1 INFO-VAX	Fri, 17 Oct 2003	Volume 2003 : Issue 576       Contents:, Re: 3-6 month Northern NJ VMS/C upgrade work, Re: 3-6 month Northern NJ VMS/C upgrade work Re: 8 way IBM power5 picture!  Re: 8 way IBM power5 picture!  Re: Clustering over GB Switches  Re: Clustering over GB Switches  Re: Clustering over GB Switches  Re: Clustering over GB Switches  Re: Clustering over GB Switches $ Re: Cost of process creation on Unix$ Re: Cost of process creation on Unix$ Re: Cost of process creation on Unix$ Re: Cost of process creation on Unix- Re: DECUS: "C" stands for Computer, not Corp. - Re: DECUS: "C" stands for Computer, not Corp. - Re: DECUS: "C" stands for Computer, not Corp.  Error installing JRE V1.4.1-2 < Good Bye Polycenter Suite, Hello Cockpit Manager for OpenVMS@ Re: Good Bye Polycenter Suite, Hello Cockpit Manager for OpenVMS@ RE: Good Bye Polycenter Suite, Hello Cockpit Manager for OpenVMS? Re: Itanium/Integrity & EV7/Marvel leadership benchmark results  LGI Callout & SSH  Re: LGI Callout & SSH  Re: LGI Callout & SSH  Mail text truncated  RE: Mail text truncated  Re: Mail text truncated C Re: Morgan Stanley analyst releases positive research report on HPQ * Re: New OpenVMS system with no root passwd2 NYMLUG NYC Encompass meeting "VMS Update" 10/29/03 OpenVms 7.3.1 ODBC Driver 2 Re: OpenVMS I64 ISV application count now over 500! Re: OpenVMS in an HP advertisment  Re: PL/1 users, where are you? Re: PL/1 users, where are you?! Problem with UCX QIOW IO$_SETMODE % Re: Problem with UCX QIOW IO$_SETMODE % Re: Problem with UCX QIOW IO$_SETMODE K Re: Q: Product Install of "older" kits warns of deleting "UNDO" directories  rwast ) Re: Shared distribution of hobbyist media ) Re: Shared distribution of hobbyist media ) Re: Shared distribution of hobbyist media  Size of a global page  Re: Size of a global page  Re: Size of a global page ) SMTP server VRFY command vs. mail aliases  VMS 7.3-1 patches. Re: VMS 7.3-1 patches. Re: VMS 7.3-1 patches. Re: VMS defeated by wind Re: VMS defeated by wind Re: VMS defeated by wind Re: VMS defeated by wind Re: VMS defeated by wind RE: VMS defeated by wind RE: VMS defeated by wind Re: VMS defeated by wind Re: VMS defeated by wind Re: VMS defeated by wind Re: VMS defeated by wind VMSTAR problem Re: W2K patches  Re: W2K patches  Re: W2K patches  Re: W2K patches  Re: W2K patches  Re: W2K patches  Re: W2K patches  Re: W2K patches  Re: W2K patches  Re: W2K patches  Re: W2K patches  Re: W2K patches  Re: W2K patches  Re: What is ACME_SERVER? Re: What is ACME_SERVER?% Re: What is the performance of iVMS ? % Re: What is the performance of iVMS ? 8 Re: White paper: CHARON-VAX cluster w/ shared SCSI disks8 Re: White paper: CHARON-VAX cluster w/ shared SCSI disks8 Re: White paper: CHARON-VAX cluster w/ shared SCSI disks8 Re: White paper: CHARON-VAX cluster w/ shared SCSI disks! Writing flat text file from DEC C % Re: Writing flat text file from DEC C % Re: Writing flat text file from DEC C % Re: Writing flat text file from DEC C % Re: Writing flat text file from DEC C % Re: Writing flat text file from DEC C > Re: [Plug] using x-windows to display vms decterms on Mac OS X  F ----------------------------------------------------------------------  % Date: Fri, 17 Oct 2003 22:47:07 +0800 , From: Paul Repacholi <prep@prep.synonet.com>5 Subject: Re: 3-6 month Northern NJ VMS/C upgrade work - Message-ID: <87y8vjhqfo.fsf@prep.synonet.com>   + Steven Bitgood <steve@goodbits.com> writes:   A > Apparently some place in NJ has been sitting at VMS 5.5-2 for a 6 > while, and is working on a plan to upgrade to 7.3 orB > thereabouts. Key words that I heard were "DEC C," "distributed,"# > "multi-threaded" and "real-time."   C I suspect the words `device driver' and porting to Alpha would also  feel in good company.   1 > If you're interested, tell Kenneth Quartarone ( 3 > kenq@designstrategy.com or (888) 425-2404 x302 ).   D > Note that, while I'm not associated with Kenneth or his company, IC > might get some referral $ if you tell him that I sent you and you  > get the work.    > Hope this helps someone.  < Nice to know there is some one not moving to the darkside...   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Fri, 17 Oct 2003 17:04:19 GMT " From:   VAXman-  @SendSpamHere.ORG5 Subject: Re: 3-6 month Northern NJ VMS/C upgrade work 0 Message-ID: <00A27837.ABCF9165@SendSpamHere.ORG>  \ In article <87y8vjhqfo.fsf@prep.synonet.com>, Paul Repacholi <prep@prep.synonet.com> writes:, >Steven Bitgood <steve@goodbits.com> writes: > B >> Apparently some place in NJ has been sitting at VMS 5.5-2 for a7 >> while, and is working on a plan to upgrade to 7.3 or C >> thereabouts. Key words that I heard were "DEC C," "distributed," $ >> "multi-threaded" and "real-time." > D >I suspect the words `device driver' and porting to Alpha would also >feel in good company.  C Perhaps you know something that the rest of use are not privied to?    --  L VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)COM             5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------  % Date: Fri, 17 Oct 2003 14:50:28 +0200 , From: "Rob Turk" <_wipe_me_r.turk@chello.nl>& Subject: Re: 8 way IBM power5 picture!9 Message-ID: <3f8fe595$0$667$c4fe714e@dreader11.xs4all.nl>   5 "Bob Ceculski" <bob@instantwhip.com> wrote in message 7 news:d7791aa1.0310161159.5ea96d5f@posting.google.com... ! > to bad it can't run OpenVMS ...  >  > + > http://www.theinquirer.net/?article=12145   I Doesn't look like it's 8-way. It looks like 4 CPU cores in the middle and " maybe 4 cache chips on the outside   Rob    ------------------------------  % Date: Fri, 17 Oct 2003 12:59:10 -0400 * From: "Bill Todd" <billtodd@metrocast.net>& Subject: Re: 8 way IBM power5 picture!2 Message-ID: <X5ydneoa7-9Hgg2iU-KYuA@metrocast.net>  7 "Rob Turk" <_wipe_me_r.turk@chello.nl> wrote in message 3 news:3f8fe595$0$667$c4fe714e@dreader11.xs4all.nl... 7 > "Bob Ceculski" <bob@instantwhip.com> wrote in message 9 > news:d7791aa1.0310161159.5ea96d5f@posting.google.com... # > > to bad it can't run OpenVMS ...  > >  > > - > > http://www.theinquirer.net/?article=12145  > K > Doesn't look like it's 8-way. It looks like 4 CPU cores in the middle and $ > maybe 4 cache chips on the outside  5 Er, that would be 4 *dual*-CPU chips in the middle...    - bill   ------------------------------  % Date: Fri, 17 Oct 2003 09:31:46 +0100 * From: Nic Clews <sendspamhere@[127.0.0.1]>( Subject: Re: Clustering over GB Switches' Message-ID: <bmo9b7$n5q$1@lore.csc.com>    Dave Baxter wrote: > D > I currently have a 2-node cluster of ES40's running OpenVMS 7.3-1.B > They are directly connected to each other via redundant GB FibreH > adapters (EW_DEGPA)using two Fibre crossover cables.     This was doneE > for expediency a few months ago, and was fine as long as I only had  > two nodes. > G >      I now need to add a third node!   To facilitate this I purchased D > two Cisco Catalyst 3508G-XL Gigabit switches and all the necessary
 > GBIC's etc.  > D >      I now need to move my existing cluster members off the Direct# > connection and onto the Switches.  > G >      I have not configured the new switches in anyway, (figuring that F > the DEFAULT settings would probably work).   They didn't!!    When IC > brought up the second node it attempted to create its own cluster D > (which caused me to do an immediate "HALT" on the system to try toD > forstall any partitioning issues).     This clearly indicated thatB > NODE_2 could not communicate with the "up" node, NODE_1, via the	 > switch.  > F >      The new switches are dedicated to the task of SCS comms and areF > not network-attached.   For this reason I have not configured any IP > info on the switches.  > H >      I am really a novice with network switches etc., so if anyone canE > give me some guidelines on how the GB switch and/or ports should be + > configured, I would really appreciate it.   9 Sorry for quoting all the text, but I think its relevant.   E Introducing any hardware between systems makes that a single point of G failure. The crossover cable is in a two node situation the best way to  go.   E Now, unless you have slot availability constraints in your servers, I E would have advised buying an extra gigabit card each for the existing E servers, and two for your new server, and more crossover cables. Each : system then can have its own dedicated link to the other.#  B Can you get gigabit hubs? That would be what you want really*. TheH switch is probably going to try to be too clever, and (personal hearsay)A CISCO gear has never had the best reputation where Digital/Compaq  protocols are involved.   B * When I say what you want, it is still a single point of failure.  H The point about a switch, is each cable is its own LAN, and its probablyG overkill in this situation, where you're expecting a mere three devices B on a protocol which my its description is "Carrier Sense, MultipleE Access, Collision Detect". Just at what point is the switch improving B data transfer between two points, when so few points are involved?  G Put this into English: Packet to be sent, "Carrier sense", "send packet H (or delay)". With a switch it is "carrier sense", "send packet", "switchD relays the packet somewhere", does it's own "carrier sense", finally "send packet".  G # From your description, there is a hint you've got the right number of H GBIT cards, and you bought a pair of switches. In this case DO NOT cableA the switches together because VMS will "think" there are multiple F redundant paths with the switches providing the illusion. The "CLUSTERC CONFIG" manual is quite clear, each member of a cluster must have a H direct connection path (i.e. not via another system). In this case, takeH out the switches, and use crossover cables A to B and C, B to A and C, C: to A and B, like three people in a ring all holding hands.  C Footnote: If you also need outside world connectivity, assuming the G reason is we have a duff port configuration for things not happening as D expected, there is nothing wrong taking *one* set of the gigabit lanA connections and putting them to a switch, and hopefully, SCS will C realize there is a "second path" should the direct system to system H crossovers get yanked out. You may also have a third network connection.   Hope this helps, --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot com    ------------------------------  % Date: Fri, 17 Oct 2003 10:58:05 +0100 * From: "Richard Brodie" <R.Brodie@rl.ac.uk>( Subject: Re: Clustering over GB Switches+ Message-ID: <bmoefe$j2a@newton.cc.rl.ac.uk>   Y "Nic Clews" <sendspamhere@[127.0.0.1]> wrote in message news:bmo9b7$n5q$1@lore.csc.com...     M > In this case, take out the switches, and use crossover cables A to B and C, L > B to A and C, C to A and B, like three people in a ring all holding hands.  K That seems less robust to me; you have eliminated the switches as points of P failure but now every NIC and cable is a single point of failure, unless you putC four NICs in each machine. I don't see anything wrong with a simple  'A-bus/B-bus' arrangement.  4 > "Carrier Sense, Multiple Access, Collision Detect"  R Is pretty much irrelevant for Gigabit Ethernet. It was largely there for politicalL reasons, made little sense technically, and hardly anyone uses it. There wasJ a feeling that "it wasn't Ethernet if it didn't support CSMA/CD". When the4 10GbE standard was published, it was quietly buried.   ------------------------------  % Date: Fri, 17 Oct 2003 11:57:25 +0100 * From: Nic Clews <sendspamhere@[127.0.0.1]>( Subject: Re: Clustering over GB Switches' Message-ID: <bmohsb$qcb$1@lore.csc.com>    Richard Brodie wrote:  > [ > "Nic Clews" <sendspamhere@[127.0.0.1]> wrote in message news:bmo9b7$n5q$1@lore.csc.com...  > O > > In this case, take out the switches, and use crossover cables A to B and C, N > > B to A and C, C to A and B, like three people in a ring all holding hands. > M > That seems less robust to me; you have eliminated the switches as points of R > failure but now every NIC and cable is a single point of failure, unless you putE > four NICs in each machine. I don't see anything wrong with a simple  > 'A-bus/B-bus' arrangement.  G Yes you're right actually, the SPOF is the card itself. Or the cable. I  stand corrected.   6 > > "Carrier Sense, Multiple Access, Collision Detect" > T > Is pretty much irrelevant for Gigabit Ethernet. It was largely there for politicalN > reasons, made little sense technically, and hardly anyone uses it. There wasL > a feeling that "it wasn't Ethernet if it didn't support CSMA/CD". When the6 > 10GbE standard was published, it was quietly buried.  / So what is the principle, "send and be damned"?   H I'm willing to be educated on gigabit, from what you're saying it varies@ from the standards laid down in February 1980 (speeds excepted).   --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot com    ------------------------------  % Date: Fri, 17 Oct 2003 13:32:27 +0100 * From: "Richard Brodie" <R.Brodie@rl.ac.uk>( Subject: Re: Clustering over GB Switches+ Message-ID: <bmongs$s7g@newton.cc.rl.ac.uk>   Y "Nic Clews" <sendspamhere@[127.0.0.1]> wrote in message news:bmohsb$qcb$1@lore.csc.com...   1 > So what is the principle, "send and be damned"?  > J > I'm willing to be educated on gigabit, from what you're saying it variesB > from the standards laid down in February 1980 (speeds excepted).  A To do CSMA/CD arbitration at higher speeds you need to shrink the ? network: from ~2km for 10Mbit to ~200m for 100Mbit. For Gigabit J Ethernet,  rather than shrinking again, they invented 'Carrier Extension',G where each frame is padded out to a minimum of 512 bytes instead of 64. = That gets you the diameter back, at some cost in utilisation.   D However, for a point-to-point connection between two systems, or viaA a switch, you can run in full duplex mode. Then, yes, you disable D CSMA/CD and send whenever you like, receive whatever you can buffer.J Gigabit is almost always used in full-duplex mode; and 10 Gigabit will not support half-duplex at all.    ------------------------------    Date: 17 Oct 2003 08:20:44 -07000 From: dave.baxter@bannerhealth.com (Dave Baxter)( Subject: Re: Clustering over GB Switches= Message-ID: <a3c44af1.0310170720.75df9cd3@posting.google.com>   ] "Richard Brodie" <R.Brodie@rl.ac.uk> wrote in message news:<bmoefe$j2a@newton.cc.rl.ac.uk>... [ > "Nic Clews" <sendspamhere@[127.0.0.1]> wrote in message news:bmo9b7$n5q$1@lore.csc.com...  >  > O > > In this case, take out the switches, and use crossover cables A to B and C, N > > B to A and C, C to A and B, like three people in a ring all holding hands. > M > That seems less robust to me; you have eliminated the switches as points of R > failure but now every NIC and cable is a single point of failure, unless you put > four NICs in each machine.    F This would be my thought too.    As far as the switches are concerned,? I don't see them as an SPOF since, they will be used to provide E redundant paths, i.e. each node has two (2) GB Adapters and therefore A has a connection to each of the switches.    The switches are NOT B going to be connected to each other.   As I see it, the failure of< either switch still leaves the path through the other switch# available.   (am I wrong here ????)   B      In principle, all I am trying to do is cluster over a private segment, so why won't it work.  D      Are there some basic port configuration commands that I need toB execute to allow the SCS type of communication ??    Or, Are there+ "Default: features that I should disable ??    Dave.    ------------------------------  # Date: Fri, 17 Oct 2003 12:19:44 GMT 3 From: wallacethinmintr@eircom.net (Russell Wallace) - Subject: Re: Cost of process creation on Unix / Message-ID: <3f8fddc2.66509348@news.eircom.net>   , On Thu, 16 Oct 2003 15:02:43 -0400, JF Mezei" <jfmezei.spamnot@istop.com> wrote:  M >Am about to start an argument with the owner of my ISP service. He is of the G >opinion that breaking the 10 minute minimum timeout for POP servers is N >beneficial (he set his to 15 seconds) because it means that processes go awayN >really fast, freeing resources, reducing paging/swapping, memory use etc etc.  F Doesn't make sense to me. Let's suppose the server doesn't share pagesF between processes (I don't know whether it does or not). Still, havingD an idle process around does no great harm, even if memory's in shortA supply - they'll get swapped out once and then just left on disk.   C Forcing users to sent a stream of pings to their process to stop it B being killed will add it to the working set, however. And when theD working set exceeds available RAM, that's when the machine starts to	 sit down.    --   "Sore wa himitsu desu."  To reply by email, remove  the small snack from address. ! http://www.esatclear.ie/~rwallace    ------------------------------    Date: 17 Oct 2003 07:33:08 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) - Subject: Re: Cost of process creation on Unix 3 Message-ID: <uaA216rEBqQh@eisner.encompasserve.org>   V In article <3F8EEB4A.AC3E14FA@istop.com>, JF Mezei <jfmezei.spamnot@istop.com> writes: > P > does unix have the equivalent of shareable images where you only have one copyO > in memory shared by all similar processes ? (eg: does the creation of another P > POP server process using the same image as an existing one really take up that > much more memory ?)   B   Older UNIX did not, but just about anything shipping today does.  L > I realise that process creation is a lot faster on unix than VMS. However,O > does a process hanging around waiting for input (possibly being paged out due P > to inactivity) cost more then process deletion followed a while later by a newC > process creation when the remote user attempts to connect again ?       YMMV.   ------------------------------  # Date: Fri, 17 Oct 2003 16:43:20 GMT 4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk>- Subject: Re: Cost of process creation on Unix 0 Message-ID: <3F901ABC.4B3CE3C8@blueyonder.co.uk>   Bob Koehler wrote: > X > In article <3F8EEB4A.AC3E14FA@istop.com>, JF Mezei <jfmezei.spamnot@istop.com> writes: > > R > > does unix have the equivalent of shareable images where you only have one copyQ > > in memory shared by all similar processes ? (eg: does the creation of another R > > POP server process using the same image as an existing one really take up that > > much more memory ?)  > D >   Older UNIX did not, but just about anything shipping today does.  D Would you care to expand on that? Shareable image like functionality@ has been removed from Unix? Or just to complicated to make work?        --   tim.llewellyn@blueyonder.co.uk   ------------------------------    Date: 17 Oct 2003 12:09:29 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) - Subject: Re: Cost of process creation on Unix 3 Message-ID: <pMqMBZSc6HMV@eisner.encompasserve.org>   g In article <3F901ABC.4B3CE3C8@blueyonder.co.uk>, Tim Llewellyn <tim.llewellyn@blueyonder.co.uk> writes:  >  >  > Bob Koehler wrote: >>  E >>   Older UNIX did not, but just about anything shipping today does.  > F > Would you care to expand on that? Shareable image like functionalityB > has been removed from Unix? Or just to complicated to make work?  ?    I don't get the question.  Did you read my comment backward?   @    My ULTIRX system did not have shareable libraries.  UNIX willI    automatically share image code pages (text in UNIX lingo) mapped from  J    the same image file, but every time you linked a C program, the linker J    found the object implimenting printf() in libc.a and copied it into theI    new image.  Every image file ended up with a complete copy of printf() H    somewhere in it's disk blocks (images were huge).  Two users running K    /usr/bin/vi share all the code in that file, but if one is running diff  K    and one is running vi then each gets his own copy of printf() in memory.   E    Modern "UNIX" like Tru64, HP-UX, and Linux all have shared object  F    libraries, so everybody's printf() maps back to the same blocks in C    libc.so, unless the user foces the link to be to libc.a.  A user H    running vi will map its printf() to the same memory as a user running4    diff since they both map to the same disk blocks.   ------------------------------  % Date: Fri, 17 Oct 2003 14:17:48 +0200 " From: Didier Morandi <no@spam.com>6 Subject: Re: DECUS: "C" stands for Computer, not Corp.4 Message-ID: <3f8fddf2$0$27013$626a54ce@news.free.fr>   Jean-Franois,  E May I suggest that we together "purchase" the trade mark from HP and   (re)create DECUS?    :-)    D.   JF Mezei wrote:   P > When Compaq took over Digital, many individual DECUS chapter, in an attempt toL > suck up to the new owner and portray itself as relevant, decided to changeM > names to remove the "Digital specific" nature and make it possible to adopt % > tandem (Itug) and compaq customers.  > 
 > However:H > I think that Itug never quite merged fully with what used to be DECUS.R > Interex seems to be here to stay and not interested in Digital's legacy productsL > What was once DECUS is now relegated to supporting legacy Digital products& > (PDP, VAX, Alpha,  RSX, VMS, Tru64). > P > It would therefore make a lot of sense if all of what used to be DECUS were to4 > reunite again under the DECUS name since it would: >  > -1 unify the world. I >   Does anyone know what "canacu" is for ? Does anyone if it has similar L > purposes to Encompass ? Does anyone know what countries each operates in ?? > DECUS USA and DECUS Canada are far more representative names.  > L > -2 "DECUS" is far more representative of the remaining membership of thoseO > groups. has any chapter truly been succesful at expanding beyond its original  > Digital-centric purposes ? > P > -3 By focusing on its core competence (Digital products) DECUS can be far moreF > relevant to the remaining customers than if it dilutes itself into aN > meaningless user group which disenfranchises Digital-centric members becauseJ > they don't get the contents they want, and doesn't succeed in attracting; > others because it doesn't excell in anything windows etc.  > K > It is far better to be excellent in one area and attract at least a small P > number of very loyal members who will greatly benefit from the user group than6 > to be mediocre in many areas and not attract anyone. > J > There are plenty of windows user groups. There was only one Digital userM > group. Focus back on Digital where DECUS serves a purpose nobody else does. ; > Widnows customers can find support in much better groups.    ------------------------------    Date: 17 Oct 2003 09:04:35 -07001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) 6 Subject: Re: DECUS: "C" stands for Computer, not Corp.= Message-ID: <cf15391e.0310170804.463f16c2@posting.google.com>   [ JF Mezei <jfmezei.spamnot@istop.com> wrote in message news:<3F8ECD55.32CD1FDA@istop.com>... P > When Compaq took over Digital, many individual DECUS chapter, in an attempt toL > suck up to the new owner and portray itself as relevant, decided to changeM > names to remove the "Digital specific" nature and make it possible to adopt % > tandem (Itug) and compaq customers.   > The US DECUS Chapter changed its name to Encompass, perhaps toC encompass a wider variety of users or to be an ENterprise COMPuting 7 ASSociation, I don't know.  See http://encompassus.org/   A In an ill-fated decision, the DECUS organization covering Europe, > Middle East and Africa (EMEA) changed its name to Compaq UsersB Organization just 3 days before the HP merger was announced.  ThisC organization subsequently merged with Interex in their geographical = area to form HP-Interex EMEA.  See http://www.hp-interex.org/   
 > However:H > I think that Itug never quite merged fully with what used to be DECUS.  D ITUG has remained independent, although it has been cooperating withF DECUS and now HP-Interex EMEA in holding joint conferences in Europe. ? In the US, they have a tradition of holding ITUG conferences in A Cupertino near the NonStop engineering organization (like how our B OpenVMS Advanced Technical Bootcamp is always held in Nashua to beD near VMS Engineering), so there's a strong disincentive to hold thatC conference jointly with HP World or HP-ETS at a different location.   " > Interex seems to be here to stay   True.   1 > and not interested in Digital's legacy products   < This is not true at all.  Interex has been very welcoming ofB pre-merger Digital, Compaq, and Tandem constituencies.  They added@ pre-merger Digital and Tandem session content to HP World in LosA Angeles in 2002 (and of course even more this year at HP World in B cooperation with Encompass).  The excellent HP World Magazine from? Interex includes pre-merger Digital content. Just check out the C Interex website at http://interex.org/ -- they even have an OpenVMS . Portal with a pointer from their main webpage.  I >   Does anyone know what "canacu" is for ? Does anyone if it has similar L > purposes to Encompass ? Does anyone know what countries each operates in ?  A Canadian Association of Compaq Users -- see http://canacu.org/ -- E obviously another victim of a name change to accommodate Compaq.  But D whatever the name, it seems to be quite an active user group, with a? regular newsletter, and having organized quite a few activities  lately.   F I don't think the names matter anymore -- in my experience, the spirit0 of DECUS lives on in all of these organizations.   ------------------------------  % Date: Fri, 17 Oct 2003 13:05:53 -0400 * From: JF Mezei <jfmezei.spamnot@istop.com>6 Subject: Re: DECUS: "C" stands for Computer, not Corp.) Message-ID: <3F902170.5A9F8F31@istop.com>    Keith Parris wrote: H > I don't think the names matter anymore -- in my experience, the spirit2 > of DECUS lives on in all of these organizations.  N But returning to DECUS would help end that ighmare during the compaq era. WhatN we, the users, need, is a user group that focuses on Digital products.  I haveJ no problems with DECUS cooperating with ITUG and Interex. But I don't wantQ some global user group whose contents is diluted amongst too many constituencies.   N If I have an interest in HP-UX, then I'll become a member of Interex. But if I. have an interest in VMS, I want my DECUS back.   ------------------------------    Date: 17 Oct 2003 09:53:28 -0700$ From: gspamtackett@yahoo.com (Galen)& Subject: Error installing JRE V1.4.1-2= Message-ID: <bdc65a53.0310170853.281daf2d@posting.google.com>   B I tried to install JRE V1.4.1-2 from the savesets you get when you@ expand the DEC-AXPVMS-JRE-V0104-12-1 kit. VMSINSTAL failed after
 reporting:    %BACKUP-E-OPENOUT, error opening: DKA0:[SYS0.SYSUPD.JRE-V141-2]JAVA$MAWT_SHR.EXE;1 as output/ -RMS-E-FEX, file already exists, not superseded     %BACKUP-E-OPENOUT, error opening; DKA0:[SYS0.SYSUPD.JRE-V141-2]LIBJAVAPLUGIN_OJI.SO as output / -RMS-E-FEX, file already exists, not superseded   < I assume this indicates a problem with the product savesets.  $ Has anyone else gotten this to work?   ------------------------------  % Date: Fri, 17 Oct 2003 09:39:13 +0200 " From: Didier Morandi <no@spam.com>E Subject: Good Bye Polycenter Suite, Hello Cockpit Manager for OpenVMS 4 Message-ID: <3f8f9ca8$0$27014$626a54ce@news.free.fr>   FYI.; http://h71000.www7.hp.com/openvms/journal/v1/ns3/mgclus.htm   % Managing Mission-Critical VMSclusters   8 Johan Michiels  Nand Staes  Gerrit Woertman  Jan Knol HP Consulting & Integration    Abstract  C In this paper we present the VMS cockpit concept. The cockpit is a  E dedicated VMS system that centralizes all management operations. The  F main task of this cockpit is to monitor the entire OpenVMS production I environment and to take care of event notification in a uniform way. The  B cockpit assists the system manager 24 hours per day and automates  routine tasks.   Introduction  G The main responsibility of the VMS system manager is to make sure that  I his OpenVMS environment always delivers the performance and availability  B levels the business demands. To achieve this, he needs tools that @ support and automate his job, and that cover all aspects of the C production environment. In many ways this can be compared with the   cockpit of a plane or boat.   B To build such a cockpit, HP Consulting & Integration has made the F toolkit CockpitMgr for OpenVMS available to customers. This toolkit F has a number of unique and valuable features every VMS system manager C needs. The present paper gives a technical overview of those tools.   C The CockpitMgr toolkit has already been deployed at many important  F OpenVMS customer sites, where it is well appreciated. CockpitMgr runs C entirely on OpenVMS, provides a solution for the entire production  D environment, and substitutes for the well-known POLYCENTER products.   ../..   G The next version of the CockpitMgr toolkit, to be released early 2003,  ? will contain its own built-in console management functionality.    ../..    D. --  ;           Read the latest VAX/VMS to Itanium Migration News >         http://www.didiermorandi.com/vaxvms2itanium_200310.pdf  F Didier Morandi  ~ sarl au capital de 8 000 euros ~  Revendeur agr HPC      Expertise en environnement DIGITAL ~ Formation ~ Programmation D    5 avenue Albert Durand 31700 Blagnac France Tl: 33(0)5 6131 6287D    SIRET 448 694 851 00016 RCS Toulouse http://www.didiermorandi.com   ------------------------------    Date: 17 Oct 2003 08:57:41 -0700+ From: seanobanion@attbi.com (Sean O'Banion) I Subject: Re: Good Bye Polycenter Suite, Hello Cockpit Manager for OpenVMS < Message-ID: <f883d5a4.0310170757.23ba9b1@posting.google.com>  1 I can't find any costs, purchase or maintenence:  : is this cheaper than CA's version of the Polycenter suite?     Sean    ^ Didier Morandi <no@spam.com> wrote in message news:<3f8f9ca8$0$27014$626a54ce@news.free.fr>... > FYI.= > http://h71000.www7.hp.com/openvms/journal/v1/ns3/mgclus.htm  > ' > Managing Mission-Critical VMSclusters  > : > Johan Michiels  Nand Staes  Gerrit Woertman  Jan Knol > HP Consulting & Integration  > 
 > Abstract > E > In this paper we present the VMS cockpit concept. The cockpit is a  G > dedicated VMS system that centralizes all management operations. The  H > main task of this cockpit is to monitor the entire OpenVMS production K > environment and to take care of event notification in a uniform way. The  D > cockpit assists the system manager 24 hours per day and automates  > routine tasks. >  > Introduction > I > The main responsibility of the VMS system manager is to make sure that  K > his OpenVMS environment always delivers the performance and availability  D > levels the business demands. To achieve this, he needs tools that B > support and automate his job, and that cover all aspects of the E > production environment. In many ways this can be compared with the   > cockpit of a plane or boat.  > D > To build such a cockpit, HP Consulting & Integration has made the H > toolkit CockpitMgr for OpenVMS available to customers. This toolkit H > has a number of unique and valuable features every VMS system manager E > needs. The present paper gives a technical overview of those tools.  > E > The CockpitMgr toolkit has already been deployed at many important  H > OpenVMS customer sites, where it is well appreciated. CockpitMgr runs E > entirely on OpenVMS, provides a solution for the entire production  F > environment, and substitutes for the well-known POLYCENTER products. >  > ../..  > I > The next version of the CockpitMgr toolkit, to be released early 2003,  A > will contain its own built-in console management functionality.  >  > ../..  >  > D.   ------------------------------  % Date: Fri, 17 Oct 2003 09:11:52 -0700 # From: "Tom Linden" <tom@kednos.com> I Subject: RE: Good Bye Polycenter Suite, Hello Cockpit Manager for OpenVMS 9 Message-ID: <CIEJLCMNHNNDLLOOGNJICEHBIEAA.tom@kednos.com>    You can send mail directly to    Gerrit.Woertman@hp.com Johan.Michiels@hp.com    >-----Original Message-----03 >From: Sean O'Banion [mailto:seanobanion@attbi.com]o' >Sent: Friday, October 17, 2003 8:58 AM- >To: Info-VAX@Mvb.Saic.ComB >Subject: Re: Good Bye Polycenter Suite, Hello Cockpit Manager for >OpenVMS >l >e1 >I can't find any costs, purchase or maintenence:c; >is this cheaper than CA's version of the Polycenter suite?B >i >e >Sean  >s >i. >Didier Morandi <no@spam.com> wrote in message1 >news:<3f8f9ca8$0$27014$626a54ce@news.free.fr>...e >> FYI. > >> http://h71000.www7.hp.com/openvms/journal/v1/ns3/mgclus.htm >>( >> Managing Mission-Critical VMSclusters >>; >> Johan Michiels  Nand Staes  Gerrit Woertman  Jan Knols >> HP Consulting & Integration >> >> AbstractJ >>E >> In this paper we present the VMS cockpit concept. The cockpit is a G >> dedicated VMS system that centralizes all management operations. ThedH >> main task of this cockpit is to monitor the entire OpenVMS productionK >> environment and to take care of event notification in a uniform way. The D >> cockpit assists the system manager 24 hours per day and automates >> routine tasks.e >> >> Introduction  >>I >> The main responsibility of the VMS system manager is to make sure that K >> his OpenVMS environment always delivers the performance and availability D >> levels the business demands. To achieve this, he needs tools thatB >> support and automate his job, and that cover all aspects of theE >> production environment. In many ways this can be compared with theO >> cockpit of a plane or boat. >>D >> To build such a cockpit, HP Consulting & Integration has made theH >> toolkit CockpitMgr for OpenVMS available to customers. This toolkitH >> has a number of unique and valuable features every VMS system managerF >> needs. The present paper gives a technical overview of those tools. >>E >> The CockpitMgr toolkit has already been deployed at many importantrH >> OpenVMS customer sites, where it is well appreciated. CockpitMgr runsE >> entirely on OpenVMS, provides a solution for the entire productiondG >> environment, and substitutes for the well-known POLYCENTER products.e >> >> ../.. >>I >> The next version of the CockpitMgr toolkit, to be released early 2003,eB >> will contain its own built-in console management functionality. >> >> ../.. >> >> D.e >2 >---' >Incoming mail is certified Virus Free. ; >Checked by AVG anti-virus system (http://www.grisoft.com). A >Version: 6.0.524 / Virus Database: 321 - Release Date: 10/6/2003  >  ---e& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.524 / Virus Database: 321 - Release Date: 10/6/2003   ------------------------------  % Date: Fri, 17 Oct 2003 05:54:54 -0400a* From: "Bill Todd" <billtodd@metrocast.net>H Subject: Re: Itanium/Integrity & EV7/Marvel leadership benchmark results2 Message-ID: <soWdnZ3sDq33IRKiU-KYhw@metrocast.net>  D "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> wrote in message, news:%9Djb.7213$gg5.3765@news.cpqcorp.net... >l7 > "Bill Todd" <billtodd@metrocast.net> wrote in message . > news:HfmdnXE3xPhzsROiU-KYuQ@metrocast.net... > >PB > > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message: > > news:cf15391e.0310141031.edc022d@posting.google.com...G > > > Some here have contended that HP would never tout Alpha benchmarkp > > > results. > >jK > > They have?  Could you provide specific citations for these contentions?s IrC > > don't remember seeing them, but of course I'm hardly omniscient  >aB > Really?  I think you pretty much want to present yourself as our all-seeing,k$ > all-knowing guardian of the light.  I No, Fred:  you're just stuck in such a hole that *any* illumination seemsr@ blinding.  These days I don't really have to do much to keep theG conversations here honest:  plenty of other people are in possession ofeJ sufficient clue to do so - and appear to be sufficiently motivated *to* do so.3  I If you find such honesty tiresome, why don't you try to take steps to fixII the corporate problems that bring it out?  That would seem to be a prettye- optimal solution from *everyone's* viewpoint.i   - bill   ------------------------------  % Date: Fri, 17 Oct 2003 19:19:18 +0400 : From: "Ruslan R. Laishev" <Laishev{at}DeltaTelecom{dot}RU> Subject: LGI Callout & SSH3 Message-ID: <85D8F3A6D73895A24376567B7F243739@nntp>    Hi All!u; 	We have tried to use a LGI callout (OVMS 7.3-1 & UCX 5.1):,  8 %%%%%%%%%%%  OPCOM  17-OCT-2003 09:45:37.14  %%%%%%%%%%%# Message from user INTERnet on ROKOT,E INTERnet ACP SSH Accept Request from Host: 999.242.10.999 Port: 54441   8 %%%%%%%%%%%  OPCOM  17-OCT-2003 09:45:37.14  %%%%%%%%%%%' Message from user AUDIT$SERVER on ROKOTaH Security alarm (SECURITY) and security audit (SECURITY) on ROKOT, system id: 104o 3f8 Auditable event:          Detached process login failure1 Event time:               17-OCT-2003 09:45:37.14e" PID:                      0000047E) Process name:             TCPIP$SSH_BG115w" Username:                 INTERnet/ Process owner:            [TCPIP$AUX,TCPIP$SSH]I Image name:D0 ROKOT$DKA0:[SYS0.SYSCOMMON.][SYSEXE]LOGINOUT.EXEE Status:                   %LIB-F-ACTIMAGE, error activating image !ASh  8 %%%%%%%%%%%  OPCOM  17-OCT-2003 09:45:37.14  %%%%%%%%%%%# Message from user INTERnet on ROKOTr7 INTERnet ACP detected TCPIP$SSH exiting before 'socket'M  8 %%%%%%%%%%%  OPCOM  17-OCT-2003 09:45:37.15  %%%%%%%%%%%# Message from user INTERnet on ROKOTsE INTERnet ACP AUXS error during process exit  Status = %LIB-F-ACTIMAGE1    6   Logins with the "SET H 0" works normaly. Any ideas ?     -- :F + WBR, OpenVMS [Sys|Net] HardWorker .................................+D Delta Telecom Inc., NMT-450i, IMT-MC-450(CDMA2000) cellular operatorE Russia,191119,St.Petersburg,Transportny per. 3 Cel: +7 (812) 116-3222bF +http://starlet.deltatelecom.ru ............. Frying on OpenVMS only +   ------------------------------    Date: 17 Oct 2003 10:55:38 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)o Subject: Re: LGI Callout & SSH3 Message-ID: <ZN0pfNrmDbR5@eisner.encompasserve.org>0  p In article <85D8F3A6D73895A24376567B7F243739@nntp>, "Ruslan R. Laishev" <Laishev{at}DeltaTelecom{dot}RU> writes:	 > Hi All!e= > 	We have tried to use a LGI callout (OVMS 7.3-1 & UCX 5.1):   2 > ROKOT$DKA0:[SYS0.SYSCOMMON.][SYSEXE]LOGINOUT.EXEG > Status:                   %LIB-F-ACTIMAGE, error activating image !ASs  < Use Auditing to determine what image could not be activated.  8 >   Logins with the "SET H 0" works normaly. Any ideas ?  @ Consider whether this TCPIP$SSH process might be running without necessary privileges.n   ------------------------------    Date: 17 Oct 2003 12:23:42 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)- Subject: Re: LGI Callout & SSH3 Message-ID: <WtK2GNJW0ML7@eisner.encompasserve.org>s  p In article <85D8F3A6D73895A24376567B7F243739@nntp>, "Ruslan R. Laishev" <Laishev{at}DeltaTelecom{dot}RU> writes:  G > Status:                   %LIB-F-ACTIMAGE, error activating image !ASp  @    analyze/image/interactive will tell you all the images you'reB    image is using.  Your image or one of the images it's linked to&    is too protected or can't be found.  E    Make no assumptions about how the network layer finds your images.    ------------------------------  % Date: Fri, 17 Oct 2003 06:23:12 -0700 # From: "Tom Linden" <tom@kednos.com>h Subject: Mail text truncated9 Message-ID: <CIEJLCMNHNNDLLOOGNJIMEGIIEAA.tom@kednos.com>e  D This on a MX5.3 on 7.3-1 system,  The text of incoming mail is being
 truncated,+ for example reading directly form VMS mail:u   ---- start sample ----5 Two Exciting Stay and Play Packages in Palm Springs!!I  H Take a break this Fall and enjoy the finest golfing experience that PALM SPRINGSeF  has to offer. The weather is already fantastic. Rated 4 stars by Golf	 Digest, T J ahquitz Creek is the only Arnold Palmer managed property in the desert and is knu own for its excv  + Visit Our Web Site at www.tahquitzcreek.com    --- end sample ---    J You will note that there do not appear to be any carriage returns and that3 the truncation seems to occur after 256 characters.   J Now I haven't changed anything so I don't understand how this can suddenlyH happen.  Now I am using MX smtp server and I have been playing with spam
 filtering.     ---e& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.524 / Virus Database: 321 - Release Date: 10/6/2003   ------------------------------  % Date: Fri, 17 Oct 2003 07:02:55 -0700 # From: "Tom Linden" <tom@kednos.com>s  Subject: RE: Mail text truncated9 Message-ID: <CIEJLCMNHNNDLLOOGNJIKEGLIEAA.tom@kednos.com>   @ Never mind, Matt helped me out, there was a setting override the default behaviour.     >-----Original Message-----e) >From: Tom Linden [mailto:tom@kednos.com]w' >Sent: Friday, October 17, 2003 6:23 AMh >To: Info-VAX@Mvb.Saic.Com >Subject: Mail text truncatede >a > E >This on a MX5.3 on 7.3-1 system,  The text of incoming mail is beingt >truncated,o, >for example reading directly form VMS mail: >  >---- start sample ----.6 >Two Exciting Stay and Play Packages in Palm Springs!! > I >Take a break this Fall and enjoy the finest golfing experience that PALM  >SPRINGSG > has to offer. The weather is already fantastic. Rated 4 stars by Golf 
 >Digest, TK >ahquitz Creek is the only Arnold Palmer managed property in the desert andw >is kn >own for its exc >o, >Visit Our Web Site at www.tahquitzcreek.com >e >--- end sample ---r >n >,K >You will note that there do not appear to be any carriage returns and thatt4 >the truncation seems to occur after 256 characters. >cK >Now I haven't changed anything so I don't understand how this can suddenly I >happen.  Now I am using MX smtp server and I have been playing with spams >filtering.h >c >  >---' >Outgoing mail is certified Virus Free. ; >Checked by AVG anti-virus system (http://www.grisoft.com).iA >Version: 6.0.524 / Virus Database: 321 - Release Date: 10/6/2003  >a >---' >Incoming mail is certified Virus Free.t; >Checked by AVG anti-virus system (http://www.grisoft.com). A >Version: 6.0.524 / Virus Database: 321 - Release Date: 10/6/2003  >n ---C& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.524 / Virus Database: 321 - Release Date: 10/6/2003   ------------------------------  % Date: Fri, 17 Oct 2003 12:43:27 -0400t* From: JF Mezei <jfmezei.spamnot@istop.com>  Subject: Re: Mail text truncated) Message-ID: <3F901C2D.BBD57E9C@istop.com>a   Tom Linden wrote: L > You will note that there do not appear to be any carriage returns and that5 > the truncation seems to occur after 256 characters.a    Ahhh, that damned limit again !   M When going through DECNET, each record is truncated at 256 bytes because that-H is the original limit of DECNET mail transport. If the mail is delivered! locally, the limit doesn't exist.n  H Remember that microsoft mailers are not compliant and send text with theK <cr><lf> delimiting paragraphs instead of lines. You get one very long linem per paragraph.   ------------------------------  % Date: Fri, 17 Oct 2003 13:43:48 +0100 O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com>tL Subject: Re: Morgan Stanley analyst releases positive research report on HPQ0 Message-ID: <bmoo6k$1o6$1@new-usenet.uk.sun.com>   Bill Gunshannon wrote:> > In article <cf15391e.0310141329.e75cac8@posting.google.com>,6 > 	keithparris_NOSPAM@yahoo.com (Keith Parris) writes: >  >>From CBS Marketwatch,t >>@ >>"Rising expectations of a strong holiday shopping season aidedD >>Hewlett-Packard (HPQ), which has embarked on a well-timed consumer >>electronics push." e >  > D > I certainly hope you don't think this is good news about VMS.  AllC > their saying is everyone wants a digital camera and photo-quality-B > printer this year.  Everybody (including Walter Hewlett) already > knew that. >  > bill >     H In fact if true it could be terrible news for OpenVMS, a storming resultE for the printing and imaging business without any sign of improvementr? in the Enterprise and PC businesses will only add weight to the  arguments for breaking up HP.t   Regardse Andrew Harrisonv   ------------------------------  # Date: Fri, 17 Oct 2003 13:08:46 GMTl, From: Mark Jilson <jilly@clarityconnect.com>3 Subject: Re: New OpenVMS system with no root passwds@ Message-ID: <032b52adbdfa2968b5ed0f9b5368dd43@news.teranews.com>   Keith A. Lewis wrote:y  1 > hoff@hp.nospam (Hoff Hoffman) writes in articlehH > <Ovijb.7103$AZ3.5859@news.cpqcorp.net> dated Wed, 15 Oct 2003 20:56:14 > GMT:F >>In article <bmhs49$2gm$1@newslocal.mitre.org>, lewis@PROBE.mitre.org >>(Keith A. Lewis) writes:K >>:There's a project called GNV which I intend to build for myself tonight.nH >>:The only releases so far are alpha-test, but it does claim to include >>:bash. >>E >>  There are solid GNV kits available.  You can download a kit from:  >>  4 >>    http://h71000.www7.hp.com/portability/gnv.html > F > My apologies, Hoff.  I was misreading the version number in the PCSIK > filename.  Now I just need to borrow the BLISS distribution (I'm dead-set 0 > on building GNV from source before I install). > K > BTW, I used "prod extract file *" to get the sources out, and it put thempL > all in the top-level directory.  Is that expected behavior?  "prod list *" > has the directory names. > 2 > --Keith Lewis              klewis {at} mitre.org@ > The above may not (yet) represent the opinions of my employer.  H You need to use PRODUCT COPY kitname /SOURCE=disk:[dir] /DEST=disk:[dir]1 /FORMAT=REFERENCE to get the directory structure.s -- pC Jilly - Working from Home in the Chemung River Valley - Waverly, NY H       - jilly@clarityconnect.com                      - Brett Bodine fanH       - Mark.Jilson@hp.com                            - since 1975 or soH       - http://www.jilly.baka.com           - http://www.brettbodine.com   ------------------------------    Date: 17 Oct 2003 06:23:57 -0700+ From: spamdump@mccready.com (Gary McCready)a; Subject: NYMLUG NYC Encompass meeting "VMS Update" 10/29/03h= Message-ID: <ffd79a6c.0310170523.700b1960@posting.google.com>   A Please join us for the next NY Metro Local Users Group meeting.   = Note: you must RSVP in advance for entrance to the building. l   "OpenVMS Update"   Wednesday, October 29th, 2003. 12:30-7 PM f HP Offices  . 2 Penn Plaza 8th Floor ( 7th Ave and 32nd St) , Directly above the entrance to Penn Station.  : 12:30 - 1:00     Check in and old-style DECUS Networking. / 1:00 - 1:10     Welcome and announcements, HP. a> 1:10 - 1:15     Gary McCready: LUG Business and Announcements : 1:15 - 2:30     Greg Jordan, HP "OpenVMS Technical Update"= 2:45 - 4:00     Greg Jordan, HP "OpenVMS Performance Update" j? 4:15 - 5:00	    Edgar Zamora, HP "OpenVMS Implementation Update15 5:00 - 5:30     Break, Demonstrations and networking. B 5:30 - 7:00	    Greg Jordan, HP "OpenVMS Technical Update" (repeat	 from 1pm)B  0  OpenVMS Technical Update and Product Road Maps   @ This session includes key highlights of the OpenVMS Alpha V7.3-1F release, the OpenVMS system support rollout plan for the EV7 platform,E a preview of upcoming OpenVMS releases, the rollout plans for OpenVMSrD on ItaniumR architecture-based HP systems and the road map and plansB for the OpenVMS layered product suite. Discussion of the operating> system's new features will include such topics as performance,. security and system availability enhancements.   OpenVMS Performance Update    E This session discusses details of the performance enhancements to thet= OpenVMS V7.3-1 operating system. A number of examples showingh? application performance improvements from both software and newrD hardware are presented. Information on performance work for the next1 release of the operating system is also provided.t     OpenVMS ImplementationsI  F Case studies of current and planned local OpenVMS implementations will
 be discussed.e    F Greg Jordan, Senior Member of Technical Staff, OpenVMS Engineering HP   C Greg Jordan: Greg is a Sr. Member of Technical staff in the OpenVMS.B Base OS Engineer Group. He has been working on the performance andC scaling of OpenVMS on large SMP systems for the last several years.tE Greg joined Digital Equipment Corporation in 1981 and has been a partA" of OpenVMS Engineering since 1991.    ) Edgar Zamora, BCS Solutions Architect, HPi  A Edgar is an HP VMS Ambassador and serves the NYC Metro area as anb OpenVMS resource for HP.  C To RSVP: send email to  NYMLUG-RSVP (at) yahoogroups.com, includingnF company name, when you plan to arrive and a way to contact you, if notB by your sending email address. You response, including the sending6 email address, will not be visible in any public list.  E You will be met by a HP employee at the security desk at the entrancesE to the building (now at the 7th Avenue side). Should you arrive late, @ please call the HP reception desk at 212-856-2000 for an escort.E Please plan to arrive promptly as there is limited seating available.t  C HP will provide light refreshments for the afternoon, and pizza foro those staying after 5pm.  F For further information on the group you may contact Gary McCready at E NYMLUG(at)McCready.com. If you are interested in suggesting a meeting-F topic please contact Gary to be placed on the Steering Committee email list.e  E Apologies are in order if you have received multiple versions of thiseD message. Feel free to forward this message to others who may have anD interest in attending - you do not have to be an Encompass member to attend.o  D If you did not receive an "original" version of this message, please? join our mailing list at  http://groups.yahoo.com/group/NYMLUG/    ------------------------------  % Date: Fri, 17 Oct 2003 09:27:49 -0500i+ From: Chuck Aaron <caaron@ceris.purdue.edu>y" Subject: OpenVms 7.3.1 ODBC Driver- Message-ID: <bmouko$1ak$1@mozo.cc.purdue.edu>e   Hello,  ' Can anyone recommend a good ODBC driverh& that will run on OpenVMS 7.3.1. I need$ one that will access an Adabas 4.1.1  database that is not relational.   Thanks,o Chuck    ------------------------------  % Date: Fri, 17 Oct 2003 05:41:54 -0400h* From: "Bill Todd" <billtodd@metrocast.net>; Subject: Re: OpenVMS I64 ISV application count now over 500 2 Message-ID: <4qednSKfDbTAJBKiU-KYuA@metrocast.net>  D "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> wrote in message, news:X2Djb.7211$v55.5586@news.cpqcorp.net... >> > Give it up Bill.  K Are you truly illiterate, Fred?  2+ years ago I stated that I would be heresJ for the duration (until Alpha got resurrected or those responsible for itsJ demise, and the lies surrounding it, were purged from the corporation).  IL remember reminding you of this personally some time during the past year, so@ the problem presumably isn't simply failing memory on your part.  2 Perhaps you can get a friend to explain it to you.   - bill   ------------------------------  % Date: Fri, 17 Oct 2003 22:42:37 +0800n, From: Paul Repacholi <prep@prep.synonet.com>* Subject: Re: OpenVMS in an HP advertisment- Message-ID: <873cdshqn6.fsf@prep.synonet.com>-  # Nigel Barker <nigel@hp.com> writes:   F > The 'New Scientist' is a popular science magazine i.e. you will find? > it on sale in most good news vendors in the UK. It's far morep; > mainstream than most IT publications. It would be read byCF > scientists, technologists & intelligent lay people. I used to have aF > cheap subscription via the high school that I attended.  It's like aD > rather more popular 'Scientific American. Check out their web site > http://www.newscientist.com/  E It is available in many newsagents all over the world EXCEPT the landtE of the single digit IQ president. NS is also printed in AU, I'll findn9 a copy of the Oct 11 issue and see if the add came south.l     --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.r@                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Fri, 17 Oct 2003 12:57:04 +0200n" From: Didier Morandi <no@spam.com>' Subject: Re: PL/1 users, where are you?g4 Message-ID: <3f8fcb06$0$27020$626a54ce@news.free.fr>   Peter Weaver wrote:X  0 > That got me curious, so I took a quick look at( > http://netscan.research.microsoft.com/  < Thanks for the pointer, Mr CHARON-VAX reseller in Canada :-)   D.   ------------------------------    Date: 17 Oct 2003 09:43:47 -07001 From: keithparris_NOSPAM@yahoo.com (Keith Parris)t' Subject: Re: PL/1 users, where are you?-= Message-ID: <cf15391e.0310170843.777a42f7@posting.google.com>i  \ David Froble <davef@tsoft-inc.com> wrote in message news:<3F8DBB26.2080407@tsoft-inc.com>...& > My concern is that HP, at least the B > enterprise and PC business, is going to be a short lived entity.  F In the enterprise area, HP has been taking marketshare away from Sun. E In the PC area, HP is very close to and is regaining marketshare from - Dell.  From the HP World Newsletter, Oct. 16:d   "HP Leads PC Sales Surge  F IDC reported yesterday that global PC shipments jumped 15.7 percent in? the third quarter to 38.4 million units, dramatically exceeding B expectations. IDC had projected 10.4 percent global growth for the quarter.  E HP led the increase, driven by notebook PC sales and bargain hunting,t according to IDC.s  > "The race between HP and Dell just got a lot closer," said Jim@ McDonnell, vice president of marketing for HP's Personal SystemsF Group, in a statement. "HP considerably narrowed the gap with only 0.3F points separating HP from the top spot. What's more, HP's market share@ grew a solid 21 percent sequentially. That not only gives our PCF business strong momentum going into the final calendar quarter and theF traditionally positive holiday buying season, but it also continues to% help drive overall PC market growth."i  A HP's PC business grew 28 percent worldwide, and 34 percent in theu7 United States compared with the same quarter last year.o  A Dell retains the number-one worldwide market share spot with 17.4oE percent versus HP's 17.1 percent (6.67 million units and 6.55 milliona units, respectively).t  @ One admittedly unsupportable data point indicates that HP may beD extremely optimistic about overall hardware shipments for next year.D The Taiwanese Economic Daily News reported this week that HP expectsE to spend $20 billion on manufacturing in Taiwan next year compared tor1 $16 billion forecast for this year.  - Mike Elgan   3 http://hpworldnews.c.tep1.com/maabyWdaa1lPda4e8YPe/    and following that pointer:n  F "When the numbers are broken down, it becomes clear just how close theB two-horse race truly is. With more than 38.4 million units shippedE during the third calendar quarter and with only 127,000 units shippedsA separating HP from Dell, that works out to roughly seven hours of  production over 90 days."-   ------------------------------  % Date: Fri, 17 Oct 2003 14:59:42 +0200(, From: "Ferry Bolhar" <bol@adv.magwien.gv.at>* Subject: Problem with UCX QIOW IO$_SETMODE8 Message-ID: <1066395582.939477@mozart.adv.magwien.gv.at>  
 Hi to all,  I I habe a self-written TCP/IP server, listening on a specific port. BeforeCJ the server is started, a DCL scripts reads some symbols from a config fileL and then starts the server image. Within the image, I use the LIB$GET_SYMBOLI RTL function to obtain the symbol values and process the requests sent to * the server accordingly. Works as expected.  K When a special request is sent to the server, the image terminates, the DCLeL procedure gains control, reads the config file again and restarts the image.K This allow changes in the server configuration by simply editing the configiK file and sending a reconfig request. There is no need to restart the entireo	 software.   L However, when running the image, the second (and further) times, the iosb ofI the QIOW IO$_SETMODE request, which is used to set the socket parameters,nL returns in iosb[0] the return code SS$_DUPLNAM. According the documentation,E this means that a process tries to listen on a TCP port which anotheriH process already is listening on. However, this isn't the case here - theL image has terminated and the bg: device which represented the listening port0 (can be shown with UCX SHOW DEVICE) disappeared.  < However, after waiting for a minute, the QIOW call succeeds.  F Well, I could insert a WAIT 0:1 in the script, but this would make theK server unusable for a minute after each reconfiguration request. This isn'ty what I'm looking for.S  > Can some explain me what happens here? And provide a solution?  : BTW: It's OpenVMS V6.2, UCX V4.2. Occurs on VAX and Alpha.  $ MTIA and kind greetings from Vienna,   Ferryr   -- Ing. Ferry Bolharr% Municipality of Vienna, Department 14c A-1010 Vienna / AUSTRIAg E-mail: bol@adv.magwien.gv.ato     -- Ing. Ferry Bolharn% Municipality of Vienna, Department 14t A-1010 Vienna / AUSTRIA  E-mail: bol@adv.magwien.gv.ath   ------------------------------   Date: 17 Oct 03 08:31:48 PST From: mckinneyj@cpva.saic.comh. Subject: Re: Problem with UCX QIOW IO$_SETMODE( Message-ID: <h0WQfrWkQog8@cpva.saic.com>  8 In article <1066395582.939477@mozart.adv.magwien.gv.at>,/  "Ferry Bolhar" <bol@adv.magwien.gv.at> writes:  > Hi to all, > K > I habe a self-written TCP/IP server, listening on a specific port. BeforeaL > the server is started, a DCL scripts reads some symbols from a config fileN > and then starts the server image. Within the image, I use the LIB$GET_SYMBOLK > RTL function to obtain the symbol values and process the requests sent toq, > the server accordingly. Works as expected. > M > When a special request is sent to the server, the image terminates, the DCLaN > procedure gains control, reads the config file again and restarts the image.M > This allow changes in the server configuration by simply editing the config:M > file and sending a reconfig request. There is no need to restart the entirep > software.t > N > However, when running the image, the second (and further) times, the iosb ofK > the QIOW IO$_SETMODE request, which is used to set the socket parameters,oN > returns in iosb[0] the return code SS$_DUPLNAM. According the documentation,G > this means that a process tries to listen on a TCP port which anothermJ > process already is listening on. However, this isn't the case here - theN > image has terminated and the bg: device which represented the listening port2 > (can be shown with UCX SHOW DEVICE) disappeared. > > > However, after waiting for a minute, the QIOW call succeeds. > H > Well, I could insert a WAIT 0:1 in the script, but this would make theM > server unusable for a minute after each reconfiguration request. This isn'tg > what I'm looking for.3 > @ > Can some explain me what happens here? And provide a solution? > < > BTW: It's OpenVMS V6.2, UCX V4.2. Occurs on VAX and Alpha. > & > MTIA and kind greetings from Vienna, >  > FerryP >  > -- > Ing. Ferry Bolhar0' > Municipality of Vienna, Department 14  > A-1010 Vienna / AUSTRIAr > E-mail: bol@adv.magwien.gv.atB >  >   H Sounds like you need to set the SO_REUSEADDR socket option when you bindF your port (usually a good idea for any server process). I suspect thatG you're waiting for the TIMED_WAIT state for your socket to expire priortF to being allowed to re-use the socket. Setting this option will permit you to re-use it immediately.i   -- n - Jims   ------------------------------   Date: 17 Oct 2003 11:01 CDTe. From: carl@nospam.gerg.tamu.edu (Carl Perkins). Subject: Re: Problem with UCX QIOW IO$_SETMODE4 Message-ID: <17OCT200311014224@nospam.gerg.tamu.edu>  0 "Ferry Bolhar" <bol@adv.magwien.gv.at> writes...L }When a special request is sent to the server, the image terminates, the DCLM }procedure gains control, reads the config file again and restarts the image.iL }This allow changes in the server configuration by simply editing the configL }file and sending a reconfig request. There is no need to restart the entire
 }software. } M }However, when running the image, the second (and further) times, the iosb of2J }the QIOW IO$_SETMODE request, which is used to set the socket parameters,M }returns in iosb[0] the return code SS$_DUPLNAM. According the documentation,iF }this means that a process tries to listen on a TCP port which anotherI }process already is listening on. However, this isn't the case here - theeM }image has terminated and the bg: device which represented the listening porta1 }(can be shown with UCX SHOW DEVICE) disappeared.a } = }However, after waiting for a minute, the QIOW call succeeds.n } G }Well, I could insert a WAIT 0:1 in the script, but this would make the L }server unusable for a minute after each reconfiguration request. This isn't }what I'm looking for. } ? }Can some explain me what happens here? And provide a solution?o } ; }BTW: It's OpenVMS V6.2, UCX V4.2. Occurs on VAX and Alpha.m } % }MTIA and kind greetings from Vienna,k }  }Ferry   See the example file:o> SYS$COMMON:[SYSHLP.EXAMPLES.TCPIP]TCPIP$TCP_SERVER_QIO_2ND.C;1  > Particularly the parts after the comment that looks like this:   *t ** ** Set Socket Options: **J ** Inform the network driver that we want to reuse this port address right ** away. **I ** There might be some old messages still on the network destined for themH ** last server that used this port. To avoid delivering these stragglingG ** messages to the wrong server the network imposes a timeout that mustgN ** expire between the time a server finishes using a port and that port can beL ** can be assigned to another server. The following call will override this. **L ** The benifit of overriding this time out is, we do not have to wait beforeI ** we can reuse the port address. The down side is, we might get messages & ** that were not ment for this server. ** */  L After that is the setup for a QIO call that uses the TCPIP$M_REUSEADDR item.   --- Carl   ------------------------------  # Date: Fri, 17 Oct 2003 13:50:06 GMTt3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond)?T Subject: Re: Q: Product Install of "older" kits warns of deleting "UNDO" directories2 Message-ID: <isSjb.7266$jT5.4939@news.cpqcorp.net>  , In article <bmn5mg$5i4$1@news01.intel.com>, . Ken Fairfield <My.Full.Name@intel.com> writes: ..A >When I do an, e.g., PROD INSTALL FORTRAN /SAVE_RECOVERY_DATA ...t? >     ... or even without the /SAVE qualifier, it warns me thattE >(a) there are 6 "undo" directories and (b) that they will be deleteda >if I continue.   E This has nothing to do with "older" kits, but with the fact that you hI can use /SAVE_RECOVERY_DATA _only_ on patch kits.  FORTRAN is a full kit.i  ' $ help prod install /save_recovery_datas   PRODUCTy  	   INSTALL2       /SAVE_RECOVERY_DATAe             /SAVE_RECOVERY_DATAd)           /NOSAVE_RECOVERY_DATA (default)e  D  --->  This qualifier applies only to patch and mandatory update kitC        installation. It is ignored when other types of products ares        installed.v         ...e  B Any PRODUCT operation that modifies the PCSI database and does notB (or can not) use /SAVE_RECOVERY_DATA renders any existing recoveryF data unusable -- because PCSI cannot rollback older operations withoutB rolling back the most recent operation.   Doing so would leave the$ database in an inconstent condition.  M Workaround?  Well, you could PRODUCT UNDO the 6 patches, install FORTRAN, andeI then re-install the patches.  However, if the patches have been installedtJ for a while and produced no problems that might require rolling them back,. I would suggest that you not worry about this.     -- iJ       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Fri, 17 Oct 2003 13:47:52 +0100l0 From: Chris Sharman <chris.sharman@sorry.nospam> Subject: rwast4 Message-ID: <bmoodo$nm9$1$830fa7a5@news.demon.co.uk>  2 We had a process (a listener) go into RWAST today.& It showed an astlm of 0, astcnt of 99.% I believe the original astlm was 100.DC I tried raising pcb$l_astcnt & phd$l_astlm (using a variant of Jim  E McKinney's bump_bytlm code posted here a few weeks ago - thanks Jim).-  F It worked for astcnt, but not for astlm, and the process stayed stuck.  F I've rebooted, with astlm 250, but the process doesn't ever appear to G use more than 2-3 asts or so. It does have up to 40 or so subprocesses V (servers for clients).  , Questions: how do I unjam an RWAST process ?  What's with astlm 0, astcnt 99 ?H How do I find out what's going on with it ? (force_process_dump doesn't  work, since it relies on ASTs).$   Thanks,a Chrisw   ------------------------------  % Date: Fri, 17 Oct 2003 14:14:00 +0200s" From: Didier Morandi <no@spam.com>2 Subject: Re: Shared distribution of hobbyist media4 Message-ID: <3f8fdd0e$0$27013$626a54ce@news.free.fr>   Yeah Wayne,   E I shall not compete with ya on the old DEC days :-) but I agree with .B Dave Smith (CSC) that we used to use SPL (or in French "distrib").  H I'm in the DEC business since 1981 (good old PDP 11/34 at Dassault) and  *never* heard of "condist".v   D.  B PS: why do you post the full headers of the messages you reply to?       Wayne Sewell wrote:a  P > No, it's not new.  This term for the package of cds containing all the layeredL > products goes back many, many years, back to the digital days and earlier.   ------------------------------  % Date: Fri, 17 Oct 2003 09:15:48 -0500 ( From: Wayne Sewell <wayne@tachysoft.com>2 Subject: Re: Shared distribution of hobbyist media0 Message-ID: <00A27817.B790DC76.17@tachysoft.com>   >o >Yeah Wayne, > F >I shall not compete with ya on the old DEC days :-) but I agree with C >Dave Smith (CSC) that we used to use SPL (or in French "distrib").e >eI >I'm in the DEC business since 1981 (good old PDP 11/34 at Dassault) and   >*never* heard of "condist". >u    M There was never anything official about the term.  But definitely some peoplea used it.      C >PS: why do you post the full headers of the messages you reply to?t  M I don't always.  It's a side-effect of my unusual configuration: info-vax fedtK into anu news.  The regular anu reply doesn't work right, so I have to edito everything manually.  M Of course, I have no control over what info-vax does header-wise when I post.n   WayneoO =============================================================================== N Wayne Sewell, Tachyon Software Consulting  (281)812-0738   wayne@tachysoft.com; http://www.tachysoft.com/www/tachyon.html and wayne.html   iO =============================================================================== H Randolph Duke (in Trading Places): "Mother always said you were greedy."1    Mortimer Duke: "She meant it as a compliment!"i   ------------------------------  % Date: Fri, 17 Oct 2003 12:36:19 -04003& From: David M Smith <dsmit115@csc.com>2 Subject: Re: Shared distribution of hobbyist media8 Message-ID: <nb60pv418t2dqgl7aejps6geq03cpc3ll3@4ax.com>  G On Fri, 17 Oct 2003 14:14:00 +0200, Didier Morandi <no@spam.com> wrote:U   >Yeah Wayne, >>F >I shall not compete with ya on the old DEC days :-) but I agree with C >Dave Smith (CSC) that we used to use SPL (or in French "distrib").l >dI >I'm in the DEC business since 1981 (good old PDP 11/34 at Dassault) and . >*never* heard of "condist".  N Well, to be honest, I used the term CONDIST up until a few months ago, and nowP I've gotten used to SPL. All the current literature refers to it as SPL, but theF term CONDIST is found on some of the older stuff. I happen to have theK CD_CONTENTS.DAT files from all the CD's I have in my library, going back tonJ September, 1989 (for VAX), and in those files I see that the "name change"E appears to have taken place in about November, 1994. If I look at theuP immediately preceding file (September, 1994) the first few lines look like this:   !t, %DISC_PRODUCT_NAME1 OpenVMS VAX Consolidated) %DISC_PRODUCT_NAME2 Software Distributionn! %KIT_PART_NUMBER QA-VWJ8A-A8.AK01p %SPINE_PART_NUMBER AV-PR58P-RE !tP %DISC_PART_NUMBER AG-PZ3PH-XE,AG-Q6PAC-XE,AG-QAN5A-XE,AG-PJ4YU-RE,AG-PNTPS-RE,A  %DFARS Y !sP %LABEL VAXBINSEP941,VAXBINSEP942,VAXBINSEP943,VAXBINSEP944,VAXBINSEP945,VAXBINS 
 %TYPE CONDISTT
 %PLATFORM VAXt !e %RELEASE_DATE September 1994  / where the November, 1994 file begins like this:-   !-( %DISC_PRODUCT_NAME1 OpenVMS VAX Software# %DISC_PRODUCT_NAME2 Product Libraryh! %KIT_PART_NUMBER QA-VWJ8A-A8.AL01> %SPINE_PART_NUMBER AV-PR58Q-RE !gP %DISC_PART_NUMBER AG-PZ3PJ-XE,AG-Q6PAD-XE,AG-QAN5B-XE,AG-PJ4YV-RE,AG-PNTPT-RE,A  %DFARS Y ! P %LABEL VAXBINNOV941,VAXBINNOV942,VAXBINNOV943,VAXBINNOV944,VAXBINNOV945,VAXBINN 
 %TYPE CONDISTm
 %PLATFORM VAXo !O %RELEASE_DATE November 1994   A Notice that "internal" to the file, you see the line which reads:   
 %TYPE CONDIST:  N in both. In fact, the most recent files I have (September, 2003 for both Alpha& and VAX) still have this line in them.I -------------------------------------------------------------------------nI David M. Smith 302.391.8533                       dsmit115 at csc dot com I Computer Sciences Corporation     (Opinions are those of the writer only) I -------------------------------------------------------------------------    ------------------------------    Date: 17 Oct 2003 02:43:40 -0700- From: XNCGULCTLUMA@spammotel.com (Roose Chua)v Subject: Size of a global page; Message-ID: <6273a2.0310170143.25e0a8c7@posting.google.com>e   Hi,   % I'm just a little bit consfused here:$  ? We have an AlphaServer DS20 with 2GB of memory. When I do a sho4 mem/phy:  @               System Memory Resources on 17-OCT-2003 04:33:29.63  C Physical Memory Usage (pages):     Total        Free      In Use     ModifiedF   Main Memory (2048.00Mb)         262144       49243      190673       22228.  F Of the physical pages in use, 88418 pages are permanently allocated to OpenVMS.  A However, when I try to check on the free contiguous global pages:l  . $ write sys$output f$getsyi("contig_gblpages") 908992( $ write sys$output f$getsyi("page_size") 8192  F Based on the sho mem/phys command above, the total pages for this nodeC is supposed to be 262144. How come that the contiguous global pagesgD here is much larger than that at 908992? Also, how can I convert the2 contiguous global page value into, say, megabytes?   Thanks,} Roose d' confused>   ------------------------------  % Date: Fri, 17 Oct 2003 12:15:04 +0200C: From: Karl Rohwedder <extern.karl.rohwedder@volkswagen.de>" Subject: Re: Size of a global page, Message-ID: <bmofe7$3fs8@doiweb4.wob.vw.vwg>   Roose Chua wrote:n   > Hi,  > ' > I'm just a little bit consfused here:  > A > We have an AlphaServer DS20 with 2GB of memory. When I do a sho$
 > mem/phy: > B >               System Memory Resources on 17-OCT-2003 04:33:29.63 > E > Physical Memory Usage (pages):     Total        Free      In Use   .
 > ModifiedH >   Main Memory (2048.00Mb)         262144       49243      190673       > 22228n > H > Of the physical pages in use, 88418 pages are permanently allocated to
 > OpenVMS. > C > However, when I try to check on the free contiguous global pages:  > 0 > $ write sys$output f$getsyi("contig_gblpages") > 908992* > $ write sys$output f$getsyi("page_size") > 8192 > H > Based on the sho mem/phys command above, the total pages for this nodeE > is supposed to be 262144. How come that the contiguous global pageseF > here is much larger than that at 908992? Also, how can I convert the4 > contiguous global page value into, say, megabytes? > 	 > Thanks,s > Roose d' confused   f CONTIG_GBLPAGES delivers pagelets, i.e. 512 byte pages, whereas SH MEM uses PAGESIZE pages, here 8192.! So 908002 pages is roughly 450MB.b   -- r  + mit freundlichen Gren | with best regards     Karl Rohwedder		iT-Ingenieurteam$ Ellernbruch 11		D-38112 Braunschweig  E mailto:it-ingteam(at)t-online.de | mailto:rohwedder(at)decus.decus.dee- mailto:extern.karl.rohwedder(at)volkswagen.dew   ------------------------------  % Date: Fri, 17 Oct 2003 11:21:10 +0100e* From: "Richard Brodie" <R.Brodie@rl.ac.uk>" Subject: Re: Size of a global page+ Message-ID: <bmofqo$mki@newton.cc.rl.ac.uk>z  : "Roose Chua" <XNCGULCTLUMA@spammotel.com> wrote in message5 news:6273a2.0310170143.25e0a8c7@posting.google.com...   H > Based on the sho mem/phys command above, the total pages for this nodeE > is supposed to be 262144. How come that the contiguous global pagess* > here is much larger than that at 908992?  A Units confusion. The global page value is in pagelets, not pages..A Each pagelet corresponds to 512 bytes, 1 disk block,  1 VAX page.o  H It's somewhat confusing at first but can make things more simpler if youJ have multiple page sizes in a cluster (as was expected for Alpha, and does( happen for mixed architecture clusters).   ------------------------------  + Date: Fri, 17 Oct 2003 15:55:03 +0000 (UTC) , From: lewis@PROBE.mitre.org (Keith A. Lewis)2 Subject: SMTP server VRFY command vs. mail aliases- Message-ID: <bmp3cn$gc$1@newslocal.mitre.org>e  J In order to track and manage who gives my e-mail address to whom, I set upJ forwarding addresses from numbered aliases to my main account.  I used the commands   $ set proc/priv=alle $ mail$ MAIL> set forward/user=kl98581 lewis  0 It works.  However, if I VRFY the address as so:   $ telnet localhost 25 ` 220 host1.mitre.org V5.3-18, OpenVMS V7.3-1 Alpha ready at Fri, 17 Oct 2003 15:44:55 +0000 (GMT) VFRY kl98581 250 kl98581 <LEWIS>   L The SMTP server will give out my destination account name.  I know there areH other ways to forward e-mail on VMS.  Do any of them mask the forwarding- address so that TCPIP$SMTP won't give it out?w  7   Compaq TCP/IP Services for OpenVMS Alpha Version V5.3u8   on a AlphaServer 1200 5/533 4MB running OpenVMS V7.3-1  0 --Keith Lewis              klewis {at} mitre.org> The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Fri, 17 Oct 2003 17:47:17 +0200 9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>- Subject: VMS 7.3-1 patches.e' Message-ID: <3F900F05.B246E3CE@aaa.com>s   Hi. 2 I have downloaded a couple of PCSI patch-sets from  3 ftp://ftp.itrc.hp.com/openvms_patches/alpha/V7.3-1/I  5 Two of the pathsets gives this when doing PROD LIST :n  < %PCSI-E-PDFERR, error in processing product description fileB DSA400:[UTIL.]<FETCH_HTTP>DEC-AXPVMS-VMS731_UPDATE-V0100--4.PCSI;1A   297       execute post_undo "@pcsi$source:[sysupd]postundo.com"p' interactive uses [SYSUPD]POSTUNDO.COM ;V- %PCSI-E-PDFUDK, undefined keyword \post_undo\ = %PCSI-E-PROCTERM, processing terminated due to previous error2" %PCSI-E-S_OPFAIL, operation failedC %PCSIUI-E-ABORT, operation terminated due to an unrecoverable error1	 conditionC $    anda  < %PCSI-E-PDFERR, error in processing product description file? DSA400:[UTIL.]<FETCH_HTTP>DEC-AXPVMS-VMS731_LAN-V0700--4.PCSI;1cA    30       execute post_undo "@pcsi$source:[sysupd]postundo.com"h' interactive uses [SYSUPD]POSTUNDO.COM ;r- %PCSI-E-PDFUDK, undefined keyword \post_undo\i= %PCSI-E-PROCTERM, processing terminated due to previous erroro" %PCSI-E-S_OPFAIL, operation failedC %PCSIUI-E-ABORT, operation terminated due to an unrecoverable errort	 condition>    + Any known problems with these two patches ?o* Or am I fetching from a wrong FTP server ?  	 Jan-Erik.-   ------------------------------  % Date: Fri, 17 Oct 2003 13:01:09 -0400  From: norm.raphael@metso.com Subject: Re: VMS 7.3-1 patches.jQ Message-ID: <OF63BE50A5.D5790033-ON85256DC2.005CFE1B-85256DC2.005D2311@metso.com>D  H Either those patches need(ed) to be applied before the PCSI eco, as the= y<H do not understand SAVE and UNDO functionality introduced in the PCSI ec= o.E (The procedure may be explained in the release notes for that ECO, or  someoneoC else can detail the process, but it is not pretty.  Some ECO's were, reissued to beH compatible with installation after the PCSI eco, so there may be a newe= ro versionnH available that helps.), or you may have the reverse problem, that you n= eedd to doc6 the PCSI ECO in order to do the ones that are balking.   Prenez Garde  @ From:  Jan-Erik S=F6derholm <aaa@aaa.com> on 10/17/2003 11:47 AM  4 Please respond to Jan-Erik S=F6derholm <aaa@aaa.com>   To:    Info-VAX@Mvb.Saic.Com cc:s   Subject:    VMS 7.3-1 patches.     Hi.t2 I have downloaded a couple of PCSI patch-sets from  3 ftp://ftp.itrc.hp.com/openvms_patches/alpha/V7.3-1/n  5 Two of the pathsets gives this when doing PROD LIST :   < %PCSI-E-PDFERR, error in processing product description fileB DSA400:[UTIL.]<FETCH_HTTP>DEC-AXPVMS-VMS731_UPDATE-V0100--4.PCSI;1A   297       execute post_undo "@pcsi$source:[sysupd]postundo.com" ' interactive uses [SYSUPD]POSTUNDO.COM ;m- %PCSI-E-PDFUDK, undefined keyword \post_undo\0= %PCSI-E-PROCTERM, processing terminated due to previous error " %PCSI-E-S_OPFAIL, operation failedC %PCSIUI-E-ABORT, operation terminated due to an unrecoverable error 	 conditiona $o   andl  < %PCSI-E-PDFERR, error in processing product description file? DSA400:[UTIL.]<FETCH_HTTP>DEC-AXPVMS-VMS731_LAN-V0700--4.PCSI;1>A    30       execute post_undo "@pcsi$source:[sysupd]postundo.com"a' interactive uses [SYSUPD]POSTUNDO.COM ;.- %PCSI-E-PDFUDK, undefined keyword \post_undo\o= %PCSI-E-PROCTERM, processing terminated due to previous error " %PCSI-E-S_OPFAIL, operation failedC %PCSIUI-E-ABORT, operation terminated due to an unrecoverable error 	 conditions    + Any known problems with these two patches ?o* Or am I fetching from a wrong FTP server ?  	 Jan-Erik.l       =c   ------------------------------  % Date: Fri, 17 Oct 2003 19:01:28 +0200.9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>  Subject: Re: VMS 7.3-1 patches. ' Message-ID: <3F902068.E6A4EAB1@aaa.com>t  5 I see, yes, the PCSI patch should be installed first.[1 And then these two can be LIST'ed. Seems logical.u  9 Now, I'v been browsing the patch site accessable from thel@ VMS homepage. I can't see that I can ge a "pure" URL to download from.   D I usualy copies from URL from the PC-browser into a DCL command file- that used FETCH_HTTP to download the patches.b   Isn't that possible any more ?  	 Jan-Erik.    norm.raphael@metso.com wrote:  > J > Either those patches need(ed) to be applied before the PCSI eco, as theyJ > do not understand SAVE and UNDO functionality introduced in the PCSI ecoG > (The procedure may be explained in the release notes for that ECO, ort	 > someoneME > else can detail the process, but it is not pretty.  Some ECO's weree > reissued to beJ > compatible with installation after the PCSI eco, so there may be a newer	 > versionsL > available that helps.), or you may have the reverse problem, that you need > to do 8 > the PCSI ECO in order to do the ones that are balking. >  > Prenez Garde > @ > From:  Jan-Erik Sderholm <aaa@aaa.com> on 10/17/2003 11:47 AM > 4 > Please respond to Jan-Erik Sderholm <aaa@aaa.com> >  > To:    Info-VAX@Mvb.Saic.Com > cc:5 >   > Subject:    VMS 7.3-1 patches. >  > Hi.U4 > I have downloaded a couple of PCSI patch-sets from > 5 > ftp://ftp.itrc.hp.com/openvms_patches/alpha/V7.3-1/  > 7 > Two of the pathsets gives this when doing PROD LIST :c > > > %PCSI-E-PDFERR, error in processing product description fileD > DSA400:[UTIL.]<FETCH_HTTP>DEC-AXPVMS-VMS731_UPDATE-V0100--4.PCSI;1C >   297       execute post_undo "@pcsi$source:[sysupd]postundo.com"r) > interactive uses [SYSUPD]POSTUNDO.COM ;p/ > %PCSI-E-PDFUDK, undefined keyword \post_undo\n? > %PCSI-E-PROCTERM, processing terminated due to previous errors$ > %PCSI-E-S_OPFAIL, operation failedE > %PCSIUI-E-ABORT, operation terminated due to an unrecoverable error  > conditionf > $c >  > andy > > > %PCSI-E-PDFERR, error in processing product description fileA > DSA400:[UTIL.]<FETCH_HTTP>DEC-AXPVMS-VMS731_LAN-V0700--4.PCSI;1sC >    30       execute post_undo "@pcsi$source:[sysupd]postundo.com"n) > interactive uses [SYSUPD]POSTUNDO.COM ;o/ > %PCSI-E-PDFUDK, undefined keyword \post_undo\ ? > %PCSI-E-PROCTERM, processing terminated due to previous errora$ > %PCSI-E-S_OPFAIL, operation failedE > %PCSIUI-E-ABORT, operation terminated due to an unrecoverable errorn > conditiona > - > Any known problems with these two patches ? , > Or am I fetching from a wrong FTP server ? >  > Jan-Erik.a   ------------------------------  % Date: Fri, 17 Oct 2003 09:41:01 +0100r* From: Nic Clews <sendspamhere@[127.0.0.1]>! Subject: Re: VMS defeated by wind ' Message-ID: <bmo9si$n7c$1@lore.csc.com>    Fred Kleinsorge wrote: >  ...r > K > And yes, USB and TCPIP are the way to go with intellegent power supplies.   > It just makes much more sense.  H Oh, I can just see it. Before the systems running Windoze XYZ gets takenH out with yet another virus/worm, the power supplies die with exactly theA same virus/worm because it's "running" the same operating system.a   -- n? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciencesc nclews at csc dot comi   ------------------------------  % Date: Fri, 17 Oct 2003 12:01:33 +0200H> From: "Winfried Bergmann" <winfried.bergmannNosPAM@empuron.de>! Subject: Re: VMS defeated by windu9 Message-ID: <bmoeip$pdtad$1@ID-170759.news.uni-berlin.de>2  9 "Jan-Erik Sderholm" <aaa@aaa.com> schrieb im Newsbeitrag ! news:3F8E8608.3DF012C2@aaa.com...t > Nic Clews wrote: > > I > > Which begs the question, will VMS have a DCL command to switch on/off H > > and select power supplies, based on the (different) power companies'( > > current rate for the electricity ;-) >n@ > And the rates should of course be automaticly fetched from theA > power companies web sites and the switching beeing automated...o  K Funny, my company tries to introduce software solutions for power utilitiesmB to provide services like this. Unfortunately, they are still quite conservative in this area...   ------------------------------  % Date: Fri, 17 Oct 2003 12:05:13 +0200a> From: "Winfried Bergmann" <winfried.bergmannNosPAM@empuron.de>! Subject: Re: VMS defeated by wind-9 Message-ID: <bmoepk$p4uro$1@ID-170759.news.uni-berlin.de>   ; "David Froble" <davef@tsoft-inc.com> schrieb im NewsbeitragO& news:3F8EC544.4090105@tsoft-inc.com... > JF Mezei wrote:  >eL > > Tonight,  VMS , the supposedly unstoppable operating system was defeated by( > > wind. It took 8 hours to boot again. > >aL > > Shirley VMS should be able to widthstand a 20cm diameter twig falling on a-2 > > power line and bringing down 3 poles with it ? > > K > > I think that the VMS engineers have overlooked something in their code.t VMSo= > > is too important to be defeated by a mere twig... :-) :-)  > >  >iJ > Well, I'd contend that it's not a VMS problem.  If the power company had usedH > more robust poles and wire the 'twig' would have been less successful. Building8 > a shelter over the lines would also have been prudent. >(G > However, a more farsighted power company would have buried lines with  backhoeo' > stopping reinforced concrete linings.o  G Actually, the power company should have had a backup line, if the break K occoured in a transportation network. The switch over should have been done H automatically by the NA (network analysis) system of the power utilities EMS.   ------------------------------   Date: 17 Oct 2003 06:30 CDTa. From: carl@nospam.gerg.tamu.edu (Carl Perkins)! Subject: Re: VMS defeated by winde4 Message-ID: <17OCT200306304777@nospam.gerg.tamu.edu>  $ John Santos <JOHN@egh.com> writes...7 }I'm not sure if IDE/ATA is considered "legacy" or not.l }--  }John Santos  C Not yet, but in a couple of years it will be. The replacement isn't  SCSI, it is Serial ATA.w   --- Carl   ------------------------------    Date: 17 Oct 2003 07:27:32 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)d! Subject: Re: VMS defeated by windu3 Message-ID: <VpxHEkT8Sj+g@eisner.encompasserve.org>   V In article <3F8EE413.8F55727D@istop.com>, JF Mezei <jfmezei.spamnot@istop.com> writes: > N > There has been a "message" that serial ports are now considered "legacy" forO > some time. And there has been a message that the builder fo hardware on whichc1 > VMS runs wants to build "legacy-free" hardware.  > O > What is the future like with regards to a data interface between a UPS system P > and a host ? Will UPS continue to use serial ports to communicate with a host,J > or are they already switching to USB or eithernet/TCPIP or some other  ?  #     The S in USB stands for serial.b   ------------------------------  % Date: Fri, 17 Oct 2003 09:55:58 -040005 From: "Bochnik, William J" <William_Bochnik@acml.com>/! Subject: RE: VMS defeated by windDQ Message-ID: <2D75787AAF09C64481BDFD89113BE6D502D06852@ac2kama0102.ac.lp.acml.com>   K Third party, but DEC... DIGITAL... Compaq... HP is suing over the copyright " infringment over the interface....   -----Original Message-----< From: Andy Bustamante [mailto:a_c_bustamante@earthlink.net] ( Sent: Thursday, October 16, 2003 9:47 PM To: Info-VAX@Mvb.Saic.Com:! Subject: Re: VMS defeated by winde    K Were the power lines and poles genuine DEC (HP/Compaq) parts or third partye substitutes?   -- Andy Bustamantes remove the ASCII 95s to replym    7 "JF Mezei" <jfmezei.spamnot@istop.com> wrote in messaged# news:3F8E4DF1.1DF32707@istop.com...lB > Tonight,  VMS , the supposedly unstoppable operating system was 
 > defeated by& > wind. It took 8 hours to boot again. > H > Shirley VMS should be able to widthstand a 20cm diameter twig falling 5 > on a power line and bringing down 3 poles with it ?d >tD > I think that the VMS engineers have overlooked something in their  > code.o VMSt; > is too important to be defeated by a mere twig... :-) :-)0      ) -----------------------------------------l The information contained in this transmission may contain privileged and confidential information and is intended only for the use of the person(s) named above. If you are not the intended recipient, or an employee or agent responsible for delivering this message to the intended recipient, any review, dissemination, distribution or duplication of this communication is strictly prohibited. If you are not the intended recipient, please contact the sender immediately by reply e-mail and destroy all copies of  the original message. Please note that we do not accept account orders and/or instructions by e-mail, and therefore will not be responsible for carrying out such orders and/or instructions.d   ------------------------------  % Date: Fri, 17 Oct 2003 11:00:59 -0400 5 From: "Bochnik, William J" <William_Bochnik@acml.com>p! Subject: RE: VMS defeated by wind Q Message-ID: <2D75787AAF09C64481BDFD89113BE6D502D06856@ac2kama0102.ac.lp.acml.com>e  G emphasis on crap though.  Going from proprietary hardware, to commodity J brought down the price of a lot of systems, and made new things available,L this was a good thing.  Using some BAD technology did nothing to promote the systems, however.t   -----Original Message-----( From: John Santos [mailto:JOHN@egh.com] ( Sent: Thursday, October 16, 2003 6:52 PM To: Info-VAX@Mvb.Saic.Comr! Subject: Re: VMS defeated by windt   [snip]I The advantages of these for PC manufacturers are huge.  Since USB devicesuI can identify themselves, autoconfiguration (plug-and-play) is much easieraL and more reliable.  PCI deals with interrupts and DMA infinitely better thanH the retarded ISA bus, and is many times faster.  (Don't know if it is as/ clever a design as the Unibus or Qbus, though.)e  L They only need to put a USB interface chip on the motherboard to replace allI the random things in an old-style PC, and then the customer can configurePL the system as he desires.  (Ever try to put 2 PS2 mice on a PC?)  Everything+ else is just a small matter of programming.   < Getting rid of all the PC crap hardware is a *GOOD* *THING*.   [/snip]L) -----------------------------------------a The information contained in this transmission may contain privileged and confidential information and is intended only for the use of the person(s) named above. If you are not the intended recipient, or an employee or agent responsible for delivering this message to the intended recipient, any review, dissemination, distribution or duplication of this communication is strictly prohibited. If you are not the intended recipient, please contact the sender immediately by reply e-mail and destroy all copies of  the original message. Please note that we do not accept account orders and/or instructions by e-mail, and therefore will not be responsible for carrying out such orders and/or instructions.e   ------------------------------  % Date: Fri, 17 Oct 2003 12:18:36 -0400e* From: JF Mezei <jfmezei.spamnot@istop.com>! Subject: Re: VMS defeated by wind ) Message-ID: <3F90165C.346F07ED@istop.com>a  L The thing about a UPS that connects to a computer via TCPIP/ethernet is thatE you now need to keep your ethernet "powered" by that UPS. (sufficientu! hubs/switches to reach the host).o  I For a large datacentre with a turbine generator and banks of batteries onmK their own floor, this is not an issue. But for a small business, this would4L mean ensuring that all the critical hardware is essentially plugged into theN same plug that is backed up by the UPS. USB and Serial may be superior in that	 respect. g  M Of course, since USB isn't available on VAX (is it available on Alpha-VMS yetn ?), this makes it rather moot.   ------------------------------  # Date: Fri, 17 Oct 2003 16:47:31 GMT-4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk>! Subject: Re: VMS defeated by windm0 Message-ID: <3F901BB6.5A8FDB1C@blueyonder.co.uk>   JF Mezei wrote:   nO > Of course, since USB isn't available on VAX (is it available on Alpha-VMS yetm  > ?), this makes it rather moot.  4 USB to serial adaptors and vice versa are available. --   tim.llewellyn@blueyonder.co.uk   ------------------------------  # Date: Fri, 17 Oct 2003 17:10:15 GMTn" From:   VAXman-  @SendSpamHere.ORG! Subject: Re: VMS defeated by windB0 Message-ID: <00A27838.8032AD61@SendSpamHere.ORG>  V In article <3F90165C.346F07ED@istop.com>, JF Mezei <jfmezei.spamnot@istop.com> writes:M >The thing about a UPS that connects to a computer via TCPIP/ethernet is thatdF >you now need to keep your ethernet "powered" by that UPS. (sufficient" >hubs/switches to reach the host). >iJ >For a large datacentre with a turbine generator and banks of batteries onL >their own floor, this is not an issue. But for a small business, this wouldM >mean ensuring that all the critical hardware is essentially plugged into theSO >same plug that is backed up by the UPS. USB and Serial may be superior in that]
 >respect.  >eN >Of course, since USB isn't available on VAX (is it available on Alpha-VMS yet >?), this makes it rather moot.e  J But... there are USB to ethernet converters.  I saw several advertised in H a PeeCee rag that, for some strange reason, keeps arriving in my mailboxJ addressed to some person named "resident/current occupant".  I am thinkingI about giving one a shot so I can see when I can do with UPShot on APC UPSeJ with USB.  I can support the ethernet now as long as SNMP is a viable sol- ution. -- rL VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)COM            n5   "Well my son, life is like a beanstalk, isn't it?" T   ------------------------------  % Date: Fri, 17 Oct 2003 13:14:18 -0400n* From: JF Mezei <jfmezei.spamnot@istop.com>! Subject: Re: VMS defeated by windi) Message-ID: <3F902368.A0D84DFA@istop.com>V   Tim Llewellyn wrote:6 > USB to serial adaptors and vice versa are available.  L Careful about those. USB is not peer to perr, it is master-slave (one has to provide power).r  J Most USB-Serial adaptors are slaves, expecting to be plugged into a master (eg: a PC).p  3 Having [UPS]---<usb>---[Adaptor]---<serial>---[VAX]i  L May not work if both the UPS and the Adaptor are slaves on the USB. The vastN majority of adaptors are designed to allow you to connect a serial device onto a computer, not the reverse.   ------------------------------    Date: 17 Oct 2003 06:24:07 -0700- From: vpalmboom@vxcompany.com (V.C. Palmboom)" Subject: VMSTAR problemw= Message-ID: <4c0eaea1.0310170524.56738fea@posting.google.com>a   Hello, e  D We've got a problem with VMSTAR V3.3-4. I try to create a tarfile ofF two files and the first time it takes both and the second time it onlyD takes one file.  We think it has something to do with the filenames,
 but why???! Does anyone have an explanation?    . $ tar/create/verb   t.tar   1_00.lis,1_00.txt ) Oct 15 15:19:05 2003        2 ./1_00.lis O $ / $ tar/create/verb   t.tar   a1_00.lis,a1_0.txt r* Oct 15 15:21:53 2003        2 ./a1_00.lis ) Oct 15 15:20:50 2003        2 ./a1_0.txt i  / $ tar/create/verb   t.tar   a1_0.txt,a1_00.lis C) Oct 15 15:20:50 2003        2 ./a1_0.txt i  / $ tar/create/verb   t.tar   a1_00.lis,a1_0.txt e* Oct 15 15:21:53 2003        2 ./a1_00.lis ) Oct 15 15:20:50 2003        2 ./a1_0.txt   $ / $ tar/create/verb   t.tar   a1_0.txt,a1_00.lis w) Oct 15 15:20:50 2003        2 ./a1_0.txt c $ . $ tar/create/verb   t.tar   a_0.txt,a1_00.lis ( Oct 15 15:24:47 2003        2 ./a_0.txt * Oct 15 15:21:53 2003        2 ./a1_00.lis    With kind regards, h   V.C. Palmboom ! System and Network Administrator i9 vpalmboom_NO@SPAM_vxcompany.com (replace _NO@SPAM_ for @)     P --------------------------------------------------------------------------------: "In a world without walls, who needs Windows or Gates..." N ------------------------------------------------------------------------------   ------------------------------  + Date: Fri, 17 Oct 2003 09:53:54 +0000 (UTC)* From: david20@alpha2.mdx.ac.uk Subject: Re: W2K patches) Message-ID: <bmoe7i$bj2$1@news.mdx.ac.uk>t  g In article <3F8EE74A.A432BAC9@blueyonder.co.uk>, Tim Llewellyn <tim.llewellyn@blueyonder.co.uk> writes:d >v >n  >david20@alpha2.mdx.ac.uk wrote: >  >> gN >> Yes but Microsoft is taking away the organisations choice of when to patch.Q >> I'd rather get the patches as soon as possible. Apply them on our test systemsv< >> and then schedule when to apply them on the live systems. >> vP >> With patches only appearing once a month you have a greater number of patchesM >> to test, with probably a greater urgency since the worm writers have knownhN >> about the vulnerabilities longer. You can't just put all the patches on theR >> test systems since if it causes problems you then don't know which patch caused >> the problem.M >> rM >> The organisation will still have to apply multiple patches to fix the sametO >> (or extremely closely related) vulnerability (like the RPC/ Blaster patches)$O >> since until Microsoft release the first patch the security testing companiesa. >> won't have seen what Microsoft have missed. >>  D >> This once a month approach just makes the patch/boot cycle worse. >> e > R >Surely doing all this stuff at least vaguely sensibly and methodically is causing\ >M$ TCO to skyrocket? Even without the colander-as-ship approach to security and reliability9 >people with power surely must notice this at some point.  >l  @ Yes. Which is why Microsoft is coming out with things like this.N Since they can't seem to reduce the number of bugs they just reduce the numberN of patch announcements to once a month. So instead of people talking about the= new bugs once a week they just talk about them once a month.  M Doesn't improve security or TCO of Microsoft products just makes it look likesN problems are occuring less frequently to the casual observer (or probably moreG to the point Senior Management who don't have to directly deal with thee	 patches).a  
 David Webb VMS and Unix team leader CCSS Middlesex University       >--  >tim.llewellyn@blueyonder.co.ukn   ------------------------------  + Date: Fri, 17 Oct 2003 10:23:59 +0000 (UTC)y From: david20@alpha2.mdx.ac.uk Subject: Re: W2K patches) Message-ID: <bmofvv$cb1$1@news.mdx.ac.uk>   V In article <3F8EE697.8D337ABE@istop.com>, JF Mezei <jfmezei.spamnot@istop.com> writes:  >david20@alpha2.mdx.ac.uk wrote:D >> This once a month approach just makes the patch/boot cycle worse. >eG >Nop. Instead of scheduling 5 hours per day of downtime to test/install.M >patches, you will schedule 2 days per month of downtime to test the patches. N >And it is then more likely that Microsoft will release a single monthly patch2 >that contains all the fixes. (if they are smart). >   L At the moment an organisation can decide when it wants to apply the patches.C If it wants to save them all up and apply them once a month it can.a* If it wants to do them once a week it can.  ( Microsoft are removing that flexibility.  H A fixed schedule of a month also means that if Microsoft developers are I creating a patch and for whatever reason they miss the deadline then the o( patch won't come out for another month.       M >My worry about this is when a new virus spreads, will customers have to waitr) >up to one month before getting a patch ?  >eM >My feeling is that customers who purchase some sort of support contract fromnN >Microsoft will get the patches right away, but those without any support will >have to wait one month. >p  2 Haven't seen anything suggesting that is the case.  L >If that is the case, then one should be worried. It could mean that virusesM >will wreak havok on the internet and essentially result in denial of serviceiK >due to overloaded servers. (that is the case witght now to SMTP servers at> >Bell Canada). > O >Of course, on the upside, perhaps ISPs will start denying service to customershW >who run any Microsoft software since they are much more likely to be causing problems.   
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  + Date: Fri, 17 Oct 2003 10:34:10 +0000 (UTC)  From: david20@alpha2.mdx.ac.uk Subject: Re: W2K patches) Message-ID: <bmogj2$cb1$2@news.mdx.ac.uk>e  l In article <zhDjb.739$oP4.13808@news.uswest.net>, "Michael D. Ober" <obermd-.@.-alum-mit-edu-nospam> writes:F >I read the entire posting at Microsoft, and buried in the middle is aB >statement that "emergency" patches will be released out of cycle.F >Basically, there's no change to how MS is releasing critical securityJ >patches.  However, the monthly patches will probably supercede any out ofI >cycle patches that were made, similar to how Windows 2000 Service Pack 4eA >removed almost all the critical updates that came out before it.l >aF >In other words, status quo on the security patches, but mini-security+ >rollups each month for those who can wait.o >aK Depends what they mean by Emergency. If they mean all critical patches theneN yes it won't be much different (though a number of people have complained thatL many of Microsoft's "important" patches should really be critical). However M the way I read it was  - If they were being roasted in the press about a new -H worm already out in the wild then they would release an emergency patch " before the normal monthly release.P So expect to see even more systems being infected since those who were patching G before the hackers could develop the exploit will no longer be able to.   
 David Webb VMS and Unix team leader CCSS Middlesex University     >Mike. >  >a8 >"JF Mezei" <jfmezei.spamnot@istop.com> wrote in message$ >news:3F8EE697.8D337ABE@istop.com..." >> david20@alpha2.mdx.ac.uk wrote:F >> > This once a month approach just makes the patch/boot cycle worse. >>I >> Nop. Instead of scheduling 5 hours per day of downtime to test/installtF >> patches, you will schedule 2 days per month of downtime to test the	 >patches.2J >> And it is then more likely that Microsoft will release a single monthly >patch4 >> that contains all the fixes. (if they are smart). >>J >> My worry about this is when a new virus spreads, will customers have to >waitI+ >> up to one month before getting a patch ?D >>J >> My feeling is that customers who purchase some sort of support contract >fromtK >> Microsoft will get the patches right away, but those without any supportA >willc >> have to wait one month. >>F >> If that is the case, then one should be worried. It could mean that >virusesG >> will wreak havok on the internet and essentially result in denial of  >serviceM >> due to overloaded servers. (that is the case witght now to SMTP servers ate >> Bell Canada). >>G >> Of course, on the upside, perhaps ISPs will start denying service to(
 >customersG >> who run any Microsoft software since they are much more likely to beM >causing problems. >S >R   ------------------------------  + Date: Fri, 17 Oct 2003 10:56:54 +0000 (UTC)R From: david20@alpha2.mdx.ac.uk Subject: Re: W2K patches) Message-ID: <bmohtl$ct3$1@news.mdx.ac.uk>R  a In article <aaE1eSCapY9H@eisner.encompasserve.org>, young_r@encompasserve.org (Rob Young) writes: K >In article <bmmngm$oag$1@news.mdx.ac.uk>, david20@alpha2.mdx.ac.uk writes: d >> In article <bnSk4y4K9inw@eisner.encompasserve.org>, young_r@encompasserve.org (Rob Young) writes:M >>>In article <bmmi2d$mc7$1@news.mdx.ac.uk>, david20@alpha1.mdx.ac.uk writes:Ad >>>> In article <CIEJLCMNHNNDLLOOGNJIIEEJIEAA.tom@kednos.com>, "Tom Linden" <tom@kednos.com> writes:L >>>>>I saw a news story this morning that MS had announced 5 new patches andO >>>>>they further stated that in the future they were going to post new patches 4 >>>>>on the first Tuesday or Wednesday of each month >>>>>a >>>>  R >>>> Yes, Microsoft's great idea. People are complaining about how frequently theyQ >>>> have to patch Microsoft systems. So Microsoft decide they will only release - >>>> patches once a month. - >>>> -	 >>>> See i >>>> iP >>>> http://www.microsoft.com/technet/treeview/default.asp?url=/technet/security >>>> /bulletin/revsbwp.asp >>>>  C >>>> They claim this is in response to extensive customer feedback.-M >>>> I don't know who the hell they talked to but I'd much rather know about 0J >>>> vulnerabilities, workarounds and patches at the earliest opportunity. >>>> a >>>a; >>>	Overheard a conversation that runs along these lines...e >>>e> >>>	"I'm sick of it, I'm recommending we don't put anything on
 >>>	Windows."2 >>>hK >>>	"You mean to tell me we should cluster everything , twice the licenses,o >>>	twice the patches?   >>>3< >>>	"We cluster so we can failover?  That's how you maintain? >>>	application availability?  That's your idea of Enterprise?"  >>>2 >>>--- >>>3I >>>	I know folks that have had to reboot servers after patching twice in  8 >>>	the last two weeks and now will have to do it again. >>>rC >>>	So yes, I bet it is customer feedback driven - very much so.  IbD >>>	know some very angry folks.  Can't speak for them but I'd assume< >>>	(at this point) they'd much rather have it all melt down8 >>>	totally instead of *wasting* their lives (literally)< >>>	in a never-ending patch/boot/patch/boot/patch/boot cycleA >>>	on many servers.  Ask around, it is the same thing with a lot 9 >>>	of Weendoze folks.  It wasn't ever funny - now it is - >>>	disgusting (patch/boot). >>>c >> 3N >> Yes but Microsoft is taking away the organisations choice of when to patch.Q >> I'd rather get the patches as soon as possible. Apply them on our test systemso< >> and then schedule when to apply them on the live systems. >> e > * >	That is you.  I'm assuming a few things. > ? >		1) You don't have hundreds of severs and 4 support personnelMH >                2) Because of 1) you aren't performing many patches andF >                   reboots in your spare time (dozens/hundreds of man >			hours).  >   J We have tons and tons of servers and desktops. We are a University and ourK department directly looks after the majority of machines in the University.,G We are using Tivoli and SUS to automate pushing patches out on the live L systems. But before doing that we need to do lots of testing on our test and reference network.         > >	If Microsoft goes to a monthly model, when/if things go veryF >	bad the support personnel can say:  "We haven't received that patch"> >	and it will be the truth.  As it stands now (and what I haveD >	personally seen) is the patch becomes available , you schedule theC >	patch for a weekend and a worm comes in and wrecks havoc.  SeniorlH >	management says:  "Why wasn't the patch applied?"  "Exhaustion" is an E >	inadequate answer.  Now you must patch CRITICAL as soon as you can. - >	So if a monthly cycle comes along, not onlye= >	can you plan for it but you can also quit becoming a Zombiea9 >	(work during the day, patch at night) in many cases andlC >	maybe miss a patch cycle or two.  In other words, from a support -1 >	standpoint it is a very good thing.  BUT as yourB >	point out it is a very bad thing as it stretches your timelines. > C >	Anyhow, if the patch is of a CRITICAL nature it will get releasedN@ >	immediately.  Hopefully, the only thing this new release cycle5 >        does do is get management calmed down a bit.m >                        >	P >> With patches only appearing once a month you have a greater number of patchesM >> to test, with probably a greater urgency since the worm writers have knowneN >> about the vulnerabilities longer. You can't just put all the patches on theR >> test systems since if it causes problems you then don't know which patch caused >> the problem.    >y9 >	But the issue is exhaustion.  Ask around.  I know folksfB >	that are exhausted.  Sure you are stretching timelines with your= >	testing cycle on top of monthly releases of patches, but it09 >	is a resource issue.  Patch Mules.  I saw a Dice Ad foru  >	a Patch Mule (just kidding).   >a   Thats bad organisation.pM There is nothing to stop a company deciding that they will only apply patchesaH once a month, once a week or whatever other period they want. If senior N management then decides that critical patches aren't being applied frequently M enough then it is upto senior management to supply extra resource to allow a s. more frequent patching schedule to be applied.H Microsoft releasing patches only once a month won't reduce the number ofG patches since the hackers will still find the holes just as frequently.   I Microsoft releasing patches once a month won't solve the exhaustion issueoN just reduce the flexibility of when companies can apply the patches which willO probably result in even more exhaustion as they try to apply then as quickly asp# possible after the monthly release.M             >	A crazy issue, eh? >  >> iM >> The organisation will still have to apply multiple patches to fix the same.P >> (or extremely closely related) vulnerability (like the RPC/ Blaster patches) P >> since until Microsoft release the first patch the security testing companies . >> won't have seen what Microsoft have missed. >>  D >> This once a month approach just makes the patch/boot cycle worse. >> / >                               @ >	Worse?  Absolutely.  But at least your staff won't be quittingB >	because they have had enough and want a life.  So when MicrosoftA >	says customer demand, I suspect they mostly interviewed/queriedxB >	support personnel THAT DESPERATLEY NEEDS THE PATCH TIMELINES TO	J >        LENGTHEN OR BECOME PREDICTABLE INSTEAD OF AD HOC.  Certainly not  >	management's wishes. >g; >	A closing ... these are my observations, I'm not patchingoC >	but have had conversations.  I've been wrong in the past, I couldlB >        be off the mark on this subject too.  But this is USENET. >  >				Rob >i    
 David Webb VMS and Unix team leader Security team leader CCSS Middlesex University   ------------------------------  + Date: Fri, 17 Oct 2003 11:02:18 +0000 (UTC)S From: david20@alpha2.mdx.ac.uk Subject: Re: W2K patches) Message-ID: <bmoi7q$ct3$2@news.mdx.ac.uk>   a In article <DtKaXNLlDitQ@eisner.encompasserve.org>, young_r@encompasserve.org (Rob Young) writes: W >In article <3F8EE697.8D337ABE@istop.com>, JF Mezei <jfmezei.spamnot@istop.com> writes: " >> david20@alpha2.mdx.ac.uk wrote:E >>> This once a month approach just makes the patch/boot cycle worse.  >> 0I >> Nop. Instead of scheduling 5 hours per day of downtime to test/installtO >> patches, you will schedule 2 days per month of downtime to test the patches.tP >> And it is then more likely that Microsoft will release a single monthly patch4 >> that contains all the fixes. (if they are smart). >> dO >> My worry about this is when a new virus spreads, will customers have to waitr+ >> up to one month before getting a patch ?  >> s >b? >	No - from what I understand that will be a critical patch ande >        released immediately. >c >				Rob  O If they are still going to realease critical patches as they currently do then  K this announcement is pointless - just look at the last few months Microsoft G patches the majority have been critical. Of the rest a fair number were N classified as important and many people from the testing companies have arguedB that many of those should really have been classified as critical.  I My reading is that patches will only be released out of cycle if there is.L already a worm exploiting the vulnerability out in the wild and Microsoft is' being crucified in the computing press.   
 David Webb VMS and Unix team leader CCSS Middlesex University   >E   ------------------------------  % Date: Fri, 17 Oct 2003 12:26:28 +0100iO From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com>- Subject: Re: W2K patches/ Message-ID: <bmojlk$5r$1@new-usenet.uk.sun.com>a   Tim Llewellyn wrote: > ! > david20@alpha2.mdx.ac.uk wrote:  >  U > M >>Yes but Microsoft is taking away the organisations choice of when to patch.nP >>I'd rather get the patches as soon as possible. Apply them on our test systems; >>and then schedule when to apply them on the live systems.  >>O >>With patches only appearing once a month you have a greater number of patchesSL >>to test, with probably a greater urgency since the worm writers have knownM >>about the vulnerabilities longer. You can't just put all the patches on thelQ >>test systems since if it causes problems you then don't know which patch causeds >>the problem. >>L >>The organisation will still have to apply multiple patches to fix the sameN >>(or extremely closely related) vulnerability (like the RPC/ Blaster patches)N >>since until Microsoft release the first patch the security testing companies- >>won't have seen what Microsoft have missed.9 >>C >>This once a month approach just makes the patch/boot cycle worse.d >> >  >   S > Surely doing all this stuff at least vaguely sensibly and methodically is causingt] > M$ TCO to skyrocket? Even without the colander-as-ship approach to security and reliability : > people with power surely must notice this at some point. >     ; I think there is an increased appreciation of this in terms " of the effect it is having on TCO.  @ The customer I am working for is currently trialing Java DesktopB System from Sun as an alternative to Windows/Office partly because) of the cost/patching/vunerability issues.   F Most people used to rely on being able to assemble and test a completeB bundle of patches for their Wintel servers on say a quarterly/half
 yearly basis.e  F Now they are having to test and deply single patches almost on demand.  D One of the weekly IT mags in the UK recently published estimates forB the cost of this per patch per seat based on NTbugtrack data which# started at $300 per patch per seat.   @ This clearly has to be factored into windows TCO and many people+ that I talk to are increasingly doing this.a   Regardst Andrew Harrisono   ------------------------------    Date: 17 Oct 2003 07:23:33 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)  Subject: Re: W2K patches3 Message-ID: <rj+EkO+qV0gP@eisner.encompasserve.org>l  _ In article <CMqdnZP8xOpoYhOiU-KYvg@mpowercom.net>, "Jack Peacock" <peacock@simconv.com> writes:r  N > Yes Windows is a sorry excuse for an OS, yes the patches are ridiculous, butK > then no one ever had 100 million plus users for an OS before.  I'm not at:M > all convinced VMS wouldn't have similar problems if it had to meet the sameDN > hardware and services criteria as Microsoft.  If VMS Engineering expanded toL > 500 times it's current size could they guarantee the same quality level asI > now?  We'll never know, because we aren't going to see VMS and "expand" + > occur in the same line in any news story.o  F    What kind of non-logic is that?  Since when does making more copiesF    of any software wear down it's reliability?  Either the reliabilityG    is in copy number 1 or it isn't.  And selling a great many copies inP>    no ways alters the size of the staff needed to engineer it.  H    The biggest problem with your argument is that it assumes Microsoft'sC    problems are associated with it's ranking amoungst OS vendors byaH    volume.  There was a time when, if we had a software system to do, weG    went out and bought a VAX.  VMS didn't have quality problems then in-F    large part because VMS engineering cared, and so did the folks they    answered to.n  F    I know people who worked for Microsoft recently.  They do care, butB    the wrong way.  Thier business model assumes they are competingH    primarily against product they shipped last year.  Quality means not H    selling a replacement next year, so the presure is on them not to do     quality.n   ------------------------------    Date: 17 Oct 2003 08:54:25 -0600+ From: young_r@encompasserve.org (Rob Young)  Subject: Re: W2K patches3 Message-ID: <uSjURDDq2wfH@eisner.encompasserve.org>a  J In article <bmoi7q$ct3$2@news.mdx.ac.uk>, david20@alpha2.mdx.ac.uk writes:c > In article <DtKaXNLlDitQ@eisner.encompasserve.org>, young_r@encompasserve.org (Rob Young) writes:eX >>In article <3F8EE697.8D337ABE@istop.com>, JF Mezei <jfmezei.spamnot@istop.com> writes:# >>> david20@alpha2.mdx.ac.uk wrote: F >>>> This once a month approach just makes the patch/boot cycle worse. >>> J >>> Nop. Instead of scheduling 5 hours per day of downtime to test/installP >>> patches, you will schedule 2 days per month of downtime to test the patches.Q >>> And it is then more likely that Microsoft will release a single monthly patch 5 >>> that contains all the fixes. (if they are smart).r >>> P >>> My worry about this is when a new virus spreads, will customers have to wait, >>> up to one month before getting a patch ? >>>  >>@ >>	No - from what I understand that will be a critical patch and >>        released immediately.s >>	 >>				Rob- > Q > If they are still going to realease critical patches as they currently do then eM > this announcement is pointless - just look at the last few months Microsoft-I > patches the majority have been critical. Of the rest a fair number werelP > classified as important and many people from the testing companies have arguedD > that many of those should really have been classified as critical.  = 	Right.  So maybe the patch cycle shrinks by one go round ando@ 	the whole point of this announcement.  I'm relying on a source,> 	I take it he is right that CRITICAL get released ASAP.  It is 	what he does day to day.?   				Rob    ------------------------------    Date: 17 Oct 2003 08:56:28 -0600+ From: young_r@encompasserve.org (Rob Young)  Subject: Re: W2K patches3 Message-ID: <G0mumUbBwfbM@eisner.encompasserve.org>r  q In article <rj+EkO+qV0gP@eisner.encompasserve.org>, koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:la > In article <CMqdnZP8xOpoYhOiU-KYvg@mpowercom.net>, "Jack Peacock" <peacock@simconv.com> writes:w > O >> Yes Windows is a sorry excuse for an OS, yes the patches are ridiculous, butnL >> then no one ever had 100 million plus users for an OS before.  I'm not atN >> all convinced VMS wouldn't have similar problems if it had to meet the sameO >> hardware and services criteria as Microsoft.  If VMS Engineering expanded to.M >> 500 times it's current size could they guarantee the same quality level asnJ >> now?  We'll never know, because we aren't going to see VMS and "expand", >> occur in the same line in any news story. > H >    What kind of non-logic is that?  Since when does making more copiesH >    of any software wear down it's reliability?  Either the reliabilityI >    is in copy number 1 or it isn't.  And selling a great many copies ina@ >    no ways alters the size of the staff needed to engineer it. > J >    The biggest problem with your argument is that it assumes Microsoft'sE >    problems are associated with it's ranking amoungst OS vendors byeJ >    volume.  There was a time when, if we had a software system to do, weI >    went out and bought a VAX.  VMS didn't have quality problems then ineH >    large part because VMS engineering cared, and so did the folks they >    answered to.a > H >    I know people who worked for Microsoft recently.  They do care, butD >    the wrong way.  Thier business model assumes they are competingJ >    primarily against product they shipped last year.  Quality means not J >    selling a replacement next year, so the presure is on them not to do 
 >    quality.  >   ? 	This point isn't to be overlooked.  The worst thing that could C 	happen from a sales standpoint is improved quality.  Organizations<B 	could stay on Microsoft 2005 an extra year or two without a care,* 	that must be a frightening thought to MS.   ------------------------------  % Date: Fri, 17 Oct 2003 08:23:11 -0700m* From: "Jack Peacock" <peacock@simconv.com> Subject: Re: W2K patches2 Message-ID: <49SdnbTAlef9lA2iU-KYuA@mpowercom.net>  H "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in message- news:rj+EkO+qV0gP@eisner.encompasserve.org...:H >    What kind of non-logic is that?  Since when does making more copiesH >    of any software wear down it's reliability?  Either the reliabilityI >    is in copy number 1 or it isn't.  And selling a great many copies inw@ >    no ways alters the size of the staff needed to engineer it. >MH My point is that VMS has it easy, closed hardware, small but technicallyH astute customer base and a revenue cycle based on gradual updates.  WhatK would happen if VMS has to support as many varied systems as Windows, and a K customer base that did not have in depth experience and training?  It isn'tiK a matter of distributing multiple copies, its making sure those copies workgL on every possible combination of hardware and 3rd party apps, for people whoI don't even understand the need to back up.  That does take a large staff.f  G A good example is IDE support on VMS.  Windows has to support *all* theyK chipset variations, while VMS choose to offer almost no support for any.  IlF have an Alphastation 400 with an IDE bus that is an order of magnitudeK slower on VMS than contemporary PCs with Windows.  Why?  Because VMS didn'tIL have the resources or inclination to do a real DMA capable device driver (onJ their own hardware!).  It's not an OS bug, but it is a hole in the vauntedH VMS quality.  Microsoft doesn't blow off customers like that.  Maybe youK would argue differently, but I'd point to the fact that Windows as a viableo< product is far healthier, even with it's problems, than VMS.  J VMS had a clean start, good design, and 3rd party software had to fit in aC nice structure that was well-documented.  Windows has none of those8E advantages.  It grew out of crude hardware and an economic model thatgH demanded low cost over design elegance.   Should it be different?  In anG ideal world yes, but economics didn't work out that way.  PCs don't runnL TOPS, they don't run Multics, and they don't run VMS.  Linux has some of the advantages of VMS as far as     Jack Peacock    ------------------------------  + Date: Fri, 17 Oct 2003 16:26:45 +0000 (UTC)b From: david20@alpha1.mdx.ac.uk Subject: Re: W2K patches) Message-ID: <bmp585$ios$1@news.mdx.ac.uk>s  a In article <uSjURDDq2wfH@eisner.encompasserve.org>, young_r@encompasserve.org (Rob Young) writes:sK >In article <bmoi7q$ct3$2@news.mdx.ac.uk>, david20@alpha2.mdx.ac.uk writes: d >> In article <DtKaXNLlDitQ@eisner.encompasserve.org>, young_r@encompasserve.org (Rob Young) writes:Y >>>In article <3F8EE697.8D337ABE@istop.com>, JF Mezei <jfmezei.spamnot@istop.com> writes: $ >>>> david20@alpha2.mdx.ac.uk wrote:R >> If they are still going to realease critical patches as they currently do then N >> this announcement is pointless - just look at the last few months MicrosoftJ >> patches the majority have been critical. Of the rest a fair number wereQ >> classified as important and many people from the testing companies have argued E >> that many of those should really have been classified as critical.n >d> >	Right.  So maybe the patch cycle shrinks by one go round andA >	the whole point of this announcement.  I'm relying on a source,b? >	I take it he is right that CRITICAL get released ASAP.  It isi >	what he does day to day. >  >				Rob >n  : Sorry Rob I'm having trouble understanding the sentence :-  4 So maybe the patch cycle shrinks by one go round and% the whole point of this announcement.o      
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  % Date: Fri, 17 Oct 2003 13:00:33 -0400l* From: JF Mezei <jfmezei.spamnot@istop.com> Subject: Re: W2K patches) Message-ID: <3F902030.855A3634@istop.com>1   Jack Peacock wrote:F > Windows has none of those.G > advantages.  It grew out of crude hardware and an economic model thati) > demanded low cost over design elegance.n    J Consider the MAC (prior to OS-X). Right since 1984, they had established a4 documented API that has been supported ever since.    M And during the short period where clones were allowed, because the specs were G clearly laid out with clear standards, the clone makers didn't have anybF problems. I am writing this on a clone and it doesn't need any specialF proprietary drivers. (Unlike Compaq wintel boxes that needed their own8 proprietary drivers which often conflicted with others).  L Had Microsoft provided, right from the start, similar standards for hardwareD and for software, you wouldn't have the total mess they are in now.   M I have no problems with VMS requiring a certain configuration, and supporting S only certain named devices. This way, one can better control quality and stability.   K This doesn't mean that you can't write your own driver for your own specialaN device. And if the specs are clearly documented on how to write a driver, then< you shouldn't have any problems writing one which is robust.   ------------------------------    Date: 17 Oct 2003 12:15:40 -0600+ From: young_r@encompasserve.org (Rob Young)e Subject: Re: W2K patches3 Message-ID: <souJQnD0H9gE@eisner.encompasserve.org>a  J In article <bmp585$ios$1@news.mdx.ac.uk>, david20@alpha1.mdx.ac.uk writes:c > In article <uSjURDDq2wfH@eisner.encompasserve.org>, young_r@encompasserve.org (Rob Young) writes:dL >>In article <bmoi7q$ct3$2@news.mdx.ac.uk>, david20@alpha2.mdx.ac.uk writes:e >>> In article <DtKaXNLlDitQ@eisner.encompasserve.org>, young_r@encompasserve.org (Rob Young) writes:rZ >>>>In article <3F8EE697.8D337ABE@istop.com>, JF Mezei <jfmezei.spamnot@istop.com> writes:% >>>>> david20@alpha2.mdx.ac.uk wrote: S >>> If they are still going to realease critical patches as they currently do then rO >>> this announcement is pointless - just look at the last few months MicrosoftlK >>> patches the majority have been critical. Of the rest a fair number were R >>> classified as important and many people from the testing companies have arguedF >>> that many of those should really have been classified as critical. >>? >>	Right.  So maybe the patch cycle shrinks by one go round andgB >>	the whole point of this announcement.  I'm relying on a source,@ >>	I take it he is right that CRITICAL get released ASAP.  It is >>	what he does day to day.- >>	 >>				Robo >> > < > Sorry Rob I'm having trouble understanding the sentence :- > 6 > So maybe the patch cycle shrinks by one go round and' > the whole point of this announcement.  >    	That wasn't very good was it.  = 	Microsoft talks about moving to a monthly cycle for patches.f= 	BUT critical patches will be released immediately.  So maybei< 	when you had a patch come in that wasn't critical you would= 	apply it on the weekends.  Now instead of every weekend, yout; 	patch monthly UNLESS a critical patch comes in and you mayr: 	be back to patching every weekend.  The monthly cycle may: 	or may not allow you to skip a weekend or skip a go round 	on the patch merry-go-round.    				RobS 				   ------------------------------  # Date: Fri, 17 Oct 2003 14:55:14 GMTg) From: "Rick Barry" <richard.barry@hp.com> ! Subject: Re: What is ACME_SERVER? 2 Message-ID: <mpTjb.7268$pZ5.2430@news.cpqcorp.net>  7 "JF Mezei" <jfmezei.spamnot@istop.com> wrote in messageo# news:3F8E4C9A.406992AE@istop.com...g > The real question should be: >@F > Did the VMS engineers show they were significantly influenced by theJ > RoadRunner show in their life when they named it the ACME_SERVER instead of > ACM_SERVER ??????? >   L For those of you who are wondering, the ACME facility was chosen because ACM was already taken.  J The ACME_SERVER process starts as part of standard VMS startup since V7.3.J It services SYS$ACM requests. Prior to version 7.3-1, the DCOM applicationL was the only SYS$ACM consumer. In version 7.3-1, SYS$ACM became a documentedI interface for any application that wanted to do user authentication usingcG either standard SYSYUAF.DAT username and password or Microsoft NT-style19 authentication (provided by the Advanced Server product).i  I OpenVMS 7.3-2 introduces an ACME SDK for those who would like to evaluate K the ACME agent interface to ACME_SERVER (see SYS$HELP:ACME_DEV_README.TXT)./F This is an early adopter kit. Some future OpenVMS version will supportH user-written ACME agents and SYS$ACM applications will transparently use them for user authentication.a  I There is an LDAP ACME agent in the ACME SDK that demonstrates how an LDAP_L directory can be used as a common authenticate database for a network of VMS systems.  E What do you think came first: the ACME acronym or "Authentication and-" Credential Management Extensions"?  ? Let it not be said that engineers do not have a sense of humor.-  
 Rick Barry OpenVMS Engineering5 Hewlett-Packard Companyi	 Nashu, NHm   ------------------------------  % Date: Fri, 17 Oct 2003 12:46:52 -0400k* From: JF Mezei <jfmezei.spamnot@istop.com>! Subject: Re: What is ACME_SERVER?l) Message-ID: <3F901CFA.4ABECF7E@istop.com>v   Rick Barry wrote: G > What do you think came first: the ACME acronym or "Authentication andh$ > Credential Management Extensions"?  ) RoadRunner show predates VMS as I recall.e   ------------------------------  % Date: Fri, 17 Oct 2003 06:00:46 -0400 * From: "Bill Todd" <billtodd@metrocast.net>. Subject: Re: What is the performance of iVMS ?2 Message-ID: <VeOdnQW3jIFXIBKiU-KYvA@metrocast.net>  3 "Rick Jones" <foo@bar.baz.invalid> wrote in messagei, news:idFjb.7235$ie5.5393@news.cpqcorp.net...- > JF Mezei <jfmezei.spamnot@istop.com> wrote:p > > Fred Kleinsorge wrote: > >>= > >> THIS ISN'T A BETA.  THIS ISN'T EVEN AN ALPHA.  THIS IS AsE > >> PRE-RELEASE, UNTUNED, KICK-THE-TIRES, "HEY COOL LOOK - IT'S JUSTc* > >> VMS AFTER ALL" - PEEK AT OPENVMS I64. >aF > > Sorry, from my point of view, based on comments I read here, it is > > still very much beta:> >nB > I may be wrong, but I get the impression that Fred was trying toG > convey that the software was even more preliminary than beta, or even F > alpha.  Perhaps there is a disagreement in what each of you consider, > the meaning/implications of "pre-release?"  D You've got to admit that it's at least confusing.  I mean, don't youL *already* have a so-called VMS 8.0 (Itanic-specific) 'release' in the field,J limited though it may be?  And wouldn't one naturally assume that the testI system was running something at least somewhat related to that 'release'?i   - bill   ------------------------------  % Date: Fri, 17 Oct 2003 11:43:26 +0100c* From: "Richard Brodie" <R.Brodie@rl.ac.uk>. Subject: Re: What is the performance of iVMS ?+ Message-ID: <bmoh4g$oma@newton.cc.rl.ac.uk>e  b "Bill Todd" <billtodd@metrocast.net> wrote in message news:VeOdnQW3jIFXIBKiU-KYvA@metrocast.net...  F > You've got to admit that it's at least confusing.  I mean, don't youN > *already* have a so-called VMS 8.0 (Itanic-specific) 'release' in the field,L > limited though it may be?  And wouldn't one naturally assume that the testK > system was running something at least somewhat related to that 'release'?a  J I don't find it confusing. The first production release will be 8.2, whichL clearly isn't alpha yet; anything available now will obviously predate that.I That VMS engineering have cut a branch and tagged it 8.0 isn't a big dealn to me.   ------------------------------  # Date: Fri, 17 Oct 2003 13:24:28 GMTs2 From: "Robert Boers" <Robert.boers@softresint.com>A Subject: Re: White paper: CHARON-VAX cluster w/ shared SCSI disksp* Message-ID: <3f8fed89$1@news.deckpoint.ch>  1 "Michael Unger" <unger@decus.de> wrote in message 3 news:bmmj9m$nto2l$2@ID-152801.news.uni-berlin.de...t- > On 2003-10-15 23:03, "Wilm Boerhout" wrote:t   [...]aE > "VMS on IA-32" isn't the problem, but "Charon-VAX on *Windows*" ...t  K As might have shown from an earlier message, CHARON-VAX uses Windows mainlyvB as program loader, and one can disable unused services. Windows isH remarkably stable if it is not used. We had our local demo system up for6 over a year before it was time for a hardware upgrade.  L CHARON-VAX/AXP on an OpenVMS/Alpha host is a good alternative. One can build" nice clusters on a Marvel as well.   Regards, Robertt   ------------------------------  % Date: Fri, 17 Oct 2003 17:37:59 +0200n" From: Didier Morandi <no@spam.com>A Subject: Re: White paper: CHARON-VAX cluster w/ shared SCSI disksI4 Message-ID: <3f900cdd$0$27018$626a54ce@news.free.fr>   Robert Boers wrote:w  1 > Windows is remarkably stable if it is not used.t   I love it...   D.   ------------------------------  % Date: Fri, 17 Oct 2003 18:19:14 +0200a4 From: Michael Unger <db6st.rev.0131a55a@t-online.de>A Subject: Re: White paper: CHARON-VAX cluster w/ shared SCSI disksc9 Message-ID: <bmp4tu$peovt$1@ID-152801.news.uni-berlin.de>n  * On 2003-10-17 15:24, "Robert Boers" wrote:  3 > "Michael Unger" <unger@decus.de> wrote in message-5 > news:bmmj9m$nto2l$2@ID-152801.news.uni-berlin.de...U. >> On 2003-10-15 23:03, "Wilm Boerhout" wrote: >  > [...]sF >> "VMS on IA-32" isn't the problem, but "Charon-VAX on *Windows*" ... > M > As might have shown from an earlier message, CHARON-VAX uses Windows mainly:D > as program loader, and one can disable unused services. Windows isD                                                           ^^^^^^^^^^J > remarkably stable if it is not used. We had our local demo system up for%   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^i8 > over a year before it was time for a hardware upgrade.   ;-)s  N > CHARON-VAX/AXP on an OpenVMS/Alpha host is a good alternative. One can build$ > nice clusters on a Marvel as well. >  > [...]s  D Agreed. But an AlphaServer *and* an additional license to run VMS as6 "hosting OS" will probably be a bit more expensive ...  A I still don't understand why -- given the Intel based approach -- B Winwoes is used as the "hosting OS". I'd expected a small and fastD (real-time) OS as "program loader", perhaps even using the "primary"8 disk and network controllers instead of additional ones.   Michaela   -- n; Real names enhance the probability of getting real answers. @ Please do *not* send "Security Patch Notifications" or "SecurityA Updates"; this system isn't running a Micro$oft operating system. = And don't annoy me <mailto:postmaster@[127.0.0.1]> please ;-)e   ------------------------------    Date: 17 Oct 2003 12:18:07 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)tA Subject: Re: White paper: CHARON-VAX cluster w/ shared SCSI disksd3 Message-ID: <C7lJj5s4yZAZ@eisner.encompasserve.org>h  p In article <bmp4tu$peovt$1@ID-152801.news.uni-berlin.de>, Michael Unger <db6st.rev.0131a55a@t-online.de> writes: > C > I still don't understand why -- given the Intel based approach --tD > Winwoes is used as the "hosting OS". I'd expected a small and fastF > (real-time) OS as "program loader", perhaps even using the "primary": > disk and network controllers instead of additional ones.  C    Because people have them, and can get them cheaply.  If you wantrF    it to run on VxWorks or something, try to make a case to the ownersJ    of Charon-VAX that this would result in more sales than the one running    on Alpha VMS.  +    Have you priced good embedded OS lately?n   ------------------------------    Date: 17 Oct 2003 06:03:23 -0700, From: garycassidy@hotmail.com (Gary Cassidy)* Subject: Writing flat text file from DEC C= Message-ID: <d4495c56.0310170503.6e03a470@posting.google.com>c  F I am writing a small program in DEC C that will read the contents of aE sequential RMS file on a VAX and write each record in flat ASCII text>E format, such that the text file can be imported into a database table  in Windows.l  F I am dealing with 400GB of data and fast performance is essential. TheF method of reading the RMS file is set in stone, so I am looking to theB text file writing to make performance gains. I have heard that theA parameters supplied to the fopen() command can make a significantbD difference to the writing speed. I am currently using the command as follows:  C fpu = fopen (upd_par_file_name, "w", "rfm=var", "rat=cr","shr=nil",y5 "fop=tef", "deq=500", "mbf=5", "mbc=127", "ctx=rec");i  B Do these parameters look OK? Are there others I could specify that: might help, or should I take any out that may be hindering performance?  ; Any thoughts on the parameters I am using, or other helpfulg! information, greatly appreciated.s   ------------------------------  % Date: Fri, 17 Oct 2003 14:46:30 +0100e0 From: Chris Sharman <chris.sharman@sorry.nospam>. Subject: Re: Writing flat text file from DEC C4 Message-ID: <bmorrm$7nj$1$8302bc10@news.demon.co.uk>   Gary Cassidy wrote: H > I am writing a small program in DEC C that will read the contents of aG > sequential RMS file on a VAX and write each record in flat ASCII text.G > format, such that the text file can be imported into a database tableu
 > in Windows.a > H > I am dealing with 400GB of data and fast performance is essential. TheH > method of reading the RMS file is set in stone, so I am looking to theD > text file writing to make performance gains. I have heard that theC > parameters supplied to the fopen() command can make a significantaF > difference to the writing speed. I am currently using the command as
 > follows: > E > fpu = fopen (upd_par_file_name, "w", "rfm=var", "rat=cr","shr=nil",e7 > "fop=tef", "deq=500", "mbf=5", "mbc=127", "ctx=rec");  > D > Do these parameters look OK? Are there others I could specify that< > might help, or should I take any out that may be hindering > performance? > = > Any thoughts on the parameters I am using, or other helpfulw# > information, greatly appreciated.n  4 Are you changing the data, or merely the structure ?8 Would convert (or even set file/attr) serve you better ?/ And why is the method of reading set in stone ? > write-behind (and read-ahead) allow better pipelining of data.   Chris    ------------------------------   Date: 17 Oct 2003 11:22 CDTa. From: carl@nospam.gerg.tamu.edu (Carl Perkins). Subject: Re: Writing flat text file from DEC C4 Message-ID: <17OCT200311223743@nospam.gerg.tamu.edu>  0 garycassidy@hotmail.com (Gary Cassidy) writes...G }I am writing a small program in DEC C that will read the contents of anF }sequential RMS file on a VAX and write each record in flat ASCII textF }format, such that the text file can be imported into a database table }in Windows. } G }I am dealing with 400GB of data and fast performance is essential. The1G }method of reading the RMS file is set in stone, so I am looking to the C }text file writing to make performance gains. I have heard that the B }parameters supplied to the fopen() command can make a significantE }difference to the writing speed. I am currently using the command as 	 }follows:a } D }fpu = fopen (upd_par_file_name, "w", "rfm=var", "rat=cr","shr=nil",6 }"fop=tef", "deq=500", "mbf=5", "mbc=127", "ctx=rec"); } C }Do these parameters look OK? Are there others I could specify thato; }might help, or should I take any out that may be hindering 
 }performance?t } < }Any thoughts on the parameters I am using, or other helpful" }information, greatly appreciated.  H You should probably add "rop=wbh" to activate "write behind" processing.  G You should probably add an "alq=" to specify an initial disk allocationeH that is near the size of the expected output file. The maximum it can be set to is  4,294,967,295.'  H You might increase the  "deq=" to something larger, to grab biger chunksJ each time the file is extended. This is in disk blocks which are 512 bytesF in size, so 500 = 250KB. A setting of 8192 would be 4MB. 65,535 is the maximum this can be set to.a  J In some cases a slightly higher "mbf" will help. Up to 8 has been known to be helpfull.  ? There is some slight chance that "fop=sqo" might help a little.o   --- Carl   ------------------------------  % Date: Fri, 17 Oct 2003 12:40:49 -0400t* From: JF Mezei <jfmezei.spamnot@istop.com>. Subject: Re: Writing flat text file from DEC C) Message-ID: <3F901B8F.95BCFD0C@istop.com>    Gary Cassidy wrote:oE > fpu = fopen (upd_par_file_name, "w", "rfm=var", "rat=cr","shr=nil",g7 > "fop=tef", "deq=500", "mbf=5", "mbc=127", "ctx=rec");   I 1- Once your text file has been created, how will it be transfered to thea windows application ?s  L 2- If you have an indexed file on VMS and want it as a text file on Windows,J there is a very simple solution: Use FTP in text/ascii mode. VMS will thenL read the records in order of their primary keys, and windows will store them2 in whatever text file format is native to windows.C (Kermit can also be used much the same way, perhaps more reliably).s  N 3-If you need all the data and you don't need to extract only certain records,L  then the CONVERT utility will be your friend. Not only will you not need toP write a program, but CONVERT will be more efficient than your average C program.  + In terms of the parameters in your example:e? 	-you are creating a file in native VMS file format. (rfm=var). J 	-ctx=rec is needed when you need the file to be shared. Not 100% sure you
 need it here.nK 	  I *assume* that if you don't have it, you may allow VMS to use bufferingoR more efficiently. On the other hand, it may be necessary in the context of rfm=var  L If you really wanted to have utmost efficiency, you might istead want to useL RMS calls to write to the file instead of the C rtl ones. (SYS$OPEN, SYS$PUT SYS$CLOSE )c  M HELP RMS at the dollar promot gives you an idea. And the RMS manual gives youk all the details.   ------------------------------  % Date: Sat, 18 Oct 2003 01:09:03 +0800T, From: Paul Repacholi <prep@prep.synonet.com>. Subject: Re: Writing flat text file from DEC C- Message-ID: <87llrjhjv4.fsf@prep.synonet.com>a  2 Chris Sharman <chris.sharman@sorry.nospam> writes:   > Gary Cassidy wrote:e  D >> I am writing a small program in DEC C that will read the contentsB >> of a sequential RMS file on a VAX and write each record in flatD >> ASCII text format, such that the text file can be imported into a >> database table in Windows.t  # What data conversion are you doing?d  : >> I am dealing with 400GB of data and fast performance isF >> essential. The method of reading the RMS file is set in stone, so IC >> am looking to the text file writing to make performance gains. ICE >> have heard that the parameters supplied to the fopen() command can	E >> make a significant difference to the writing speed. I am currentlyh  >> using the command as follows:  F >> fpu = fopen (upd_par_file_name, "w", "rfm=var", "rat=cr","shr=nil",8 >> "fop=tef", "deq=500", "mbf=5", "mbc=127", "ctx=rec");  E >> Do these parameters look OK? Are there others I could specify that-= >> might help, or should I take any out that may be hindering  >> performance?m  E As David says, add write behind to the output file, and read ahead to$A in input file. Also prealocate the output file to be a bit biggereG that the input so you do not need to extend, or at least use a VERY big8 initial extent.t  > >> Any thoughts on the parameters I am using, or other helpful$ >> information, greatly appreciated.  = Read from one drive, write to another is the BIG win! No head  contention.a  E > Are you changing the data, or merely the structure ?  Would convert F > (or even set file/attr) serve you better ?  And why is the method ofD > reading set in stone ?  write-behind (and read-ahead) allow better > pipelining of data.i  8 You should be able to go at over 90% of raw drive speed.   -- s< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.o@                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------   Date: 17 Oct 03 10:28:54 PST From: mckinneyj@cpva.saic.comi. Subject: Re: Writing flat text file from DEC C( Message-ID: <jV1JdxiO1UoQ@cpva.saic.com>  = In article <d4495c56.0310170503.6e03a470@posting.google.com>,i/  garycassidy@hotmail.com (Gary Cassidy) writes: H > I am writing a small program in DEC C that will read the contents of aG > sequential RMS file on a VAX and write each record in flat ASCII textpG > format, such that the text file can be imported into a database tabley
 > in Windows.- > H > I am dealing with 400GB of data and fast performance is essential. TheH > method of reading the RMS file is set in stone, so I am looking to theD > text file writing to make performance gains. I have heard that theC > parameters supplied to the fopen() command can make a significant F > difference to the writing speed. I am currently using the command as
 > follows: > E > fpu = fopen (upd_par_file_name, "w", "rfm=var", "rat=cr","shr=nil",t7 > "fop=tef", "deq=500", "mbf=5", "mbc=127", "ctx=rec");i > D > Do these parameters look OK? Are there others I could specify that< > might help, or should I take any out that may be hindering > performance? > = > Any thoughts on the parameters I am using, or other helpfula# > information, greatly appreciated.   K Consider adding an initial allocation size to accompany your "deq=". If youeK know how big the output file will be just allocate ("alq=") the whole thingpL when you open the file. You might also add a "rop=wbh" and include "sqo" and) "cbt" on your "fop=" ("fop=tef,sqo,cbt").r  K If the the CRTL open routines don't fill in the FAB/RAB fields on a vanilla I fopen statement then you could likely get an improvement on your reads byh issueing a c   $ SET RMS/SEQ/BUF=8/BLO=127u  G prior to executing your program without changing the input fopen. MaybesJ someone else can speak to the behavior of the CRTL in that regard. (I seemH to recall that C does fill in these fields with it's own defaults so you may be out of luck there.)   -- s - Jimr   ------------------------------  % Date: Fri, 17 Oct 2003 14:21:24 +0200 " From: Didier Morandi <no@spam.com>G Subject: Re: [Plug] using x-windows to display vms decterms on Mac OS Xs4 Message-ID: <3f8fdecb$0$27035$626a54ce@news.free.fr>   Andy Bustamante wrote:  L > There's a some good information running DecWindows on an alternate displayL > platform in:  Linux and OpenVMS interoperability, Tricks for Old Dogs, New@ > Dogs and Hot Dogs with Open Systems by John Robert Wisniewski.   Wisniewski?  This name rings me a bell...   D.   ------------------------------   End of INFO-VAX 2003.576 ************************d at $300 per patch per seat.   @ This clearly has to be factored into windows TCO and many people+ that I talk to are increasingly doing this.a   Regardst Andrew Harrisono   ------------------------ת8|5Q\yqMV:lhkc%ǯkP[wR4-:ںap&_ne0㣭;W8&
x|,L73mbSfh^^WKLyu,ww9>OOc<>ZyuS!OJ|b~eyVDM]ubakvD--:l)W
)XlD6e\âɛuvlO`&0㣵ɻt^Ve:Md6Q-O[o
׀W-v/SmNv܁L:
x-qtkD(!lY<6=y:j֖qݫbyvӭӠצ
ۑ)`{t<:l3;NGq}bw:֬֟`[<#j1K3b1](߫\|%ZŦü?Ooy*~KjmQ?Of^u2[aGۦϓ8&kS]:lha;3l׀O--3:lHs!̼zb407mΙɟҺq*ۦz%Ze:WYJ|PS7Zh0a{k
گ[[<>ZloL[5`;b]ϓY3qMfGuѶ0bSi#&Fh^>ry:TCSw7O{[jn
2&=]y
ZՃUwugs,<>t:]TFw]Ƶ{@jy`[7z%1џyhE=:u.ĻZS\F*xڡ8Դ:,za.O{Q!waIq90񴦴1i5tѴX]tG20㣵:l|$6d6L-aSq0Nna?D}
up
uaG(=+cs_=h36Ȟ3I}mΌ-ǖشuDQ
}Z-m:ZWav8 f)ϓw2(1dbG+XlGHӥâI0
:zZbv,rA徝.<LaVa=y6ӭSV:lYd:>OF]3,ΊE{f"<3+V"J5اâҾX=9WR(gjRh4ba+}eluv,x|sUٱتDsUQ&ڬâaϓU2\.x|u>'s\wG\}D 7ާf<O䍏47.΍ŢŠZI0D^àPmWà*uv牢