1 INFO-VAX	Wed, 04 Apr 2001	Volume 2001 : Issue 189       Contents: 3180 Re: $ PRODUCT INSTALL DECSET CD on new DS10 Re: CD on new DS10 Re: CD on new DS10/ Re: Christof a bot (was: "cut" utility for VMS) ' CI based cluster for max I/O perf. How? + Re: CI based cluster for max I/O perf. How? # dir/date=m/siz/gran/since=y ... etc ' RE: dir/date=m/siz/gran/since=y ... etc ' Re: dir/date=m/siz/gran/since=y ... etc ' Re: dir/date=m/siz/gran/since=y ... etc ' Re: dir/date=m/siz/gran/since=y ... etc > Re: EV7 and EV8 for the low end (was: VMS-Related: Affordable)! Re: F$CONTEXT anomaly ? Feature ? ! Re: F$CONTEXT anomaly ? Feature ? ! Re: F$CONTEXT anomaly ? Feature ?  Firmware question..... Re: Firmware question..... Re: Firmware question..... RE: Free The OpenVMS Seven! ( Re: GS160 froze due to a bad KGPSA-CX...  Re: Hopf 6841 dtss time provider  Re: Hopf 6841 dtss time providerA Re: How to tell if a system is booting immediately after a crash.  installing pci cards on ds10  Re: installing pci cards on ds10  Re: installing pci cards on ds10  Re: installing pci cards on ds10  Re: installing pci cards on ds10% It's all a Zincuous plot I tell you!!  Meaning of RSBREFNZRO? Re: Meaning of RSBREFNZRO? Re: Merging DCLTABLES.EXE  Re: Merging DCLTABLES.EXE 3 O.T. (Was: Re: dir/date=m/siz/gran/since=y ... etc) < PCSI Patch - VMS VAX v7.2 (Was: Re: TCP/IP 5.0A ECO Appears)# Re: POSIX Streams, File Permissions # Re: POSIX Streams, File Permissions > Re: product announcement: the rebirth of DateSim (version 2.1) Re: pthreads + printf  Re: pthreads + printf  Restart TCP/IP Services  Re: Restart TCP/IP Services : Re: Seeking CD-R/CD-RW SCSI INQUIRY data / Yamaha CRW4416S: Re: Seeking CD-R/CD-RW SCSI INQUIRY data / Yamaha CRW4416S: Re: Seeking CD-R/CD-RW SCSI INQUIRY data / Yamaha CRW4416S: Re: Seeking CD-R/CD-RW SCSI INQUIRY data / Yamaha CRW4416S Re: Spice for OpenVMS Alpha? Re: Spice for OpenVMS Alpha? Re: Support for CD-RD Re: Support for CD-R (was: Re: Seeking CD-R/CD-RW SCSI INQUIRY data)D Re: Support for CD-R (was: Re: Seeking CD-R/CD-RW SCSI INQUIRY data)4 Re: TCPIP V5.0A install fails, can't proceed; ideas?4 Re: TCPIP V5.0A install fails, can't proceed; ideas?! TELNETSYM ( TCPIP Sevices x UCX ) % Re: TELNETSYM ( TCPIP Sevices x UCX ) % Re: TELNETSYM ( TCPIP Sevices x UCX )  Re: VAX Debugger command. 
 Re: VAX in 1U  Re: VMS-Related: Affordable  Re: VMS-Related: Affordable  Re: VMS-Related: Affordable  Re: VMS-Related: Affordable  Re: VMS-Related: Affordable : What to do with old CDROMS... (do NOT try this at home...)) Re: When might SYS$SCHDWK not wake me up?  ZXLp-L2 bugcheck analysis ( [Q] DCL minute of the day: semi-graphics  F ----------------------------------------------------------------------  % Date: Wed, 04 Apr 2001 08:03:24 -0400 ) From: Bob Ricci <maxx0623@concentric.net> 
 Subject: 3180 3 Message-ID: <00eb01c0bcff$4001b530$585b5cc0@Subway>   % 3180-40mb memory 5 drives (appx2.1gb) 2 free-must come and get it in Milford , Connecticut       Robert V. Ricci  Systems Manager  Drs. Associates (SUBWAY) 325 Bic Dr.  Milford, Ct 06460   tel  203 877 4281 ext 1144  fax to pc 203 783 7144  fax 203 876 6682  email ricci_r@subway.com  or     maxx0623@concentric.net  http://www.subway.com    ------------------------------  % Date: Wed, 04 Apr 2001 12:19:21 +0200 = From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com> % Subject: Re: $ PRODUCT INSTALL DECSET ) Message-ID: <3ACAF529.BFF99D00@gtech.com>    Dick Wolff wrote: ) > How do I install DECSET? Then I issue a " >         $ PRODUCT INSTALL DECSET   > dir  >  > Directory  > DQA0:[DECSET124.KIT] > K > DECSET124.A;1       DECSET124.B;1       DECSET124.C;1       DECSET124.D;1  > K > DECSET124.E;1       DECSET124.F;1       DECSET124.G;1       DECSET124.H;1  > 7 > DECSET124.I;1       DECSET124.J;1       DECSET124.K;1  > 
 > Total of 11  > files. > $ product install  > decset5 > %PCSIUI-I-NOMATCH, no products were found matching:  > DECSET  9 DECset 12.4 are not using PCSI yet. So try with good old:    $ @SYS$UPDATE:VMSINSTAL    Arne   ------------------------------  % Date: Wed, 04 Apr 2001 10:44:03 -0600  From: Kevin Handy <kth@srv.net>  Subject: CD on new DS10 ' Message-ID: <3ACB4F53.2DAA16FA@srv.net>   A I am having a problem with a new DS10 that I have recently bought > (with VMS 7.2-1 license). I am waiting for a response from the@ company I bought it from, but thought I'd ask here in case it is something well known.   > The system boots fine, and everything seems to work correctly,< except for the CD-rom drive.  If I put in one of the install9 CD's that came with the system, and try to mount it using    	mount dqa0:/over=ident   9 the system will crash with an error message that includes ; the text "exception while above ASTDEL". I'd try to include 9 more of the message, but the system clears the screen and : reboots too fast for me to get much information off of it.  6 If I put in a CD that I burned myself (on windows 95),; it mounts and reads without any problems. I also received a 0 manditory update CD, and it mounts and reads OK.6 It seems that only the VMS install CD's (even some old9 VAX/VMS ones) will cause it to crash when I try to mount.   5 Sometimes, it will mount, then die when I try to do a  directory "DIR DQA0:[000000]".  ; I have re-installed VMS (from the same CD that causes it to 6 crash) and did not have any problems with the install.> Tried to mount it afterwards, and it crashed just like before.  : I haven't tried a previous version of VMS yet (hoping I'll! hear from someone how to fix it).    --  E If they're not putting secret messages to me in their music, then why > do they keep putting my picture on the other side of the CD's?   ------------------------------    Date: 05 Apr 2001 01:15:50 +0800, From: Paul Repacholi <prep@prep.synonet.com> Subject: Re: CD on new DS10 - Message-ID: <87n19w37sp.fsf@prep.synonet.com>   ! Kevin Handy <kth@srv.net> writes:   C > I am having a problem with a new DS10 that I have recently bought @ > (with VMS 7.2-1 license). I am waiting for a response from theB > company I bought it from, but thought I'd ask here in case it is > something well known.  > @ > The system boots fine, and everything seems to work correctly,> > except for the CD-rom drive.  If I put in one of the install; > CD's that came with the system, and try to mount it using  >  > 	mount dqa0:/over=ident  > ; > the system will crash with an error message that includes * > the text "exception while above ASTDEL".  ; Ah, that sounds like a 7.2-1 system mounting an ODS-2 CD :) ; Yes, it does that... There is fix in one of the patch kits, % and adding the /nowrite help as well.   8 > If I put in a CD that I burned myself (on windows 95),= > it mounts and reads without any problems. I also received a 2 > manditory update CD, and it mounts and reads OK.8 > It seems that only the VMS install CD's (even some old; > VAX/VMS ones) will cause it to crash when I try to mount.   E The ex-Windows CD is ISO9660, and uses a different ACP, not the ODS-2  XQP.  7 > Sometimes, it will mount, then die when I try to do a   > directory "DIR DQA0:[000000]". > = > I have re-installed VMS (from the same CD that causes it to 8 > crash) and did not have any problems with the install.@ > Tried to mount it afterwards, and it crashed just like before.  C Get all the current patches and apply. I think there is a DQ driver  patch or two as well.    --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  # Date: Wed, 04 Apr 2001 17:39:30 GMT 1 From: "Mark D. Jilson" <jilly@clarityconnect.com>  Subject: Re: CD on new DS10 2 Message-ID: <3ACB5CB2.57601EE1@clarityconnect.com>  C Have you pulled the patches for V7.2-1, especially the SYS$DQDRIVER  ones, from  ; http://ftp.support.compaq.com/patches/.new/openvms.shtml ?? H Search on VMS721 at that page.  You especially want VMS721_DRIVER-V0200,- VMS721_UPDATE-V0100 & VMS721_DQCONFIG-V0200 .    Kevin Handy wrote: > C > I am having a problem with a new DS10 that I have recently bought @ > (with VMS 7.2-1 license). I am waiting for a response from theB > company I bought it from, but thought I'd ask here in case it is > something well known.  > @ > The system boots fine, and everything seems to work correctly,> > except for the CD-rom drive.  If I put in one of the install; > CD's that came with the system, and try to mount it using  >   >         mount dqa0:/over=ident > ; > the system will crash with an error message that includes = > the text "exception while above ASTDEL". I'd try to include ; > more of the message, but the system clears the screen and < > reboots too fast for me to get much information off of it. > 8 > If I put in a CD that I burned myself (on windows 95),= > it mounts and reads without any problems. I also received a 2 > manditory update CD, and it mounts and reads OK.8 > It seems that only the VMS install CD's (even some old; > VAX/VMS ones) will cause it to crash when I try to mount.  > 7 > Sometimes, it will mount, then die when I try to do a   > directory "DIR DQA0:[000000]". > = > I have re-installed VMS (from the same CD that causes it to 8 > crash) and did not have any problems with the install.@ > Tried to mount it afterwards, and it crashed just like before. > < > I haven't tried a previous version of VMS yet (hoping I'll# > hear from someone how to fix it).  >  > --G > If they're not putting secret messages to me in their music, then why @ > do they keep putting my picture on the other side of the CD's?   --  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: Wed, 04 Apr 2001 12:16:31 +0200 = From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com> 8 Subject: Re: Christof a bot (was: "cut" utility for VMS)) Message-ID: <3ACAF47F.B7414BC2@gtech.com>    Dave Greenwood wrote: 0 > Tim Llewellyn <tim.llewellyn@bbc.co.uk> wrote:7 > > yeah, lexicals are a subset of RTL/system services.  > > A > > Maybe there is the odd exception but I can't think of it now.  > & > F$EDIT and F$ELEMENT come to mind...  : True, but it is usually very easy to do similar operations in any HLL.    Arne   ------------------------------  % Date: Wed, 04 Apr 2001 10:09:48 -0500 . From: Lyndon Bartels <lbartels@pressenter.com>0 Subject: CI based cluster for max I/O perf. How?. Message-ID: <3ACAF2EC.621EFB22@pressenter.com>  
 Good Morning,   E I'd like to start a discussion on how to configure a CI based cluster  for F maximum I/O performance, redundancy, and availability. But Performance is the	 key goal.   B Why? The application I have on this cluster is more I/O and memory
 intensive,H not CPU intensive. Memory and CPU don't seem to be limiting factors. But
 I/O might be.   G Anyway. I've been reading the cluster configuration part of the manuals  and I'm D thinking of using multiple CI (Star couplers). Each node member will have3 two(2) CIPCAs. And the HSJs will be dual-redundant.   H For CI, should I split the paths between couplers, or split by node? How shouldF the nodes be numbered? I didn't get a clear understanding of that from my reading.  F For disks, considering the discussion on loosing pagedisks. Giving I'm using F the same disks would it be better to have it connected to a host based SCSIH bus, or run it through the CI? Should I be running disks in a controller based D RAID scenario? RAID 0+1? How many members of the Stripe set should I	 have. HOw = should I have the disks arranged amoung the six SCSI shelves?   C Also for disks, what would the best disk layout be. How should I be  layingH these disks out amoungst the controller shelves for optimum performance?  H Almost nothing is out of the realm of possible discussion here. Fibre is out.  C This is conceptual in nature only. And the concept is limited to CI  based 
 storage only.   B I will of course continue my reading of both manuals and any white papers I can find.    Thanks in advance,   Lyndon --  G My opinions are mine and mine alone. They seldom align with those of my 	 employer.    ------------------------------    Date: 05 Apr 2001 00:29:55 +0800, From: Paul Repacholi <prep@prep.synonet.com>4 Subject: Re: CI based cluster for max I/O perf. How?- Message-ID: <87elv84oho.fsf@prep.synonet.com>   0 Lyndon Bartels <lbartels@pressenter.com> writes:    ? > I'd like to start a discussion on how to configure a CI based 6 > cluster for maximum I/O performance, redundancy, and0 > availability. But Performance is the key goal.    What are the CPUs? and how many?  D > Why? The application I have on this cluster is more I/O and memory? > intensive, not CPU intensive. Memory and CPU don't seem to be % > limiting factors. But I/O might be.   A > Anyway. I've been reading the cluster configuration part of the E > manuals and I'm thinking of using multiple CI (Star couplers). Each ; > node member will have two(2) CIPCAs. And the HSJs will be  > dual-redundant.   8 Get more memory! The fastest IO is the one you don't do!  F > For CI, should I split the paths between couplers, or split by node?F > How should the nodes be numbered? I didn't get a clear understanding > of that from my reading.  F Each Cipca has 2 circuits. So that is a total of 4. A SC-xx contains 2D couplers, so you need 2 SCs, or stick two more transformers into the
 existing cab.   D > For disks, considering the discussion on loosing pagedisks. GivingE > I'm using the same disks would it be better to have it connected to F > a host based SCSI bus, or run it through the CI? Should I be running? > disks in a controller based RAID scenario? RAID 0+1? How many F > members of the Stripe set should I have. HOw should I have the disks' > arranged amoung the six SCSI shelves?   7 What are your controller? HSJ?? from the shelf comment.   E > Also for disks, what would the best disk layout be. How should I be D > laying these disks out amoungst the controller shelves for optimum > performance?  ; > Almost nothing is out of the realm of possible discussion A > here. Fibre is out.  This is conceptual in nature only. And the . > concept is limited to CI based storage only.     --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  $ Date: Wed, 4 Apr 2001 15:43:40 +0200) From: "Pbo" <philippe.bocher@euriware.fr> , Subject: dir/date=m/siz/gran/since=y ... etc& Message-ID: <3acb1803$1@news.euriware>  
     Hello,  L     Do you have any idea to learn the number of files and size modified eachJ days on big volumes without dir/since....etc etc and without 1 h CPU on 90 Go disks   thanks   ------------------------------  % Date: Wed, 04 Apr 2001 10:19:46 -0400 2 From: "Kent, Philip  JW1811" <kent@jwfc.jfcom.mil>0 Subject: RE: dir/date=m/siz/gran/since=y ... etcH Message-ID: <5B57189920E7D41190B500606D210686011A69D4@mailsvr.jfcom.mil>   >Hello, I >Do you have any idea to learn the number of files and size modified each K >days on big volumes without dir/since....etc etc and without 1 h CPU on 90 	 >Go disks  >thanks   G I would suggest you install DFU off the Freeware CD-ROM.  It works much L faster and has other nice features like finding large files, fast deletes of& directory trees, and undeleting files.  ! Then you could issue the command: 3 	$DFU SEARCH/MODIFIED=SINCE=TODAY/SORT DEVICE_NAME: - to find all the files modified for each disk.   I Then you could write a command file using f$device("*","device_name") and G f$getdvi(device_name,"MNT") to get the mounted disks to run against the  above command.  	 Blessings   
 Phil Kent  kent@jwfc.jfcom.mil  philip.kent@compaq.com   ------------------------------  % Date: Wed, 04 Apr 2001 16:30:18 +0200 , From: Didier Morandi <Didier.Morandi@gmx.ch>0 Subject: Re: dir/date=m/siz/gran/since=y ... etc% Message-ID: <3ACB2FFA.F3924BE@gmx.ch>n  P Unless you write this information every day somewhere where you can read it fromJ the day after, I don't see how you could get it but from the header of the, files, which is what the dir/blablabla does.  L What is the problem with 1h CPU? You can have a batch job doing this at 5am,
 can't you.   Il fait beau  La Hague ?a   D.  
 Pbo wrote: >  >     Hello, > N >     Do you have any idea to learn the number of files and size modified eachL > days on big volumes without dir/since....etc etc and without 1 h CPU on 90
 > Go disks >  > thanks   -- m6 MORANDI Consultants, Swiss Quality Computer Consulting6 avenue de Granges-Paccot 2, 1700 Fribourg  Switzerland1     Tel: +41.79.705.46.70 - Fax: +41.26.465.13.58 4  Visit our Web site at http://Didier.Morandi.Free.fr   ------------------------------  % Date: Wed, 04 Apr 2001 15:15:18 +0000a- From: Peter Harding <harding@herald.ox.ac.uk> 0 Subject: Re: dir/date=m/siz/gran/since=y ... etc/ Message-ID: <3ACB3A86.40B2D3B3@herald.ox.ac.uk>e   Didier Morandi wrote: N > What is the problem with 1h CPU? You can have a batch job doing this at 5am, > can't you.  G Some of us already have all CPUs thrashing away on batch jobs all nightn already.   > Il fait beau  La Hague ?=  
 Il pleut ici.y   -- c http://i.am/getting_married , ICQ 40628243 Tel 07092057581 Fax 08707345230   ------------------------------  % Date: Wed, 04 Apr 2001 19:38:11 +0200c, From: Didier Morandi <Didier.Morandi@gmx.ch>0 Subject: Re: dir/date=m/siz/gran/since=y ... etc& Message-ID: <3ACB5C03.88DE1466@gmx.ch>   Il pleut aussi  Berne :-(   Peter Harding wrote: >  > Didier Morandi wrote:e > >k > > Il fait beau  La Hague ?a >  > Il pleut ici.e   ------------------------------  $ Date: Wed, 4 Apr 2001 08:40:24 -04005 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>oG Subject: Re: EV7 and EV8 for the low end (was: VMS-Related: Affordable) 2 Message-ID: <TCEy6.518$fB6.14875@news.cpqcorp.net>  $ Larry Kilgallen wrote in message ...G >> At the simplest level, EV8 can be treated by the OS as 4 independent, CPUs.p6 >> With each "thread" appearing as an independent CPU. >dB >But wouldn't such use suffer from the various threads not sharing= >an address space (tromping on cache in different fashions) ?A  J I wasn't commenting on how the EV8 might be best used, but to the questionH if the threads had to execute within a process.  AFAIK the answer is no.I There is enough state to allow each thread to act as a virtual CPU.  To arK O/S that does no additional work, a single chip could look like a 4-way SMPe system.i   ------------------------------  $ Date: Wed, 4 Apr 2001 11:44:02 +0200- From: "Roland Hauk" <roland.hauk@indramat.de> * Subject: Re: F$CONTEXT anomaly ? Feature ?- Message-ID: <9aeq2b$4508@sunny.mannesmann.de>:   Please try it in octal   $....03 $ TEMP = F$CONTEXT("PROCESS",CTX,"GRP",%o111,"EQL")  $...   --  5 Roland Hauk        ( Email: roland.hauk@indramat.de ) , Rexroth Indramat  ( http://www.indramat.de )   ------------------------------  # Date: Wed, 04 Apr 2001 11:03:00 GMTt1 From: CSABA  HARANGOZO   <csabah@zipworld.com.au> * Subject: Re: F$CONTEXT anomaly ? Feature ?9 Message-ID: <EbDy6.1155$CN.194479@nostril.pacific.net.au>-  , Roland Hauk <roland.hauk@indramat.de> wrote: > Please try it in octal   > $....h5 > $ TEMP = F$CONTEXT("PROCESS",CTX,"GRP",%o111,"EQL")T > $...  > 	Yes, that's what I did ( I mentioned it in my post ). Thanks.   				Cheers,		Csaba  I    ---------------------------------------------------------------------- E    * Csaba I. Harangozo     |    'To err is human', said the hedgehogSE    * csabah@zipworld.com.au |           as he dismounted a wirebrush. I    ----------------------------------------------------------------------2;    EARTH::AUSTRALIA:[SYDNEY]HARANGOZO.CSABA;1, delete? [N]:f   ------------------------------  # Date: Wed, 04 Apr 2001 11:12:45 GMTr1 From: CSABA  HARANGOZO   <csabah@zipworld.com.au>.* Subject: Re: F$CONTEXT anomaly ? Feature ?9 Message-ID: <NkDy6.1156$CN.194489@nostril.pacific.net.au>-  * Adam Maulis <maulis@ludens.elte.hu> wrote:o > In article <Odiy6.1133$CN.190227@nostril.pacific.net.au>, CSABA  HARANGOZO   <csabah@zipworld.com.au> writes: @ >> 	This bit me today. I was building up a context list with the( >> 	above-mentioned lexical, like this : >> a	 >> 	$....r7 >> 	$ TEMP = F$CONTEXT("PROCESS",CTX,"GRP","111","EQL")e >> 	$... >>  C >> 	( 111 is a UIC group number. ) Fire it off, no go. Zilch, nada. C >> 	It took me a while and a bit of experimenting before I realizedt4 >> 	that F$CONTEXT wanted the value as _decimal_(!).    3 > No. f$context wanted it as integer. Not a string.m    > from the help lex f$cont argu:  ; >          GRP        Integer  GEQ,       UIC group number.t    = > When you specify string, then normal DCL string to integer nG > translation is performed. ( "true" -> 1, "111" -> 111, "false" -> 0 )t  : > when you specify integer, there is no rational of radix.  J > I mean when AAA symbol has a value 111 it has a value %d111, %x6F, %o157 > %b01101111 in same time.  A 	Thanks, Adam. I think what confused me is that I used the doubleu@ 	quotation marks ("), not realising that these make it a string.8 	Tomorrow I will try without them, and see what happens.> 	( Still, this way it works : "%O111", even with the quotation
 	marks, hmm )d  						   Cheers,   Csabat  I    ----------------------------------------------------------------------.E    * Csaba I. Harangozo     |    'To err is human', said the hedgehogoE    * csabah@zipworld.com.au |           as he dismounted a wirebrush.gI    ----------------------------------------------------------------------o;    EARTH::AUSTRALIA:[SYDNEY]HARANGOZO.CSABA;1, delete? [N]:a   ------------------------------  % Date: Wed, 04 Apr 2001 07:03:13 -0500>( From: Alva Butler <Alva.Butler@noaa.gov> Subject: Firmware question.....m/ Message-ID: <3ACB0D80.404D85E3@nesdis.noaa.gov>m  
 Good morning,   E      Do you know what is the most current version of firmware for the  following systems ?s        VAX 4000/600p      VAX 4000/200m   Regards, John Colhoune   ------------------------------  # Date: Wed, 04 Apr 2001 14:30:12 GMTd From: Dirk Munk <munk@home.nl># Subject: Re: Firmware question..... ' Message-ID: <3ACB2FF3.BE68D233@home.nl>a  H I don't think a Vax has replacable firmware. Adapters in a Vax may have, but not the Vax itself AFAIK.r   Alva Butler wrote:   > Good morning,  >hG >      Do you know what is the most current version of firmware for thes > following systems ?D >  >      VAX 4000/600. >      VAX 4000/200o >h > Regards, John Colhoung   ------------------------------  % Date: Wed, 04 Apr 2001 08:43:03 -0600 % From: Dan O'Reilly <dano@process.com>h# Subject: Re: Firmware question..... A Message-ID: <5.0.2.1.2.20010404084229.06af7390@ntbsod.psccos.com>   M It does to the extent that some have ROM chips that can be replaced.  There'sw( not a flash firmware like the Alpha has.  & At 08:30 AM 4/4/2001, Dirk Munk wrote:I >I don't think a Vax has replacable firmware. Adapters in a Vax may have,  >but not the Vax itself AFAIK. >  >Alva Butler wrote:l >  > > Good morning,C > >tI > >      Do you know what is the most current version of firmware for the  > > following systems ?r > >n > >      VAX 4000/600b > >      VAX 4000/200u > >  > > Regards, John Colhount   ------I +-------------------------------+---------------------------------------+ I | Dan O'Reilly                  |                                       |KI | Principal Engineer            |  "Why should I care about posterity?  |6I | Process Software              |   What's posterity ever done for me?" |-I | http://www.process.com        |                    -- Groucho Marx    |lI +-------------------------------+---------------------------------------+.   ------------------------------  % Date: Wed, 04 Apr 2001 09:32:26 -0500e+ From: Christopher Smith <csmith@amdocs.com>/$ Subject: RE: Free The OpenVMS Seven!L Message-ID: <3B55D7F383B0D31197D9009027541CBF0D9D1CFD@cmiexch1.cmi.itds.com>   > -----Original Message-----2 > From: Christof Brass [mailto:brass@infopuls.com]  A > Objection, your honour! :-) Only against the last part "nor thehA > machine for which they build". I claim that if the abstractionseB > are well taken and the language implementation is appropriate we= > shouldn't bother much about the HW details. This is exactlye@ > what's abstraction is all about: not having to take care about@ > details which shouldn't be important for the task at hand on a= > specific level. That is the idea of stepwise refinement andt > layering.y  L Ahh -- I think you're arguing for a language at a higher level than has beenH seriously attempted in anything but a script language before. :)  On theL other hand, in a few languages, you're right, it is possible to know nothingJ about the machine, and still produce workable programs.  (I find that it'sD generally more difficult.  To really do it effectively, you'd need aK "programming environment" with its own interface, maybe its own filesystem.lI Something along the lines of early Pascal, or modern Oberon/inferno/maybem java.)  @ > > Bjarne Stroustrup has said on several occasions that C++ is  > not, strictly08 > > speaking, an "object oriented" language. (Though is  > _possible_ to write OO2 > > code in C++)  People just won't listen to him.   > Objection, your honour!s> > I studied the paper by Bjarne Stroustrup which came with theA > first AT&T C++ compiler (preprocessor). I have it still and canxB > quote from it. Out of my memory it was quite amusing to read and? > culminated in the subtle statement that you can program in ooo? > way in every language but there is a big difference between aC= > language which allows you to program oo with or in it and a	" > language that *supports* it. :-)  : From http://slashdot.org/interviews/00/02/25/1034222.shtml  K If you like long words, you can say C++ is a "multi-paradigm language," but:H simply saying "C++ is an OOPL" is inaccurate. I wrote a paper about thatK "Why C++ isn't just an Object-Oriented Programming Language" (download fromCB my papers page). I presented that paper at OOPSLA - and survived.   > The papers page is http://www.research.att.com/~bs/papers.html  A > And he claimed that C++ were supporting oo quite well. Maybe hen > changed his mind later ...B > Could you give exactly a reference? I you like me to do it first@ > I'll do it because it will take me about 15 minutes of digging@ > out my material from my college time (and another 5 minutes to > store it back again).s  J He still stands by that statement as far as I've seen.  The point was thatJ even while "supporting" OO, C++ also allows writing non-OO code -- it does not enforce the OO method.  E > > > But again: I insist in my question about "knowing how computersaB > > > really work". I'm sure that it is a misconception to avoid aA > > > level of abstraction because it widens the distance betweensA > > > written source and machine instructions because it would beaD > > > difficult to define the appropriate separation line to justifyE > > > the use of an abstraction given by the PL C but not Ada and notC  F I think things of that sort should be weighed on a case-by-case basis,I personally.  There may be situations where more abstraction is desirable,lL and others where it's not.  I would not personally write a bootstrap loader,L or a real-time application (unless the compiler was astoundingly good!) in a very abstract language.s   > (Macintosh desktopH > > metaphor, anyone?  Perhaps even more sinister ... device abstraction= > > libraries.  I don't claim these things are bad -- I like   > them in many casesC > > -- but they lead to misconception.), and a larger percentage ofa; > > "programmers" these days don't even think to challenge I > those ideas in their > > own minds.  ? > Very well presented points! This seems to be also a claim for = > better documentation of the interfaces and how to use them.o  G Certainly.  Though that may be a separate problem.  Do you remember the J volume of manuals that used to come with even home computers just 15 years ago?  @ > Did you complain about the Mac desktop metaphor? I think about9 > buying one of these cool Titan PowerBooks if MacOS X is  > installed as default.n  K I wasn't complaining about the metaphor -- I was complaining that the users-D are never told "this isn't exactly how the computer works, it's justI presented this way for your convenience."  You'd be surprised -- or maybe J not -- that most people don't understand this.  Try explaining to somebodyE why, when they drag a file from a floppy onto their desktop, it stilli) disappears when the floppy is dismounted.   > > > For instance, when you have a function which is buried in  > the hierarchy of' > > objects such that you call it thus:d > > . > > MyObject.SomeInheretedClass.SomeFunction() > >   > > (oversimplification here...)< > > It is most likely compiled into just a plain jump.  The  > "hierarchy" there is8 > > purely a linguistic device.  It shows which bits of  > previous work, if youa8 > > will, we've called upon in setting this "object" up.  < > This is a good and a bad example at the same time. I would< > suggest to remove the "SomeInertedClass." part of the call@ > because exactly this is a major aspect against oo: you burn inB > the class and by that you prevent dynamic binding. The classical  G Yep -- the funny thing is that were I actually writing the program, I'dpJ never have written it that way unless there were two functions of the same= name.  I didn't think of that when I was writing the example.i  @ > > That's not a bad thing, and in fact, it's incredibly useful  > on occasion.= > > The problem comes when some person fails to realize that d > it's an "illusion,"S> > > just as macros, high-level flow control, and other things.  A > I need an explanation of "illusion". We shouldn't forget that at@ > high level PL is an abstraction per se and you even don't know= > if the HW is closer to the language than you thought - everS9 > heard about Java-In-Silicon? Do you remember the frenchd > Pascaline?  H Good point -- and it's really something close to what I was saying.  YouK probably noticed my mention above of the fact the this isn't just a featurem. of OO, but of high-level languages in general.  L "Illusion," meaning that the language -- as it should be -- is doing quite aB bit more work for us in these areas than it appears.  If you don'tI understand that, you might end up doing some things that aren't very good K for your end product.  For instance -- and I've seen this before -- leaving K the compiler optimizations set to the defaults because you don't understand H to check them! :) (Ok, the programmer who did this wasn't too smart, but0 it's the first thing that popped into my head)     > Back to nature, he? ;-)r* > I for one prefere my comfortable bed :-)  K Me too -- I'd like to think that I could survive in the wilderness if I hadiL to.  That's not true, of course, since my continued survival is dependant onL a medicine which hasn't been around for long, actually.  That aside, though,L I hope that my furnace will not prevent me from knowing how to make fire. :)  J Have you read the science-fiction stories (it's a recurring theme) where aG civilization builds a computer which will perform some pretty importantlI function, and the computer -- maybe it's a VAX -- ends up working so welltH that the people who designed it are no longer around when it breaks?  OfI course nobody's bothered to learn how it works; it wasn't necessary.  NownL society is left without <some function> because nobody can figure out how to	 do it. :)i  > > I never loved the "one million flies can't be wrong" type ofB > argument which led to "nobody got fired for buying IBM (nowadays > M$)".   9 I wonder whether anyone ever got promoted for buying IBM?e   Regards,   Chrise  ! Christopher Smith, Perl Developeru Amdocs - Champaign, IL   /usr/bin/perl -e '? print((~"\x95\xc4\xe3"^"Just Another Perl Hacker.")."\x08!\n");o 'q   ------------------------------  % Date: Wed, 04 Apr 2001 10:55:54 +01000  From: steven.reece@quintiles.com1 Subject: Re: GS160 froze due to a bad KGPSA-CX...sH Message-ID: <OFB339E912.26040BC2-ON80256A24.0035ABE9@qedi.quintiles.com>  , Not particularly neat, tidy or nice but.....I How about partitioning the GS160 into more than two instances and put thet# KGPSA cards on the new instance(s)?tG Use in-memory transports for clustering and hope/plan that any problems 6 with the KGPSAs don't cause the whole cluster to hang. I did say it wasn't neat.b Steve.   John C. Koska wrote: >>> 8 Problem:  Two bad KGPSA-CX (no green LEDs) caused system4 to hang, and worse... prevented system from init'ing6 properly after power cycle.  (Console hangs on polling of bad KGPSA cards)A  5 Solution:  Remove failed KGPSA cards and power cycle.e  9 A Better Long Term Solution:  Hardware partition GS160 in1 half, and cluster.  : Concern/Worry:  I fear that even with hardware partitioned< GS160, that if I had two KGPSA cards go bad again, each in a3 hardware partition, I would be in same situation ofg< downtime. (Granted, I might stand a better chance of winning my state lottery.) <<<3   ------------------------------  $ Date: Wed, 4 Apr 2001 11:55:04 +0200- From: "Ruzsinszky Attila" <aruzsi@mailbox.hu>n) Subject: Re: Hopf 6841 dtss time providera, Message-ID: <9aeqpe$39o$1@athena.euroweb.hu>  < > Who uses that and can help me to configure and find error?C So no one uses Hopf 6841 or any Hopf  GPS clock with GPS except me?0   RuzsiF   ------------------------------  # Date: Wed, 04 Apr 2001 14:26:08 GMT1 From: Dirk Munk <munk@home.nl>) Subject: Re: Hopf 6841 dtss time provider:' Message-ID: <3ACB2EFF.16FC523B@home.nl>s  C Could be. I have seen implementations with Hopf clocks that use thee< DCF77 timesignal from Frankfurt, but not with the GPS clock.   Ruzsinszky Attila wrote:  > > > Who uses that and can help me to configure and find error?E > So no one uses Hopf 6841 or any Hopf  GPS clock with GPS except me?d >c > Ruzsij   ------------------------------  # Date: Wed, 04 Apr 2001 12:04:06 GMTs3 From: "Tom Wade" <t.wade@vms.eurokom.ie.removespam>rJ Subject: Re: How to tell if a system is booting immediately after a crash./ Message-ID: <W4Ey6.24163$PF4.36434@news.iol.ie>a  9 Lyndon Bartels <lbartels@pressenter.com> wrote in message'( news:3AC8D949.281B9C29@pressenter.com...H > I want, during a startup, if the system has gone done as a result of a* > crash or power outage or... whatever.... >bI > Why? So if the system crashes and reboots in the middle of the night, Io0 > can tell the computer to page me or something. > D > I was thinking of having syshtdown.com write out a data file. ThenD > during subsequent boot, systartup_vms look for the file, if found,H > delete all versions of the file, and peacefully exit. If not found, so > my alarm procedures.  G Rather than generate Unix style flag files, why not use one of the usergI defined SYSGEN parameters ?  For example, set the parameter USER3 to 1 in0G SYSGEN and WRITE CURRENT. Your system shutdown command file can set theaI parameter to 0.  The startup file checks the value of USER3.  If it is 0,pG then assume a normal shutdown was done, set the parameter back to 1 andpF exit. If the parameter is 1 already you can assume a crash took place.    L ---------------------------------------------------------------------------- --A Tom Wade    | EMail: T.Wade@vms.eurokom.ie  (all domain mailers). G EuroKom     | X400:  g=tom;s=wade;o=eurokom;p=eurokom;a=eirmail400;c=ie"& 30, Dale Rd | Tel:   +353 (1) 278-7878& Stillorgan  | Fax:   +353 (1) 278-78793 Co Dublin   | Disclaimer:  This is not a disclaimer-@ Ireland     | Tip:         "Friends don't let friends do Unix !"   ------------------------------  # Date: Wed, 04 Apr 2001 12:36:05 GMTa& From: "john nixon" <jnixon@cfl.rr.com>% Subject: installing pci cards on ds10-< Message-ID: <VyEy6.25029$z4.5619716@typhoon.tampabay.rr.com>  I How difficult or tricky is it to install a PCI card in a DS10 running VMS- 7.2-1?  H Do you have to deal with CSRs or Vectors (showing my age here), or otherL interrupt addresses, or do you just stick the card in and it works.  Or doesI it depend on the card?  I need to order a bisynch card (pbxdp) for a DS10e? and don't know if we should install it or if we need to pay forc; installation.  (installation will cost more than the card).t  B I suppose when we get the card, it will come with documentation or- instructions, but I would like to know first.a   Thanks.r   ------------------------------  $ Date: Wed, 4 Apr 2001 08:43:55 -04005 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>.) Subject: Re: installing pci cards on ds10a2 Message-ID: <aGEy6.519$fB6.14956@news.cpqcorp.net>  F If it is a supported option, stick it in, power up the system, and youK should be all set.  PCI is plug and play, as long as the O/S has the driverG; support for the card it will recognize it and configure it.y  K I don't know anything about the pbxdp, but if it is in the supported optionm list for VMS, you are all set.       john nixon wrote in message ...iJ >How difficult or tricky is it to install a PCI card in a DS10 running VMS >7.2-1?  > I >Do you have to deal with CSRs or Vectors (showing my age here), or othertH >interrupt addresses, or do you just stick the card in and it works.  Or doesJ >it depend on the card?  I need to order a bisynch card (pbxdp) for a DS10@ >and don't know if we should install it or if we need to pay for< >installation.  (installation will cost more than the card). >-C >I suppose when we get the card, it will come with documentation or.. >instructions, but I would like to know first. >' >Thanks. >d >o   ------------------------------  # Date: Wed, 04 Apr 2001 13:37:41 GMTg= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)-) Subject: Re: installing pci cards on ds10A0 Message-ID: <009FA073.5B5C43F9@SendSpamHere.ORG>  j In article <aGEy6.519$fB6.14956@news.cpqcorp.net>, "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> writes:G >If it is a supported option, stick it in, power up the system, and you L >should be all set.  PCI is plug and play, as long as the O/S has the driver< >support for the card it will recognize it and configure it.  7 OK, that explains it.  The ZXLp-L2 is not supported. ;)-  J Hmm.  What I wouldn't give to see the very nice and expensive option work.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMa            hO city, n., 1. a place where trees are cut down and streets are named after them.t   ------------------------------  # Date: Wed, 04 Apr 2001 13:58:57 GMT & From: "john nixon" <jnixon@cfl.rr.com>) Subject: Re: installing pci cards on ds10 < Message-ID: <BMFy6.25321$z4.5673421@typhoon.tampabay.rr.com>  J Good.  That sounds promising.  It is a supported option, but I am not sureG if the driver is included in the base OS, or if it is part of a layeredrK software package (in this case, 2780/3780PE).  If the driver is provided bysJ the software, can I install the software before installing the device?  OrI does the device need to be installed so that the software will know which0 driver to load.t  @ "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote in message, news:aGEy6.519$fB6.14956@news.cpqcorp.net...H > If it is a supported option, stick it in, power up the system, and youF > should be all set.  PCI is plug and play, as long as the O/S has the driver= > support for the card it will recognize it and configure it.  >eF > I don't know anything about the pbxdp, but if it is in the supported option  > list for VMS, you are all set. >e >y > ! > john nixon wrote in message ... L > >How difficult or tricky is it to install a PCI card in a DS10 running VMS	 > >7.2-1?t > >6K > >Do you have to deal with CSRs or Vectors (showing my age here), or other J > >interrupt addresses, or do you just stick the card in and it works.  Or > doesL > >it depend on the card?  I need to order a bisynch card (pbxdp) for a DS10B > >and don't know if we should install it or if we need to pay for> > >installation.  (installation will cost more than the card). > >nE > >I suppose when we get the card, it will come with documentation orc0 > >instructions, but I would like to know first. > >m
 > >Thanks. > >[ > >o >a >i >p   ------------------------------  % Date: Wed, 04 Apr 2001 09:52:36 -0500h" From: Earl Lakia <lakia@ipact.com>) Subject: Re: installing pci cards on ds10e) Message-ID: <3ACB3533.5E909634@ipact.com>    john nixon wrote:u  K > How difficult or tricky is it to install a PCI card in a DS10 running VMSi > 7.2-1? >nJ > Do you have to deal with CSRs or Vectors (showing my age here), or otherN > interrupt addresses, or do you just stick the card in and it works.  Or doesK > it depend on the card?  I need to order a bisynch card (pbxdp) for a DS10lA > and don't know if we should install it or if we need to pay forn= > installation.  (installation will cost more than the card).e >gD > I suppose when we get the card, it will come with documentation or/ > instructions, but I would like to know first.l >o	 > Thanks.t  5 The POST (Power On and Self Test) code probes the PCIg@ bus and this information is then made available to OpenVMS after4 bootstrap.  There are modules ICBMs that help driver2 determine if their card is present on the bus (PCI5 standard has a set of configuration registers and oner8 of the registers identifies the vendor and the product).> Most third party suppliers I've come across (including myself)9 use a privileged program that browses the above mentionedd. tables and then loads the driver using sysman.  ; Therefore, I suspect that even if a driver is not supportedl6 by Compaq, they will provide some way to load it.  You9 should most likely install the card first and then try tot: load the driver for the card.  I would try myself and then9 think of buying the support if I couldn't get it working.    -earlo ----------------------
 Earl D. Lakiae Senior Staff Engineer)
 IPACT Inc. lakia@ipact.comi   ------------------------------    Date: 05 Apr 2001 00:42:57 +0800, From: Paul Repacholi <prep@prep.synonet.com>. Subject: It's all a Zincuous plot I tell you!!- Message-ID: <8766gk4nvy.fsf@prep.synonet.com>O  ? Sun has been working on preventing those nasty remote denial ofe? service attacks in a well tested (well, for Sun anyway) method.s  G This reduces the proportion of DOS down time for the happy owner of thea new Ultra-SparcIII.   F Yes, the cache-Gate fogs of Sillycon valley are now playing at a Blade 1000 somewhere near[1] you.e  D Oh dear. But, there is a u-code fix RSN, and the midrange boxes willE be fixed by Scotts tame vet before being alowed to wander free range.u   [1] <= 15000 NMb -- g< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.l@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  % Date: Wed, 04 Apr 2001 11:17:33 -0500c- From: GSV00001 <GSV00001@gsvms2.cc.gasou.edu>n Subject: Meaning of RSBREFNZRO?eM Message-ID: <Pine.VMS.3.91-2.1010404111402.40190A-100000@gsvms2.cc.gasou.edu>   J What is the meaning of RSBREFNZRO?  Help/message says "tried to deallocate RSB with non-zero ref. count."  J I have had two crashes with this error and a mysterious reboot last night.D The reboot shows no entries in operator or error logs except that itG is booting.  The Clue$last file says "This Crash already in this file!"sJ which tells me the system did not perform a crash dump.  However, the last' two crashes had the RSBREFNZRO message.a  I The system is a VAX 6420 with OVMS 7.1  It's been very stable for 3 yearseI with 7.1 and now I have had these events occur.  The first one in January G and the last two within 5 days of each other.  They all occurred during B off hours with no load and several non-privledged users logged on.   TIA,   Jonathan   ------------------------------  # Date: Wed, 04 Apr 2001 16:03:20 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)d# Subject: Re: Meaning of RSBREFNZRO?w0 Message-ID: <009FA087.B4618EE6@SendSpamHere.ORG>  } In article <Pine.VMS.3.91-2.1010404111402.40190A-100000@gsvms2.cc.gasou.edu>, GSV00001 <GSV00001@gsvms2.cc.gasou.edu> writes:iK >What is the meaning of RSBREFNZRO?  Help/message says "tried to deallocate  >RSB with non-zero ref. count."c  ' ... and that description is not enough?n    K >I have had two crashes with this error and a mysterious reboot last night. E >The reboot shows no entries in operator or error logs except that ithH >is booting.  The Clue$last file says "This Crash already in this file!"K >which tells me the system did not perform a crash dump.  However, the lasth( >two crashes had the RSBREFNZRO message. >oJ >The system is a VAX 6420 with OVMS 7.1  It's been very stable for 3 yearsJ >with 7.1 and now I have had these events occur.  The first one in JanuaryH >and the last two within 5 days of each other.  They all occurred duringC >off hours with no load and several non-privledged users logged on.l >l >TIA,l >i	 >Jonathanp >n  J There is likely a LKB somehow still referencing the RSB (or a bug that didJ not decrement the ref count when some LKB was tossed.  I'd log a call with support if I were you.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMh            dO city, n., 1. a place where trees are cut down and streets are named after them.u   ------------------------------  % Date: Wed, 04 Apr 2001 14:07:23 +0200l= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>p" Subject: Re: Merging DCLTABLES.EXE) Message-ID: <3ACB0E7B.59B76202@gtech.com>    JF Mezei wrote:dF > If one has two separate nodes with different software that are beingK > integrated into a homogenous cluster, what is the best way to merge the 2b > DCLTABLES.EXE ?  > N > Must one make a manual list of commands present on node 2 but not on node 1,H > use VERB to extract them and then add them to node 1's DCLTABLES.EXE ? > M > What about subcommands that might be present on node 2 but not node 1 (somee+ > subset of the SET command for instance) ?n  B I would recommend against VERB and EDIT, because you may suceed in gettingt) DCLTABLES rigth, but what about the rest.n  % You should install the missing stuff.a   Arne   ------------------------------  # Date: Wed, 04 Apr 2001 13:24:27 GMT 8 From: hammond@not@peek.ppb.cpqcorp.net (Charlie Hammond)" Subject: Re: Merging DCLTABLES.EXE2 Message-ID: <fgFy6.521$fB6.15030@news.cpqcorp.net>  . In article <87y9thpyfv.fsf@prep.synonet.com>, . Paul Repacholi <prep@prep.synonet.com> writes:  E >I have got into the habit of squileling the CLUs away in sys$update.c8 >(BTW, can this become a standard 'thing to do perhaps?)   CLU?  CLD, I presume.  yes?s  5 Open VMS Alpha does this.  It needs them when you usee' SYS$SYSTEM:AXPVSM$PCSI_INSTALL_MIN.COM.,  C Since the CLD may not be of much use without other supporting filesu< (images) the general value of this is subject to discussion. Personally, I like the idea.   -- dK     Charlie Hammond -- Compaq Computer Corporation -- Pompano Beach  FL USAmH        (hammond@not@peek.ppb.cpqcorp.net -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Wed, 04 Apr 2001 14:01:38 -0400t  From: norm.raphael@jamesbury.com< Subject: O.T. (Was: Re: dir/date=m/siz/gran/since=y ... etc)4 Message-ID: <C2256A24.00622745.00@jklh21.valmet.com>  ' Please respond to Didier.Morandi@gmx.cht  , Il ne pleut pas ici a la Novelle Angleterre.   > Il pleut aussi =E0 Berne :-( >b > Peter Harding wrote: > >t > > Didier Morandi wrote:. > > > ! > > > Il fait beau =E0 La Hague ?i > >e > > Il pleut ici.g   ------------------------------  % Date: Wed, 04 Apr 2001 17:32:18 +0100b  From: steven.reece@quintiles.comE Subject: PCSI Patch - VMS VAX v7.2 (Was: Re: TCP/IP 5.0A ECO Appears)rH Message-ID: <OF39A4DDD2.995B884C-ON80256A24.005A0D51@qedi.quintiles.com>  " Peter Langstoeger wrote / quoted : >>>sB >The long awaited VMS72_PCSI ECO for OpenVMS VAX is now officially	 availabler  ! I did read the announcement, too.-K But, after two more weeks, it is still NOT available in the FTP server areav! and NOT downloadable via DSNlink.:1 So, is it really available or do I need glasses ? " Or should I get paranoic instead ? <<<q  C It is in the UK.  VMS72_PCSI-V0100 or, more correctly for the file, # DEC-VAXVMS-VMS72_PCSI-V0100--4.PCSIe  K I got a copy via DSNlink on the 21-MAR-2001.  There's a new kit as of todaymI too, probably because the cover letter that came with it documented it asK being a kit for Alpha !!!   G The e-mail that came back to confirm I'd ordered it also told me to use: VMSINSTAL to put the kit on :-  G "If this ECO kit is for an OpenVMS system, it can be installed with theo VMSINSTAL command procedure:  F $ @SYS$UPDATE:VMSINSTAL VMS72_PCSI-V0100 DSN$COPY_DIRECTORY: OPTIONS N  = where "VMS72_PCSI-V0100" is the name of this particular ECO."S   Steve.   ------------------------------  % Date: Wed, 04 Apr 2001 16:23:53 +0100t0 From: andrew harrison <andrew.nospam@uk.sun.com>, Subject: Re: POSIX Streams, File Permissions* Message-ID: <3ACB3C89.DF4F7B73@uk.sun.com>   Christof Brass wrote:a >  > andrew harrison wrote: > >  > > Christof Brass wrote:y > > >3 > > > andrew harrison wrote: > > > >a > > > > Christof Brass wrote:2 >  > [SNIP] > @ > > > IP is a widely used standard and at the moment there is no? > > > promising way to avoid it. Besides that it is possible to 2 > > > implement an IP stack according to the RFCs. > > >n > > : > > Of course it is but the UCX stack came from UNIX where; > > it was implimented according to the RFC's. The point isp9 > > it origionated from UNIX so you need to rip it out ifc9 > > you are to deliver on your architectural purity rant.. > @ > I really don't understand the problem. You can use whatever OS= > you want to develop helpful ideas. The ideas may or may notw? > influenced by the OS. The fact that VMS has been designed andaA > implemented because the engineers at DEC knew after 10 years ofpB > bad experience with UNIX (which should be avoided) what mistakes= > to avoid led to a clean and well thought through design. Of0< > course they didn't implement all forthcoming standards and? > functions. The important thing is how a feature is integrated A > and that its API is according to the rules/standards of the OS.n >   6 The UCX IP stack was developed on Tru64 and ported to 2 OpenVMS they share code. This isn't an idea it is + actual code. The UCX utilites are the same.s  7 The same applies to Motif. You don't think that OpenVMS 5 engineering wrote the OpenVMS Motif lib and utilites  4 from scratch from the spec do you, they didn't they 
 ported it.    ? > [All the examples of standards/protocols snipped: CDE, Motif, " > DECthreads, PERL, Tcl/Tk, CORBA] >   < You don't think that perl and Tck/Tk are developed from the 0 Specs for OpenVMS, they wern't they were ported.    @ > If a lot of protocols have firstly been implemented not on VMSA > than this reflects mainly the fact that the market share of VMS   > in these areas isn't that big.  > Of course the great thing about your argument is that you have; now argued for something that you origionally claimed would  be the ruination of OpenVMS.  < COE will be implimented on OpenVMS (in the kernel) from the 8 spec which as you have now said is perfectly acceptable.  > The COE utilites will probably be ported from Tru64 to OpenVMS8 but they are utilites and should not have any impact on  OpenVMS itself..  : You appear to have forgotten what your origional argument : was since you are now agreeing that it is OK to impliment 0 from a spec provided the implimentation is done 
 properly.   3 Perhaps you should appologise to everyone for beingl a time waster.     Regardsd Andrew Harrisond Enterprise IT Architect-   ------------------------------    Date: 05 Apr 2001 01:08:53 +0800, From: Paul Repacholi <prep@prep.synonet.com>, Subject: Re: POSIX Streams, File Permissions- Message-ID: <87r8z8384a.fsf@prep.synonet.com>   2 andrew harrison <andrew.nospam@uk.sun.com> writes:  8 > The UCX IP stack was developed on Tru64 and ported to 4 > OpenVMS they share code. This isn't an idea it is - > actual code. The UCX utilites are the same.t  1 BZZZT. UCX was written before T64/DU/OSF existed.r    9 > The same applies to Motif. You don't think that OpenVMSo7 > engineering wrote the OpenVMS Motif lib and utilites (6 > from scratch from the spec do you, they didn't they  > ported it.  C BZZZT W and X where done at MIT, on a farm of Vax 750s. VMS was one " of the original systems that ran X  > > COE will be implimented on OpenVMS (in the kernel) from the : > spec which as you have now said is perfectly acceptable.  E No, part will be in the kernel, some in kernel mode code, and most in1D libraries. You know Andrew, those new fangled thing you need special  swithches to your compilers for.  D > The COE utilites will probably be ported from Tru64 to OpenVMS butE > they are utilites and should not have any impact on OpenVMS itself.   C Well, as T64 is a COE platform, that is ahead of what Sun can do...i/ Or you could buy some SGIs Andrew and use them.o     -- s< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.l@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------   Date: 4 Apr 2001 07:15:27 -0500n9 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)sG Subject: Re: product announcement: the rebirth of DateSim (version 2.1) 3 Message-ID: <JYfeSkOjf6Qw@eisner.encompasserve.org>c  n In article <aCeAXgtvmStT@tachxxsoftxxconsult>, wayne@tachysoft.xxx.412538.killspam.015a (Wayne Sewell) writes:  Q > This customer uses DateSim during remote logins to set the time for the processSL > to the remote user's local time, which is much easier on the nerves of theN > users.  As far as they are concerned, the server is running in the same time > zone they are.  E By "remote", do you mean the technical VMS definition (not local, note! dialup, not network, not batch) ?t   ------------------------------  $ Date: Wed, 4 Apr 2001 12:15:59 +0200$ From: "Jakob Erber" <erberj@post.ch> Subject: Re: pthreads + printf Message-ID: <3acaf460$1@hcwe67>i  ? Does it also happen, if you avoid sleep and use usleep instead?m    D "MrSignor" <MrSignor@nospam_bellatlantic.net> schrieb im Newsbeitrag1 news:d3ty6.6570$jL4.986859@typhoon1.ba-dsg.net...- > Alpha VMS 7.2-1T >I? > I have been writing some threaded code and seem to have hit at > wall, or at least a deadlock.  >mG > Basically, I am compiling with /reentrancy=multi, although the onlineMA > help says this is applicable to the MAIN only, and I am linking ! > with threads_enabled + upcalls.F >2H > One thread enters an infinite while loop, it calls usleep, to wait forB > 100 milli-secs (adjusted for usleep), and then when it awakes it > prints the current time. > L > The other thread (main really) loops on sleep (rather than usleep), for 30J > seconds, and it uses printf to display how many seconds remain, if it is > awoken prematurely.s >eE > Problem, once it begins, a few outputs occur - sometimes, sometimesv< > only 1. At this point I am assuming a problem with Printf. >rA > The above explanation is over simplified, but tomorrow, I am toi( > do exactly what I have explained here. >p8 > Anyone have any suggestions for further investigation. >l > TIAl >  > -FredS >s >e   ------------------------------  # Date: Wed, 04 Apr 2001 14:20:08 GMT 2 From: seibel_r@localhost.localdomain (Rich Seibel) Subject: Re: pthreads + printf; Message-ID: <slrn9cmbce.572.seibel_r@localhost.localdomain>o  N On Tue, 03 Apr 2001 23:31:21 GMT, MrSignor <MrSignor@nospam_bellatlantic.net>  wrote: Alpha VMS 7.2-1 > > >I have been writing some threaded code and seem to have hit a >wall, or at least a deadlock. >dF >Basically, I am compiling with /reentrancy=multi, although the online@ >help says this is applicable to the MAIN only, and I am linking  >with threads_enabled + upcalls. >dG >One thread enters an infinite while loop, it calls usleep, to wait forlA >100 milli-secs (adjusted for usleep), and then when it awakes it, >prints the current time.a >nK >The other thread (main really) loops on sleep (rather than usleep), for 30[I >seconds, and it uses printf to display how many seconds remain, if it is0 >awoken prematurely. > D >Problem, once it begins, a few outputs occur - sometimes, sometimes; >only 1. At this point I am assuming a problem with Printf.w >e@ >The above explanation is over simplified, but tomorrow, I am to' >do exactly what I have explained here.h > 7 >Anyone have any suggestions for further investigation.h > C printf is not thread-safe in POSIX usage.  You might try using your > own lock around it.  A thread mutex should be sufficient.  See= pthread_mutex_init, pthread_mutex_lock, pthread_mutex_unlock.t   -- wD --------------------------------------------------------------------D Rich Seibel, Software Engineer                 (314)579-0066 ext 220D Object Computing, Inc.                           seibel_r@ociweb.comD Need ACE training?                      See http://www.theaceorb.comD --------------------------------------------------------------------   ------------------------------   Date: 4 Apr 2001 09:31:40 -0500r+ From: young_r@encompasserve.org (Rob Young)s  Subject: Restart TCP/IP Services3 Message-ID: <x6UC7FL3dHi$@eisner.encompasserve.org>t  9 	Usually I stick at it until I figure it out... but givenc; 	the time spent in help and poking at com files, am tossingv; 	the towel.  For Digital/Compaq TCP/IP Services what is the # 	equivalent (short of a reboot) to:   0 	$ @multinet:start_server   ! Does restarts too!   	$ tcpip 	TCPIP> show version  9   DIGITAL TCP/IP Services for OpenVMS Alpha Version V5.0At   	Thanks.   				Rob    ------------------------------  % Date: Wed, 04 Apr 2001 19:40:30 +0200 , From: Didier Morandi <Didier.Morandi@gmx.ch>$ Subject: Re: Restart TCP/IP Services& Message-ID: <3ACB5C8E.66E69032@gmx.ch>   @sys$manager:tcpip$shutdowna then @sys$manager:tcpip$startup  O You also need to do this *each* time you change a TCP/IP parameter via the menul @sys$manager:tcpip$config4   D.   Rob Young wrote: > B >         Usually I stick at it until I figure it out... but givenD >         the time spent in help and poking at com files, am tossingD >         the towel.  For Digital/Compaq TCP/IP Services what is the, >         equivalent (short of a reboot) to: > 9 >         $ @multinet:start_server   ! Does restarts too!e   ------------------------------  " Date: Wed, 4 Apr 2001 10:41:59 GMT( From: Terry Kennedy <terry@gate.tmk.com>C Subject: Re: Seeking CD-R/CD-RW SCSI INQUIRY data / Yamaha CRW4416Sg' Message-ID: <GB9KDz.51q@spcuna.spc.edu>   * Paul A. Jacobi <nospan@nospam.com> writes:M > DKDRIVER performs a MODE SELECT to set the block size to 512 bytes, insteadWM > of the default 8192 bytes for CD-ROM drives.  The CHECK CONDITION is a good-N > indication that the device does not support 512-byte block which is required
 > by OpenVMS..  L   Minor nit - the default data sector size on CD-ROM drives is 2048, and the only music sector size is 2352..  L   SCSI CD-ROM drives on emulating controllers (as opposed to on basic SCSI -L DKDRIVER) don't generally need to support 512-byte sectors as the controllerJ will re-block the data appropriately. All of the CMD MSCP-emulating boardsM do this, for example. And I've never seen an Emulex or Dilog one that didn't,h" though I haven't seen all of them.  K   By the way, I'm sure DKDRIVER (actually, probably PKxDRIVER) on Alpha andrH just about every VAX can spare the 2K to have a local deblock buffer. ItM should be trivial to do, since we're talking about read-only media here. [ThesI commands for writing are completely different, and CD-ROMs are mounted aso read-only anyway.]  4         Terry Kennedy             http://www.tmk.com5         terry@tmk.com             Jersey City, NJ USA-   ------------------------------   Date: 4 Apr 2001 15:08:07 GMTm2 From: mathog@seqaxp.bio.caltech.edu (David Mathog)C Subject: Re: Seeking CD-R/CD-RW SCSI INQUIRY data / Yamaha CRW4416S4, Message-ID: <9afdcn$a8l@gap.cco.caltech.edu>  _ In article <OLry6.512$fB6.14650@news.cpqcorp.net>, "Paul A. Jacobi" <nospan@nospam.com> writes:  >t< >"Chuck McCrobie" <mccrobie@cablespeed.com> wrote in message) >news:3AC94E41.D8DE5937@cablespeed.com...fK >> Our SCSI bus analyer reported a CHECK CONDITION on a MODE SELECT.  Don'trI >> remember the details though.  It seemed that DKDRIVER attempts to MODEw0 >> SELECT something that the drive doesn't like. >hL >DKDRIVER performs a MODE SELECT to set the block size to 512 bytes, insteadL >of the default 8192 bytes for CD-ROM drives.  The CHECK CONDITION is a goodM >indication that the device does not support 512-byte block which is required  >by OpenVMS.  J Just for the sake of argument, what does DKDRIVER do at the 512 byte levelF that it couldn't do if it was somehow instructed that the device had a cluster size of 16?    Regards,   David Mathog mathog@seqaxp.bio.caltech.edus? Manager, sequence analysis facility, biology division, Caltech     ------------------------------  % Date: Wed, 04 Apr 2001 16:09:59 +0100 - From: Tim Llewellyn <tim.llewellyn@bbc.co.uk>cC Subject: Re: Seeking CD-R/CD-RW SCSI INQUIRY data / Yamaha CRW4416St) Message-ID: <3ACB3947.12F38E6B@bbc.co.uk>o   Terry Kennedy wrote:   >. >.M >   By the way, I'm sure DKDRIVER (actually, probably PKxDRIVER) on Alpha and-J > just about every VAX can spare the 2K to have a local deblock buffer. ItO > should be trivial to do, since we're talking about read-only media here. [TheLK > commands for writing are completely different, and CD-ROMs are mounted aso > read-only anyway.] >  >h  Q yes Terry, I've been wondering for a while why VMS can't support 2048 byte blocksh" on CD's, seems a bit silly really.   --6 Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project0 MedAS at the BBC, Whiteladies Road, Bristol, UK.A Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.ukt  A I speak for myself only and my views in no way represent those ofm MedAS or the BBC.    ------------------------------    Date: 05 Apr 2001 00:37:09 +0800, From: Paul Repacholi <prep@prep.synonet.com>C Subject: Re: Seeking CD-R/CD-RW SCSI INQUIRY data / Yamaha CRW4416Sg- Message-ID: <87ae5w4o5m.fsf@prep.synonet.com>-  4 mathog@seqaxp.bio.caltech.edu (David Mathog) writes:  E > In article <OLry6.512$fB6.14650@news.cpqcorp.net>, "Paul A. Jacobi"L > <nospan@nospam.com> writes:f  > > >"Chuck McCrobie" <mccrobie@cablespeed.com> wrote in message+ > >news:3AC94E41.D8DE5937@cablespeed.com....  F > >> Our SCSI bus analyer reported a CHECK CONDITION on a MODE SELECT.@ > >> Don't remember the details though.  It seemed that DKDRIVERC > >> attempts to MODE SELECT something that the drive doesn't like.   F > >DKDRIVER performs a MODE SELECT to set the block size to 512 bytes,B > >instead of the default 8192 bytes for CD-ROM drives.  The CHECKB > >CONDITION is a good indication that the device does not support/ > >512-byte block which is required by OpenVMS.   F > Just for the sake of argument, what does DKDRIVER do at the 512 byteA > level that it couldn't do if it was somehow instructed that the " > device had a cluster size of 16?  9 It will support it, but only a a squential device (tape).-  F If it's not 512 byte blocks, it's not a disk! Definded that way.  careC to consider how to deal with boot code, home blocks, headers, n*512s? byte buffers... through everything? Or do you want the DOS typed> evereything is a unique foobar disk/floppy/CD/DVD/tape/tape...   -- r< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.i@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  % Date: Wed, 04 Apr 2001 12:02:42 +0100 0 From: andrew harrison <andrew.nospam@uk.sun.com>% Subject: Re: Spice for OpenVMS Alpha?A* Message-ID: <3ACAFF52.7F5A96C1@uk.sun.com>   Mark Berryman wrote: >  > andrew harrison wrote: >  > > Mark Berryman wrote: > >l > >> andrew harrison wrote:t > >> > >> > >>> Tim Llewellyn wrote: > >>>k > >>>s > >>>> Bill Gunshannon wrote:  > >>>> > >>>> > >>>>1 > >>>>> In article <3AC49C9F.B90D2FF4@bbc.co.uk>,e8 > >>>>>  Tim Llewellyn <tim.llewellyn@bbc.co.uk> writes:
 > >>>>> |>G > >>>>> |> and btw I have beeing doing some research in comp.unix.* :-)- > >>>>>S > >>>>> And??? > >>>>>  > >>>>> bill > >>>>L > >>>> well, they don't seem to have a high opinion of Solaris security over > >>>> in comp.security.unix.  > >>>> > >>>t > >>>e3 > >>> I am not sure what the point of your post is.  > >>>i2 > >>> Posters on this group do have a high opinion4 > >>> of OpenVMS security. This high opinion however8 > >>> appears to be based on a lack of information about2 > >>> the actual risks to OpenVMS fostered by what1 > >>> is a culture of security through obscurity.P; > >>> The CERT advisories for OpenVMS are a perfect example  > >>> of this. > >>>-9 > >>> Can you say with certainty that the differences your5 > >>> see in each newsgroup are anything other than aS2 > >>> manifestation of the difference in approach. > >>>34 > >>> Various posters have claimed on this newsgroup3 > >>> that OpenVMS pretty much clean as far as CERT 7 > >>> advisories are concerned only to find that Compaqt/ > >>> had not posted responses when OpenVMS was  > >>> vunerable. > >>>a7 > >>> Various posters have also claimed that OpenVMS is.6 > >>> immune to the stack overflow exploits attributed3 > >>> to UNIX and NT only to find that this is alsou > >>> not the case.  > >>K > >> As desperate as Andrew is for OpenVMS to turn out to be no more securetC > >> than Solaris, or as desperate as he appears to be to show some I > >> conspiracy of silence within VMS engineering I don't think the factst > >> bear him out. > >> > >n > >w" > > To answer your points in turn. > > 9 > > 1.    I am not and have not made any claims about theh> > >       relative security of OpenVMS vs any other OS Solaris > >       included.e > I > Let's see, you go out of your way to try to assert the VMS's reputation F > for security is either undeserved or overinflated and remain totally; > silent when pressed for details.  Sounds desperate to me.d >   8 Hang on it is the OpenVMS advocates that make the claim : that OpenVMS is better/more secure etc etc and frequently 8 they use the CERT advisories as proof of this. I am not 6 arguing that OpenVMS is less secure than UNIX but I am5 arguing that OpenVMS is considerably less secure thans most OpenVMS posters think.   ; If you want conclusive proof of this look at all the posts o; from OpenVMS advocates citing CERT advisories (or the lack e9 of them) as proof that OpenVMS is more secure than other d OS's.   ; These posters were clearly unaware of the gap between what ]= is in CERT and what OpenVMS is actually vunerable to, if they = had been why would they have opened the OpenVMS/CERT pandorasf box ?o     > >l> > > 2.    The Cert advisories prove beyond doubt that security> > >       information released by Compaq relating to OpenVMS's@ > >       vunerability to common exploits is either non existant< > >       when OpenVMS was vunerable, incorrect when OpenVMSB > >       was vunerable, or unhelpfull when OpenVMS was vunerable. > I > Cite specifics.  Oh, and remember, as much as you might wish otherwise,pD > UCX does *not* equal VMS.  During the timeframe to which we are inJ > reference there were no less than 5 IP stacks available for VMS.  If youF > are able to come up with any specifics, it/they will need to includeA > which stack had the vulnerability - it will NOT be VMS, itself.  >   G POD, Land/Teardrop, Smurf where do you want to stop. The Smurf responseuH is unhelpfull since it recommends hiding OpenVMS boxes behind a FirewallH and punishing violators inside the firewall, this is when other vendors ! have provided non firewall fixes.r  H The POD response was orgionally wrong, and the correct response (OpenVMSG was vunerable) and the fix arrived 4 months after the exploits started.rA OpenVMS is not listed as being vunerable to Land/Teardrop at all dD when in fact according to the Wizard it is and there is now a patch B for it. How much evidence do you need that the Compaq responses to" CERT advisories cannot be trusted.  ? Of course you may not have UCX but if you have an IP stack fromaA another vendor do you know that they are also not vunerable, none(A of the OpenVMS IP vendors post CERT responses but most of the DOSi attacks are generic ones.o   > >I< > >       This may be for any one of a number of reasons but= > >       which would you prefer ? Incompetence, arrogance orr@ > >       a conspiracy, how do you explain the fact that Compaqs: > >       responses for OpenVMS leave a lot to be desired. > F > They may leave a lot to be desired for you.  They reflect accurately& > what I have found in my own testing. >  > >. > > K > >> For instance, part of my job is to verify what attacks our systems mayeH > >> be vulnerable to - said systems including VMS, NT, and various UnixJ > >> flavors.  Not once, *not once*, have the VMS systems I've tested beenK > >> vulnerable to any of the DoS or exploit attacks that are documented on K > >> CERT (and I've tested pretty much all of them).  Solaris, on the othermM > >> hand, has been vulnerable to most of them (I don't have my list in frontC7 > >> of me but there were one or two that it survived).- > >> > >- > > J > > Sorry but this is simply wrong. If you had a current unpatched version > > ofH > > UCX at the time of each of the major DOS attacks then you would haveI > > been vunerable to most of them on OpenVMS from POD onwards, it is notcL > > even worth contesting this point because its documented either belatedly2 > > in CERT or in Compaqs own patch documentation. > >dC > > Of course if you had applied the latest patches you would be OK ? > > but you would have had no idea that you needed to apply thes; > > patches to protect yourself from DOS attacks by lookingt> > > in CERT or Auscert. You could have gleened the information? > > from the patch reports or you could have unwittingly filledt; > > the hole by keeping UCX to the current patch level. YouA9 > > would however in the case of POD for example have had>< > > a window of 4 months between the origional attacks being: > > reported and a patch being made available for UCX that > > fixed the problem. > A > Um, excuse me, we were talking about VMS, not UCX.  There is notI > relationship between the two other than one is a layered product of thee > other. >   ; But the majority of OpenVMS boxes do have an IP stack, your @ systems apparently fall into this category. UCX is also bundled = with OpenVMS so trying to hide behind the UCX isn't actually  ' part of OpenVMS argument is just spin. a  J > My statement is accurate not, as you assert, wrong.  I had responsiblityG > for literally hundreds of VMS systems running IP and everyone of themwH > tested immune and did so at the time the attack became known, not some > undefineable time later. >   F I am sorry but this simply cannot be tha case. If you were running theA current release of UCX at the time of the POD exploits your would A have been vunerable. The same for Land/Teardrop. This is because t? the patches to fix these exploits were released some time after ? the exploits themselves. Were you running UCX, did you actuallytB test Land/Teardrop/Smurf, POD is very easy to do from the command  line, the other three arn't.    C > > And Solaris has generally not been vunerable to the DOS attackse; > > like POD and in addition Sun has supplied configurationo= > > changes (not patches) to stop some of the performance DOS > > > attacks, not suggestions that you should hide your OpenVMS# > > boxes safely behind a FireWall.n > @ > I've never seen this suggestion.  Can you give me a reference?  = Ist in the Smurf advisory allong with the social engineering l3 suggestion (discipline people inside the FireWall).   E > > of OpenVMS is its invunerablility to stack overflow exploits whent! > > this isn't actaully the case.t > . > Back up your claim.  Show me one.  Just one. >  > >iD > > Could it be that as I have proved conclusively over the IP issue? > > that the rest of OpenVMS's reputation for security is basediE > > on assumptions that don't actually stand too much close scrutiny.s > H > You have proven squat.  You have asserted, but you have shown nothing. >   B So POD, Land/TearDrop/Smurf don't provide all the proof you need ?     Regards  Andrew Harrisonm Enterprise IT Architectt   ------------------------------   Date: 4 Apr 2001 10:14:03 -0500 - From: koehler@encompasserve.org (Bob Koehler)c% Subject: Re: Spice for OpenVMS Alpha?A3 Message-ID: <m2xaPdMp15a3@eisner.encompasserve.org>h  ] In article <3ACAFF52.7F5A96C1@uk.sun.com>, andrew harrison <andrew.nospam@uk.sun.com> writes:A > Mark Berryman wrote: >  >>  I had responsiblityeH >> for literally hundreds of VMS systems running IP and everyone of themI >> tested immune and did so at the time the attack became known, not somer >> undefineable time later.1 >> - > 1 > I am sorry but this simply cannot be tha case.      G The facts don't match the preconceived notions so the messenger must be/	 at fault?o  F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences Corporation = NASA GSFC Flight Software       | Federal Sector, Civil GroupRE                                 | please remove ".aspm" when replyingo   ------------------------------  " Date: Wed, 4 Apr 2001 10:52:37 GMT( From: Terry Kennedy <terry@gate.tmk.com> Subject: Re: Support for CD-Rr' Message-ID: <GB9Kvp.5E5@spcuna.spc.edu>h  " steven.reece@quintiles.com writes:@ > But there are two Ss in SCSI and neither of them are Standard.E > How can you drive a Standard when it's really a group of optionallye > implemented features?   K   Everyone else seems to manage pretty well. I have generic code for disks, J tapes, and "everything else" that provides a standardized API and talks to9 something like 2 dozen widely divergent controller types.s  F   But don't take my word for it - it is perfectly reasonable to have aI table-based capability/feature matrix for CD-R drives. Look at the devices support listed in:P http://download.gearsoftware.com/pub/updates/windows/recorders/310008_readme.txt     and then download:Y http://download.gearsoftware.com/pub/updates/windows/recorders/310008_recorder_update.exe   F   In which you'll find the aforementioned readme as well as the tablesE defining all the models. Hint: the readme is 29KB, the table is 28KB.s  4         Terry Kennedy             http://www.tmk.com5         terry@tmk.com             Jersey City, NJ USA.   ------------------------------    Date: 04 Apr 2001 04:51:57 +0800, From: Paul Repacholi <prep@prep.synonet.com>M Subject: Re: Support for CD-R (was: Re: Seeking CD-R/CD-RW SCSI INQUIRY data)S- Message-ID: <874rw5puz6.fsf@prep.synonet.com>a  " steven.reece@quintiles.com writes:  F > Reading between the lines, I guess one of the issues is that as soonE > as a target device has been qualified for use on VMS, it's gone and F > has been replaced by a later variant (at least two or three times in > some cases!!)c  @ > Would it make any sense at all for Compaq to release their ownC > version of a CD-R or CD-RW drive that could be qualified and thenp= > made available in sufficient numbers to satisfy the (as yeta= > unquantified) demand?  This would perhaps make it a similarr# > proposition to the RRD-4x drives.e  B No, IMO, the thing to do is set up the specs and tests to show youC conform, minimum docs, support etc, then label it 'VMS Able' Oh and ? send at least one to Hoff ;) The only way to go is to drive thec$ standard, not be driven under by it.   -- w< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.a@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  % Date: Wed, 04 Apr 2001 10:58:34 +01001  From: steven.reece@quintiles.comM Subject: Re: Support for CD-R (was: Re: Seeking CD-R/CD-RW SCSI INQUIRY data)tH Message-ID: <OF57C58C67.08F3FA1C-ON80256A24.00369C37@qedi.quintiles.com>  > But there are two Ss in SCSI and neither of them are Standard.C How can you drive a Standard when it's really a group of optionallyd implemented features?   (And yes, I have been watching). Steve.  @ Paul Repacholi (prep at prep dot synonet dot com) wrote/quoted : >>>o@ > Would it make any sense at all for Compaq to release their ownC > version of a CD-R or CD-RW drive that could be qualified and then = > made available in sufficient numbers to satisfy the (as yety= > unquantified) demand?  This would perhaps make it a similare# > proposition to the RRD-4x drives.o  B No, IMO, the thing to do is set up the specs and tests to show youC conform, minimum docs, support etc, then label it 'VMS Able' Oh anda? send at least one to Hoff ;) The only way to go is to drive theu' standard, not be driven under by it.<<<o   ------------------------------  # Date: Wed, 04 Apr 2001 08:34:10 GMTe1 From: CSABA  HARANGOZO   <csabah@zipworld.com.au>l= Subject: Re: TCPIP V5.0A install fails, can't proceed; ideas?09 Message-ID: <60By6.1144$CN.193691@nostril.pacific.net.au>c  + Lawrence Bleau <bleau@umtof.umd.edu> wrote:Q  D 	I was bitten by this as well. What solved the problem is an upgradeF 	to the latest PCSI from the CD set. It can be done without a shutdown- 	( you just have to log out and back again ).u   	I hope this helps...a 					Cheers,		Csabao  J > I ran into a problem installing (rather, upgrading to) TCPIP V5.0A.  I'mK > running OpenVMS AXP V7.1-2 with UCX V4.2 ECO 1.  I read the instyallationaN > notes, nothing seemed unusual, so I followed them.  The part that caused theA > problem was the PRODUCT INSTALL TCPIP command.  The output was:C   > -----a > Execution phase starting ...  9 > The following product will be installed to destination:eJ >     DEC AXPVMS TCPIP V5.0-11               DISK$DISKF0_SYS:[VMS$COMMON.]    > UCX product already installed.  I > ***********************************************************************BH > Another version of TCP/IP is installed. You must execute the following: > three commands before continuing with this installation:  5 >  $ BACKUP SYS$COMMON:[SYSMGR]UCX$LPD_STARTUP.COM; - 0 >      SYS$COMMON:[SYSMGR]TCPIP$LPD_STARTUP.COM;6 >  $ BACKUP SYS$COMMON:[SYSMGR]UCX$LPD_SHUTDOWN.COM; -1 >      SYS$COMMON:[SYSMGR]TCPIP$LPD_SHUTDOWN.COM;, >  $ PRODUCT REMOVE UCXsI > ***********************************************************************r     > Portion done: 0%$ > %PCSI-E-S_OPFAIL, operation failedO > %PCSIUI-E-ABORT, operation terminated due to an unrecoverable error conditionn > -----e  M > This looked strange, but okay, let's assume they (VMS and TCPIP folks) knowtL > what they're doing.  I entered the above commands.  The PRODUCT REMOVE UCX. > failed, however, with the following message:   > -----1 > $ PRODUCT REMOVE UCX  * > The following product has been selected:< >     DEC AXPVMS UCX V4.2-21                 Layered Product    > Do you want to continue? [YES]  [ > %PCSI-W-OPTREF, product DEC AXPVMS UCX V4.2-21 is referenced by DEC AXPVMS OPENVMS V7.1-2iN > -PCSI-W-OPTRF1, this software dependency is expressed within a configurationQ > -PCSI-W-OPTRF2,   option that is currently selected for the referencing product,N > -PCSI-W-OPTRF3, if you want to reconfigure the referencing product or reviewN > -PCSI-W-OPTRF4,   its configuration requirements, answer NO to the followingR > -PCSI-W-OPTRF5,   question to terminate this operation; if you are sure you wantP > -PCSI-W-OPTRF6,   to remove the referenced product, then answer YES to proceed  > Do you want to continue? [YES]  ; > The following products will be removed from destinations:uJ >     DEC AXPVMS UCX V4.2-21                 DISK$DISKF0_SYS:[VMS$COMMON.]J >     DEC AXPVMS UCXECO_42_1 V1.0            DISK$DISKF0_SYS:[VMS$COMMON.]   > Portion done: 0%7 > %PCSI-I-PRCOUTPUT, output from subprocess follows ...sO > %DCL-E-OPENIN, error opening PCSI$SOURCE:[000000]CLEAN_SDA_HELP.COM; as inputiM > -RMS-F-DEV, error in device name or inappropriate device type for operation  > Portion done: 10%l  F > %PCSI-E-EXERMVFAIL, product supplied EXECUTE REMOVE procedure failedM > -RMS-F-DEV, error in device name or inappropriate device type for operationL$ > %PCSI-E-OPFAILED, operation failedG > Terminating is strongly recommended.  Do you want to terminate? [YES] V > %PCSI-E-CANCEL_WIP, termination resulted in an incomplete modification to the system1 > %PCSI-E-S_OPCAN, operation cancelled by requestsO > %PCSIUI-E-ABORT, operation terminated due to an unrecoverable error conditionn > -----h  O > So, what gives here?  It won't let me install TCPIP V5.0A until I remove UCX,o= > and it won't let me remove UCX.  Any hints are appreciated.    > Lawrence Bleau > University of Maryland$ > Physics Dept., Space Physics Group > 301-405-6223 > bleau@umtof.umd.eduu    I    ---------------------------------------------------------------------- E    * Csaba I. Harangozo     |    'To err is human', said the hedgehognE    * csabah@zipworld.com.au |           as he dismounted a wirebrush.oI    ---------------------------------------------------------------------- ;    EARTH::AUSTRALIA:[SYDNEY]HARANGOZO.CSABA;1, delete? [N]:t   ------------------------------  # Date: Wed, 04 Apr 2001 13:32:06 GMTd8 From: hammond@not@peek.ppb.cpqcorp.net (Charlie Hammond)= Subject: Re: TCPIP V5.0A install fails, can't proceed; ideas?t2 Message-ID: <qnFy6.522$fB6.15030@news.cpqcorp.net>  * In article <9addb8$4io$1@hecate.umd.edu>, , bleau@umtof.umd.edu (Lawrence Bleau) writes: .. >$ PRODUCT REMOVE UCX  ..6 >%PCSI-I-PRCOUTPUT, output from subprocess follows ...N >%DCL-E-OPENIN, error opening PCSI$SOURCE:[000000]CLEAN_SDA_HELP.COM; as inputL >-RMS-F-DEV, error in device name or inappropriate device type for operation >Portion done: 10% >iE >%PCSI-E-EXERMVFAIL, product supplied EXECUTE REMOVE procedure failed L >-RMS-F-DEV, error in device name or inappropriate device type for operation# >%PCSI-E-OPFAILED, operation failedrF >Terminating is strongly recommended.  Do you want to terminate? [YES]U >%PCSI-E-CANCEL_WIP, termination resulted in an incomplete modification to the systemn0 >%PCSI-E-S_OPCAN, operation cancelled by requestN >%PCSIUI-E-ABORT, operation terminated due to an unrecoverable error condition ..  = Have you installed the VMS721_PCSI patch?  If not, please do.l   -- yK     Charlie Hammond -- Compaq Computer Corporation -- Pompano Beach  FL USA H        (hammond@not@peek.ppb.cpqcorp.net -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Wed, 04 Apr 2001 10:14:59 -0300e) From: fabio_compaq@ep-bc.petrobras.com.bre* Subject: TELNETSYM ( TCPIP Sevices x UCX )L Message-ID: <OF350139BD.B79E710F-ON03256A24.00486A22@ep-bc.petrobras.com.br>  G I am having the following message below in my UCX$TELNETSYM_printer.LOGe  O Wed Apr  4 09:14:59 2001, open_socket_ast invoked with bad IOSB 660: connect to  network object rejectedeO Wed Apr  4 09:24:59 2001, open_socket_ast invoked with bad IOSB 660: connect to  network object rejectedp     My UCX version is :a  ?  DIGITAL TCP/IP Services for OpenVMS Alpha Version V4.2 - ECO 1 :   on a AlphaServer 4X00 5/600 8MB running OpenVMS V7.1-1H2    R The problem is ... when I print from my OpenVMS 7.2-X servers using TCPIP ServicesS to the Lantronix printserver , the printing  goes OK, but  when I try to print from C the OpenVMS 7.1-1H2 server using UCX 4.2-ECO 1 the printing becomest stalled....h   Any suggestion ? Upgrade ?   Regardsu   FC   ------------------------------  % Date: Wed, 04 Apr 2001 15:05:23 +0100,  From: steven.reece@quintiles.com. Subject: Re: TELNETSYM ( TCPIP Sevices x UCX )H Message-ID: <OFDDAFE47B.F1152547-ON80256A24.004D0CEB@qedi.quintiles.com>   Fabio asked:J >>>I am having the following message below in my UCX$TELNETSYM_printer.LOG  D Wed Apr  4 09:14:59 2001, open_socket_ast invoked with bad IOSB 660:
 connect to network object rejectedaD Wed Apr  4 09:24:59 2001, open_socket_ast invoked with bad IOSB 660:
 connect to network object rejecteda     My UCX version is :i  ?  DIGITAL TCP/IP Services for OpenVMS Alpha Version V4.2 - ECO 1e:   on a AlphaServer 4X00 5/600 8MB running OpenVMS V7.1-1H2    I The problem is ... when I print from my OpenVMS 7.2-X servers using TCPIPP ServicesH to the Lantronix printserver , the printing  goes OK, but  when I try to
 print fromC the OpenVMS 7.1-1H2 server using UCX 4.2-ECO 1 the printing becomese stalled....    Any suggestion ? Upgrade ?<<<n  F Depends.  Is something else trying to print when the 7.1-1H2 system isH trying to print?  Does anything else have the port open on the Lantronix
 print server? A Does the output from the 7.2-1 system actually print?  Correctly?h> How are the queues configured?  Are they on the right port(s)?9 Is it any size of job or is it only above a certain size?    Steve.   ------------------------------  % Date: Wed, 04 Apr 2001 15:24:24 +0100 - From: Tim Llewellyn <tim.llewellyn@bbc.co.uk> . Subject: Re: TELNETSYM ( TCPIP Sevices x UCX )) Message-ID: <3ACB2E98.ECEA0391@bbc.co.uk>n  * fabio_compaq@ep-bc.petrobras.com.br wrote:  I > I am having the following message below in my UCX$TELNETSYM_printer.LOGo >hQ > Wed Apr  4 09:14:59 2001, open_socket_ast invoked with bad IOSB 660: connect to' > network object rejected Q > Wed Apr  4 09:24:59 2001, open_socket_ast invoked with bad IOSB 660: connect to  > network object rejectedf >u > My UCX version is :d >rA >  DIGITAL TCP/IP Services for OpenVMS Alpha Version V4.2 - ECO 1t< >   on a AlphaServer 4X00 5/600 8MB running OpenVMS V7.1-1H2 > T > The problem is ... when I print from my OpenVMS 7.2-X servers using TCPIP ServicesU > to the Lantronix printserver , the printing  goes OK, but  when I try to print fromcE > the OpenVMS 7.1-1H2 server using UCX 4.2-ECO 1 the printing becomeso
 > stalled....d >  > Any suggestion ? Upgrade ? >s  / Can you ping the printer from the problem node?b  O Does the Lantronix have a mechanism for rejecting jobs depending on IP address?m     --6 Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project0 MedAS at the BBC, Whiteladies Road, Bristol, UK.A Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.uk   A I speak for myself only and my views in no way represent those oft MedAS or the BBC.g   ------------------------------  # Date: Wed, 04 Apr 2001 10:43:18 GMT,= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)h" Subject: Re: VAX Debugger command.0 Message-ID: <009FA05A.FF5F1FCD@SendSpamHere.ORG>  ] In article <nauy6.471069$w35.72679943@news1.rdc1.nj.home.com>, "Fatz" <fatz@fatz.com> writes:yM >I can't remember this.  It's a command you do in debug to examine the entireu' >argument list and goes something like:s >. >DBG> EX @(AP:@AP)AP@AP)   EXAMINE .AP:.AP+..AP*4     --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMy             O city, n., 1. a place where trees are cut down and streets are named after them.i   ------------------------------  % Date: Wed, 04 Apr 2001 09:15:52 +0100 % From: Alan Greig <a.greig@virgin.net>f Subject: Re: VAX in 1U8 Message-ID: <80mlctkdkbcj14c1sciloe2209dbmn1q3r@4ax.com>  A On 4 Apr 2001 00:23:46 +0200, maulis@ludens.elte.hu (Adam Maulis)  wrote:  U >In article <Drqy6.181583$Z2.2166666@nnrp1.uunet.ca>, <chris.moore@stelco.ca> writes: O >> An interesting alternative to any of this is the CHARON-VAX emulator runninge >> on a Wintel box.  >n; >or intel like machie with linux... may not full supported,e# >may not stable but an alternative.s  C And the beta quality port of the emulator to run under Alpha/VMS is # now available for download I think.g   >o >t	 >Regards,u >Adam Maulis   -- Alan   ------------------------------    Date: 04 Apr 2001 04:47:58 +0800, From: Paul Repacholi <prep@prep.synonet.com>$ Subject: Re: VMS-Related: Affordable- Message-ID: <878zlhpv5t.fsf@prep.synonet.com>n  " steven.reece@quintiles.com writes:  F > I wonder how the lifetime for VMS compares with the lifetime for theB > Alpha architecture in the light of the DII COE work that's beingF > done....?  Presumably X years (minimum) for VMS would also equate to9 > X years (minimum) for Alpha in the present environment.e  @ And that, acording to the first gospel of Alpha, is after it hasF turned in its chips. The vax ran from 77-00. 23 years. Alpha, 92-17 is1 25. And COE calls for a '20 eol at the earliest. D  < As to Davids enthiusiasm for big hammers, look at the memory managment! yuck...   -- e< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.B@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  # Date: Wed, 04 Apr 2001 13:01:30 GMTs4 From: "Terry C. Shannon" <terryshannon@mediaone.net>$ Subject: Re: VMS-Related: Affordable< Message-ID: <KWEy6.32859$Wz.9165720@typhoon.ne.mediaone.net>  9 "Paul Repacholi" <prep@prep.synonet.com> wrote in messages' news:878zlhpv5t.fsf@prep.synonet.com... $ > steven.reece@quintiles.com writes: >eH > > I wonder how the lifetime for VMS compares with the lifetime for theD > > Alpha architecture in the light of the DII COE work that's beingH > > done....?  Presumably X years (minimum) for VMS would also equate to; > > X years (minimum) for Alpha in the present environment.e >aB > And that, acording to the first gospel of Alpha, is after it hasH > turned in its chips. The vax ran from 77-00. 23 years. Alpha, 92-17 is2 > 25. And COE calls for a '20 eol at the earliest. >t  G The gospel, like most religious tracts, is subject to interpretation bysG various theologians and processor priests n' priestesses. I believe the @ Alpha lifespan will be extended due to the slippage in processorI generations. If you go back to the Alpha rollout and look at the original L processor roadmap, we'd be at EV8 right now. Since we are not, it would seem7 likely that Alpha has gained a de facto life extension.P   ------------------------------   Date: 4 Apr 2001 15:04:37 GMTt2 From: mathog@seqaxp.bio.caltech.edu (David Mathog)$ Subject: Re: VMS-Related: Affordable, Message-ID: <9afd65$a8l@gap.cco.caltech.edu>  X In article <3ACB9D13.692BA33B@infopuls.com>, Christof Brass <brass@infopuls.com> writes:9 >If VMS is behind wrt I/O speed it will catch up. I wouldpA >recommend not to through the baby out with the bath because thise< >aspect is temporarily (yes, I know, for some years now, but2 >better performance to come soon) disadvantagous.   B Ever read "Waiting for Godot"?  I too expect VMS IO performance toJ improve, and in fact, XFC in 7.3 helps a lot with reads (but does nothing G for writes).  Unfortunately even XFC does nothing at all to improve the4K 2-3X factor buried in RMS or the C RTL, and I've heard nothing credible to o- suggest that that factor will soon disappear.N  < >VMS app should stay with VMS and take full advantage of the >superior design.l  E You mean, those few remaining VMS apps?  That's the other reason for i: leaving VMS - there's practically no native software left.  . >But of course there are areas where VMS could9 >not completely excel because its advantages are not thats >important.o  C Yes and no.  The VMS security model is much, much, much better than J Unix's.  I'd be a happy camper indeed if I could get Unix performance withJ VMS security.  DII/COE?  Who knows, but I'm old enough to know better than to bet the farm on vapor.   1 >We might state an analogy of VMS and UNIX (whichM> >should be avoided) to Rolls-Royce and a formula 1 car: nobody: >would like to drive it for a long time but for short time! >exploits it may be appropriate. )  I VMS is a Rolls only if you buy a high end system. On low end systems it's F more like buying a Mercedes without a gas tank. What you've got is putH together well, but you can't do much with it and it costs you more than * cars which will get you to work every day.  H >> And the only reason this isn't a Linux system running on high qualityN >> Athlon hardware (which would have cost $500 less and been twice as fast) isJ >> that we have 1 piece of licensed software that requires Solaris or Irix0 >> (and I'm not stepping into the SGI quagmire.) >0? >My experience is that PC HW isn't that reliable and versatile.i  D My experience is that if you carefully select components PCs are as H reliable as low end Alphas.  That assumes you aren't running Windows on F them though - windows has historically been reliably unreliable.  (The) jury's still out on W2K in this regard.) s   >aN >> Which all goes to show that technical credentials, both for an OS (VMS) andJ >> for the hardware (Alpha) count for squat if the company controlling the@ >> technology does not market it well.  Or in this case, at all. >i@ >Or doing the opposite: actively damaging the business image and >future prospects. :-(  K Yup.  Another reason to shun the company.  Who wants to deal with guys thatS# keep wrecking their own business?  H   >)M >> I don't have Terry Shannon's credentials, but my "fearful prognostication" J >> is that the eventual winner of the 64 bit wars will be AMD.  Unlike theO >> itanic, the "Hammer" will actually run 32 bit x86 code at a reasonable speedl >> AND offer 64 bits as well.  >s1 >I always thought that tItanic is x86 compatible.u  A The x86 mode/emulation currently runs so slow as to be unusuable.y  = >I read that in some Intel, AMD and other compatibles (Cyrix,e> >ITCWinChip (?) whatever) in fact RISC kernels are working and> >with some even the RISC instruction set can directly be used.  I I believe that part may be true for the Hammer, it's upwardly compatible aI with the x86, but may have extra 64bit instructions and other extensions.1  D >Is there any problem to provide a low level x86 emulation on Alpha?  J Yes.  It was called FX!32 on WNT/Alpha and it never ran very well.  It wasF an amazing piece of technology, but in the end what it gave you was anJ Alpha that cost 3X as much running at 1/3 the speed.   Platforms live and K die by native applications and hardware/software purchase costs.  (NOT TCO,  duh.)   A >Some HyperVMWare which allows to run VMS and e.g. something like*? >SoftWindows, RealPC or WINE which is then capable of executing  >native Intel apps?o  C Most of the Linux Windows compatibility projects go directly to x86$H code and execute it on the processor.  They are Intel specific and won't work on the Alpha. .   > = >The 68k -> PPC move was similar to the VAX -> Alpha move butS3 >less VAX apps moved to Alpha than 68k apps to PPC.U  K Because the market for the new Macs was larger than the market for the new *L Alphas.  Because people knew that the Mac existed but were forgetting about  VMS at the time.   >What about an6 >Intel -> Alpha move supported by emulation technique?  L I don't think so.  A complete cross compilation suite might do it (treating I the x86 binary as a source code), but I don't think that there's anything  around like that.   > >What about a low level x86 emulation to directly run x86 OSs?  E That was FX!32.  Forget about it.  It wasn't enough to keep WNT/Alpha / afloat and it won't do anything for VMS either.o  @ >What about a low level SPARC emulation to directly run Solaris?  I There's one for Tru64 called FreePort Express.  I don't know how well it   works.  You need to visit:  <   http://www.support.compaq.com/amt/tools/migrate-cover.html  @ >I personally think this is the way to go for Alpha. The ballast< >of the compatibility instructions should be put into the SW@ >layer instead like Intel keeping all the old instructions along* >for decades burning 50% of power by that.  A Actually Intel apps are a pain when emulated (by any technology, mG including softwindows).  They tend to use timing loops and when you runc& them on an Alpha they eat CPU cycles.    Regards,   David Mathog mathog@seqaxp.bio.caltech.eduX? Manager, sequence analysis facility, biology division, Caltech     ------------------------------   Date: 4 Apr 2001 09:06:33 -0700 1 From: nothome@spammers.are.scum (Malcolm Dunnett)s$ Subject: Re: VMS-Related: Affordable, Message-ID: <r9j1P9fFFy6K@malvm1.mala.bc.ca>  = In article <KWEy6.32859$Wz.9165720@typhoon.ne.mediaone.net>, r:     "Terry C. Shannon" <terryshannon@mediaone.net> writes: > I > The gospel, like most religious tracts, is subject to interpretation byFI > various theologians and processor priests n' priestesses. I believe thePB > Alpha lifespan will be extended due to the slippage in processorK > generations. If you go back to the Alpha rollout and look at the originaluN > processor roadmap, we'd be at EV8 right now. Since we are not, it would seem9 > likely that Alpha has gained a de facto life extension.e > D    So you're saying that as long as they never get around to rolling> out the EV8 ( or EV7 even ) the Alpha could live forever?  :-)  G    Course one could read the gospels to say that the failure to delivereE these follow-ons on time means the Alpha's lifespan will be seriously F shortened as competing processors eliminate any performance advantagesC the Alpha might have. A quick rule-of-thumb calculation from postedz SPECint 2000 benchmarks at:   = (http://www.compaq.com/alphaserver/performance/spec2000.html)   A   suggests that a Pentium needs about a 50% greater clock rate to-A match alpha performance, which suggests there's really no lead inZ> currently shipping products for integer performance. Of courseA floating performance is a different story, but how many loads areo! that dependant on FP performance?R   ------------------------------    Date: 05 Apr 2001 01:00:55 +0800, From: Paul Repacholi <prep@prep.synonet.com>$ Subject: Re: VMS-Related: Affordable- Message-ID: <87vgok38hk.fsf@prep.synonet.com>   4 mathog@seqaxp.bio.caltech.edu (David Mathog) writes:  = > In article <3ACB9D13.692BA33B@infopuls.com>, Christof BrassU > <brass@infopuls.com> writes:  E > >If VMS is behind wrt I/O speed it will catch up. I would recommendfC > >not to through the baby out with the bath because this aspect isp; > >temporarily (yes, I know, for some years now, but better , > >performance to come soon) disadvantagous.  : The poor performance of C IO has been there since day one.  tD > Ever read "Waiting for Godot"?  I too expect VMS IO performance toC > improve, and in fact, XFC in 7.3 helps a lot with reads (but doesnE > nothing for writes).  Unfortunately even XFC does nothing at all topD > improve the 2-3X factor buried in RMS or the C RTL, and I've heardC > nothing credible to suggest that that factor will soon disappear.   h> > >VMS app should stay with VMS and take full advantage of the > >superior design.n  oF > You mean, those few remaining VMS apps?  That's the other reason for< > leaving VMS - there's practically no native software left.  F And what little there is is dying off a a huge rate. Even sadder, much: code that is no longer available was born and bred on VMS.  D The nursery is now Sun/M$ owned and mind locked. So people end up inD the situation David is in. Rewrite and re-engineer a huge code base,@ suffer abismal performance, or get with a second rate compatible6 system.  Using VMS with that software is third rate...   -- r< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.j@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  $ Date: Wed, 4 Apr 2001 14:21:51 +0100= From: Nicholas Mark de Smith <Nicholas.MarkdeSmith@liffe.com>oC Subject: What to do with old CDROMS... (do NOT try this at home...)oM Message-ID: <78E5B8E274DBD1118D6800805FE60E7775B203@ntprdex4.admin.liffe.com>e   I just couldn't resist this...  E I am throwing away most of my DEC stuff, accumulated over 20 years or G so. Amongst the detritus (5 3100s, 4000/76, several AXPs etc...) are 10c years worth of CONDIST.S  D Many discussions over the years have suggested uses for the probably1 1000s of CDs, but I have now found the answer(s):e  C http://www.hamjudo.com/notes/cdrom.html (good results with DEC CDs)0  = http://www.netcomuk.co.uk/~wwl/cdzap.html (a complete nutter)   = http://margo.student.utwente.nl/el/microwave/ (not just CDs!)e  C http://www.amasci.com/weird/microexp.html (very unwise experiments)   H http://www.powerlabs.org/uwavexp.htm (never, not in a million years am I going to do this)   H http://www.physics.ohio-state.edu/~maarten/microwave/microwave.html (has movies!)  B http://www.sci.tamucc.edu/~pmichaud/grape/ (not CDs, but "fun with grapes"!!!)t  > http://members.tripod.com/~hochwald/microwave/ (has section on. "reasonably safe experiments"! - includes CDs)  F Thanks enough ideas for now. I accept ABSOLUTELY NO RESPONSIBILITY for- any consequence of the use of this posting...e   I'm off to play...   Nick   ------------------------------  % Date: Wed, 04 Apr 2001 16:07:01 +0100r- From: Tim Llewellyn <tim.llewellyn@bbc.co.uk>l2 Subject: Re: When might SYS$SCHDWK not wake me up?) Message-ID: <3ACB3894.FD32C7EB@bbc.co.uk>n   David Jones wrote:  M > I Suppose this 'trivial processing' did a hiber()/wake() sequence somewhere L > inside a library call.  If the timing was such that on rare occasions bothE > the $SCHDWK and latent $WAKE/$SCHDWK fired before the latent $HIBER9K > was called, then the $HIBER in the main loop would miss the intended wakemL > and block forever.  (It would also probably mean that the internal processJ > that depended upon the wake would from time to time resume prematurely).  ) Agreed, it is a race condition minefield.0  C I never used $HIBER and $WAKE much except in programming tutorials. O Isn't it a bit like using event flag zero, its OK if none of the other code youc call does any synchronization?   P Personally I'd use event flags and/or AST's or threads, but then I do appreciate  * the concern for the high value peripheral.     > --  6 Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project0 MedAS at the BBC, Whiteladies Road, Bristol, UK.A Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.uko  A I speak for myself only and my views in no way represent those of  MedAS or the BBC.o   ------------------------------  # Date: Wed, 04 Apr 2001 15:59:19 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)I" Subject: ZXLp-L2 bugcheck analysis0 Message-ID: <009FA087.24FB95FF@SendSpamHere.ORG>   OK,l  0 Here's what I've found in my ZXLp-L2 crash dump.  G The system crashes with a DECWINDOWS bugcheck which corresponds to lineeH 14191 in the V7-2.1 listing of DUBDRIVER.  The reason it crashes is that  IOC$MAP_IO returns SS$_BADPARAM.  E I computed the address of UCB$L_GUB_BASE_ADDRESS(R5) and examined its D contents and it has: 07000000(16).  The driver called the IOC$MAP_IO? requesting:  #<1024*1024*16*2>   ; Map 16Mbytes in sparse spacet< (This looks like 32 Meg to me)  and #IOC$K_BUS_MEM_BYTE_GRAN  C The MAP_IO routine in IO_SUPPORT_1605 checks the following for thisn
 attribute:  2 3   40220         case IOC$K_BUS_MEM_BYTE_GRAN : {	 3   40221mC 3   40222             /* Sparse Memory Space is from 0 to 128 MB */o	 3   40223 K 3   40224             if ((bus_address + num_bytes) > ONE_TWENTY_EIGHT_MB ) 2 3   40225                     return SS$_BADPARAM;   So, let me see...0  E 1024*1024*16*2 is 02000000(16)  plus 07000000(16) yields 09000000(16)i? or 01000000(16) more than ONE_TWENTY_EIGHT_MB 08000000(16).  :(    The question now is WHY???   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMr             O city, n., 1. a place where trees are cut down and streets are named after them.    ------------------------------  % Date: Wed, 04 Apr 2001 19:42:29 +0200m, From: Didier Morandi <Didier.Morandi@gmx.ch>1 Subject: [Q] DCL minute of the day: semi-graphicsa& Message-ID: <3ACB5D05.42D57DF4@gmx.ch>  P I don't remember the escape characters to toggle semi-graphics without having toD embed an <ESC>(0 <ESC>(B pair every x character in a column drawing.  . I don't have any terminal book under the hand.   Thanks,o   D.   ------------------------------   End of INFO-VAX 2001.189 ************************