFINE_QUEUE 1A1_HPLJ5MP_1 + > """ip_rawtcp/1a1_hplj5mp_1.si.com:9100""" # > $ CALL DEFINE_QUEUE 1A3_HPLJ4PL_1 + > """ip_rawtcp/1a3_hplj4pl_1.si.com:9100""" % > $ CALL DEFINE_QUEUE 1A5_HPLJ4000N_1 - > """ip_rawtcp/1a5_hplj4000n_1.si.com:9100""" # > $ CALL DEFINE_QUEUE 1A5_HPLJ4PL_2 + > """ip_rawtcp/1A5_HPLJ4PL_2.si.com:9100"""  > .... > $DEFINE_QUEUE: SUBROUTINE  > $!< > $ DEFINE/SYSTEM/EXECUTIVE/NOLOG DCPS$'P1'_DEVICE_NAME 'P2'= > $ IF F$GETQUI("DISPLAY_QUEUE","QUEUE_PRINTER",P1) THEN EXIT  > $!7 > $ WRITE SYS$OUTPUT "Initializing Printer Queue ''P1'"  > $!L > $ INIT/QUE/BASE_PRIORITY=4 /DEFAULT=(FORM=DCPS$DEFAULT)/NOENABLE_GENERIC -D >    /LIBRARY=DCPS$DEVCTL /OWNER=[SYSTEM]/PROCESSOR=DCPS$SMB/START -H >    /PROTECTION=(S:M,O:D,G:R,W:RS) /SCHEDULE=(NOSIZE)/SEPARATE=(FLAG) -= >    /AUTOSTART_ON=(FMC1::DCPS$,FMC2::DCPS$,CASS::DCPS$) 'P1'  > $ EXIT > $! > $ ENDSUBROUTINE  > --C > Brian Tillman                   Internet: tillman_brian at si.com C > Smiths Aerospace                          tillman at swdev.si.com ? > 3290 Patterson Ave. SE, MS      Addresses modified to prevent > > Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@": >        This opinion doesn't represent that of my company >  >    ------------------------------  % Date: Fri, 03 Aug 2001 08:33:06 +0200 = From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>  Subject: Re: absolute beginner) Message-ID: <3B6A45A2.EFAE9045@gtech.com>    Luca Balzano wrote: K > it is a lot that I want to put my hands on a Vax machine, last week I had M > the opportunity to get a vaxstation 3100 for a very cheap price and now I'm H > trying to understand how to get an hobbyist (cheap) release/license ofI > OpenVMS. Somebody told me that since I live in Italy this could be more  > difficult.  % What you need is a DECUS member-ship.   : If DECUS Italy does no longer exist, then it is a problem.   Arne   ------------------------------   Date: 3 Aug 2001 08:23:30 -0700 1 From: nothome@spammers.are.scum (Malcolm Dunnett)  Subject: Re: absolute beginner, Message-ID: <sAuT3YaBfqD9@malvm5.mala.bc.ca>  * In article <3B6A45A2.EFAE9045@gtech.com>, B    Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com> writes:   > Luca Balzano wrote: L >> it is a lot that I want to put my hands on a Vax machine, last week I hadN >> the opportunity to get a vaxstation 3100 for a very cheap price and now I'mI >> trying to understand how to get an hobbyist (cheap) release/license of J >> OpenVMS. Somebody told me that since I live in Italy this could be more
 >> difficult.  > ' > What you need is a DECUS member-ship.  > < > If DECUS Italy does no longer exist, then it is a problem. > -    Encompass takes memberships from anywhere.   6    see http://www.encompassus.org/membership/join.html  C    Once you join Encompass you can use your Encompass membership id G to get a hobbyist license ( select U.S. Chapter Decus as the chapter ).   K   disclaimer: I don't speak for Encompass or the hobbyist program, I'm just L an Encompass member ( from outside the US ) who has gotten hobbyist licenses	 this way.    ------------------------------  % Date: Fri, 03 Aug 2001 17:17:51 +0100 0 From: andrew harrison <andrew.nospam@uk.sun.com>1 Subject: Re: Alpha:  an invitation to communicate * Message-ID: <3B6ACEAF.C993CE31@uk.sun.com>   jlsue wrote: > A > I see.  So if you document a particular behavior, and then some G > competitor complains that it does exactly what it's documented to do, ( > then you should fix it.  Yeah.  Right. >   = Sorry that isn't what you did, try again. You documented the  9 fact of the 3x difference and then went on to claim that  6 this was near UMA and that it would not be a problem.   B You then backed these claims up with just enough pseudo technical   guff to hoodwink most customers.  > If you had said its 3x and you will need to make the follwing ; changes to your apps like using OPS because 3x is an issue  9 then you would have a case, you did not and so you don't.    Not your best try.  ; If you are in any doubt about any of this read the WildFire 6 architecture "white paper" which you clearly have not.    C > >Of course you subsequently had to disprove this claim yourselves E > >when you published the second TPC-C result which was configured to ? > >ensure that there was an absolute minimum of remote traffic.  > H > Perhaps, but then TPC-C isn't necessarily a real-world app.  It's onlyE > a benchmark.  And in any case, it's only one instance and certainly @ > doesn't claim that most of the accesses (for most of the apps,; > obviously) don't exhibit the behavior in the white paper.  > E > There are apps out there in the real world using the GS systems who E > are pretty happy with the performance.  I guess that's what matters  > the most.  > G > Er... aside from TPC-C, do you have other examples?  If not, then you , > haven't come close to disprove any claims. >   D This is getting boring, how about Oracle apps, WildFire performance A was dreadfull for Oracle Apps (you were unable to use OPS) a much D smaller and less expensive Sun and a much smaller and less expensive$ IBM box trashed the WildFire result.  H > Well, if we tell people that this is the performance characteristic ofG > the system, then there's nothing to fix.  They already know when they C > buy them that this *could* be an issue in their environment.  And F > guess what?  We have architecture services labs where they can setup@ > their application and try them out before buying the hardware. >   ) This is not what your "white paper" says.   2 Read the white paper, please do yourself a favour 3 you are only makeing yourself look more ridiculous.     8 > >Interesting, I have come up with concrete examples of6 > >where you have missled your customers. All you seem- > >to be able to do in response is type BAHH.  > G > You have?  Which customer have I misled?  The fact is that you choose E > to pick nits so fine that most people didn't even notice, much less E > care.  Even the ones that may be true are very insignificant, and I G > doubt that they caused someone to spend multi-bucks on something that * > didn't perform to customer expectations. >   1 Umm dropping Alpha in favour of Intel is hardly a / small nit the people havn't even noticed or had . you forgotten what this thread is about ??????   You really are shameless.   > > All I got that comes close is two entries.  Both were in theG > fa.alpha-osf-managers newsgroup.  And neither of them had a "Wait for A > Marvel" comment from Rob Young.  His comments in that vein were G > directed at your comments, which, on analysis, were nothing but smoke  > from a non-existent fire.  >   0 Do some more research or check with Rob remember0 you arn't generally good at this so a second try might be a good idea.    regards  Andrew Harrison  Enterprise IT Architect    ------------------------------  # Date: Fri, 03 Aug 2001 10:39:08 GMT * From: "Bill Bradley" <wdbradley3@home.com>0 Subject: Any way to get a BA350 to work with NT?= Message-ID: <gbva7.7670$c8.1782706@news1.denver1.co.home.com>   - OT, but, I need this working for a project...   G We're running into a problem trying to get a StorageWorks array cabinet I working.  I'm setting up NT4 Enterprise Sever w/SP6a, and have a BA350-SB D cabinet with four SWXD3-WE 4.3 GB drives that I would setup in an NTK software RAID 5 array as drive D.  The PC is a standard Dell GX200 with PCI E slots, so, I cannot use the existing SCSI (DSSI?) cards from an Alpha  server, which are EISA.   L  I've tried the following Adaptec SCSI adapters, but, none of them can "see" the drives at the BIOS level:    3940   2930  L 2944 (this is a HV Differential SCSI, but, wasn't sure WHAT kind the cabinet
 supports).  8 Any ideas on how to hook the cabinet up to a regular PC?   Thanks!    ------------------------------  % Date: Fri, 03 Aug 2001 12:32:14 +0100 5 From: "Steeples, Oliver" <Oliver.Steeples@compaq.com> 4 Subject: RE: Any way to get a BA350 to work with NT?N Message-ID: <F498D199EDB12D468CD2C66680D3080116D60B@reoexc04.emea.cpqcorp.net>  H A BA350 has a 50 pin high-density connector and the disks are FNSE (fast narrow single ended).    3940-should work  K 2930-ultra card should work but need to terminate top 8 bits (bit overkill)    2944-HVD thus won't work    H I guess it is a BA350 and not a 356 with IO module.  Also have you triedG different slots as it may be configured for slpit bus or the terminator J jumper maybe attached.  The adapter scsi ID could be clashing with a drive on the shelf.....    Some ideas to think about.   	Oliver    -----Original Message-----/ From: Bill Bradley [mailto:wdbradley3@home.com] & Sent: Friday, August 03, 2001 11:39 AM To: Info-VAX@Mvb.Saic.Com 0 Subject: Any way to get a BA350 to work with NT?    - OT, but, I need this working for a project...   G We're running into a problem trying to get a StorageWorks array cabinet I working.  I'm setting up NT4 Enterprise Sever w/SP6a, and have a BA350-SB D cabinet with four SWXD3-WE 4.3 GB drives that I would setup in an NTK software RAID 5 array as drive D.  The PC is a standard Dell GX200 with PCI E slots, so, I cannot use the existing SCSI (DSSI?) cards from an Alpha  server, which are EISA.   L  I've tried the following Adaptec SCSI adapters, but, none of them can "see" the drives at the BIOS level:    3940   2930  L 2944 (this is a HV Differential SCSI, but, wasn't sure WHAT kind the cabinet
 supports).  8 Any ideas on how to hook the cabinet up to a regular PC?   Thanks!    ------------------------------  # Date: Fri, 03 Aug 2001 11:58:07 GMT * From: "Bill Bradley" <wdbradley3@home.com>4 Subject: Re: Any way to get a BA350 to work with NT?= Message-ID: <jlwa7.7891$c8.1811379@news1.denver1.co.home.com>    Thanks!   L I've got the drives in the four slots farthest away from the two internal 50J pin SCSI connectors (there are two power modules in this cabinet).  TakingC the fans off, there are two, I guess terminators, on the backplane.   H Connecting the 3940 with a 68 to 50 pin adapter cable to either internal! (cabinet) connector doesn't work.   E I heard about something called a Tri-Link adapter--would I need that? @ "Steeples, Oliver" <Oliver.Steeples@compaq.com> wrote in messageH news:F498D199EDB12D468CD2C66680D3080116D60B@reoexc04.emea.cpqcorp.net...J > A BA350 has a 50 pin high-density connector and the disks are FNSE (fast > narrow single ended).  >  > 3940-should work > C > 2930-ultra card should work but need to terminate top 8 bits (bit 	 overkill)  >  > 2944-HVD thus won't work >  > J > I guess it is a BA350 and not a 356 with IO module.  Also have you triedI > different slots as it may be configured for slpit bus or the terminator L > jumper maybe attached.  The adapter scsi ID could be clashing with a drive > on the shelf.....  >  > Some ideas to think about. >  > Oliver >  > -----Original Message-----1 > From: Bill Bradley [mailto:wdbradley3@home.com] ( > Sent: Friday, August 03, 2001 11:39 AM > To: Info-VAX@Mvb.Saic.Com 2 > Subject: Any way to get a BA350 to work with NT? >  > / > OT, but, I need this working for a project...  > I > We're running into a problem trying to get a StorageWorks array cabinet K > working.  I'm setting up NT4 Enterprise Sever w/SP6a, and have a BA350-SBOF > cabinet with four SWXD3-WE 4.3 GB drives that I would setup in an NTI > software RAID 5 array as drive D.  The PC is a standard Dell GX200 withe PCIeG > slots, so, I cannot use the existing SCSI (DSSI?) cards from an Alphao > server, which are EISA.T > H >  I've tried the following Adaptec SCSI adapters, but, none of them can "see"w > the drives at the BIOS level:3 >t > 3940 >  > 2930 >nF > 2944 (this is a HV Differential SCSI, but, wasn't sure WHAT kind the cabinet  > supports). >M: > Any ideas on how to hook the cabinet up to a regular PC? > 	 > Thanks!m >( >    ------------------------------  $ Date: Fri, 3 Aug 2001 13:09:01 +01005 From: "Steeples, Oliver" <Oliver.Steeples@compaq.com>r4 Subject: RE: Any way to get a BA350 to work with NT?N Message-ID: <F498D199EDB12D468CD2C66680D3080116D60C@reoexc04.emea.cpqcorp.net>   Right, a BA350 is like so:   	slot  	7	6	5	4	3	2	1	0	scsit connse 	PSU	o  ! Single	6	5	4	3	2	1	0	use top scsi 	 connector  busu  ' split 	6	5	(4)	3	(2)	1	(0)	(x)=2nd bus,  use other scsi slot	 bus.    K For single bus the jumper needs to be behind slot 5 and the terminator (has  chips on) behind slot 1   . For split bus jumper=slot 1, terminator slot 5    F The shelf is terminated by the teminator, depending on how the adapterL terminates itself it may need a y-cable, not really a trilink.  Normally youK run a bios utility to set termination on the scsi card and ID it is set to..J If termination on the card is enabled then a straight cable system-storage should work.   	Oliver    -----Original Message-----/ From: Bill Bradley [mailto:wdbradley3@home.com] & Sent: Friday, August 03, 2001 12:58 PM To: Info-VAX@Mvb.Saic.Comi4 Subject: Re: Any way to get a BA350 to work with NT?     Thanks!   L I've got the drives in the four slots farthest away from the two internal 50J pin SCSI connectors (there are two power modules in this cabinet).  TakingC the fans off, there are two, I guess terminators, on the backplane.a  H Connecting the 3940 with a 68 to 50 pin adapter cable to either internal! (cabinet) connector doesn't work.E  E I heard about something called a Tri-Link adapter--would I need that?p@ "Steeples, Oliver" <Oliver.Steeples@compaq.com> wrote in messageH news:F498D199EDB12D468CD2C66680D3080116D60B@reoexc04.emea.cpqcorp.net...J > A BA350 has a 50 pin high-density connector and the disks are FNSE (fast > narrow single ended).s >n > 3940-should work >sC > 2930-ultra card should work but need to terminate top 8 bits (bitu	 overkill)u >i > 2944-HVD thus won't work >e >?J > I guess it is a BA350 and not a 356 with IO module.  Also have you triedI > different slots as it may be configured for slpit bus or the terminatortL > jumper maybe attached.  The adapter scsi ID could be clashing with a drive > on the shelf.....n >  > Some ideas to think about. >e > Oliver >s > -----Original Message-----1 > From: Bill Bradley [mailto:wdbradley3@home.com]t( > Sent: Friday, August 03, 2001 11:39 AM > To: Info-VAX@Mvb.Saic.Come2 > Subject: Any way to get a BA350 to work with NT? >t >]/ > OT, but, I need this working for a project...o >uI > We're running into a problem trying to get a StorageWorks array cabinetoK > working.  I'm setting up NT4 Enterprise Sever w/SP6a, and have a BA350-SBAF > cabinet with four SWXD3-WE 4.3 GB drives that I would setup in an NTI > software RAID 5 array as drive D.  The PC is a standard Dell GX200 withh PCI G > slots, so, I cannot use the existing SCSI (DSSI?) cards from an Alphas > server, which are EISA.  >oH >  I've tried the following Adaptec SCSI adapters, but, none of them can "see"h > the drives at the BIOS level:o >e > 3940 >e > 2930 >eF > 2944 (this is a HV Differential SCSI, but, wasn't sure WHAT kind the cabinet- > supports). >-: > Any ideas on how to hook the cabinet up to a regular PC? >r	 > Thanks!o >  >Z   ------------------------------  # Date: Fri, 03 Aug 2001 16:19:31 GMTt2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)4 Subject: Re: Any way to get a BA350 to work with NT?2 Message-ID: <naAa7.584$Yx2.11448@news.cpqcorp.net>  j In article <gbva7.7670$c8.1782706@news1.denver1.co.home.com>, "Bill Bradley" <wdbradley3@home.com> writes:. :OT, but, I need this working for a project...  F   Waaay off topic...  (No offense intended.)   Please simply consider F   who you are asking, and particularly consider the question that you    are asking of the audience.   F   Put another way, I would encourage asking Windows NT and Windows NT F   hardware support questions in a newsgroup that targets Windows NT.  F   I would not assume that OpenVMS folks would be familiar with WindowsG   NT, nor would I assume that Compaq folks would be familiar with Dell..  E   These disks are SCSI disks and this is a SCSI disk enclosure.  The .F   BA350 series itself is fairly old SCSI.  Specifically 8-bit (narrow)G   single-ended SCSI.  Most of the SWXD3 disks spin-up on power-up, and "    this can snarl some systems.    G   The Windows NT folks should be fairly more experienced at connecting sE   SCSI widgets to their hardware, once you get the cabling and power 3-   worked out with the target SCSI controller.U  C   In other words, tell the Windows NT newsgroup folks what the SCSI E   connector is, and that these are single-ended narrow SCSI widgets, 1H   and ask the folks what SCSI connector and what (PCI?) SCSI controller E   they would use for j-random SCSI disks and ask how these disks and ,@   this PCI SCSI controller would be configured under Windows NT.  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   Date: 3 Aug 2001 02:46 PST+ From: rankin@eql14.caltech.edu (Pat Rankin)o, Subject: Re: C v6.2 vs v6.4 volatile keyword0 Message-ID: <3AUG200102461572@eql14.caltech.edu>  1 In article <9kc78g$i7q@dispatch.concentric.net>,\a*  "Paul Dembry" <pade@trifox.com> writes... [...]"3 > I found out something about this problem.  We usec >a > a) volatile char *p; > as opposed to- > b) char * volatile p;0 >0F > a) used to work but no longer does.  b) works now.  Is it correct toL > interpret a) as saying that the data that *p points to is volatile whereasE > b) says that the value stored in p (the pointer value) is volatile?   ?      Yes.  C declarations generally make more sense if you readtA them from right to left.  The declaration of (a) is equivalent toa   a') char volatile *p;   @ and I think that form might make this situation clearer.  In (a)B and (a'), p is an ordinary pointer to volatile char; in (b), it isB a volatile pointer to ordinary char.  Since your code is trying to@ retain that pointer value in a local variable around a longjmp()A call, rather than worrying about whatever contents of memory it'soB pointing at, you want (b).  It was just coincidence--or deliberate@ but perhaps unwarranted conservatism on the part of the compilerB developers--that (a) worked as you wanted with the older compiler.  2                 Pat Rankin, rankin@eql.caltech.edu   ------------------------------   Date: 3 Aug 2001 01:43:45 -0700i. From: mistdragon@zdnetonebox.com (mist dragon)K Subject: Re: Comments from a VMS customer (and I would like to stay as one)e= Message-ID: <7500353b.0108030043.2db79a0b@posting.google.com>   a JF Mezei <jfmezei.spamnot@videotron.ca> wrote in message news:<3B6A30A8.D2935D73@videotron.ca>...S > Paul Dembry wrote:A > > Oracle owns Rdb, why would CPQ have anything to say about it?o > > Paul > P > Compaq doesn't hesitate to brag about how Oracle has a great realtionship withO > Compaq and how Oracle has great coomitment to Tru64. (as was witnessed duringtH > the wildfire launch some time ago). If Compaq and Oracle are such goodL > buddies, then Compaq should have convinced Oracle to stand behind Compaq'sT > decision to kill Alpha and announce the porting of the rdb legacy product to IA64. >   ? The interest between Compaq and Oracle (and any Oracle platformaB partner) is purely business based, .e.g Compaq pays Oracle to portF it's software. In this case, it has obviously decided it does not want to pay..   ------------------------------  % Date: Fri, 03 Aug 2001 11:16:31 +0100m% From: Alan Greig <a.greig@virgin.net>oK Subject: Re: Comments from a VMS customer (and I would like to stay as one)i8 Message-ID: <e7ukmt8ksu0h2fgvasb08pi8gu26383rb4@4ax.com>  C On 02 Aug 2001 22:23:10 GMT, "Paul Dembry" <pade@trifox.com> wrote:    >h> >Oracle owns Rdb, why would CPQ have anything to say about it? >Paulj  C Because RDB runtime and DBMS runtime were once part of BASE VMS anduC many(most?) of the large commercial VMS applications need this evenyB though it now comes from Oracle. VMS engineering is also currentlyD entirely dependent on VMS although I suppose they could unpick this.  F Lack of RDB and, more importantly, DBMS would be a showstopper for us.     -- Alan   ------------------------------   Date: 3 Aug 2001 07:19:20 -0500s9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)oK Subject: Re: Comments from a VMS customer (and I would like to stay as one)e3 Message-ID: <Jti5TE+AUNYw@eisner.encompasserve.org>   ` In article <e7ukmt8ksu0h2fgvasb08pi8gu26383rb4@4ax.com>, Alan Greig <a.greig@virgin.net> writes:E > On 02 Aug 2001 22:23:10 GMT, "Paul Dembry" <pade@trifox.com> wrote:r >  >>? >>Oracle owns Rdb, why would CPQ have anything to say about it?p >>Paul > E > Because RDB runtime and DBMS runtime were once part of BASE VMS and   ? Rdb was never part of VMS.  For some machines, for a while, Rdbr- licenses were bundled with new machine sales.   & For DBMS, not even that much was done.   ------------------------------  % Date: Fri, 03 Aug 2001 13:46:04 +0100o, From: Alan Greig <alan.greig@intl.fmcti.com>K Subject: Re: Comments from a VMS customer (and I would like to stay as one)l8 Message-ID: <qh6lmt44ft9kidad7ljo2adhjkls765hm2@4ax.com>  F On 3 Aug 2001 07:19:20 -0500, Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) wrote:r  a >In article <e7ukmt8ksu0h2fgvasb08pi8gu26383rb4@4ax.com>, Alan Greig <a.greig@virgin.net> writes: F >> On 02 Aug 2001 22:23:10 GMT, "Paul Dembry" <pade@trifox.com> wrote: >>   >>> @ >>>Oracle owns Rdb, why would CPQ have anything to say about it? >>>Paulo >>  F >> Because RDB runtime and DBMS runtime were once part of BASE VMS and >e@ >Rdb was never part of VMS.  For some machines, for a while, Rdb. >licenses were bundled with new machine sales. >o' >For DBMS, not even that much was done.w   Larry,  E I respectfully disagree. DBMS and RDB runtime were both licensed with C base VMS for a time in exactly the same way that he Fortran runtimecE libraries are still today. In fact Oracle took DEC to court over this E or at least threatened to. If it was not true to say that RDB runtime B was included with base VMS then it is not true to say that FortranC runtime is currently part of base VMS. Although you could argue thepF definition of "base vms".  You did not have to install RDB runtime. It was just there.   E It was this bundling that caused a lot of ISVs to make use of RDB andiD DBMS rather than competing products. Even DEC built DFG on top of an? RDB database on the assumption that RDB runtime would always befA available on any supported VMS system. ManMan was built on top oflF RDB/DBMS because it was bundled with VMS.  After DEC sold RDB and DBMSD all the sites making use of runtime calls to these components had toC pay a considerable amount of money to Oracle if upgrading as Oraclee? dropped the concept of runtime-only versions for both products.e  ( Development licenses were never bundled.   -- Alan   ------------------------------  % Date: Fri, 03 Aug 2001 13:42:28 -04001- From: JF Mezei <jfmezei.spamnot@videotron.ca>hK Subject: Re: Comments from a VMS customer (and I would like to stay as one)7, Message-ID: <3B6AE283.8564521C@videotron.ca>   Alan Greig wrote:sH > Lack of RDB and, more importantly, DBMS would be a showstopper for us.  N But you can stay on VMS/ALPHA for another 10 years right ? And since they saidN it would be a single code base, it is a given (yeah, right !) that all new VMS" features will benefit Alpha users.  L Think about this: what if the vast majority of VMS sites decide to stay withK Alpha and don't bother migrating to IA64-VMS and instead put in place *longI( term* migration to some other platform ?  J But putting new applications on a new platform, it means that the good oldN Alpha machines are not overworked and hence don't really need extra horsepowerM (since Alpha won't be increased in speed once its fabbing is done and a largeh stockpile made).  L Think about it: if you need to add stuff to your machine and the only way isK to go from Alpha to IA64 for the whole kit and caboodle, isn't it easier to I decentralise your stuff and add the new application on a separate box andcJ leave the Alpha-VMS alone in its corner to continue doing what it has beenN doing for years ? This way, you don't have to worry about porting your apps to< IA64, having dual platforms during a testing period etc etc.   ------------------------------  # Date: Fri, 03 Aug 2001 17:35:18 GMTiB From: Simon Clubley <simon_clubley@remove_me.excite.com-Earth.UFP>K Subject: Re: Comments from a VMS customer (and I would like to stay as one) 6 Message-ID: <qhBa7.16515$ar1.60633@www.newsranger.com>  - On Thu, 2 Aug 2001 23:56:28 -0500, in article-= <m9qa7.20197$j02.286338@news.goodnet.com>, Rich Jordan wrote:@ >eE >Paul Dembry wrote in message <9kcjse$7gu@dispatch.concentric.net>...oM >>> 5) Rdb. I do not use Rdb, and at my current employer, it is unlikely thatw >>IgI >>> ever will. As a long standing VMS customer however, I am aware of theoC >>> importance that other customers place on it been available. TheP >>announcementM >>> that Rdb will be ported to IA64 VMS should have been made on the same day J >>> of the porting announcement, and it should certainly have been made by >>now.M >>> This _is_ very damaging as it gives the image that the VMS port will be a F >>> port in name only, with few layered products actually been ported. >>>  ....... >>? >>Oracle owns Rdb, why would CPQ have anything to say about it?  >>Paul >> >aJ >Because Compaq should have made sure to talk to Oracle about it, gotten aF >porting agreement, and arranged for Oracle to announce that fact very& >shortly after the Q dropped the bomb. >1  I Yes, this was the point that I was trying to get across. I didn't make ityK clear in my original posting that I already knew that Oracle owns Rdb aftern it was sold off.  E What I also am hearing is that a lot of questions will be answered attC CETS2001. I look forward to seeing the announcements following thisiJ event. They will hopefully answer a lot of the questions that we all have.   Simon.   --  ; Simon Clubley, simon_clubley@remove_me.excite.com-Earth.UFPeK In the task of removing Microsoft from the marketplace, I have discovered a E truly remarkable plan, but this signature is too small to contain it.v   ------------------------------  # Date: Fri, 03 Aug 2001 13:42:07 GMTD1 From: "Mark D. Jilson" <jilly@clarityconnect.com>  Subject: Re: DCL... Oops...n2 Message-ID: <3B6AAAEA.55BED825@clarityconnect.com>  B My understanding of this issue is that it was fixed when initiallyH reported as it doesn't fail on OpenVMS VAX of recent vintage but the bugG has been reintroduced in some manner for OpenVMS Alpha V6.2 and above. sG Again if this is more than idle curiosity please report it to the local H CSC and you will be given a fix for it.  When a downloadable kit becomes@ available it will be published via the normal patch notification process.   "Kenneth H. Fairfield" wrote:i > N > One last note on this bug.  There is a DSNlink article that acknowledges the) > problem in the OpenVMS database titled,M > I >    "DEFINE Command Corrupts Image Context Preventing CONTINUE Commands"  > M > I have a copy of it with "Last Technical Review: 23-DEC-1996", although thenN > copy currently in the database has a revew date of 8-MAY-1997.  In any case,B > VMS Engineering have been sitting on this one for 4 1/2 years... > 	 >    -Kena > --8 > I don't speak for Intel, Intel doesn't speak for me...   -- sD Jilly	- Working from Home in the Chemung River Valley - Lockwood, NY0 	- jilly@clarityconnect.com			- Brett Bodine fan. 	- Mark.Jilson@Compaq.com			- since 1975 or so, 	- http://www.jilly.baka.com               -   ------------------------------   Date: 3 Aug 2001 10:36:07 -0500n- From: koehler@encompasserve.org (Bob Koehler)w0 Subject: DECset enhancements for ODS-5 and Java?3 Message-ID: <XQ$Fu1k0jC5$@eisner.encompasserve.org>g  A   The only enhancement I know of to DECset recently was the OSD-5 F   support in CMS.  It's incomplete.  SHOW HISTORY can't find lowercase>   or mixed case elements unless the element name is in quotes.  F   Recently we've been trying to write some MMS scripts to support JavaB   work we've been doing.  Simply trying to fetch elements from CMS@   has been problematic since MMS macros all expand in uppercase.A   (Our solution takes about 12 lines to fetch and 11 to compile).      Enhancements would be nice.i  ?   For anyone else trying to use MMS with Java, here's examples:a   fetch_java.mms:d .suffixes .java .java~  
 .java~.java :t&       	$open/write temp fetch_java.com       	$purge fetch_java.com 	$write temp "$set verify"-         $cmsfile = "$(mms$cms_library)" - "]" 9       	$cmsfile = cmsfile + "...]" + "$(mms$cms_element)"t!         $name = f$search(cmsfile)eN       	$line = "$cms fetch " + f$parse(name,,,"name") + f$parse(name,,,"type")       	$write temp linee       	$close temp 	$rename fetch_java.com ;1         $define dcl$path []c 	$fetch_java   A.java : A.java~     compile_java.mms:X .suffixes .class .java  
 .java.class :e(       	$open/write temp compile_java.com       	$purge compile_java.com 	$write temp "$set verify")         $name = f$search("$(mms$source)")eJ       	$line = "$javac " + f$parse(name,,,"name") + f$parse(name,,,"type")       	$write temp line        	$close temp 	$rename compile_java.com ;1         $define dcl$path []e 	$compile_java       	$deassign dcl$path    A.class : A.java    F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences CorporationE= NASA GSFC Flight Software       | Federal Sector, Civil Group E                                 | please remove ".aspm" when replying    ------------------------------  % Date: Fri, 03 Aug 2001 02:29:29 -0600u$ From: Lee Y T Mah <lytmah@cha.ab.ca>  Subject: Re: Dismounting a drive) Message-ID: <3B6A60E8.B1055315@cha.ab.ca>   R My problem was caused by the 12 LAT print queues spooling to device DSA6 through aS logical device name.  The logical name had been revised to a new drive, but the LATeR queues were still going to DSA6.  The cleanest way to break the link to DSA6 is toT delete each LAT print queue and recreate it.  Using our standard procedures, it tookO less than 30 seconds to delete/recreate j$ltax, k$ltax, l$ltax, m$ltax.  Again,V  thanks for all your suggestions.     John Santos wrote:  ( > On Thu, 2 Aug 2001, Lee Y T Mah wrote: >iQ > > Compaq Support had recommended the ANA/SYS procedure.  I did it to one of thekR > > nodes (M).  There was a print symbiont linked to DSA6.  I deleted the symbiontM > > and there were no more references to DSA6 under another ANA/SYS.  Made no Q > > difference to the "Trans Count".  It still showed as 13 on node M.  Just now, > > > I tried dismounting DSA6 on node M.  Here are the results. > >l& > > M _J$DKB0:[RAHSEC] $ sho dev dsa6: > > P > > Device                  Device           Error    Volume         Free  Trans > > MnthP > >  Name                   Status           Count     Label        Blocks Count > > CntsP > > DSA6:                   Mounted              0  SS6            3712400    13 > > 4tE > > $2$DUA14:          (K)  ShadowSetMember      0  (member of DSA6:)-E > > $2$DUA36:     (HSC004)  ShadowSetMember      0  (member of DSA6:)o > >n- > > M _J$DKB0:[RAHSEC] $ dismount/abort dsa6:e1 > > %DISM-W-CANNOTDMT, DSA6: cannot be dismountedY3 > > %DISM-W-SPOOLEDEV, 12 devices spooled to volumec > >2K > > Thanks to all who replied.  I guess I'll just have to reboot the nodes.  >,G > Stopping the symbiont is irrelevent to freeing the channels.  I thinkfF > you need to stop the queues (stop/queue/next) and change the devices8 > to use a different spool disk (set device/nospool, set? > device/spool=(queue-name, disk:) and then restart the queues.i6 > (Stopping the queues may not be strictly necessary.) > E > If any processes currently have any of the 12 spooled devices open, F > they have channels open to temp files on the disk, so they will haveC > to be stopped as well.  How to do that depends on the the programsF > that the process is running.  However, I think this would have shownD > up on the show device/files list, so I don't think it is an issue. > G > Be sure to change your system startup files the same way or rebooting. > wont help! >w > > "Mark D. Jilson" wrote:o > >pB > > > For all nodes that have this volume mounted do the following > > >e > > > $ ANA/SYSr > > > SET OUTPUT CHAN.LISa > > > SHOW PROCESS/CHANNEL ALL
 > > > EXIT; > > > $ SEARCH CHAN.LIS "Extended PID: ","DSA6:"/OUTPUT=X.Xo$ > > > $ SEARCH X.X "DSA6:"/WIN=(1,1)J > > > If there are no matches here then you may dismount DSA6 on this nodeN > > > otherwise the output should show what process has an open channel to theL > > > disk and you'll need to get that process to close the channel.  How toE > > > do this will depend on what application the process is running.- > > >-J > > > Remember to DELETE CHAN.LIS.,X.X. before moving on to the next node. > > >  > > > Lee Y T Mah wrote: > > > >eO > > > > I have an old shadow set (DSA6) which no longer fits in our plans.  AllnM > > > > the data files on it have been deleted and no users are accessing theeN > > > > drive anymore.  I am ready  to dismount it but the "Trans Count" stillO > > > > shows "13".  Following is the info on the drive.  Can anyone suggest asdM > > > > to what (phantom) files may still be lingering on the drive and beingt > > > > accessed?  TIA.. > > > >o9 > > > > K _DSA6:[000000] $ dir/size/date dsa6:[000000...]l > > > >r > > > > Directory DSA6:[000000]  > > > >eB > > > > 000000.DIR;1                    3  24-MAY-1990 13:00:09.21> > > > > BACKUP.SYS;1                0  24-MAY-1990 13:00:09.21> > > > > BADBLK.SYS;1                5  24-MAY-1990 13:00:09.21> > > > > BADLOG.SYS;1                0  24-MAY-1990 13:00:09.21> > > > > BITMAP.SYS;1              185  24-MAY-1990 13:00:09.21@ > > > > CONTIN.SYS;1                  0  24-MAY-1990 13:00:09.21? > > > > CORIMG.SYS;1                 0  24-MAY-1990 13:00:09.21n= > > > > INDEXF.SYS;1           46135  24-MAY-1990 13:00:09.21b? > > > > SECURITY.SYS;1               1   5-DEC-1994 02:14:22.44sA > > > > VOLSET.SYS;1                   0  24-MAY-1990 13:00:09.21l > > > > ( > > > > Total of 10 files, 46329 blocks. > > > >o, > > > > K _DSA6:[000000] $ show device dsa6: > > > > I > > > > Device               Device           Error   Volume         Freet > > > > Trans  MntM > > > >  Name                 Status            Count   Label          Blocks  > > > > Count  Cnt; > > > > DSA6:              Mounted              0       SS6  > > > > 3712400    13    4I > > > > $2$DUA14:   (HSC001)   ShadowSetMember      0   (member of DSA6:)aO > > > > $2$DUA36:                (L)  ShadowSetMember      0  (member of DSA6:)3 > > > >@4 > > > > K _DSA6:[000000] $ analyze/disk/repair dsa6:G > > > > Analyze/Disk_Structure/Repair for _DSA6: started on  2-AUG-2001a > > > > 11:08:36.88  > > > >n6 > > > > %ANALDISK-I-OPENQUOTA, error opening QUOTA.SYS* > > > > -SYSTEM-W-NOSUCHFILE, no such file > > > >o. > > > > K _DSA6:[000000] $ show dev/file dsa6: > > > >OA > > > > Files accessed on device DSA6: on  2-AUG-2001 11:09:42.47h > > > >e+ > > > > Process name      PID     File namea6 > > > >                 00000000  [000000]INDEXF.SYS;1 > > > > K _DSA6:[000000] $ > > > > 
 > > > > -- > > > > Leed > > > > C > > > > Lee Y T Mah                        Capital Health AuthoritytG > > > > Email: lytmah@cha.ab.ca            Information Systems, RAH CSCn< > > > > Phone:  (780) 477-4725, 477-4233   10240 Kingsway NWG > > > > Fax:      (780) 491-5119, 491-5619    Edmonton, AB, CAN  T5H3V9- > > >- > > > --L > > > Jilly   - Working from Home in the Chemung River Valley - Lockwood, NYP > > >         - jilly@clarityconnect.com                      - Brett Bodine fanP > > >         - Mark.Jilson@Compaq.com                        - since 1975 or so9 > > >         - http://www.jilly.baka.com               -  > >l > > -- > > Leeo > >P? > > Lee Y T Mah                        Capital Health AuthoritynC > > Email: lytmah@cha.ab.ca            Information Systems, RAH CSCr8 > > Phone:  (780) 477-4725, 477-4233   10240 Kingsway NWC > > Fax:      (780) 491-5119, 491-5619    Edmonton, AB, CAN  T5H3V9p > >t > >M > >h > >I >p > --
 > John Santosk > Evans Griffiths & Hart, Inc. > 781-861-0670 ext 539   -- Leeo  ; Lee Y T Mah                        Capital Health Authority ? Email: lytmah@cha.ab.ca            Information Systems, RAH CSCg4 Phone:  (780) 477-4725, 477-4233   10240 Kingsway NW? Fax:      (780) 491-5119, 491-5619    Edmonton, AB, CAN  T5H3V9J   ------------------------------  % Date: Fri, 03 Aug 2001 13:32:59 -0400p- From: JF Mezei <jfmezei.spamnot@videotron.ca>i  Subject: Re: Dismounting a drive, Message-ID: <3B6AE04A.F8AF8A5D@videotron.ca>   Lee Y T Mah wrote:T > queues were still going to DSA6.  The cleanest way to break the link to DSA6 is to/ > delete each LAT print queue and recreate it. M  N If you stop the queue completely, you should then be able to SET DEV/NOSPOOLED on the LTA devices.n   ------------------------------   Date: 3 Aug 2001 08:24:09 -0700-' From: c_goudy@hotmail.com (Chris Goudy)  Subject: Re: f$parse glitchm= Message-ID: <48f4addd.0108030724.16462d46@posting.google.com>    You guys are geniuses!!   F I don't know how in the world this logical got set up, but it did. The# logical LDR20_NAV_DEV is defined as-E "FMS_FMS20:[FMS20.CONFIG]LDR00_NAV_DEV.CONFIG", so naturally when youI< parse the NAME from it you are going to get "LDR00_NAV_DEV".  / Many thanks to everyone who helped me on this. 0  1 Now I have to figure out why that logical exists.:   ------------------------------  $ Date: Fri, 3 Aug 2001 08:41:34 -0400+ From: "Main, Kerry" <Kerry.Main@compaq.com>a- Subject: RE: Few People in DEC Understood.... R Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF4D4954F@kaoexc01.americas.cpqcorp.net>   Steve,  K As a fyi (and for the benefit of others), ALL-IN-1 was upgraded and renamed F to OfficeServer awhiles back. Current version of OfficeServer is V6.0.  / Beware - semi-marketing statement to follow :-)h  J For those that want a clustered, fully load balanced (across multi-site ifI required), 64bit very secure email system that is free from the x86 based K server virus's of the day, supports latest email standards, allows multiple L different types of clients (including MS Outlook), can be managed by WindowsL NT/9x clients  (help desk folks) and that can also run other applications onK the same cluster (e.g. Secure Web Server, File-n-Print) - OfficeServer is a  very good solution.e  
 Reference:A http://www.compaq.com/info/SP6149/SP6149PF.PDF (OfficeServer SPD)a  ) ok, ok, enough of the marketing stuff ...a   :-)o   Regards,  
 Kerry Main Senior Consultant  Compaq Canada Corp.a Professional Servicest Voice: 613-592-4660B Fax  :  819-772-7036 Email: Kerry.Main@Compaq.com     -----Original Message-----8 From: Steve Reece [mailto:SYSTEM@ipl.demon.co.nospam.uk] Sent: July 31, 2001 5:42 PMi To: Info-VAX@Mvb.Saic.Comr- Subject: Re: Few People in DEC Understood....d    > (This may be where ALL-IN-1 came into this discussion but....)H Isn't ALL-IN-1 still winning new customers like the US Government in theF form of the White House mail system?  I thought they'd installed a newF GS series system running ALL-IN-1 for handling mail since this does itC more reliably than the Microsoft system they migrated onto and thenr  migrated off to return to VMS...  ? You don't get much more real than the US Government, do you Mr.  President (Clinton, that is)?    Steve.     Fred Kleinsorge wrote: > L > It was... ugly.  Especially in the beginning.  At least one of the reasonsK > was that it had to accomplish everything it did without any (or grudging)d > help from engineering. > L > But frankly I don't care much about how elegant it might or might not have4 > been - it solved real problems for real customers. > ? > JF Mezei wrote in message <3B6590FB.4F1408D9@videotron.ca>...s > >Fred Kleinsorge wrote:eL > >> Was it ugly?  You bet.  But what it did was solve a lot of problems for a I > >> lot of customers.  *It* sold the systems.  Yup... JUST LIKE WINDOWS.m > >aI > >Beauty is in the eye of the beholder. ALL-IN-1 was a large applicationy withE > >lots and lots of files. By VMS standards, it was huge, compared to- > compilers,J > >TPU etc. So I can understand that many would have found A1 to be a huge > ugly monster.o > >bL > >But when those engineers didn't realise is that part of the reason A1 was > soL > >big is that it shipped with about half of the source code which customersJ > >could then copy or modify. The A1 engineers also devised a logical nameI > >structure that allowed you to store your changed modules in a separategJ > >directory which would not be overwritten the next time you upgraded A1, and0F > >you could then run a report that would show you how each customized module# > >would work with the new version.9 > >$J > >However, to the casual observer, the large number of directories seemed > like aJ > >big mess, but once you understood why they were there, it made a lot of > sensev! > >and was in fact quite elegant.n   --  G "A shadow fell over her face; clear, as if the composure were rent liketE a veil.  And her lips parted, but only with a short intake of breath. A Then she said, 'Well, then you are right.  Indeed, we are even.'"m% 		Louis, "Interview with the Vampire"g   ------------------------------  $ Date: Fri, 3 Aug 2001 12:02:35 -0400- From: "Peter Weaver" <peter.weaver@stelco.ca>r- Subject: Re: Few People in DEC Understood....w0 Message-ID: <BWza7.3871$Z2.28817@nnrp1.uunet.ca>  6 "Main, Kerry" <Kerry.Main@compaq.com> wrote in messageL news:BE56C50EA024184DAF48F0B9A47F5CF4D4954F@kaoexc01.americas.cpqcorp.net... >...1 > Beware - semi-marketing statement to follow :-). >1L > For those that want a clustered, fully load balanced (across multi-site ifK > required), 64bit very secure email system that is free from the x86 basednD > server virus's of the day, supports latest email standards, allows multipleF > different types of clients (including MS Outlook), can be managed by WindowsiK > NT/9x clients  (help desk folks) and that can also run other applications  onK > the same cluster (e.g. Secure Web Server, File-n-Print) - OfficeServer isn ae > very good solution.s >  > Reference:C > http://www.compaq.com/info/SP6149/SP6149PF.PDF (OfficeServer SPD)r >s+ > ok, ok, enough of the marketing stuff ...- >...  J No Kerry, it is not enough. Please do more! But not to us, send that to myI boss' boss' boss (or even to his boss) on a nice colour glossy. Make sure.L that the "free from the X86 based server virus's" part is in extremely large print.   -- Peter WeaverJ Using a WIN NT/WIN 2000 box to manage your VMS systems is like towing your7 mechanic in a 5th wheel motor home behind your Porsche.b   ------------------------------  % Date: Fri, 03 Aug 2001 13:51:57 -0400o- From: JF Mezei <jfmezei.spamnot@videotron.ca>W- Subject: Re: Few People in DEC Understood....l, Message-ID: <3B6AE4BB.7F5FE182@videotron.ca>   "Main, Kerry" wrote:L > For those that want a clustered, fully load balanced (across multi-site ifK > required), 64bit very secure email system that is free from the x86 based  > server virus's of the day,  N Sorry, but ALL-IN-1 or PMDF might be immune from DOS/NT viruses, but they willN happily deliver messages to users on PCs and if those message contain a virus,N tough luck. Furthermore, when your server runs on a non-standard platform suchN as VMS, the odds of having the latest and greatest virus detection for inboundI and outbound emails are next to nil, so it is in fact harder to intercepteJ viruses at the server level before thei are delivered to the end users and ruin their PC.  M > the same cluster (e.g. Secure Web Server, File-n-Print) - OfficeServer is a  > very good solution.-  N ALL-IN-1 has very neat features, such as WEB access to your file cabinet, veryM powerful scripting language with integrated to VMS features such as access toDH queues, SYSUAF fields etc. HOWEVER, it lacks one big basic thing: directL access to SMTP servers. The concept of Message Router or Mailbus400 is a bitM dated since most everything is now transfered between hosts (even internally)a
 with SMTP.   ------------------------------  % Date: Fri, 03 Aug 2001 08:31:50 +0200a= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>r1 Subject: Re: How2 DECnet-copy with two passwords? ) Message-ID: <3B6A4556.CB535A5F@gtech.com>    "Holi - Holitska Andrs" wrote:o: >   Maybe a "tipical user question": how can I decnet-copy= >   files, when I don't have decnet-proxy on the target node,a( >   and have not one, but two passwords?  
 Not possible.p  5 It is documented in the DECnet manual as not possiblem> (or at least it was 15 years ago in the DECnet phase IV docs).   Arne   ------------------------------  # Date: Fri, 03 Aug 2001 14:06:03 GMT 3 From: Tim Llewellyn <tim.llewellyn@cableinet.co.uk>h? Subject: Re: IPF Console Bootstrap (was: No chance for OpenVMS)t/ Message-ID: <3B6AAE31.D323FCB3@cableinet.co.uk>j   Fred Kleinsorge wrote: > M > Nope.  The console ROM has full support for network booting.  It uses "PXE" L > (preboot execution environment) which uses things like DHCP, and MTFTP.  IK > am by no means an expert, but our expert tells me that this will not be ar
 > problem.  ; that was my point, Fred, it was JF who was being skeptical.$   regards, > G > Tim Llewellyn wrote in message <3B682EC4.4ED4B9AB@cableinet.co.uk>.... > >1 > >  > >JF Mezei wrote: > >>I > >> OK, so Fred has explained that the IFE console will get all its cool- > stuff-M > >> (utilities/applications) from a FAT partition that lives as an ODS2 file  > on > >> the system disk.p > >>L > >> This means that upon power up, if I want to check system sanity, format > drivesH > >> etc, the console will have to access the system disk to fetch those
 > "programs".a > >>M > >> How will that translate where there is no system disk and the machine isR( > >> expected to boot from the network ? > >1J > >presumably if there is no network boot support in ROM you run a network > >bootv > >program from floppy or CD.i > >--.  > >Tim.Llewellyn@cableinet.co.uk > >aE > >Standard disclaimer applies. My views in no way represent those oft$ > >my employers or service provider.   -- e Tim.Llewellyn@cableinet.co.uk  i  C Standard disclaimer applies. My views in no way represent those of e! my employers or service provider.    ------------------------------  * Date: Fri, 3 Aug 2001 10:31:06 +0100 (GDT)) From: Andy Harper <Andy.Harper@kcl.ac.uk> : Subject: Re: Looking for a Pine.exe for TCPIP 5.x services: Message-ID: <SIMEON.10108031006.B@kings-ut-srv1.kcl.ac.uk>  E On Thu, 02 Aug 2001 18:08:50 -0500 Thomas G Wirt <twirt@kittles.com> e wrote:  I > I am trying to use pine to send attachments from a VMS 7.1 Alpha systemuJ > with TCPIP 5.0a.  I have a working copy of Pine, but it does not seem toI > want to work with my TCP stack.  I do not have a C compiler and am verydJ > interested in finding a Pine executable that is ready to work with TCPIP% > 5.x that is available for download.m  C   If you're using the special vms version of PINE (latest of which tA   is 3.95-2) - it runs on top of MadGoat's netlib software which  !   supports all known TCP stacks.       So you may need to6      (a) rebuild it with netlib if you haven't already"      (b) install the latest netlib  E I am assuming that the latest netlib supports tcp 5 here. Maybe matt sB or hunter can confirm?  If it doesn't, then you're probably stuck " without doing some code rewriting.   Regards,   ----------------------! Andy Harper B.Sc., M.B.C.S, C.Enge Systems and Mail Manager Kings College London   ------------------------------  % Date: Fri, 03 Aug 2001 09:45:47 -0500e' From: Thomas G Wirt <twirt@kittles.com>s: Subject: Re: Looking for a Pine.exe for TCPIP 5.x services+ Message-ID: <3B6AB91B.D50368E3@kittles.com>e   Andy Harper wrote: > F > On Thu, 02 Aug 2001 18:08:50 -0500 Thomas G Wirt <twirt@kittles.com> > wrote: > K > > I am trying to use pine to send attachments from a VMS 7.1 Alpha systemeL > > with TCPIP 5.0a.  I have a working copy of Pine, but it does not seem toK > > want to work with my TCP stack.  I do not have a C compiler and am veryiL > > interested in finding a Pine executable that is ready to work with TCPIP' > > 5.x that is available for download.s > D >   If you're using the special vms version of PINE (latest of whichB >   is 3.95-2) - it runs on top of MadGoat's netlib software which" >   supports all known TCP stacks.  E I am using Version 3.91-2 for VMS.  This is the newest version that IAF could find for VMS.  I am using NETLIB 2.3.  If anyone can point me toF new versions of these (I need .exe or at lest complete object files) IC would appreciate it.  I have attempted to link Pine with the NETLIBuH option.  It looked like it worked, but I still get "[Error connecting toD mail server: Can't connect to worf,25: 0x0000022c]" message (after aB time out of about 60 seconds) when I try to send a simple message.  F I would really like this to work and would appreciate any help at all.   Thanks,    Thomas WIrt    <snip>   ------------------------------  % Date: Fri, 03 Aug 2001 17:54:31 +0200 < From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <noone@home.com>: Subject: Re: Looking for a Pine.exe for TCPIP 5.x services( Message-ID: <3B6AC937.A9B04DF3@home.com>   Hi.s" Not an direct answer maybe, but...  ? I'v been sending MIME attachements from VMS 7.2-1 in production = environment for a couple of years using the MIME utility that < is delivered with VMS for the MIME encoding, and a tool from: Arne Vajhoej, called NBL (No Blank Line) to send it. Works> in an "lights-out" 24*7 enviroment like a sharm. This is BATCH( oriented which may not be what you want.   Regards  Jan-Erik Sderholm.A   ------------------------------  $ Date: Fri, 3 Aug 2001 09:05:57 -07000 From: "Hank Vander Waal" <hvanderw@novagate.com>F Subject: Marketing - any kind !  was  Few People in DEC Understood....; Message-ID: <000801c11c36$2ed92900$9c96a8c6@manufact5l8vs8>A  L Don't apologize Kerry   for some of us little guys this is the only place we hear about this stuff!  F Personally I'll take the info anyway I can get it 'cause it sure isn'tG getting out any other way !!   This is the first I ever heard of Office  Server.M   Hank Vander Waal Mansco Hudsonville MI  / Beware - semi-marketing statement to follow :-)D  J For those that want a clustered, fully load balanced (across multi-site ifI required), 64bit very secure email system that is free from the x86 based K server virus's of the day, supports latest email standards, allows multiplenL different types of clients (including MS Outlook), can be managed by WindowsL NT/9x clients  (help desk folks) and that can also run other applications onK the same cluster (e.g. Secure Web Server, File-n-Print) - OfficeServer is a0 very good solution.   
 Reference:A http://www.compaq.com/info/SP6149/SP6149PF.PDF (OfficeServer SPD)i  ) ok, ok, enough of the marketing stuff ...A   :-)a   Regards,  
 Kerry Main Senior ConsultantA Compaq Canada Corp.r Professional Services  Voice: 613-592-4660H Fax  :  819-772-7036 Email: Kerry.Main@Compaq.com     -----Original Message-----8 From: Steve Reece [mailto:SYSTEM@ipl.demon.co.nospam.uk] Sent: July 31, 2001 5:42 PM  To: Info-VAX@Mvb.Saic.Com - Subject: Re: Few People in DEC Understood....0    > (This may be where ALL-IN-1 came into this discussion but....)H Isn't ALL-IN-1 still winning new customers like the US Government in theF form of the White House mail system?  I thought they'd installed a newF GS series system running ALL-IN-1 for handling mail since this does itC more reliably than the Microsoft system they migrated onto and then   migrated off to return to VMS...  ? You don't get much more real than the US Government, do you Mr.o President (Clinton, that is)?l   Steve.     Fred Kleinsorge wrote: > L > It was... ugly.  Especially in the beginning.  At least one of the reasonsK > was that it had to accomplish everything it did without any (or grudging)  > help from engineering. > L > But frankly I don't care much about how elegant it might or might not have4 > been - it solved real problems for real customers. > ? > JF Mezei wrote in message <3B6590FB.4F1408D9@videotron.ca>...  > >Fred Kleinsorge wrote: L > >> Was it ugly?  You bet.  But what it did was solve a lot of problems for aSI > >> lot of customers.  *It* sold the systems.  Yup... JUST LIKE WINDOWS.  > > I > >Beauty is in the eye of the beholder. ALL-IN-1 was a large application. withE > >lots and lots of files. By VMS standards, it was huge, compared to2 > compilers,J > >TPU etc. So I can understand that many would have found A1 to be a huge > ugly monster.- > >-L > >But when those engineers didn't realise is that part of the reason A1 was > soL > >big is that it shipped with about half of the source code which customersJ > >could then copy or modify. The A1 engineers also devised a logical nameI > >structure that allowed you to store your changed modules in a separateuJ > >directory which would not be overwritten the next time you upgraded A1, andlF > >you could then run a report that would show you how each customized module# > >would work with the new version.r > >_J > >However, to the casual observer, the large number of directories seemed > like aJ > >big mess, but once you understood why they were there, it made a lot of > senseh! > >and was in fact quite elegant.t   --G "A shadow fell over her face; clear, as if the composure were rent likeOE a veil.  And her lips parted, but only with a short intake of breath. A Then she said, 'Well, then you are right.  Indeed, we are even.'" % 		Louis, "Interview with the Vampire"t   ------------------------------  % Date: Fri, 03 Aug 2001 13:54:52 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>iJ Subject: Re: Marketing - any kind !  was  Few People in DEC Understood...., Message-ID: <3B6AE56A.9A9BE593@videotron.ca>   Hank Vander Waal wrote:iH > Personally I'll take the info anyway I can get it 'cause it sure isn'tI > getting out any other way !!   This is the first I ever heard of Offices	 > Server.c  K When you consider the total silence about ALL-IN-1 (aka Office Server) fromsM Compaq, and not even a commitment to port it to IA64, I wouldn't really touch6K that product for a new install right now for as much as I love playing withsD ALL-IN-1. If Compaq were to start to use its own products instead ofL Microsoft's I might consider pushing A1 again. But the last time I suggestedK ALL-IN-1 to a potential custoemr, I was laughed at because even Digital hadaI interanlly dumped the product and stopped using it. Why should a customery5 invest in a platform that even its maker has dumped ?n   ------------------------------  % Date: Fri, 03 Aug 2001 11:18:56 +0100n% From: Alan Greig <a.greig@virgin.net>r: Subject: Re: Media Kits (was Re: IPF Console Bootstrap...)8 Message-ID: <ghukmtk0h7j10eovd9bgcc20pjhbd4siaq@4ax.com>  , On Thu, 02 Aug 2001 17:59:03 -0400, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote::   >Hoff Hoffman wrote:M >>   That won't be a problem, since we will be shipping each IPF distributionFL >>   CD-ROM pre-loaded into a hermetically sealed CD-ROM drive, and each IPFN >>   system will include a processor-integrated wireless LAN (WLAN) connection >rL >Why not just integrate an Iridium handset into every VMS system and have itI >load the latest and greatest console from a satellite link directly from 	 >Compaq ?l  C Hmm, I'm sure there are battlefield tanks configured something likeoF that. Info given at a VMS presentation in Edinburgh several years ago.   -- Alan   ------------------------------   Date: 3 Aug 2001 10:02:51 -0500 - From: koehler@encompasserve.org (Bob Koehler)e: Subject: Re: Media Kits (was Re: IPF Console Bootstrap...)3 Message-ID: <8r$Bl04LNWUs@eisner.encompasserve.org>i  g In article <nqja7.559$Yx2.10217@news.cpqcorp.net>, hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) writes:    >  The Tempest oM >   environment variant of the WLAN media distribution kit will be available FL >   at a small additional cost, though the media self-destruction mechanism ' >   mandates ground shipment only.  :-)o  D Why not simply use an 8 1/2 inch floppy?  Security by obscurity and  self destructing!S  F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences Corporation = NASA GSFC Flight Software       | Federal Sector, Civil GrouprE                                 | please remove ".aspm" when replyingb   ------------------------------  $ Date: Fri, 3 Aug 2001 10:14:39 -07000 From: "Hank Vander Waal" <hvanderw@novagate.com>: Subject: RE: Media Kits (was Re: IPF Console Bootstrap...); Message-ID: <000e01c11c3f$c7c4b7c0$9c96a8c6@manufact5l8vs8>r  ( just please please don't suggest a tu-58     -----Original Message-----4 From: Bob Koehler [mailto:koehler@encompasserve.org]% Sent: Friday, August 03, 2001 8:03 AM  To: Info-VAX@Mvb.Saic.Comw: Subject: Re: Media Kits (was Re: IPF Console Bootstrap...)    2 In article <nqja7.559$Yx2.10217@news.cpqcorp.net>,4 hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) writes:   >  The TempestL >   environment variant of the WLAN media distribution kit will be availableK >   at a small additional cost, though the media self-destruction mechanismm' >   mandates ground shipment only.  :-)l  C Why not simply use an 8 1/2 inch floppy?  Security by obscurity ande self destructing!g  F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences Corporation4= NASA GSFC Flight Software       | Federal Sector, Civil Group:E                                 | please remove ".aspm" when replyingr   ------------------------------   Date: 3 Aug 2001 08:40:41 -0700e- From: afeldman@gfigroup.com (Alan E. Feldman)a1 Subject: Mystery jumper on MicroVAX 3100 Model 80m< Message-ID: <af1e4ce6.0108030740.49c7032@posting.google.com>  ? NOTE: This is cross-posted to comp.sys.dec.micro, comp.sys.dec,d) comp.os.vms. Reply to comp.sys.dec.micro.r  C We have obtained some MicroVAX 3100 Model 80's. Some of them have aaB jumper in a certain place and others don't have the jumper in that? same place. We have tried it with and without and don't see anyeC difference. This jumper is located just behind and on the MMJ-portsaC side of the 25-pin asynchronous modem control port. There are three F pins and this jumper currently connects the two innermost pins (in the one currently on my desk).  : Can anyone please tell us what this jumper is for? Thanks.  B If it matters, we are running VMS 6.1 on some nodes and VMS 6.2 on others.t  ? NOTE: This is cross-posted to comp.sys.dec.micro, comp.sys.dec,s) comp.os.vms. Reply to comp.sys.dec.micro.    Disclaimer: JMHO   &-) Alan E. Feldmanl afeldman@gfigroup.come   ------------------------------  $ Date: Fri, 3 Aug 2001 13:26:06 +0200$ From: "Jakob Erber" <erberj@post.ch> Subject: openssh Message-ID: <3b6a8a4f$1@hcwe67>-   Hello,  0 is there a port of OpenSSH (client) for OpenVMS?   best regards   Jakob@   ------------------------------  . Date: Fri, 3 Aug 2001 08:12:05 +0200 (MET DST)& From: Rudolf Wingert <win@fom.fgan.de> Subject: Re: OpenVMS  + Itanium 6 Message-ID: <200108030612.IAA06100@sinet1.fom.fgan.de>   Hello,  J I think that there are a lot of freeware tools, which should be integratedH into OpenVMS, such as a fast DELETE. Also I would like to be able to setK the onboard write cache of to disk to the enable state during MOUNT (I knowsH that this will be to my own risk, but we need it). A GUI driven user in-E terface for DCPS would be nessesary. This should use the standard PPD E (like Windows), so that the user can use all the features of the usedtJ printers. The performance management should be based by Compaq. A standardH within OpenVMS. An easy way to use Windows typical standard device, suchG as ZIP, MO, DVD, CD-RW, DVD-RW, ... Also should we be able to read (andeH write) NTFS and FAT formated disk under OpenVMS, as we can (with PCDISK)G with DOS formated diskettes. Support for the extended filespecificationoD format on CDs (e.g. Rock Rich). Better symbol transfer out of a PIPE# command. Boot via GIGAbit Ethernet.    Regards Rudolf Wingert   ------------------------------  . Date: Fri, 3 Aug 2001 08:41:14 +0200 (MET DST)& From: Rudolf Wingert <win@fom.fgan.de> Subject: Re: OpenVMS  + Itanium-6 Message-ID: <200108030641.IAA06179@sinet1.fom.fgan.de>   Hello,  E a veryf nice feature will be, that a cluster client copies the systemHE to a local disk (e.g. as memory dump). The next time it must boot, iteD boots from the local disk and, if neccessary, it updates the changedF software from the server. This would allow a fast boot and low network traffic.   Regards Rudolf Wingert   ------------------------------  % Date: Fri, 03 Aug 2001 10:35:36 +0100h% From: Alan Greig <a.greig@virgin.net>t Subject: Re: OpenVMS  + Itanium 8 Message-ID: <qnrkmts2e632isceo48vaakd1pr2fo32eu@4ax.com>  ; On Fri, 3 Aug 2001 08:12:05 +0200 (MET DST), Rudolf Wingertl <win@fom.fgan.de> wrote:   >Hello,  >jK >I think that there are a lot of freeware tools, which should be integrated6I >into OpenVMS, such as a fast DELETE. Also I would like to be able to seteL >the onboard write cache of to disk to the enable state during MOUNT (I know  D You can do this already with sys$etc:scsi_mode. It will allow you to@ set the SCSI mode bit to turn on drive write-back cache. I can'tF recall the specific incantation you need though. A glance through SCSI* docs in conjunction with $ scsi_mode -help9 should give the info you need if nobody can post it here.      -- Alan   ------------------------------  % Date: Fri, 03 Aug 2001 10:28:39 +0100i% From: Alan Greig <a.greig@virgin.net>t9 Subject: Re: Outsourcing of VMS support to Digital India! 8 Message-ID: <0hrkmtgh3dvhe3ei2k0nur3nu0hmdf87rn@4ax.com>  , On Thu, 02 Aug 2001 13:30:57 -0400, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:    >"Alan B." wrote: A >> system and application engineers.  This team has been actively>H >> involved in the development, enhancement, maintenance and support forC >> applications and products on the VMS platform since 1995 and haseA >> played a significant role in migrating applications for Compaq < >> customers worldwide to newer technologies and platforms.  >cN >It is the last part of the sentence that Compaq should learn NEVER to includeN >in text, especially now that it has embarked on what seems to be a mission toN >kill off the remainder of Digital's products in favour of newer techonologies >and platforms.c  E Don't jump to conclusions. When Digital India talk about migration to0E "newer technologies and platforms" they mean to Alpha/VMS ! Obviously:4 they haven't been indoctrinated in Compaq-speak yet.  ( From a migration case study on the site:   Project Description = The Project involved Migration of around 30 Batch and on-line D Applications from IBM 4361 to ALPHA Server and carryout the requiredC Year 2000 fixes to the code for the Programs to behave correctly int Year 2000 and thereafter.t The Project scope covered:4 	Migration of 1029 IBM COBOL programs to DEC COBOL ; 	Re-writing of 16 IBM CICS programs using DEC COBOL SCREENP SECTIONh9 	Carryout Year 2000 fixes as necessary in the above 1029c
 programs.   B In both Migration & Year 2000 Fixing Processes, tools were used to4 ensure maximum automation in executing the Project.        -- Alan   ------------------------------  % Date: Fri, 03 Aug 2001 10:23:13 +0100 % From: Alan Greig <a.greig@virgin.net> 9 Subject: Re: Outsourcing of VMS support to Digital India!b8 Message-ID: <s3rkmtseri6173idqtsc4ba9jeur60tqe8@4ax.com>  / On Thu, 2 Aug 2001 10:58:19 -0500, WILLIAM WEBBn <WWEBB1@email.usps.gov> wrote:   >- >Holy COW?  Digital India? >  >Surely you jest.   # Go to http://www.digitalindiasw.com-  A It's a bit like traveling back in time to the good old days. AlsoeB notice that, by concentrating on the VMS/Enterprise market they've6 boosted profits to record levels. Compaq US take note.       >WWWebbl >c >> -----Original Message----- 2 >> From: Info-VAX-Request@Mvb.Saic.Com at INTERNET+ >> Sent: Thursday, August 02, 2001 10:14 AMyG >> To: Webb, William W - Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNETh< >> Subject: RE: Outsourcing of VMS support to Digital India! >> >>B >> On 2 Aug 2001 04:21:40 -0700, alanb@cloud9.net (Alan B.) wrote: >>G >> >Holy Cow!...I received an Email from Compaq yesterday entitled "LowiF >> >Cost VMS Support Services from Compaq". Here's some of the text... >>H >> When Digital India recently advertised for VMS staff the ad said: "IfG >> VMS is your religion, here is a chance to work with God" So far theyaE >> seem to be the only Compaq subsidiary still allowed to call itselft. >> "Digital" and use the "d|i|g|i|t|a|l" logo. >> >> >> --c >> Alanm >>   -- Alan   ------------------------------   Date: 3 Aug 2001 04:47:24 -0700p  From: alanb@cloud9.net (Alan B.)9 Subject: Re: Outsourcing of VMS support to Digital India!e= Message-ID: <88599d89.0108030347.6958342b@posting.google.com>n  _ WILLIAM WEBB <WWEBB1@email.usps.gov> wrote in message news:<0033000031129758000002L082*@MHS>...s > Holy COW?  Digital India?l >  > Surely you jest.  A Negative. This was no April fool's joke. This was a genuine Emaili message from Compaq.   Alan     >  > WWWebb >  > > -----Original Message-----3 > > From: Info-VAX-Request@Mvb.Saic.Com at INTERNETo, > > Sent: Thursday, August 02, 2001 10:14 AMH > > To: Webb, William W - Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNET= > > Subject: RE: Outsourcing of VMS support to Digital India!n > >o > > C > > On 2 Aug 2001 04:21:40 -0700, alanb@cloud9.net (Alan B.) wrote:E > >iH > > >Holy Cow!...I received an Email from Compaq yesterday entitled "LowG > > >Cost VMS Support Services from Compaq". Here's some of the text...u > >*I > > When Digital India recently advertised for VMS staff the ad said: "Ifu >  .H > > VMS is your religion, here is a chance to work with God" So far theyF > > seem to be the only Compaq subsidiary still allowed to call itself/ > > "Digital" and use the "d|i|g|i|t|a|l" logo.s > >  > >e > > -- > > Alan > >    ------------------------------  % Date: Fri, 03 Aug 2001 09:06:37 -0400g2 From: rdeininger@mindspring.com (Robert Deininger)9 Subject: Re: Outsourcing of VMS support to Digital India!eL Message-ID: <rdeininger-0308010906380001@user-2iveaut.dialup.mindspring.com>  = In article <88599d89.0108030347.6958342b@posting.google.com>,V! alanb@cloud9.net (Alan B.) wrote:   7 > WILLIAM WEBB <WWEBB1@email.usps.gov> wrote in messageh) news:<0033000031129758000002L082*@MHS>...c > > Holy COW?  Digital India?r > >  > > Surely you jest. > C > Negative. This was no April fool's joke. This was a genuine Emailo > message from Compaq. >  > Alan  J I recieved the email as well.  I thought it strange that it originated outH of New York City.  Actually I got 2 messages; the second corrected a URL
 in the first.h  G From my reading, it was simply Digital India trying to drum up some new A business.  In a sensible world, that would not seem very unusual.4   -- 9 Robert Deininger rdeininger@mindspring.como   ------------------------------   Date: 3 Aug 2001 14:06:12 GMTr1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)r9 Subject: Re: Outsourcing of VMS support to Digital India!K+ Message-ID: <9keb4k$ps0$2@info.cs.uofs.edu>g  9 In article <Xns90F1965E1B68Efalkarcabca@205.233.108.180>,u%  Alfred Falk <falk@arc.ab.ca> writes:n |> XA |> And PDP-1 through PDP-9, but not all were actually marketed.     < Wow, looks like they have more in common with the Alpha than anyone thought.i   bill   -- cJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   o   ------------------------------   Date: 3 Aug 2001 14:04:58 GMTm1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)d9 Subject: Re: Outsourcing of VMS support to Digital India!t+ Message-ID: <9keb2a$ps0$1@info.cs.uofs.edu>b  L In article <OFB61D0A3A.9E62FC3A-ON03256A9C.006F0B41@ep-bc.petrobras.com.br>,.  <fabio_compaq@ep-bc.petrobras.com.br> writes: |> i |> Answered... |> r
 |> 10,11,....a |> c& |> PS: I never saw a PDP in my life.     You have my deepest sympathy.r   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   u   ------------------------------  $ Date: Fri, 3 Aug 2001 11:36:19 -0500* From: WILLIAM WEBB <WWEBB1@email.usps.gov>9 Subject: RE: Outsourcing of VMS support to Digital India!r- Message-ID: <0033000031243652000002L022*@MHS>r  - =0Anononono.   You miss the point completely.   : I've seen this information before, either in this forum or via e or snailmail.t  A I wasn't saying "surely you jest" with respect to the informationt being valid.  B I was saying "surely you jest" with respect to the use of the termB "Holy Cow" in a message about India (where there are "sacred cows" that tread the streets.)  4 I'm sure there's an .au file of a rimshot somewhere.   WWWebb   > -----Original Message-----1 > From: Info-VAX-Request@Mvb.Saic.Com at INTERNETu' > Sent: Friday, August 03, 2001 7:55 AMnF > To: Webb, William W - Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNET; > Subject: RE: Outsourcing of VMS support to Digital India!n >s >d7 > WILLIAM WEBB <WWEBB1@email.usps.gov> wrote in messagewH > news:<0033000031129758000002L082*@MHS>... > Holy COW?  Digital India?=   > >c > > Surely you jest. > C > Negative. This was no April fool's joke. This was a genuine Email  > message from Compaq. >n > Alan >g >  > >e
 > > WWWebb > >   > > > -----Original Message-----5 > > > From: Info-VAX-Request@Mvb.Saic.Com at INTERNET0. > > > Sent: Thursday, August 02, 2001 10:14 AM> > > > To: Webb, William W - Raleigh, NC; Info-VAX@Mvb.Saic.Com
 > at INTERNETe? > > > Subject: RE: Outsourcing of VMS support to Digital India!c > > >a > > >eE > > > On 2 Aug 2001 04:21:40 -0700, alanb@cloud9.net (Alan B.) wrote:u > > >u< > > > >Holy Cow!...I received an Email from Compaq yesterday > entitled "Lowf= > > > >Cost VMS Support Services from Compaq". Here's some ofL
 > the text...i > > >e> > > > When Digital India recently advertised for VMS staff the > ad said: "If > >c> > > > VMS is your religion, here is a chance to work with God"
 > So far theysH > > > seem to be the only Compaq subsidiary still allowed to call itsel= fs1 > > > "Digital" and use the "d|i|g|i|t|a|l" logo.s > > >I > > >t > > > --
 > > > Alan > > >u >=   ------------------------------  $ Date: Fri, 3 Aug 2001 11:11:05 -0600- From: "Dan Notov" <dannoHATES_SPAM@large.com>y9 Subject: Re: Outsourcing of VMS support to Digital India! / Message-ID: <tmlmorq8n06985@corp.supernews.com>u  H DIGITAL India exists as such due to Indian commerce laws regarding stockE ownership of foreign corporations. I do not remember the details, butjH DIGITAL India stock remains majority owned by Indians, or something like that...j  7 The gist of what I read means the name cannot change...:  2 "Alan Greig" <a.greig@virgin.net> wrote in message2 news:m7himt8um48q6a5v29ftjlnqrcg6tfoj3l@4ax.com...A > On 2 Aug 2001 04:21:40 -0700, alanb@cloud9.net (Alan B.) wrote:i >iF > >Holy Cow!...I received an Email from Compaq yesterday entitled "LowE > >Cost VMS Support Services from Compaq". Here's some of the text...  >aG > When Digital India recently advertised for VMS staff the ad said: "IfCF > VMS is your religion, here is a chance to work with God" So far theyD > seem to be the only Compaq subsidiary still allowed to call itself- > "Digital" and use the "d|i|g|i|t|a|l" logo.- >- >- > -- > Alan   ------------------------------  # Date: Fri, 03 Aug 2001 10:46:19 GMTu' From: Tim Ellis <nospam@newsranger.com>eE Subject: Re: Printer settings was - Re: Print Queue's on IP Addresses 6 Message-ID: <%hva7.16068$ar1.58002@www.newsranger.com>  C In article <3B6936B2.A1A1CED@notrash.csc.com>, Carl Hampson says...  >n  @ > What sort of printer is it? (I missed the start of the thread) >FO One of them is a Facit LC404  - the other one is a similar sort of lineprinter, % but older (and may not be a facit...)    -- o$ Tim Ellis  CSC Computer Sciences Ltd% Speaking for himself, not the companyi   ------------------------------  % Date: Fri, 03 Aug 2001 01:17:15 -0400-' From: Howard S Shubs <howard@shubs.net>p< Subject: Re: Problem with floppy drive on an Aplastation 200< Message-ID: <howard-C1F1AE.01171503082001@enews.newsguy.com>  & In article <GHH53L.2Ir@world.std.com>,9  moroney@world.std.spaamtrap.com (Michael Moroney) wrote:1  H > BTW, the 2.88MB DEC drives work fine on an AS200.  That is, if you can > find one (or the media).  K Found one yesterday in a to-be-scrapped 2100, and installed it in the (get  N this) SCSI-based desktop floppy box (model SZ03A-AA) I liberated months ago.  M This thing contains, in addition to a power supply and the (scavenged) drive  O itself, a board labeled "SCSI to FDI Controller".  Bizzare.  Haven't yet tried e out the new drive.  M Don't know where to get -media-!  ED floppies appear to be VERY scarce these  3 days, not that they were ever -plentiful-, exactly.m -- a Howard S ShubsD "Run in circles, scream and shout!"  "I hope you have good backups!"   ------------------------------  % Date: Fri, 03 Aug 2001 11:20:36 +0100p% From: Alan Greig <a.greig@virgin.net>a< Subject: Re: Problem with floppy drive on an Aplastation 2008 Message-ID: <2mukmto16uu1tktm0u6hofac2n0rbmsife@4ax.com>  4 On Fri, 03 Aug 2001 01:49:19 GMT, "John Fredrickson"  <jafred@bellatlantic.net> wrote:  J >MaizeI recently purchased an Alphastation 200 4/233 which lacked a floppyL >drive. When I installed one and booted up, the light came on and stayed on.K >I double checked the ribbon cable and it was indeed installed correctly. I   A Are you absolutely sure? This is normally a symptom of the ribbonm) cable connected back to front at one end.p     -- Alan   ------------------------------  % Date: Fri, 03 Aug 2001 11:49:12 +0100o( From: Nic Clews <sendspamhere@127.0.0.1>" Subject: Re: Quorum Disk Questions) Message-ID: <3B6A81A8.DAFD6933@127.0.0.1>i   Rob Young wrote: > Z > In article <3B6A129D.217A8079@home.com>, "Richard L. Dyson" <rickdyson@home.com> writes:S > >> This is not cool.  It makes it hard to start one node at a time after shuttingoE > >> down all the nodes OR to shut down more than one node at a time.d > E >         There is a better technique, see below.  Expected Votes arelH >         adjusted down by REMOVE_NODE when a node is properly shutdown. > @ > http://www.mi.infn.it/~calcolo/OpenVMS/ssb71/6015/6017p008.htm  ( The description of REMOVE_NODE is wrong.  F The code actually broke for earlier versions of VMS (pre 6?), but yes,B if you use the REMOVE_NODE option, then during shutdown the systemE enables itself to run without QUORUM, removes its vote, then forces a:H recalculation of quorum clusterwide. The classic situation of being in a@ two node cluster (Q=2, V=2) the total votes drops to 1, then theC 'leaving' system tells the remaining node to work out quorum, which-A automagically becomes 1. This is contrary to what is stated; that F "Specifying REMOVE_NODE will not decrease the EXPECTED_VOTES below the quorum value. "0 The docs are wrong.o   (Code is in OPCCRASH.EXE)n   Extract:  8 4674 100$:	TSTL	G^CLU$GL_CLUB		; Are we in a VAXcluster? 4675 	BEQL	109$			; if EQL, no.. 4676 ;F 4677 ; Allow the process running OPCCRASH to run without requiring the QUORUME 4678 ; capability.  If this nodes loses quorum it will still completee
 the shutdown.i 4679 ; ...  [cut]  ...CH 5961 ; We have located the CSB for this node. Use its votes to calculate& 5962 ; a new value for expected votes. 5963 ;3 5964 	MOVZWL	CSB$W_VOTES(R1),R0	; This node's votese, 5965 	BNEQ	204$			; If NEQ, not yet retrying1 5966 	MOVL	R8,R0			; CSB was zero, may be a retry B 5967 204$:	CMPW	R0,CLUB$W_VOTES(R2)	; Are we the last voting node?- 5968 	BEQL	208$			; If eql, yes, don't adjustnD 5969 	CLRW	CSB$W_VOTES(R1)		; Zero our votes and don't count 'em any more4 5970 	MOVL	R0,R8			; Save our votes in case we retry3 5971 	MOVZWL	CLUB$W_CEVOTES(R2),R1	; Expected voteso  5972 	SUBL2	R0,R1			; Difference 5973 ;C 5974 ; Tell the connection manager to set a new expected votes. TheA
 connectionA 5975 ; manager must handle this since it is done in a coordinatedc fashion across the9 5976 ; cluster. The expected votes value is passed in R1.d 5977 ;7 5978 	JSB	G^CNX$ADJ_EXPT_VOTES	; Set new expected votesnA 5979 	MOVL	R0,R3			; Save status -- failure means try again laterx 5980 ;; 5981 ; Wait for the connection manager to effect the changey (CLUB$V_ADJ_QUORUM flag) 5982 ;1 5983 208$:	UNLOCK	LOCKNAME=SCS,-		; Desynchronizel 5984 		NEWIPL=#0,- 5985 		PRESERVE=NO  5986 	BRB	210$			; Test for done 5987 5988 ;H 5989 ; At this point, we wait in kernel mode for a short time, then test to see8 5990 ; if the connection manager has completed its work. 5991 ;( 5992 210$:	$SETIMR_S -			; Define a tick9 5993 		efn	= S^#EXE$C_SYSEFN - ; ...use system event flag 9 6001 		daytim	= ONE_TENTH	  ; ...shouldn't be perceptiblee 6016 	BLBC	R0,220$			; ... 6017 	$WAITFR_S - * 6018 		efn	= S^#EXE$C_SYSEFN	; Wait a tick 6024 	BLBC	R0,220$			; ... 60254 6026 	BLBC	R3,230$			; Branch if a retry is required6 6027 	BBS	#CLUB$V_ADJ_QUORUM, -	; If BS, not done yet. 6028 		CLUB$L_FLAGS(R2),210$< 6029 	BBS	#CLUB$V_LK_SHUTDOWN, -	; Branch if lckmgr shutdown3 6030 		CLUB$L_LK_FLAGS(R2),210$ ; not complete yet..  6031 220$:	BRW	109$			; BugCheck 6032 230$:	BRW	200$			; Retryp  G (sorry about formatting, its too wide. I also hope I've not contravened 9 copyright, but I've quoted it in context of explanation).1  H This however does leave a potential hole for a partition, this is why itG is absolutely essential that EXPECTED_VOTES is correctly calculated anduH NOT hardcoded, system managers are expected to know what they are doing!  = Personnaly I think it is WRONG that SHUTDOWN.COM in a clustersC environment does not default to the REMOVE_NODE option, perhaps forlH those that thought it did nothing, this will change your mind. Not usingA the option makes the remaining members of the cluster 'think' theaE leaving system has crashed (which it ultimately does of course at thee end of shutdown).e  D A cluster is a most beautiful and elegant thing, and the nice men inE white coats will be along shortly to put me in my padded room for thea weekend.   --  ( Regards, Nic Clews CSC Computer Sciences nclews at csc dot comt   ------------------------------  % Date: Fri, 03 Aug 2001 13:41:26 +0100u( From: Nic Clews <sendspamhere@127.0.0.1>" Subject: Re: Quorum Disk Questions) Message-ID: <3B6A9BF6.948576A4@127.0.0.1>    "Richard L. Dyson" wrote:e > Q >         Would it be better to simply choose an important node (I have one!) andjH > give it four (4) votes [Quorum=(4+1+1+1+2)/2=4]?  Then, as long as the > important nodeO > is working, the "cluster" can run/start.  This does away with the performancee > issues- > I have read about when using a quorum disk.  > S >         I don't think this could lead to a partitioning of the cluster in anyway,EM > either.  Correct?  Even if all the other three (3) nodes got together for ah5 > mutiny, they would be one vote short of a quorum...1  C Let me place the context, I teach a cluster course internally and Il? labour the points of voting primarily because they are so oftentH misunderstood. A number of our systems managers have (on the strength ofG my arugments) reconfigured clusters which now (in their words) shutdown:+ and boot more smoothly than they ever have.b  H Setting up voting depends what you want to achieve, my advice is to workE out (on paper) the failure modes that are acceptable, and the service- levels that are acceptable.   F QUORUM_DISK is a bad name, it ought to be called VOTING_DISK. It givesG rise to mythconceptions. QUORUM is half the sum of all votes, plus one,b
 rounded down.   A Your method of giving one node all the votes (casting votes) is as@ technique which I would only advocate if the loss of that systemE prevented real work being done by the cluster, (e.g. boot node and/orT serving data).  @ Using a 'quorum disk' is only useful is more than one node has aG 'direct' (non-MSCP served) connetion to that disk, if only one node hasiB a direct connection then the QD votes should be given to that node' (similar to above). This is documented..  A You mention performance of quorum disk in clusters. Why is this aS concern? Where did you read it?  C Every QDSKINTERVAL seconds, each DISK_QUORUM serving node READS the E QUORUM.DAT file, checking for changes (reconfigurations). The file iscH ONLY WRITTEN at RECONFIGURATION times (these times include nodes leavingC and joining, and virtual circuit failures, should be rare). In this|D case, the QUORUM.DAT file is (possibly written,) read and compared 4E times, i.e. 4 x QDSKINTERVAL seconds, so you can work out the effectsnA (or tradeoffs) of lowering or increasing this figure. It also hasdH nothing to do with if QUORUM is achieved or not, but in most cases it isC lost, or 'questioned' which is why it does in fact become a factor.2  H Small values speed up transitioning times, but places a 'higher' load on/ the IO. Large values delay transitioning times.i  E Example of a 4 node cluster, 1 vote each plus a 1 vote 'quorum' disk, D you will not lose quorum if you lose the disk, but a reconfigurationG will be forced, which could incurr some delay (IPL8?) I'll welcome moree7 detail in this area however, I don't profess to know itEG [reconfiguration] in intimate detail, but I know theres around 40 stepso	 involved.   4 To sum up,you need to know what you want to achieve.  H In the following examples, at least one of the voting nodes has a direct? connection to the quorum disk, but ideally (and properly) more.   # 4 node cluster, FORCE a quorum diskg  
 1+1+1+1+4. expected votes is 8. Quorum is 5aC Lose quorum disk, then hang the cluster. (node votes insufficient).kD In this situation when booting or shutting down, you can have just 1G node and the quorum disk. This disk may also be the location of clusters< common data files, controller mirrored or raided for safety.    @ 4 node cluster, use a quorum disk as an ENABLER, but not forced.  	 1+1+1+1+3  expected votes is 7. quorum is 4tF Lose quorum disk, and cluster will still continue, total systems votesH equal the quorum value, but this also assumes you can 'live' without theH data on the quorum disk. You can have just one node and the quorum disk,/ so booting and shutting down is OK to one node.e  @ 4 node cluster, use a quorum disk but minimum TWO systems needed  	 1+1+1+1+1I expected votes is 5e quorum is 3 G lose quorum disk and the cluster still runs (may be an issue?) but when H booting, and NORMAL running, there must be at least two nodes as well asG the quorum disk available. when booting, the first system will wait for G the second system as well as the quorum disk. Shtting down is dependent,D on useage of the REMOVE_NODE qualifiers as I've discussed in another post.s  E If you were to give a single node enough votes to run by itself, then B you may as well give this system just 1 vote and set all the otherA systems votes at 0. This also have lock manager implications thatuH LOCKDIRWT is implied 0 (satellite), and the voting node also handles allF distributed (not local) locking, but you can override the LOCKDIRWT in MODPARAMS of course.  @ I guess in summary, many people cause themselves problems by notC understanding the voting system, what quorum is, and using SHUTDOWNs7 incorrectly and wrongly setting/forcing EXPECTED_VOTES.e  C The connection manager makes a reasonable attempt at protecting theiD integrity of your data, even in the adversity of EXPECTED_VOTES=1 in
 MODPARAMS.  A I like your description of 'mutiny'. Say in a 4 node cluster, twoaH systems and the quorum disk have 3 votes, and quorum. What would cause a mutiny?a@ Why, any other system coming up with EXPECTED_VOTES=1. In a SCSIA connected hardware context, if no other interconnect is availableID [broke], this system will propose formation of (an already existing)D cluster and walk right over the data it can see down the SCSI paths.G What might help is if the quorum disk is down that path, this node will1B (helpfully) indicate (write) its intent to join the cluster in theH QUORUM.DAT file, which will be observed by the other two systems as theyG read the file. Reconfiguration battles commence, but the node trying to ; join will lose in this case. Mutiny foiled. (CLUEXIT irrc!)   G Now, if you have a broken path to the SCSI disk, or no quorum disk, I'd-' like to quote from Howard S Shub's sig:,  D "Run in circles, scream and shout!"  "I hope you have good backups!"  G You're in trouble! SCSI is what I describe as a 'special' interconnect,3, SCS (cluster) traffic does not pass down it.  B My closing line is, if you've _understood_ this, then go forth and reconfigure!  D Disclaimer: This is not the documentation, if you're adding a quorumH disk for the first time, there are issues surrounding that not mentioned< here. There's way more detail that I've skipped for clarity.   -- b( Regards, Nic Clews CSC Computer Sciences nclews at csc dot com)   ------------------------------   Date: 3 Aug 2001 09:56:17 -0500 + From: young_r@encompasserve.org (Rob Young)X" Subject: Re: Quorum Disk Questions3 Message-ID: <uZAN6IgWjbfG@eisner.encompasserve.org>   T In article <3B6A9BF6.948576A4@127.0.0.1>, Nic Clews <sendspamhere@127.0.0.1> writes: > "Richard L. Dyson" wrote:a >> sR >>         Would it be better to simply choose an important node (I have one!) andI >> give it four (4) votes [Quorum=(4+1+1+1+2)/2=4]?  Then, as long as theu >> important nodeiP >> is working, the "cluster" can run/start.  This does away with the performance	 >> issuesl. >> I have read about when using a quorum disk. >> aT >>         I don't think this could lead to a partitioning of the cluster in anyway,N >> either.  Correct?  Even if all the other three (3) nodes got together for a6 >> mutiny, they would be one vote short of a quorum... > E > Let me place the context, I teach a cluster course internally and I2A > labour the points of voting primarily because they are so often J > misunderstood. A number of our systems managers have (on the strength ofI > my arugments) reconfigured clusters which now (in their words) shutdown.- > and boot more smoothly than they ever have.d > J > Setting up voting depends what you want to achieve, my advice is to workG > out (on paper) the failure modes that are acceptable, and the servicel > levels that are acceptable.e > H > QUORUM_DISK is a bad name, it ought to be called VOTING_DISK. It givesI > rise to mythconceptions. QUORUM is half the sum of all votes, plus one,-I                                                                ^^^^^^^^^^P > rounded down.u >    	Ooops!o   	Q = (EV + 2 ) / 2      (1)l 	Rounded Down.   	Plus two, not "plus one"p  9 	But.... get rid of the Quorum Disk, adjust EV on the waymB 	up as outlined earlier.  Pick up Roy Davis' VAXCluster Principles8 	to give insight to all this (Yes, it is a bit dated but 	Principles are Principles!)   				Rob      (1)tF Ref:  http://www.openvms.compaq.com:8000/72final/4477/4477pro_002.html   ------------------------------  # Date: Fri, 03 Aug 2001 13:53:29 GMTj1 From: "Mark D. Jilson" <jilly@clarityconnect.com>-" Subject: Re: Quorum Disk Questions2 Message-ID: <3B6AAD94.D847BBA1@clarityconnect.com>  : In a cluster with more than 2 voting nodes you can use SETF CLUSTER/EXPECTED to reduce the current cluster expected votes and thus change quorum.   SETa  	   CLUSTERd       /EXPECTED_VOTESe         Descriptionn  G            The SET CLUSTER/EXPECTED_VOTES command enables you to adjustt thelH            total number of expected votes in the cluster. Set this valueE            equal to the number of votes contributed by each node plus  the1F            number of votes contributed by the cluster quorum disk. TheG            system will automatically calculate the value of the cluster7@            quorum from the total number of expected votes in the cluster.  H            You can specify the expected total votes value as part of theF            SET CLUSTER/EXPECTED_VOTES command string. If you enter theE            command without specifying a value for expected votes, the  systemE            calculates the value for you, using the following formula:   
           > EXPECTED_VOTES=(NODE1_VOTES+NODE2_VOTES+...)+QUORUM_DISK_VOTES  E            NODEn_VOTES is the value of the system parameter VOTES foro eachH            node in the cluster and QUORUM_DISK_VOTES is the value of the&            system parameter QDSKVOTES.  H            When you enter the SET CLUSTER/EXPECTED_VOTES command withoutE            specifying a value, the system assumes that all nodes thata are ?            expected to be in the cluster are currently members.a  E            In general, you use the SET CLUSTER/EXPECTED_VOTES commands onlyG            when a node is leaving the cluster for an extended period of,G            time. Under normal circumstances, quorum is not reduced whenr aaG            node leaves the cluster, because it is assumed that the nodee mayvG            be rebooted and rejoin the cluster. If a node is removed andr isE            unable to rejoin the cluster within a reasonable period ofa timeH            (for example, if a node crashes due to a hardware problem andF            cannot rejoin the cluster for several days), the quorum for the G            cluster can safely be reduced by lowering the total expectedd)            votes until that node rejoins.u  >            The purpose of a cluster quorum is to eliminate any possibilityJA            of the cluster partitioning into separate clusters andtG            simultaneously accessing the same resources (such as disks).nD            If the sum of the votes for all members of the cluster is smallercG            than the cluster quorum, all nodes will block activity until  newuE            nodes join to increase the vote total. Lowering the quoruma valuerG            (by reducing the value of the total expected votes) when one  orH            more nodes leave the cluster for long periods of time reduces            this possibility.  A            Note that no matter what value you specify for the SET H            CLUSTER/EXPECTED_VOTES command, you cannot increase quorum toH            a value that is greater than the number of the votes present,H            nor can you reduce quorum to a value that is half or fewer of thea            votes present.   H            When you enter the SET CLUSTER/EXPECTED_VOTES command, eitherH            with or without an expected votes value specified, the systemD            responds with a message indicating the new value that wasF            actually set. Note that you need to enter this command only onC            one node in the cluster, because the new value for totalo expectedE            votes is propagated through the cluster. This new expected  voteseH            value should then be stored in the system parameter EXPECTED_E            VOTES on each node, so that it remains in effect after the1 nodest            reboot.  E            When a node that was previously a member of the cluster iseD            ready to rejoin, you should increase the system parameterH            EXPECTED_VOTES to its original value before bringing the nodeH            back to the cluster. Note that you do not need to use the SETC            CLUSTER/EXPECTED_VOTES command to increase the number ofL expectedD            votes, because the expected votes value will be increased;            automatically when the node rejoins the cluster.m     Nic Clews wrote: >  > "Richard L. Dyson" wrote:  > > S > >         Would it be better to simply choose an important node (I have one!) and2J > > give it four (4) votes [Quorum=(4+1+1+1+2)/2=4]?  Then, as long as the > > important nodeQ > > is working, the "cluster" can run/start.  This does away with the performancem
 > > issues/ > > I have read about when using a quorum disk.t > >sU > >         I don't think this could lead to a partitioning of the cluster in anyway,RO > > either.  Correct?  Even if all the other three (3) nodes got together for a 7 > > mutiny, they would be one vote short of a quorum...1 > E > Let me place the context, I teach a cluster course internally and IhA > labour the points of voting primarily because they are so often J > misunderstood. A number of our systems managers have (on the strength ofI > my arugments) reconfigured clusters which now (in their words) shutdown7- > and boot more smoothly than they ever have.p > J > Setting up voting depends what you want to achieve, my advice is to workG > out (on paper) the failure modes that are acceptable, and the serviceC > levels that are acceptable.  > H > QUORUM_DISK is a bad name, it ought to be called VOTING_DISK. It givesI > rise to mythconceptions. QUORUM is half the sum of all votes, plus one,s > rounded down.n > C > Your method of giving one node all the votes (casting votes) is anB > technique which I would only advocate if the loss of that systemG > prevented real work being done by the cluster, (e.g. boot node and/or  > serving data). > B > Using a 'quorum disk' is only useful is more than one node has aI > 'direct' (non-MSCP served) connetion to that disk, if only one node hasRD > a direct connection then the QD votes should be given to that node) > (similar to above). This is documented.B > C > You mention performance of quorum disk in clusters. Why is this aD
 > concern? > Where did you read it? > E > Every QDSKINTERVAL seconds, each DISK_QUORUM serving node READS the G > QUORUM.DAT file, checking for changes (reconfigurations). The file istJ > ONLY WRITTEN at RECONFIGURATION times (these times include nodes leavingE > and joining, and virtual circuit failures, should be rare). In thisfF > case, the QUORUM.DAT file is (possibly written,) read and compared 4G > times, i.e. 4 x QDSKINTERVAL seconds, so you can work out the effectscC > (or tradeoffs) of lowering or increasing this figure. It also hasaJ > nothing to do with if QUORUM is achieved or not, but in most cases it isE > lost, or 'questioned' which is why it does in fact become a factor.a > J > Small values speed up transitioning times, but places a 'higher' load on1 > the IO. Large values delay transitioning times.u > G > Example of a 4 node cluster, 1 vote each plus a 1 vote 'quorum' disk,eF > you will not lose quorum if you lose the disk, but a reconfigurationI > will be forced, which could incurr some delay (IPL8?) I'll welcome moreN9 > detail in this area however, I don't profess to know it-I > [reconfiguration] in intimate detail, but I know theres around 40 stepsm > involved.s > 6 > To sum up,you need to know what you want to achieve. > J > In the following examples, at least one of the voting nodes has a directA > connection to the quorum disk, but ideally (and properly) more.v > % > 4 node cluster, FORCE a quorum diskr >  > 1+1+1+1+4. > expected votes is 8.
 > Quorum is 5oE > Lose quorum disk, then hang the cluster. (node votes insufficient). F > In this situation when booting or shutting down, you can have just 1I > node and the quorum disk. This disk may also be the location of clusterr> > common data files, controller mirrored or raided for safety. > B > 4 node cluster, use a quorum disk as an ENABLER, but not forced. >  > 1+1+1+1+3e > expected votes is 7.
 > quorum is 4eH > Lose quorum disk, and cluster will still continue, total systems votesJ > equal the quorum value, but this also assumes you can 'live' without theJ > data on the quorum disk. You can have just one node and the quorum disk,1 > so booting and shutting down is OK to one node.e > B > 4 node cluster, use a quorum disk but minimum TWO systems needed >  > 1+1+1+1+1  > expected votes is 5m
 > quorum is 3cI > lose quorum disk and the cluster still runs (may be an issue?) but whenaJ > booting, and NORMAL running, there must be at least two nodes as well asI > the quorum disk available. when booting, the first system will wait forrI > the second system as well as the quorum disk. Shtting down is dependenthF > on useage of the REMOVE_NODE qualifiers as I've discussed in another > post.o > G > If you were to give a single node enough votes to run by itself, then D > you may as well give this system just 1 vote and set all the otherC > systems votes at 0. This also have lock manager implications thattJ > LOCKDIRWT is implied 0 (satellite), and the voting node also handles allH > distributed (not local) locking, but you can override the LOCKDIRWT in > MODPARAMS of course. > B > I guess in summary, many people cause themselves problems by notE > understanding the voting system, what quorum is, and using SHUTDOWNu9 > incorrectly and wrongly setting/forcing EXPECTED_VOTES.e > E > The connection manager makes a reasonable attempt at protecting theuF > integrity of your data, even in the adversity of EXPECTED_VOTES=1 in > MODPARAMS. > C > I like your description of 'mutiny'. Say in a 4 node cluster, two J > systems and the quorum disk have 3 votes, and quorum. What would cause a	 > mutiny?nB > Why, any other system coming up with EXPECTED_VOTES=1. In a SCSIC > connected hardware context, if no other interconnect is available F > [broke], this system will propose formation of (an already existing)F > cluster and walk right over the data it can see down the SCSI paths.I > What might help is if the quorum disk is down that path, this node willuD > (helpfully) indicate (write) its intent to join the cluster in theJ > QUORUM.DAT file, which will be observed by the other two systems as theyI > read the file. Reconfiguration battles commence, but the node trying tou= > join will lose in this case. Mutiny foiled. (CLUEXIT irrc!)e > I > Now, if you have a broken path to the SCSI disk, or no quorum disk, I'd ) > like to quote from Howard S Shub's sig:n > F > "Run in circles, scream and shout!"  "I hope you have good backups!" > I > You're in trouble! SCSI is what I describe as a 'special' interconnect,u. > SCS (cluster) traffic does not pass down it. > D > My closing line is, if you've _understood_ this, then go forth and > reconfigure! > F > Disclaimer: This is not the documentation, if you're adding a quorumJ > disk for the first time, there are issues surrounding that not mentioned> > here. There's way more detail that I've skipped for clarity. >  > --* > Regards, Nic Clews CSC Computer Sciences > nclews at csc dot comu   -- lD Jilly	- Working from Home in the Chemung River Valley - Lockwood, NY0 	- jilly@clarityconnect.com			- Brett Bodine fan. 	- Mark.Jilson@Compaq.com			- since 1975 or so, 	- http://www.jilly.baka.com               -   ------------------------------   Date: 3 Aug 2001 08:35:36 -0700 1 From: nothome@spammers.are.scum (Malcolm Dunnett)e" Subject: Re: Quorum Disk Questions, Message-ID: <XK0jGEzVFZ6L@malvm5.mala.bc.ca>  * In article <3B6A9BF6.948576A4@127.0.0.1>, .     Nic Clews <sendspamhere@127.0.0.1> writes: > C > I like your description of 'mutiny'. Say in a 4 node cluster, twoeJ > systems and the quorum disk have 3 votes, and quorum. What would cause a	 > mutiny?vB > Why, any other system coming up with EXPECTED_VOTES=1. In a SCSIC > connected hardware context, if no other interconnect is availablemF > [broke], this system will propose formation of (an already existing)F > cluster and walk right over the data it can see down the SCSI paths.I > What might help is if the quorum disk is down that path, this node willcD > (helpfully) indicate (write) its intent to join the cluster in theJ > QUORUM.DAT file, which will be observed by the other two systems as theyI > read the file. Reconfiguration battles commence, but the node trying to/= > join will lose in this case. Mutiny foiled. (CLUEXIT irrc!)o > C     What is the purpose of the SCSI_CHECK that is run when you booteG a SCSI connected node? I thought it detected other hosts on the bus andu; refused to start if it couldn't also seem them through SCS.d   ------------------------------  % Date: Fri, 03 Aug 2001 17:38:20 +0100s( From: Nic Clews <sendspamhere@127.0.0.1>" Subject: Re: Quorum Disk Questions) Message-ID: <3B6AD37C.DC652161@127.0.0.1>o   Malcolm Dunnett wrote: > E >     What is the purpose of the SCSI_CHECK that is run when you bootaI > a SCSI connected node? I thought it detected other hosts on the bus andd= > refused to start if it couldn't also seem them through SCS.s  D I dunno if it "refused to start" I'm not sure where that would be inF connection manager code. However I know on a SCSI system, you can fromF the SH DEV at the console prompt, 'see' another booted VMS system as aH SCSI target. I don't think it is the same thing, I'm certainly not aware of it, anyone?  G This is most evident on a Galaxy system sharing SCSI, with one instance,F up, the other console reports the existance of the other instance as aE SCSI target, I can't remember exactly how it appears. I know the SCSI F reset forces mount verification on the 'remote' node, but I believe it. can be controlled (stopped?) from the console. -- m( Regards, Nic Clews CSC Computer Sciences nclews at csc dot comk   ------------------------------    Date: 03 Aug 2001 11:27:40 +0200G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>a( Subject: Re: Reboot to close a file ????H Message-ID: <y48zh1ijcj.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  / JF Mezei <jfmezei.spamnot@videotron.ca> writes:b  O > -if new location is smaller, then first update the size of file in the memorymK > structure, then update the LBN to point to the new location. (otherwise, a > switch the other way around).a   So far, so good.  J > -then, you open the new file and lock it, and if succesful, you can then > close/release the old file.4   Wrong order.N The first step has to be opening the new file. Then you update the pointers inM the HWRPB (I'm assuming it's stored there) as you describe. Now you can closeIB the old file and allow the user to delete it or whatever he wants.  L I can't remember whether dump files must be contiguous or just single-header; files. For non-contiguous files, things are more difficult.f   	Jan   ------------------------------  % Date: Fri, 03 Aug 2001 10:44:54 +0100e% From: Alan Greig <a.greig@virgin.net>r( Subject: Re: Reboot to close a file ????8 Message-ID: <f4skmt825efadkv6giiv63c1go89tnfjlt@4ax.com>  @ On Thu, 02 Aug 2001 19:08:05 GMT, system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) wrote:   g >In article <3B699448.A636D40B@clarityconnect.com>, "Mark D. Jilson" <jilly@clarityconnect.com> writes:eF >>There have been functionality requests in the past that VMS create aJ >>utility that will redo the system dump mapping routines so that a rebootD >>isn't required when deleting, reducing, increasing the system dumpI >>file.  If you would like this to get a higher priority (if it exists onvG >>a wish list at all) then contact your local CSC and request that they  >>log an enhancement request.o >u > @ >But there already is a set of tools to do this:  SDA and DELTA.  ? Hehehe... Poking around with SDA/XDELTA to unmap the dump file.j) Hehehehe. I'll leave that to the brave :).   -- Alan   ------------------------------  % Date: Fri, 03 Aug 2001 10:49:23 +0100b% From: Alan Greig <a.greig@virgin.net>s( Subject: Re: Reboot to close a file ????8 Message-ID: <mpskmtoc1pfp4pad6o1pnvcanuhk1rqfbo@4ax.com>  " On Thu, 2 Aug 2001 14:20:59 -0300,, <fabio_compaq@ep-bc.petrobras.com.br> wrote:   >l >Sorry ! >l >tJ >I didnt explain right:  I renamed the SYSDUMP.DMP to SYSDUMP.DMP3 after a
 >forced crashoH >for analysis purposes and created a new SYSDUMP.DMP using the MC SYSGEN >CREATEX! >SYSDUMP.DMP /SIZE=size /NOCONTIGr >pL >The problematic file is SYSDUMP.DMP3. I deleted after the completion of the
 >analysis. >It still allocating space.T  F Because the real sysdump file is still the one mapped at boot. If yourD system crashed now it would dump to sysdump3.dmp. You need to rebootD *after* you rename the file. Any new sysdump.dmp you create will not be mapped until a reboot   -- Alan   ------------------------------  # Date: Fri, 03 Aug 2001 13:28:54 GMTl1 From: "Mark D. Jilson" <jilly@clarityconnect.com>q( Subject: Re: Reboot to close a file ????2 Message-ID: <3B6AA7D1.23D0F4FE@clarityconnect.com>  G Dump files are single header but don't have to be contiguous.  Also theEE code has to be bullet proof in the case where the system bugchecks ine- the middle of updating the dumpfile pointers.m   Jan Vorbrueggen wrote: > 1 > JF Mezei <jfmezei.spamnot@videotron.ca> writes:  > Q > > -if new location is smaller, then first update the size of file in the memorynL > > structure, then update the LBN to point to the new location. (otherwise,! > > switch the other way around).S >  > So far, so good. > L > > -then, you open the new file and lock it, and if succesful, you can then > > close/release the old file.  >  > Wrong order.P > The first step has to be opening the new file. Then you update the pointers inO > the HWRPB (I'm assuming it's stored there) as you describe. Now you can closeCD > the old file and allow the user to delete it or whatever he wants. > N > I can't remember whether dump files must be contiguous or just single-header= > files. For non-contiguous files, things are more difficult.c > 
 >         Jani   -- eD Jilly	- Working from Home in the Chemung River Valley - Lockwood, NY0 	- jilly@clarityconnect.com			- Brett Bodine fan. 	- Mark.Jilson@Compaq.com			- since 1975 or so, 	- http://www.jilly.baka.com               -   ------------------------------  $ Date: Fri, 3 Aug 2001 11:29:33 -04005 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>c( Subject: Re: Reboot to close a file ????2 Message-ID: <Isza7.578$Yx2.11351@news.cpqcorp.net>  L And don't even think about changing the dump *disk* - it better have been inH dump_dev when you booted, or the console may not be able to write to it.  $ Jan Vorbrueggen wrote in message ...0 >JF Mezei <jfmezei.spamnot@videotron.ca> writes: > I >> -if new location is smaller, then first update the size of file in the  memoryK >> structure, then update the LBN to point to the new location. (otherwise,s  >> switch the other way around). >e >So far, so good.a >rK >> -then, you open the new file and lock it, and if succesful, you can thenr >> close/release the old file. >l
 >Wrong order.pL >The first step has to be opening the new file. Then you update the pointers inH >the HWRPB (I'm assuming it's stored there) as you describe. Now you can closenC >the old file and allow the user to delete it or whatever he wants.t > ? >I can't remember whether dump files must be contiguous or justu
 single-header < >files. For non-contiguous files, things are more difficult. >  > JanW   ------------------------------  $ Date: Fri, 3 Aug 2001 11:26:14 -04005 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>y( Subject: Re: Reboot to close a file ????2 Message-ID: <ypza7.577$Yx2.11214@news.cpqcorp.net>  < JF Mezei wrote in message <3B69BA47.9C6F43F@videotron.ca>... >"Mark D. Jilson" wrote:H >> The way to request these things is, currently, the CSC and any directG >> contact a customer may have with VMS Engineering in a formal settingo
 >> like CETS.r >iG >1- Why can't engineers remport to management that they have received ai request G >for feature XX from customers. Why force the customers to repeat theire stories 1 >through all sorts of layers of heavy management.r >y    F They do.  But you get what you pay for.  A random note or request fromH someone on usenet isn't a business justification, and doesn't convey howJ many customers/users need or want a feature/change/"fix".  Remember only aH handful of customers read or write in this forum, and there are a lot of# people who aren't customers at all.   E So.  The weight of a request for something is different for differentmK forums.  If the engineers who read this forum see a lot of noise about someo@ subject, you can usually be assured that it is thought about and considered - on it's merits.  L >2- Not everyone has access to the CSC. I remember trying to report a bug on aiF >product and for the privilege of doing so, the canadian CSC wanted to chargeG >me 1 hour of consulting at $250.00 for the privilege of allowing me to  report >a bug.c >d  L The CSC is in business to handle problems for people with support contracts, or who are under warrenty.  K >3- Please stop seing that CETS thing as the only way for customers to meet.K >engineers. comp.os.vms reaches engineers YEAR ROUND, and reaches customershI >from around the world. That CETS thing is just a local event in the USA.u  L But CETS is a well established forum where face-to-face discussions can takeJ place.  But it's not the only one, and not all of them are in the US.  TheD "Technical Days" meetings that were held is another example of this.  J The other thing you get at CETS is that you can talk to people who are notH just engineering, like marketing and management - so that they can get a feel for what is on your mind.   ------------------------------   Date: 3 Aug 2001 08:39:27 -0700i1 From: nothome@spammers.are.scum (Malcolm Dunnett)s( Subject: Re: Reboot to close a file ????, Message-ID: <qoJTbJP386RL@malvm5.mala.bc.ca>  3 In article <3B6AA7D1.23D0F4FE@clarityconnect.com>, s6    "Mark D. Jilson" <jilly@clarityconnect.com> writes:  I > Dump files are single header but don't have to be contiguous.  Also theiG > code has to be bullet proof in the case where the system bugchecks ine/ > the middle of updating the dumpfile pointers.  > C    If your purpose was only to free up disk space being used by thesB dump file you could define a value for dumpstyle that meant "don'tK do dumps". Setting this value dynamically would clear the dumpfile pointersdE and close the file. You'd have the restriction that once set this wayE7 you'd need to reboot in order to re-enable crash dumps.i   ------------------------------  $ Date: Fri, 3 Aug 2001 10:49:51 +0200$ From: "Jakob Erber" <erberj@post.ch> Subject: ssh port forwarding Message-ID: <3b6a65b0$1@hcwe67>    Hello,  E ist there an SSH implementation (client) for VMS, which supports this G feature (I hope I named it correctly). Port forwarding allows to securefI arbitary ports by SSH without applications like mail or http noticing it.h   best regards   Jakob Erberb   ------------------------------  % Date: Fri, 03 Aug 2001 06:49:33 -0600t% From: Dan O'Reilly <dano@process.com>s  Subject: Re: ssh port forwardingB Message-ID: <5.1.0.14.2.20010803064832.00af7cf8@ntbsod.psccos.com>  K Both Multinet and TCPware provide an SSH1 implementation that supports full  port forwarding.  ( At 02:49 AM 8/3/2001, Jakob Erber wrote: >Hello,u >iF >ist there an SSH implementation (client) for VMS, which supports thisH >feature (I hope I named it correctly). Port forwarding allows to secureJ >arbitary ports by SSH without applications like mail or http noticing it. >d
 >best regardsm >, >Jakob Erber   ------I +-------------------------------+---------------------------------------+tI | Dan O'Reilly                  |                                       |rI | Principal Engineer            |  "Why should I care about posterity?  |.I | Process Software              |   What's posterity ever done for me?" | I | http://www.process.com        |                    -- Groucho Marx    |hI +-------------------------------+---------------------------------------+    ------------------------------  . Date: Fri, 3 Aug 2001 08:34:54 +0200 (MET DST)& From: Rudolf Wingert <win@fom.fgan.de>+ Subject: Re: Sun goes after Alpha user base 6 Message-ID: <200108030634.IAA06157@sinet1.fom.fgan.de>   Hello,  G other then Sun I did not see, that Compaq did sold machines not having.hC But Sun do. They did offer last year a new one Sun (buy now and getaJ them next year), which the would like to build in the middle of this year.J Offering a computer realized in the feature I did see the last five years.J What's happen if this computer can never be build in case of technological problem?   Regards Rudolf Wingert   ------------------------------  % Date: Fri, 03 Aug 2001 08:34:27 +0200,= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>s+ Subject: Re: Sun goes after Alpha user baset) Message-ID: <3B6A45F3.1E5820F7@gtech.com>d   Alan Greig wrote:eG > Yesterday I received a "Sun Offers" brochure from Sun UK addressed toq  > "A Greig, VMS Systems Manager" > A > Today several of us, including the Europe/Eastern Hemisphere IT 4 > Director, received a postcard from Sun which says: > 	 > DUMPED?a" > Alpha recovery campaign from Sun  	 Ofcourse.a  - SUN may make lousy hardware/software/support.I  ( But they sure know how to do marketing !   Arne   ------------------------------  % Date: Fri, 03 Aug 2001 08:55:37 +0200 = From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com> ' Subject: Re: SYS$LOGIN:[some.directory]n) Message-ID: <3B6A4AE9.933E2A41@gtech.com>    JF Mezei wrote:gL > I beleive that Unix has a construct which uses the tilde ~ to point to theM > login directory so that something such as ~/some/directory/myfile.txt wouldt5 > work no matter what the currect directory would be.= > . > Shouldn't VMS have some way to do the same ? > A > I realise that one can define a rooted logical that would allow O > MY$LOGIN:[some.directory] but it would be nice if it were to become something  > that is standard.E  > Send a suggestion to Compaq VMS Engineering for such a logical% (I think it should be SYS$LOGINROOT).i   Arne   ------------------------------  % Date: Fri, 03 Aug 2001 19:00:47 +0010o% From: paddy.o'brien@zzz.tg.nsw.gov.aua' Subject: Re: SYS$LOGIN:[some.directory]d5 Message-ID: <01K6PHQRAHJM003IO3@tgmail.tg.nsw.gov.au>n   Arne wrote:i   >JF Mezei wrote:M >> I beleive that Unix has a construct which uses the tilde ~ to point to thesN >> login directory so that something such as ~/some/directory/myfile.txt would6 >> work no matter what the currect directory would be. >> X/ >> Shouldn't VMS have some way to do the same ?t > B >> I realise that one can define a rooted logical that would allowP >> MY$LOGIN:[some.directory] but it would be nice if it were to become something >> that is standard. >r? >Send a suggestion to Compaq VMS Engineering for such a logicalC& >(I think it should be SYS$LOGINROOT).   I support Arne's proposal.  P Yep, I did jump in and throw all of my n feet into my mouth in an earlier reply.  R Where I did get confused was because I make a rooted logical in my LOGIN.COM from * SYS$LOGIN to generate other path logicals.  
 Apologies.   Regards, Paddy   ------------------------------   Date: 3 Aug 2001 09:53:42 -0500C- From: koehler@encompasserve.org (Bob Koehler)-' Subject: Re: SYS$LOGIN:[some.directory]u3 Message-ID: <$HbYMuR1wlZU@eisner.encompasserve.org>c  V In article <3B6A1FB2.2F2FCECB@bigfoot.com>, Hamlyn Mootoo <univms@bigfoot.com> writes: > D >> Why?  It's not standard on UNIX.  (Try it on Sun's Bourne shell). > D > All that means is that Sun's implementation of sh is not standard. >   C No, it means it's a shell standard, not a UNIX standard.  I've alsom. tried it on Tru64, AIX, and Irix, same result.  1 Do you know of anyone's Bourne shell that has ~ ?e  F OBTW, VMS did have it when it had the POSIX shell.  IIRC EUNICE had it> in it's csh for VMS, too.  A shell feature, not an OS feature.  F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences Corporationn= NASA GSFC Flight Software       | Federal Sector, Civil GroupoE                                 | please remove ".aspm" when replyinga   ------------------------------  $ Date: Fri, 3 Aug 2001 11:37:06 -04005 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>a' Subject: Re: SYS$LOGIN:[some.directory] 2 Message-ID: <Jzza7.579$Yx2.11377@news.cpqcorp.net>  H Well, when the COE stuff makes it's way into the mainstream, you will beD able to set up logicals so that the Motif/CDE stuff can present UNIXE pathnames, and will gladly accept the tilde as in ~/subdir/myfile.dat   H But I think it is a resonable suggestion that it would be nice to have aJ rooted logical equivalent to SYS$LOGIN created (automatically) so that you3 can have something like SYS$HOME:[SUBDIR]MYFILE.DAT   L However, it is also something that you can do today with a minimal amount of DCL in your login.come   _Frede  G Bob Koehler wrote in message <$HbYMuR1wlZU@eisner.encompasserve.org>... : >In article <3B6A1FB2.2F2FCECB@bigfoot.com>, Hamlyn Mootoo <univms@bigfoot.com> writes: >>E >>> Why?  It's not standard on UNIX.  (Try it on Sun's Bourne shell).  >>E >> All that means is that Sun's implementation of sh is not standard." >> >tD >No, it means it's a shell standard, not a UNIX standard.  I've also/ >tried it on Tru64, AIX, and Irix, same result.  >s2 >Do you know of anyone's Bourne shell that has ~ ? >tG >OBTW, VMS did have it when it had the POSIX shell.  IIRC EUNICE had it,? >in it's csh for VMS, too.  A shell feature, not an OS feature.  >hG >----------------------------------------------------------------------i@ >Bob Koehler                     | Computer Sciences Corporation> >NASA GSFC Flight Software       | Federal Sector, Civil GroupF >                                | please remove ".aspm" when replying   ------------------------------   Date: 3 Aug 2001 16:37:13 GMTo1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)3' Subject: Re: SYS$LOGIN:[some.directory]u+ Message-ID: <9kejvp$vg9$1@info.cs.uofs.edu>b  + In article <3B6A1FB2.2F2FCECB@bigfoot.com>, +  Hamlyn Mootoo <univms@bigfoot.com> writes:u |> o |> k |> -F |> > Why?  It's not standard on UNIX.  (Try it on Sun's Bourne shell). |> >    . The shell is no more "Unix" than DCL is "VMS".   |> nE |> All that means is that Sun's implementation of sh is not standard.t  F Wrong again.  The Bourne Shell has never supported the ~ abbreviation.G The Bourne Shell was the very first shell released with Unix.  How manyhG features of current day DCL were not there in the first version??  I'll F give you a hint, I use DCL on the PDP-11 and any resemblance to DCL on# the Alpha is purely accidental. :-)e  D The ~ abbreviation is available in KSH, CSH, BASH and I am sure someE other shells as well.  It was never available in the Bourne Shell ando< the functions built into the shell don't define Unix anyway.   bill   -- oJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  % Date: Fri, 03 Aug 2001 13:31:15 -0400e- From: JF Mezei <jfmezei.spamnot@videotron.ca>.' Subject: Re: SYS$LOGIN:[some.directory]l+ Message-ID: <3B6ADFE3.DA84485@videotron.ca>h   Arne Vajhj wrote:@ > Send a suggestion to Compaq VMS Engineering for such a logical' > (I think it should be SYS$LOGINROOT).t  < I would much rather have a shorter notation to acheive this.  F The idea is to make is quicker to refer to files below your sys$login.   ------------------------------  % Date: Fri, 03 Aug 2001 13:47:55 -0400 # From: sol gongola <sol@adldata.com>e% Subject: tcpip 5.1 dhcp error messagey+ Message-ID: <3B6AE3CB.9D8CBD35@adldata.com>.  4 I have an alpha with VMS 7.2-1 and TCPIP v5.1 eco 1 G with a primary ip address in 206.234.147.3 and an alias of 192.168.1.3.-C DHCP was configured to serve addresses for the 192.168.1.0 network.b  / When a PC (win9x) tries to get an ip address,  R5 the following error message shows up in the dhcp log:f  H DHCPDISCOVER from HW xx:xx:xx:xx:xx : network not administered by server  D The compaq manual (chap 7.9) only said that the NETMASKS. table may D need to be fixed. The join manual (page 118) also mentioned several # other possibilities including this:s  E    3. The owner of a valid IP range is not the server node on whichP JOINF       is running. Check that the official internet address of serverG       corresponds to the owner of IP range, as shown in debug level 5e&       JOIN log file at server startup.  A I am guessing that "owner" and "official" refer to the primary ipd address.G Does this mean that I can't use it to serve addresses for the IP alias?c  	 thank yout -- m' Sol Gongola           (sol@adldata.com)-.    IBM Certified Specialist-AIX System Support,    COMPAQ ASE Alphaserver/OpenVMS Specialist ADL Data Systems Inc 20 livingstone ave Dobbs Ferry, NY 10522 ' phone: 914 591 1800   fax: 914 591 1818n   ------------------------------  % Date: Fri, 03 Aug 2001 14:09:29 +0100A, From: Alan Greig <alan.greig@intl.fmcti.com>  Subject: Third postcard from Sun8 Message-ID: <l78lmt0hmop1ulutsufpfrralvg8sdj12o@4ax.com>  @ I've today received a third postcard from Sun. This time it saysE simply "Therapy from Sun" in huge bold letters on one side and on thefC other "It's such an uncertain world. Just when you think that maybet2 Alpha's here to stay after all, out it goes." etc.  ' Still waiting for anything from Compaq.   B Plus I was just talking to a long serving DEC/Compaq field serviceE engineer. Field service engineers are about the only Compaq employees B with regular customer contact these days (other than the excellent@ service provided in c.o.v) and he says that customer response to? recent events is near universally negative. Capellas says it iso. "unbelievably positive". I know who I believe. -- Alan   ------------------------------  % Date: Fri, 03 Aug 2001 08:02:06 -0700 ' From: David Mathog <mathog@caltech.edu> $ Subject: Re: Third postcard from Sun+ Message-ID: <3B6ABCEE.24320C6D@caltech.edu>o   Alan Greig wrote:i   >nD > Plus I was just talking to a long serving DEC/Compaq field serviceG > engineer. Field service engineers are about the only Compaq employees-D > with regular customer contact these days (other than the excellentB > service provided in c.o.v) and he says that customer response toA > recent events is near universally negative. Capellas says it is 0 > "unbelievably positive". I know who I believe. >   A Capellas's statement is a classic example of Clintonesque meaningn mangling.  It appears to sayE that the customer response is so positive that the company can hardlyr believe it - but what itE really says is that the listener should not believe that the responseeD was as positive as the speaker is insinuating.   Given the real (andE well deserved) level of customer confidence in the veracity of Compaq E management's statements (zero) a more compelling statement would havec been "we spoke to theaC CIOs at our 100 largest customers and ##% of them said they were inn favor of the change."  But theyhF can't say that, because they probably only went down a very short listC until they found somebody who said something vaguely positive, likeiE "Well, at least Intel knows how to market its products, so if you use D their CPU,  maybe the VMS group will be able to survive your moronicF management long enough to be bought by somebody who knows what they're doing."a  ? Which to Q management instantly became "unbelievably positive".o   Regards,   David Mathog mathog@caltech.edu  G ***********************************************************************M3 *                                       RIH  Compaq. * G ***********************************************************************y   ------------------------------  % Date: Fri, 03 Aug 2001 07:48:57 -0700n  From: Jon <jsmyth69@hotmail.com>$ Subject: Re: Third postcard from Sun2 Message-ID: <o7lqOwEI3ss0oPY0UGnlBRYiSAZ6@4ax.com>  F This one is an easy one to answer:  The FSE is talking to the user end/ while Capellas is talking to the executive end.v  - Alan Greig <alan.greig@intl.fmcti.com> wrote:t   >pA >I've today received a third postcard from Sun. This time it saysuF >simply "Therapy from Sun" in huge bold letters on one side and on theD >other "It's such an uncertain world. Just when you think that maybe3 >Alpha's here to stay after all, out it goes." etc.  >p( >Still waiting for anything from Compaq. >wC >Plus I was just talking to a long serving DEC/Compaq field service F >engineer. Field service engineers are about the only Compaq employeesC >with regular customer contact these days (other than the excellentdA >service provided in c.o.v) and he says that customer response to @ >recent events is near universally negative. Capellas says it is/ >"unbelievably positive". I know who I believe.r   ------------------------------  # Date: Fri, 03 Aug 2001 16:40:37 GMTt) From: "unix" <kevin.gillins@glcgroup.com>u$ Subject: Re: Third postcard from Sun9 Message-ID: <9uAa7.923$h94.129628@paloalto-snr1.gtei.net>l  F I have learned that Compaq is nearing their last Alpha series and willH change to Intel Titanium.  Not sure what they will call the server.  The/ last Alpha series is on the assembly bench now.   9 "Alan Greig" <alan.greig@intl.fmcti.com> wrote in messagef2 news:l78lmt0hmop1ulutsufpfrralvg8sdj12o@4ax.com... >uB > I've today received a third postcard from Sun. This time it saysG > simply "Therapy from Sun" in huge bold letters on one side and on the E > other "It's such an uncertain world. Just when you think that maybe 4 > Alpha's here to stay after all, out it goes." etc. > ) > Still waiting for anything from Compaq.n > D > Plus I was just talking to a long serving DEC/Compaq field serviceG > engineer. Field service engineers are about the only Compaq employeesoD > with regular customer contact these days (other than the excellentB > service provided in c.o.v) and he says that customer response toA > recent events is near universally negative. Capellas says it ise0 > "unbelievably positive". I know who I believe. > -- > Alan   ------------------------------  $ Date: Fri, 3 Aug 2001 10:27:14 -0700* From: "Jack Peacock" <peacock@simconv.com>$ Subject: trapping telnet disconnects? Message-ID: <x7Ba7.63807$w4.380904@e420r-sjo3.usenetserver.com>r  F After some experimenting it appears that an exit handler using $DCLEXHF will not trap a telnet disconnect.  I assume this means the process isC terminated by a $DELPRC instead of a $FORCEX.  Am I correct in that-E there is no mechanism for a process running on a telnet connection toAH continue running after the connection has been dropped from the user endF (i.e. PC users close the telnet window instead of logging out)?  Would> it be possible to trap this at the system level using callback facilities in LOGIN/LOGOUT?r     Jack Peacock   ------------------------------    Date: 03 Aug 2001 11:20:46 +0200G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>- Subject: Re: VAX CH Message-ID: <y4bslxijo1.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  ' GreyCloud <wholland@tscnet.com> writes:g  ; > Do you know how to get the hobbyist license to upgrade?? .  N The license covers all software up to the current versions. THus, all you haveM to do is find someone with the media for the current version of Compaq C, and 
 install that.S   	Jan   ------------------------------   Date: 3 Aug 2001 03:16 PST+ From: rankin@eql14.caltech.edu (Pat Rankin)s Subject: Re: VAX C0 Message-ID: <3AUG200103160984@eql14.caltech.edu>  P In article <D808BB825C9EAC96.F1D84EE5FD1C66E7.AA6D1902F419BBDE@lp.airnews.net>,\*  GreyCloud <wholland@tscnet.com> writes... [...]-K > Yes.  I still have to get a scsci box to put in more drives and a cd-rom. L > Do you know how to get the hobbyist license to upgrade??  I went to DECCUSK > but the site was rather obscure about it.  The VAX C is ok for now, but I I > need to know what the float range is.... I'm getting some big humongous , > number for an output that I know is wrong.  @      You could determine attributes of floats and doubles--apartA from their actual bit patterns--by examing <limits.h>.  Depending<A on a choice made during compiler installation, you might have thes> file sys$library:float.h to look at; otherwise you need to use the commandn  ? $ library/text sys$library:vaxcdef.tlb/extract=float/output=XXX   D to get a copy (pick suitable file name for XXX, such as float.h :-).A [If/when you switch to Compaq C, beware that sys$library:*.h willfA be old files from VAX C which aren't used by the new compiler, soe@ examining them for situations like this would lead to confusion.D Also, you won't need any special linking help as described next....]  =      The usual reason to get the situation you describe is torA compile code for D_float format doubles but link with the G_float:> format run-time library, or possibly vice versa, based on someC confusing advice about how to link C programs present in the onlineeD help.  I suggest that you create file sys$common:[syslib]vaxcrtl.opt
 containing   sys$share:vaxcrtl.exe/shareabler  ? in it.  Then add ``,sys$library:vaxcrtl/options'' to the end ofnA your link commands and _don't_ define LNK$LIBRARY logical name(s) B for VAXCRTL as shown in ``help cc link_libraries.  (You'll need to@ do things a little differently if you really want to use G-float instead of D-float.)  @      A possible alternate reason, which is actually even simpler? than mis-linking, is that you've failed to declare some routineo@ which returns a float or double, so it is implicitly declared asC returning an integer.  Naturally the bits in its return value won'toC mean what you intended them to mean and the results will be screwy.o  2                 Pat Rankin, rankin@eql.caltech.edu   ------------------------------  # Date: Fri, 03 Aug 2001 16:01:36 GMT?2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: VAX C2 Message-ID: <AVza7.582$Yx2.11227@news.cpqcorp.net>  w In article <D808BB825C9EAC96.F1D84EE5FD1C66E7.AA6D1902F419BBDE@lp.airnews.net>, GreyCloud <wholland@tscnet.com> writes:tJ :Yes.  I still have to get a scsci box to put in more drives and a cd-rom.K :Do you know how to get the hobbyist license to upgrade??  I went to DECCUSvK :but the site was rather obscure about it.  The VAX C is ok for now, but I O* :need to know what the float range is....   8   The floating-point formats are architecture-specified.  J :I'm getting some big humongous number for an output that I know is wrong.  B   That behaviour would imply an "application-embedded misfeature".  H   If I were guessing, I would guess you are encounting a floating point B   casting or conversion problem, or the results of floating point 4   promotions that can occur during argument passing.  K   If you are not already using it and/or you do not already use it, please iF   learn to use the OpenVMS Source Code Debugger, and then (of course) F   please use the debugger to find the triggers for these "executional    disfunctions".   :Hoff Hoffman wrote:L :>  ---------------------------- #include <rtfaq.h> ---------------------...M :>       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.come    I   For additional information on the hobbyist license, please see the FAQ.3  D   Please also see the FAQ for the information that we must have whenD   we (try to) answer questions -- background and general informationH   on this topic is in the introductory section of the FAQ.  If we don't E   have the necessary information, then we do exactly what I am doing (C   right now, which is to guess and to ask questions and to request pF   details and to request a reproducer and (of course) to point to the E   FAQ.  If we have the necessary contextual information, then we can xE   generally answer the question.  Meaning you get the answer you want ;   (need?) far more quickly.  (No offense is intended here.)C    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Fri, 03 Aug 2001 13:43:59 GMT,1 From: "Mark D. Jilson" <jilly@clarityconnect.com> / Subject: Re: Virus Alert at Compaq (2001.08.02)L2 Message-ID: <3B6AAB5A.C565E2F9@clarityconnect.com>  G I'd be tempted to report this to CA (it is a CA product isn't it?) as am possible bug in InoculateIT.   Neil Rieck wrote:t > > > "Mark D. Jilson" <jilly@clarityconnect.com> wrote in message. > news:3B6992D2.8E38BE7E@clarityconnect.com...E > > Here is the response from our security team concerning this item.a > >tL > > " We contacted OPS on this right away.  The server was scanned with bothL > > Inoculan and NAV, and nothing was found.  They have pulled the documentsL > > and scanned them, they report as being clean you may want to update your' > > database in the scanner being used.Y > >lK > > FYI: There is a 4 tier process for moving items onto the COMPAQ.COM web,J > > pages, which includes a multiple scans by both Inoculan and NAV.  This9 > > document was clean all the way through the process. "a > >n, > > Compaq's Software Security Response Team > > security-ssrt@compaq.com > > Mark Menkhus4 > > Subscribe to our Security Advisory mailing list:= > > http://www.support.compaq.com/patches/mailing-list.shtml"s > >n > > --H > > Jilly - Working from Home in the Chemung River Valley - Lockwood, NY1 > > - jilly@clarityconnect.com - Brett Bodine fanl/ > > - Mark.Jilson@Compaq.com - since 1975 or so-/ > > - http://www.jilly.baka.com               -- > N > This does seem very strange. At approximately 4:00 AM I tried downloading toN > tool 3 times and received the same error message from InoculateIT (I receiveK > a new virus signature file every 4 days and I'm sitting behind a hardware-M > firewall in my house with no other computers, so I think I'm safe). If yoursM > site wasn't infected, then I wonder if it's possible for the file to becomen2 > infected "along the way" (a very scary thought). > L > p.s. I did a complete virus scan of my system and everything is/was clean. >  > Neil Rieck > Kitchener/Waterloo/Cambridge,  > Ontario, Canada.# > http://www3.sympatico.ca/n.rieck/n   --  D Jilly	- Working from Home in the Chemung River Valley - Lockwood, NY0 	- jilly@clarityconnect.com			- Brett Bodine fan. 	- Mark.Jilson@Compaq.com			- since 1975 or so, 	- http://www.jilly.baka.com               -   ------------------------------  % Date: Fri, 03 Aug 2001 08:36:09 +0200s= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>d Subject: Re: VMS Prompt value ) Message-ID: <3B6A4659.3DA4D3CE@gtech.com>b   Nick Paszty wrote:H > we're running vms 7.1 on an alphaserver 800 5/333 and i have written aF > com file, cd.com,  that dynamically changes my prompt to the currentH > working directory.  as i get deeper into a directory tree, i sometimes& > receive the following error message. > % > $3$DRA2:.SASAPPS.TOOLS]cd utilitiese9 > %DCL-E-STRTOOLNG, string argument is too long - shorten ! > $3$DRA2:.SASAPPS.TOOLS]show def#0 >   LARDAT2DISK:[000000.SASAPPS.TOOLS.UTILITIES] > A > is there a length limit to the prompt value?  is there a way toi > increase the value?t   32 characters.   Not that I know about.   Arne   ------------------------------  % Date: Fri, 03 Aug 2001 19:24:28 +0010a% From: paddy.o'brien@zzz.tg.nsw.gov.auo Subject: Re: VMS Prompt value 5 Message-ID: <01K6PIL3GPLE003J2T@tgmail.tg.nsw.gov.au>a   Arne wrote:o  A (On Info-vax, I often do not see originals, but replies to them?)s   >Nick Paszty wrote: I >> we're running vms 7.1 on an alphaserver 800 5/333 and i have written a G >> com file, cd.com,  that dynamically changes my prompt to the currentgI >> working directory.  as i get deeper into a directory tree, i sometimesS' >> receive the following error message.a >>  & >> $3$DRA2:.SASAPPS.TOOLS]cd utilities: >> %DCL-E-STRTOOLNG, string argument is too long - shorten" >> $3$DRA2:.SASAPPS.TOOLS]show def1 >>   LARDAT2DISK:[000000.SASAPPS.TOOLS.UTILITIES]w >> rB >> is there a length limit to the prompt value?  is there a way to >> increase the value? >i >32 characters.r >  >Not that I know about..  P I am always amazed that people want to do that.  Isn't it an MS-DOS inheritance?  R I know many of my users go through this, but (no aspersions on Nick) when working S in an environment, don't people know which directory they are in??  I do, and when  S working on several applications (often 3/4 applications are active in the same day PL for user problems), several CMS, SCA and DTM libraries, and several DECterm R sessions, I can find exactly where I am within this "mess" with a few commands -- Q and labelling of the DECterm icons.  $ show default is pretty good if you suffer - from Altzeihmers.   P With the naming conventions that our users have, often only one directory level  will exceed the prompt length.  S Programs -- which I have seen examples of -- that (I consider badly written) throw 0O you from one directory to another can cause a problem, but $ show default will 1! still work at the end of the day.<   Regards, Paddy   ------------------------------  $ Date: Fri, 3 Aug 2001 08:37:27 -0400* From: "Andy Stoffel" <acs@fcgnetworks.net> Subject: Re: VMS Prompt value.6 Message-ID: <LWwa7.27732$%7.417393@news6.giganews.com>  E [Replace the $'s with the correct amounts to get the full effect :-)]J  / Compaq share price:                           $y5 Compaq CEO salary:                           $$$$$$$$t Compaq AlphaServer ES40 2    [1GB RAM/833MHz] :                        $$$$$' Compaq OpenVMS Engineer salary : $$$$$$e= OpenVMS prompt:                               priceless (!!!)    Must be a Friday :-).....n   -Andy-   ------------------------------   Date: 3 Aug 2001 09:44:38 -0500o- From: koehler@encompasserve.org (Bob Koehler)a% Subject: Re: WATCHER compile problemsi3 Message-ID: <MY6+3pguLUYf@eisner.encompasserve.org>-  _ In article <3B69ADF8.4D7E0FAA@oracle.com>, norm lastovica <norman.lastovica@oracle.com> writes:  > it is called TPAMAC... r   Yes, that one I've got.6  F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences Corporation = NASA GSFC Flight Software       | Federal Sector, Civil Group E                                 | please remove ".aspm" when replyinge   ------------------------------  $ Date: Fri, 3 Aug 2001 06:12:33 -0500' From: "Glenn Wiens" <gwiens@swbell.net>lE Subject: When/how to send answerback string during logon negotiations / Message-ID: <MLva7.477$%u4.10143@nnrp1.sbc.net>t  H When I am negotiating terminal options during logon to VMS, I am looking for:  J 1. What byte sequence signals my program to send the answerback string. IsB it part of the DO/DONT/WILL/WONT stuff or is there another signal?   ands  = 2. What bytes I should send along with the answerback string.)   ------------------------------   Date: 3 Aug 2001 07:38:18 -0500s9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) I Subject: Re: When/how to send answerback string during logon negotiationsa3 Message-ID: <ke8NUP4Umhny@eisner.encompasserve.org>a  Y In article <MLva7.477$%u4.10143@nnrp1.sbc.net>, "Glenn Wiens" <gwiens@swbell.net> writes:dJ > When I am negotiating terminal options during logon to VMS, I am looking > for: > L > 1. What byte sequence signals my program to send the answerback string. IsD > it part of the DO/DONT/WILL/WONT stuff or is there another signal? >  > andh > ? > 2. What bytes I should send along with the answerback string.i  I If, and only if, the VMS machine in question does a SET TERMINAL/INQUIRE,*F you need to send back the proper responses for whatever supported typeD of terminal you are capable of emulating.  For the terminal emulatorB that comes with Windows, that is the null set.  The responses varyA considerably depending on the family of terminals involved.  VT52m is different from VT100.  > IIRC, this has nothing at all to do with answerback.  Some VMS= utilities block answerback inquiries for reasons of security.:  > All in all, you might be better just doing a SET TERMINAL onceB you are logged in (or in your LOGIN.COM file, if the same terminal is used all the time.e   ------------------------------   Date: 3 Aug 2001 04:57:04 -0700y  From: alanb@cloud9.net (Alan B.)@ Subject: You Get What You Pay For, a.k.a., There's No Free Lunch= Message-ID: <88599d89.0108030357.18c525ca@posting.google.com>i  A A recent comment from a software vendor this week at a meeting...u  E "Alphas? We're not really investing any more in Alphas. You can't gete> a kid out of college to work with them. You would have to hireA somebody who knows what they're doing...and you would have to payd them".  D In other words, MicroS**T/Intel commodity "dime a dozen" "technical"E people to run their servers are what they prefer in order to save $$.p  , Later in the meeting, software vendor again:  E "It would be nice if we didn't have to reboot the WinTel servers once< a week".   YOU GET WHAT YOU PAY FOR!R   Alan   ------------------------------  % Date: Fri, 03 Aug 2001 09:31:02 -0400t2 From: rdeininger@mindspring.com (Robert Deininger)D Subject: Re: You Get What You Pay For, a.k.a., There's No Free LunchL Message-ID: <rdeininger-0308010931020001@user-2iveaut.dialup.mindspring.com>  = In article <88599d89.0108030357.18c525ca@posting.google.com>,e! alanb@cloud9.net (Alan B.) wrote:2  C > A recent comment from a software vendor this week at a meeting...  > G > "Alphas? We're not really investing any more in Alphas. You can't getd@ > a kid out of college to work with them. You would have to hireC > somebody who knows what they're doing...and you would have to pay  > them". > F > In other words, MicroS**T/Intel commodity "dime a dozen" "technical"G > people to run their servers are what they prefer in order to save $$.r    4 I trust you're not buying anything from these goons!   -- c Robert Deininger rdeininger@mindspring.comv   ------------------------------  $ Date: Fri, 3 Aug 2001 12:23:21 -04005 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> E Subject: Re: [OT] Market corrections (was Re: Alpha: an invitation to-2 Message-ID: <6fAa7.585$Yx2.11485@news.cpqcorp.net>  J John Santos wrote in message <1010802195754.47969B-100000@Ives.egh.com>...  G >This reminds of a Ken Olsen quote, possibly apocryphal.  Some reporterrH >asked him in an interview, "How many people work for you?"  He answered >"Oh, about half of them." >d    J Having been sith DEC (now Compaq) 22 years, I think his estimate was high.   ------------------------------   End of INFO-VAX 2001.428 ************************