0 INFO-VAX	Sat, 26 Jan 2002	Volume 2002 : Issue 49      Contents:/ Re: Advanced Server 7.3 issues when PDC reboots P Re: BLISS pros and cons, was: Re: historical evidence of what went wrong  at DECP Re: BLISS pros and cons, was: Re: historical evidence of what went wrong  at DEC Re: Buffer Overflows c.o.vms archive? Where?  Re: c.o.vms archive? Where?  RE: c.o.vms archive? Where?  Capellas conference  Re: Capellas conference  Re: Capellas conference  Re: Capellas conference  Re: Capellas conference 0 Capellas is either a "moron" or he's "paid off"!4 Re: Capellas is either a "moron" or he's "paid off"!4 Re: Capellas is either a "moron" or he's "paid off"!D RE: Capellas redefines Industry Standard to mean Windows *and* LinuxD Re: Capellas redefines Industry Standard to mean Windows *and* LinuxD Re: Capellas redefines Industry Standard to mean Windows *and* LinuxD Re: Capellas redefines Industry Standard to mean Windows *and* LinuxD Re: Capellas redefines Industry Standard to mean Windows *and* Linux! Re: Changing colors in DECWindows  Re: Compaq financial conference  Re: Compaq financial conference  Re: Compaq financial conference  Re: Compaq financial conference  Re: Compaq financial conference  Re: Compaq financial conference  Re: Compaq financial conference  RE: Compaq financial conference  Re: Compaq financial conference  Re: Compaq financial conference  RE: Compaq financial conference  RE: Compaq financial conference  Re: Compaq financial conference  Re: Compaq financial conference  Re: Compaq financial conference  Re: Compaq financial conference  Re: Compaq financial conference C Re: compiler group (was: RE: Gartner Group lies about VMS ... loses C RE: compiler group (was: RE: Gartner Group lies about VMS ... loses C Re: compiler group (was: RE: Gartner Group lies about VMS ... loses C Re: compiler group (was: RE: Gartner Group lies about VMS ... loses   DHCP as a client using Multinet?9 Re: ethernet boot diskless VAX-11/730--the question first 1 Re: Even Intel thinks Itanic isn't Proprietary...  Re: fopen crashes while in AST Re: fopen crashes while in AST; Re: Gartner Group lies about VMS ... loses all credibility! ; Re: Gartner Group lies about VMS ... loses all credibility! ; Re: Gartner Group lies about VMS ... loses all credibility! ! Re: I/O Performance on Alpha 4100 ) Re: Inquirer : "Intel *could* can Itanic" ) Re: Inquirer : "Intel *could* can Itanic" ) Re: Inquirer : "Intel *could* can Itanic"  Kerberos in VMS  Re: Kerberos in VMS F Re: MORE!!  Problems with IO$_SETMODE | IO$M_WRTATTN I/O Function Code$ new partner also sent to openvms.org RE: New VMS newsgroup  RE: New VMS newsgroup  Re: New VMS newsgroup  Re: New VMS newsgroup  Re: New VMS newsgroup  Re: New VMS newsgroup  Re: New VMS newsgroup  Re: New VMS newsgroup  Re: New VMS newsgroup  Re: New VMS newsgroup  RE: New VMS newsgroup  Re: New VMS newsgroup  Re: Newbie Help on Security G Re: Origins and dated nature of passord dictionary (was Re: bkjfn is an + Re: Problem with TSM and ETS-Terminalserver % Re: What is unix for SEARCH /WINDOW=0 1 Re: Wierd File Reading Behaviour - Dec C on Alpha 1 Re: Wierd File Reading Behaviour - Dec C on Alpha $ Re: WTB:Freeware 4 or 5 CD for sale?; Re: Younger recruits versus experienced veterans  ( was The ; Re: Younger recruits versus experienced veterans  ( was The ; Re: Younger recruits versus experienced veterans  ( was The P Re: Younger recruits versus experienced veterans ( was The demise    of      comP Re: Younger recruits versus experienced veterans ( was The demise    of      com9 Re: [Q] Why is there the limit of 8 levels of directories 9 Re: [Q] Why is there the limit of 8 levels of directories   F ----------------------------------------------------------------------  % Date: Fri, 25 Jan 2002 16:07:09 -0500 5 From: "Brad McCusker" <Brad.McCuskerNOSP@MCompaq.com> 8 Subject: Re: Advanced Server 7.3 issues when PDC reboots/ Message-ID: <RPj48.58$am1.650@news.cpqcorp.net>    Ed,   K We recently had an escalation for a problem with the failover code.  Please  contact J your CSC, and ahve them escalate this to engineering.  Have them reference the  "-89520" patches.    Regards,   Brad    / "Ed Wilts" <ewilts@ewilts.org> wrote in message 6 news:P9348.26444$yC.3433370@typhoon.mn.mediaone.net...F > Background:  we're an AS7.3 member server .  We had a case where our domainL > PDC (Win2K) took a nosedive and had to be rebooted.  Immediately followingK > that, we started to log errors from only 1 of our cluster hosts (out of 3 J > that run AS) that it could not reach a domain controller, even after the PDC L > was rebooted.  The other 2 cluster hosts continued to run and authenticateK > fine.  I ended up having to do a PWSTOP and PWSTART to get it functioning  > again.  My questions: G >     1.Is there a mechanism to determine which domain controller we're  > authenticating against? F >     2.If a PDC is rebooted, how do I tell AS to authenticate against another 3 > domain controller?  Should this not be automatic?  > I > Our domain has 1 PDC (obviously) and about 30 BDCs scattered throughout  the L > world, although mostly within the US and most not close to the VMScluster. > 	 > Thanks,  >     .../Ed > --
 > Ed Wilts > Mounds View, MN, USA > mailto:ewilts@ewilts.org >  >    ------------------------------    Date: 25 Jan 2002 15:34:23 -0500- From: Rich Alderson <alderson+news@panix.com> Y Subject: Re: BLISS pros and cons, was: Re: historical evidence of what went wrong  at DEC . Message-ID: <mdd4rla5fw0.fsf@panix2.panix.com>  ( weiner@TheWorld.com (Sam Weiner) writes:  & > Anyone know what happened to Scribe?   TeX.   --  N Rich Alderson                                          alderson+news@panix.comL   "You get what anybody gets.  You get a lifetime."  --Death, of the Endless   ------------------------------    Date: 25 Jan 2002 16:59:14 -0500& From: fdc@columbia.edu (Frank da Cruz)Y Subject: Re: BLISS pros and cons, was: Re: historical evidence of what went wrong  at DEC 1 Message-ID: <a2skfi$d52$1@watsol.cc.columbia.edu>   . In article <mdd4rla5fw0.fsf@panix2.panix.com>,/ Rich Alderson  <alderson+news@panix.com> wrote: * : weiner@TheWorld.com (Sam Weiner) writes: : ( : > Anyone know what happened to Scribe? :  : TeX. : H We still have and use Scribe here (at Columbia).  It was used to produceL all of the Kermit books.  But I don't have anything to do with the licensingK so don't know who owns it now, or who to contact.  The local copy says this  when you start it:  2   Scribe 8.1  Copyright (C) 1981, 1993  Unus, Inc.A   Distributed exclusively by Cygnet Publishing Technologies, Inc. 5   Welcome to the Scribe Document Production Software.    *    Hmmm, what's that "*"?  K Scribe, by the way, is still mighty nice for producing complicated types of I books and articles.  Its main drawback in modern times (for me at least), G is its limited ability to deal with different writing systems (accented K letters, Cyrillic, Hebrew, etc).  It can do it, but only up to a point, and  it ain't easy.  M A few years ago I posted a query to the old Scribe mailing list, what was it, F scribe-wizards@somewhere-or-other... asking if anybody had made Scribe* output HTML, and sure enough there it was.   - Frank    ------------------------------  % Date: Fri, 25 Jan 2002 14:45:58 -0500 E From: Jan C. =?iso-8859-1?Q?Vorbr=FCggen?= <jvorbrueggen@mediasec.de>  Subject: Re: Buffer Overflows + Message-ID: <3C51B5F6.59788B79@mediasec.de>   G > If they follow Bill Gates' new directive, Microsoft's developers will F > have to go back and review untold millions of lines of code, lookingJ > for buffer overflow opportunities.  There are so many such opportunities? > that finding and fixing all of them might well be impossible.   N The compiler and/or linker will find them for you. Either remove the offending? C RTL routines from the library, or have a compiler option that E makes it "forget" them (die with an error if it finds one). You couls H also remove the interface declarations from the include files, and force5 the compiler to not work without explicit interfaces.   L I do hope VMS engineering is already doing this. Does anyone have the source4 listings to search for "printf", "strcpy" and so on?   	Jan   ------------------------------    Date: 26 Jan 2002 01:13:13 +0800, From: Paul Repacholi <prep@prep.synonet.com>  Subject: c.o.vms archive? Where?0 Message-ID: <877kq6tkuu.fsf_-_@prep.synonet.com>  . Didier Morandi <Didier.Morandi@gmx.ch> writes:  E > Or read all "DCL minutes of the *.*" in an archived version of this 1 > forum. The live version purges "expired" posts.   , Is there a c.o.vms archive? Or for Info-VAX?   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. 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, 25 Jan 2002 19:20:21 GMT 2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)$ Subject: Re: c.o.vms archive? Where?/ Message-ID: <Vdi48.51$am1.483@news.cpqcorp.net>   _ In article <877kq6tkuu.fsf_-_@prep.synonet.com>, Paul Repacholi <prep@prep.synonet.com> writes: / :Didier Morandi <Didier.Morandi@gmx.ch> writes:  : F :> Or read all "DCL minutes of the *.*" in an archived version of this2 :> forum. The live version purges "expired" posts. : - :Is there a c.o.vms archive? Or for Info-VAX?   *   Search for "archive" in the OpenVMS FAQ.    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Fri, 25 Jan 2002 15:07:54 -0500 * From: WILLIAM WEBB <WWEBB1@email.usps.gov>$ Subject: RE: c.o.vms archive? Where?- Message-ID: <0033000049265907000002L072*@MHS>   ( =0AIt's not perfect but it's accessible.  3 http://groups.google.com/groups?group=3Dcomp.os.vms    WWWebb   -----Original Message-----/ From: Info-VAX-Request@Mvb.Saic.Com at INTERNET & Sent: Friday, January 25, 2002 2:57 PMB To: Webb, William W Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNET  Subject: c.o.vms archive? Where?    . Didier Morandi <Didier.Morandi@gmx.ch> writes:  E > Or read all "DCL minutes of the *.*" in an archived version of this 1 > forum. The live version purges "expired" posts.   , Is there a c.o.vms archive? Or for Info-VAX?   --< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.G EPIC, The Architecture of the future, always has been, always will be.=    ------------------------------  % Date: Fri, 25 Jan 2002 13:56:30 -0500 2 From: "Sue Skonetski" <susan.skonetski@compaq.com> Subject: Capellas conference/ Message-ID: <4Sh48.48$am1.761@news.cpqcorp.net>   I At the risk of starting another long stream I thought that you might like L this it is a good conference with Michael Capellas.  I think it is worth the time.    sue     L http://webevents.broadcast.com/wsp/build_01/english/frameset.asp?nEventID=10 50   ------------------------------    Date: 25 Jan 2002 15:41:27 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)   Subject: Re: Capellas conference3 Message-ID: <9J2fdWiE7sqM@eisner.encompasserve.org>   e In article <dPj48.443926$oj3.83891337@typhoon.tampabay.rr.com>, "John N." <JNixon@cfl.rr.com> writes: M > Has anyone else been able to connect to this?  I get a 404 error - file not  > found  > ? > "Sue Skonetski" <susan.skonetski@compaq.com> wrote in message + > news:4Sh48.48$am1.761@news.cpqcorp.net... L >> At the risk of starting another long stream I thought that you might likeK >> this it is a good conference with Michael Capellas.  I think it is worth  > the  >> time. >> >> sue >> >> >>P > http://webevents.broadcast.com/wsp/build_01/english/frameset.asp?nEventID=1050  D John, something is wrong with your browser.  I get a message that is  right on-topic for VMS advocacy:  ; 	Microsoft OLE DB Provider for SQL Server error '80040e14'    , 	Incorrect syntax near the keyword 'ORDER'.   / 	/wsp/includes/_user_driven_slides.inc, line 9     ------------------------------  # Date: Fri, 25 Jan 2002 21:08:25 GMT # From: "John N." <JNixon@cfl.rr.com>   Subject: Re: Capellas conference? Message-ID: <dPj48.443926$oj3.83891337@typhoon.tampabay.rr.com>   K Has anyone else been able to connect to this?  I get a 404 error - file not  found   = "Sue Skonetski" <susan.skonetski@compaq.com> wrote in message ) news:4Sh48.48$am1.761@news.cpqcorp.net... K > At the risk of starting another long stream I thought that you might like J > this it is a good conference with Michael Capellas.  I think it is worth the  > time.  >  > sue  >  >  > L http://webevents.broadcast.com/wsp/build_01/english/frameset.asp?nEventID=10 > 50 >  >  >  >    ------------------------------  # Date: Fri, 25 Jan 2002 21:10:31 GMT # From: "John N." <JNixon@cfl.rr.com>   Subject: Re: Capellas conference? Message-ID: <bRj48.443938$oj3.83893274@typhoon.tampabay.rr.com>   9 I got it.  The url wrapped and I missed the trailing "50" K Now it is asking me to register, and asking if I want a reminder on Jan 23.   = "Sue Skonetski" <susan.skonetski@compaq.com> wrote in message ) news:4Sh48.48$am1.761@news.cpqcorp.net... K > At the risk of starting another long stream I thought that you might like J > this it is a good conference with Michael Capellas.  I think it is worth the  > time.  >  > sue  >  >  > L http://webevents.broadcast.com/wsp/build_01/english/frameset.asp?nEventID=10 > 50 >  >  >  >    ------------------------------  # Date: Sat, 26 Jan 2002 04:28:28 GMT - From: "John E. Malmberg" <wb8tyw@qsl.network>   Subject: Re: Capellas conference* Message-ID: <3C5234B3.6090009@qsl.network>   Try:  A http://www.netroadshow.com/webcasts/lehman/CeoCompaq2002.asp?loc=   B It requires an E-mail address and name in order to start the feed./ Also either RealPlayer or Windows Media Player.   0 It will send a confirmation of the registration.  F AFAIK, Neither are available for OpenVMS.  But this page does not give, the OLE DB error when accessed from Mozilla.  E This seems to be a service of Lehman Brothers, etc, and not a Compaq   Production.    -John  wb8tyw@qsl.network Personal Opinion Only    ------------------------------    Date: 25 Jan 2002 13:57:27 -0800( From: bob@instantwhip.com (Bob Ceculski)9 Subject: Capellas is either a "moron" or he's "paid off"! = Message-ID: <d7791aa1.0201251357.49dfcc4e@posting.google.com>   E Itanic is garbage as the first spec results out shows ... Alpha still  isD the only true high end chip ... if Intel doesn't adapt it fully into itanium D and continue EV8 development, then IBM power will take over the high end, and? Capellas thinking that windows or linux are the high end future : without a good 64 bit multiprocessing platform means he is   1. a total moron  # 2. he's paid off by microsoft/intel   
 3. he's drunk    which is it?   ------------------------------  % Date: Fri, 25 Jan 2002 18:31:36 -0500 % From: JF Mezei <jfmezei@videotron.ca> = Subject: Re: Capellas is either a "moron" or he's "paid off"! , Message-ID: <3C51EAD4.87E9E3D9@videotron.ca>   Bob Ceculski wrote:  > 1. a total moron% > 2. he's paid off by microsoft/intel  > 3. he's drunk  > which is it?    I 4- A typical IT director who chooses the most popular solution because he  doesn't want to be fired.     : Remember the "nobody's ever been fired for choosing IBM" ?  N Well, these days, no CEO of a wintel company is going to be fired for focusing= on wintel and trying to make his company even more efficient.   N What this tells me is that Compaq management are somewhat insecure (since theyN have been unable to leverage the digital and tandem stuff to their fullest andH seem to be inept and rapidly fixing their PC business to compete againstJ Dell). And when you are insecure, you don't go out on a limb dropping yourC core business and going for some business you understand even less.    ------------------------------  # Date: Sat, 26 Jan 2002 04:26:41 GMTe1 From: "David J. Dachtera" <djesys.nospam@fsi.net> = Subject: Re: Capellas is either a "moron" or he's "paid off"!f' Message-ID: <3C52311C.BCD82FA2@fsi.net>m   JF Mezei wrote:  >  > Bob Ceculski wrote:  > > 1. a total moron' > > 2. he's paid off by microsoft/intelc > > 3. he's drunke > > which is it? > K > 4- A typical IT director who chooses the most popular solution because hea > doesn't want to be fired.  > < > Remember the "nobody's ever been fired for choosing IBM" ? > P > Well, these days, no CEO of a wintel company is going to be fired for focusing? > on wintel and trying to make his company even more efficient.a  D "Windows" and "efficiency" in the same sentence should constitute an
 oxymoron, no?I   --   David J. Dachteran dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/w   ------------------------------  % Date: Fri, 25 Jan 2002 14:14:28 -0500 + From: "Main, Kerry" <Kerry.Main@compaq.com>nM Subject: RE: Capellas redefines Industry Standard to mean Windows *and* LinuxeT Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF401AB1BE5@kaoexc01.americas.cpqcorp.net>   David,   >>>=20/ OpenVMS: $1000        $5000     $7000       10Xn/ Windows:  $300        $1000     $1300        1Xc >>  H That is Windows Profesional only ie. desktop flavour. If you are lookingH at home market, then VMS hobby program means OpenVMS license is free.=20  8 Where is the comparable version of Windows that is free?  G In addition, can you (or someone else) also post what the retail servere9 version of W2K Advanced Server and W2K Datacenter is ?=20r   Thanks,   
 Kerry Main Senior Consultantr Compaq Canada Corp.n Professional Services  Voice: 613-592-4660  Fax  :  819-772-7036 Email: Kerry.Main@Compaq.com     -----Original Message-----. From: David Mathog [mailto:mathog@caltech.edu] Sent: January 25, 2002 12:04 PMT To: Info-VAX@Mvb.Saic.ComnG Subject: Re: Capellas redefines Industry Standard to mean Windows *and*a Linuxr     Phillip Helbig wrote:s >=20E > > In the financial conference Capellas has just said (I paraphrase)y "thereC > > is absolutely no question that industry standard systems (sliden defines H > > this as Windows and Linux) are eviscerating proprietary systems. The erac# > > of proprietary systems is over"  > > $ > > Better open source VMS now then. >=20B > Why?  Do you think it will then be more popular because Linux is popular E > and open-source?  Consider that VMS is now proprietary and not very 4 > popular compared to Windows, which is proprietary.  B Popularity has very, very little to do with open versus closed and
 everythingC to do with purchase price.  Just for starters, the cost to purchasee8 these OS's commercially for a small machine are roughly:  H          OS Software  Hardware  Total       Relative cost of application software/ OpenVMS: $1000        $5000     $7000       10Xi/ Windows:  $300        $1000     $1300        1X / Linux       $0        $1000     $1000        0Xe  F Any way you slice it OpenVMS is an expensive solution and it's been an	 expensive C solution for way, way, way too long, and the market (except for bigh	 machines)n
 has passedC it by. It will be interesting to see if Linux can get past the huge  market leade thatD Windows holds.  I'm reasonably sure that if MS Office for Linux ever shippedaD linux usage would explode (because most office workers never use any others	 software,rD and once you're in Word it really doesn't matter which OS is running underneath).< So we can be pretty sure that this product will never exist.  @ As for VMS, well, sometimes you've just got to say good bye :-(.   Regards,   David Mathog mathog@caltech.edu   ------------------------------    Date: 25 Jan 2002 13:41:52 -0800( From: bob@instantwhip.com (Bob Ceculski)M Subject: Re: Capellas redefines Industry Standard to mean Windows *and* Linux = Message-ID: <d7791aa1.0201251341.642d2f8b@posting.google.com>e  Z David Mathog <mathog@caltech.edu> wrote in message news:<3C518FF4.9D81978F@caltech.edu>... > Phillip Helbig wrote:y > > N > > > In the financial conference Capellas has just said (I paraphrase) "thereM > > > is absolutely no question that industry standard systems (slide definesoN > > > this as Windows and Linux) are eviscerating proprietary systems. The era% > > > of proprietary systems is over"A > > > & > > > Better open source VMS now then. > > L > > Why?  Do you think it will then be more popular because Linux is popularG > > and open-source?  Consider that VMS is now proprietary and not veryg6 > > popular compared to Windows, which is proprietary. > O > Popularity has very, very little to do with open versus closed and everythingnE > to do with purchase price.  Just for starters, the cost to purchasea: > these OS's commercially for a small machine are roughly: > J >          OS Software  Hardware  Total       Relative cost of application
 > software1 > OpenVMS: $1000        $5000     $7000       10X 1 > Windows:  $300        $1000     $1300        1X 1 > Linux       $0        $1000     $1000        0X  > R > Any way you slice it OpenVMS is an expensive solution and it's been an expensiveO > solution for way, way, way too long, and the market (except for big machines)S > has passedQ > it by. It will be interesting to see if Linux can get past the huge market lead0 > thatN > Windows holds.  I'm reasonably sure that if MS Office for Linux ever shippedL > linux usage would explode (because most office workers never use any other > software, F > and once you're in Word it really doesn't matter which OS is running > underneath).> > So we can be pretty sure that this product will never exist. > B > As for VMS, well, sometimes you've just got to say good bye :-(. > 
 > Regards, >  > David Mathog > mathog@caltech.edu  J you must not know were to look for hardware ... we just bought last year aJ brand new end of life alphaserver 1200 w/licenses and vms 7.2-1 for $2600!  J I can get you brand new alphas right now and for years at great prices ...L we also got a 75% vax to alpha license trade up 4 years ago when we migrated from vax to alpha vms ...s  J factor in total cost of ownership, running several boxes instead of 80,000K windoze boxes or 768 cpu sparc junk unix/linux boxes and the money and timelI spent not having to spend 80% of your time applying patches like windoze,9K vms is as cheap as it gets!  I have been using vms for 17 years now w/out aIH crash, and no I don't buy the latest and greatest, but I am only severalL years behind, and that is how you save money buying ... in between boxes youI upgrade memory or add a processor if an smp alpha ... that is the troublesO w/most people I hear complaining about vms costs, "they don't know how to buy"!m   ------------------------------    Date: 25 Jan 2002 15:41:37 -0800( From: bob@instantwhip.com (Bob Ceculski)M Subject: Re: Capellas redefines Industry Standard to mean Windows *and* Linuxc< Message-ID: <d7791aa1.0201251541.b891ce3@posting.google.com>  e Alan Greig <a.greig@virgin.net> wrote in message news:<i6r25usbi2m6fflqfrau4sm0buhokp2hbb@4ax.com>... C > In the financial conference Capellas has just said (I paraphrase)CH > "there is absolutely no question that industry standard systems (slideA > defines this as Windows and Linux) are eviscerating proprietarye2 > systems. The era of proprietary systems is over" > " > Better open source VMS now then. > H > Interestingly Capellas seems to be defining Windows as non proprietary' > again. Someone should tell Microsoft.a  I Please, IBM, someone, anyone, buy VMS and Alpha and get the best high endr* platform out of the hands of these clowns!   ------------------------------  % Date: Fri, 25 Jan 2002 19:10:43 -0500 % From: JF Mezei <jfmezei@videotron.ca>oM Subject: Re: Capellas redefines Industry Standard to mean Windows *and* Linux , Message-ID: <3C51F3FC.8C27EC51@videotron.ca>   Bob Ceculski wrote:nK > Please, IBM, someone, anyone, buy VMS and Alpha and get the best high endv, > platform out of the hands of these clowns!  M The deed was done a decade ago. Only a few more years of pain and uncertainty0M until VMS' retirement is officially announced and after that, no more debatess on what could be done.  J This is actually great for hobbyists. Within a few years, a Wildfire classK system will be available for a few dollars (actually, the movers may chargeo  more than the machine is worth).  H There are still people who swear by their Amiga or OS2 systems and still6 defend it, and in a few years, we'll join their ranks.   ------------------------------  # Date: Sat, 26 Jan 2002 02:00:23 GMT   From: cjt <cheljuba@prodigy.net>M Subject: Re: Capellas redefines Industry Standard to mean Windows *and* Linux + Message-ID: <3C520DEA.75B090D3@prodigy.net>h   JF Mezei wrote:l >  > Bob Ceculski wrote:-M > > Please, IBM, someone, anyone, buy VMS and Alpha and get the best high endc. > > platform out of the hands of these clowns! > O > The deed was done a decade ago. Only a few more years of pain and uncertaintykO > until VMS' retirement is officially announced and after that, no more debatesd > on what could be done. > L > This is actually great for hobbyists. Within a few years, a Wildfire classM > system will be available for a few dollars (actually, the movers may charge " > more than the machine is worth). > J > There are still people who swear by their Amiga or OS2 systems and still8 > defend it, and in a few years, we'll join their ranks.    Time for the computer graveyard.   ------------------------------    Date: 25 Jan 2002 18:38:55 -0800/ From: chris@applied-synergy.com (Chris Scheers) * Subject: Re: Changing colors in DECWindows= Message-ID: <754a27c1.0201251838.17815157@posting.google.com>-  d "Tom Linden" <tom@kednos.com> wrote in message news:<CIEJLCMNHNNDLLOOGNJIOEDGEAAA.tom@kednos.com>...H > Running 6.2 on AXP,  going through the OPTIONS pull down under Session	 > ManagereN > I am able to change the background color of screen, but not windows.  When I > try,K > a message tells me that it will only affect new applications, so i open aT
 > new DECtermrJ > and, nada.  What is the trick?  Normally I wouldn't waste time with such > matters, butJ > it is difficult to see, particularly as the glasses need to be stronger. > I > Now, if I open emacs, dbg or book reader the window has the new colors,o > almost.  It doesn'te > change the highlight color.p > 6 > Do I need to muck with some X-xindows stuff or what?  > The non-obvious part here is that all DECterms run in the same5 process.  That is, they are all the same application.   = First, kill all your DECterms.  (Or you may need to log out.)   A Then when you create a new DECterm, you should get the new windowe attributes.e  @ To test window colors, I usually bring up the clock.  When I get? everything the way I want it, then I close down my DECterms andi# restart them to get the new scheme.c  
 Good luck!   ------------------------------    Date: 25 Jan 2002 13:00:18 -0600+ From: young_r@encompasserve.org (Rob Young) ( Subject: Re: Compaq financial conference3 Message-ID: <4jr2xW9avjWV@eisner.encompasserve.org>t  T In article <3C51A2F2.CB9BA157@videotron.ca>, JF Mezei <jfmezei@videotron.ca> writes: > Rob Young wrote:G >>         NSK can't be safe if the era of proprietary systems is over.  >  > M > NSK is a sacred cow that runs the stock exchanges in New York. Too visible. G > And Unix cannot replace it because Unix does not have fault tolerance8L > built-in, and more importantly, doesn't have applications that are writtenN > specifically for those customers and applications that make use of the faultR > tolerance (an app needs to be fault tolerant aware to be fault tolerant on NSK).  < 	I would assert for transactions , you certainly don't need = 	fault tolerance.  What you do need is disaster tolerance andiA 	RTR or something similar.  AOL has a very large NSK front-end... C 	I think the key there is it is highly scalable.  To think that NSKcB 	can't be replaced here there and everywhere wouldn't hold up well@ 	under examination.  (In AOL's case, box crashes sign back in...@ 	eBay finally got their infrastructure stabilized by de-couplingB 	and avoiding single points of failures all over the place... they: 	still have outages but they don't amount to much at all).   	Back to the statement.... a  = 	NSK can't be safe if the era of proprietary systems is over.r   				Robj   ------------------------------  % Date: Fri, 25 Jan 2002 14:11:24 -0500n% From: JF Mezei <jfmezei@videotron.ca>e( Subject: Re: Compaq financial conference, Message-ID: <3C51ADD3.481FA134@videotron.ca>   Rob Young wrote:D >         I would assert for transactions , you certainly don't needF >         fault tolerance.  What you do need is disaster tolerance and# >         RTR or something similar.w  J There are aspects to fault tolerance that VMS doesn't quite have 100%. For instance, cluster transitions.  K What is most important in fault tolerance is the fact that a transaction IN-N PROGRESS can be cotinued on the fallback processor on Tandem, whereas in a VMSK cluster, the  transaction would fail and only subsequent transactions would8& succeed, being routed to another node.  E The problem with RTR is that it needs to also run on a flaut tolerant > platform. If it fails, then your stream of transactions fail !  N That was the original idea for the FT-VAX. They were small 4000s if I rememberN well, but their goal was to run the RTR (or equivalent for that time) that fedM the mega cluster. So the entry point for transactions was fault tolerant, and K then use a non fault tolerant but highly available and scalable platform to/ execute the transactions.p    J NSK provides systems services that make it possible for applications to doL lots of stuff to make themselevs fault tolerant.Tandem is far more than just> having dual power supplies, dual paths to shadowed drives etc.  K In many ways, VMS is far more elegant with its clustering architecture, butmT NSK just goes that tiny extra distance between disaster tolerant and fault tolerant.   ------------------------------  # Date: Fri, 25 Jan 2002 19:20:11 GMT * From: "Bill Todd" <billtodd@metrocast.net>( Subject: Re: Compaq financial conferenceB Message-ID: <Ldi48.111258$QB1.7915541@bin3.nnrp.aus1.giganews.com>  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:4jr2xW9avjWV@eisner.encompasserve.org... 7 > In article <3C51A2F2.CB9BA157@videotron.ca>, JF Mezei- <jfmezei@videotron.ca> writes: > > Rob Young wrote:I > >>         NSK can't be safe if the era of proprietary systems is over.e > >  > >rF > > NSK is a sacred cow that runs the stock exchanges in New York. Too visible.I > > And Unix cannot replace it because Unix does not have fault tolerancetF > > built-in, and more importantly, doesn't have applications that are writtentJ > > specifically for those customers and applications that make use of the faultaK > > tolerance (an app needs to be fault tolerant aware to be fault tolerant  on NSK). > < > I would assert for transactions , you certainly don't need > fault tolerance.  D Depends on what the consequences of an inaccurate (not just aborted)J transaction are.  Real fault-tolerance doesn't just assume that a hardwareD fault will stop the system rather than produce incorrect results, itL guarantees it by performing the operation concurrently on separate hardware,D comparing the results, and disabling the faulty system (as typicallyJ determined by pair-and-spare or TMR mechanisms) while allowing the healthy system to continue.c  I True, the probability of hardware producing an incorrect, rather than no,rI result is fairly low.  But some people think the additional protection isoF well worth the expense, and VMS hasn't offered that option since ftVAX disappeared.   - bill   ------------------------------  # Date: Fri, 25 Jan 2002 19:31:16 GMT * From: "Bill Todd" <billtodd@metrocast.net>( Subject: Re: Compaq financial conference@ Message-ID: <8oi48.11597$vH6.639591@bin6.nnrp.aus1.giganews.com>  2 "JF Mezei" <jfmezei@videotron.ca> wrote in message& news:3C51ADD3.481FA134@videotron.ca... > Rob Young wrote:F > >         I would assert for transactions , you certainly don't needH > >         fault tolerance.  What you do need is disaster tolerance and% > >         RTR or something similar.a >sL > There are aspects to fault tolerance that VMS doesn't quite have 100%. For  > instance, cluster transitions. >uJ > What is most important in fault tolerance is the fact that a transaction INL > PROGRESS can be cotinued on the fallback processor on Tandem, whereas in a VMSmG > cluster, the  transaction would fail and only subsequent transactionse woulde( > succeed, being routed to another node.  K No, what is most important in true fault-tolerance is that a transaction isoD protected from producing an incorrect result due to non-catastrophicH hardware failure.  A transaction can *always* fail (due to user abort orF deadlock if for no other reasons) and software using transactions mustH always be prepared to resubmit them (or otherwise handle the situation).J Indeed, if if a system is implemented such that a 'transaction' can always; proceed to completion, then it's not a 'transaction' in anyq$ commonly-accepted sense of the term.  J I suspect that whether an NSK application elects to continue a transactionD when its underlying hardware fails is an application decision:  IIRCL process-pairs periodically checkpoint their state to their backup processes,K and if they did so within transactions rather than just between them they'dtG be able to continue - at the expense of implementing their own internalaI sub-transaction handlers, so I'd guess that this may be quite rare (givensL how much easier it is just to back out the transaction and start it again on the backup platform).f  < But I've never been an NSK guru, so corrections are welcome.   - bill   ------------------------------  % Date: Fri, 25 Jan 2002 20:08:13 +00006% From: Alan Greig <a.greig@virgin.net>i( Subject: Re: Compaq financial conference* Message-ID: <3C51BB2C.B94E6A76@virgin.net>   Rob Young wrote:  C        NSK can't be safe if the era of proprietary systems is over.R  c That's why it was singled out as an exception for its non-stop fault tolerance capabilities. And on0^ multiple occasions... Even got mentioned by Capellas himself who joked about having one in hisd basement. IIRC it was shown to be around seven times as reliable as VMS in one slide - I think Eliasa said based on information from Gartner!!!). You'll be glad to know Windows came bottom. Somethingo# called "Mainframe" came second top.?       >                         Robo   --
 Alan Greig   ------------------------------    Date: 25 Jan 2002 15:13:41 -0600+ From: young_r@encompasserve.org (Rob Young)w( Subject: Re: Compaq financial conference3 Message-ID: <wTjGUKYUOmUD@eisner.encompasserve.org>   R In article <3C51BB2C.B94E6A76@virgin.net>, Alan Greig <a.greig@virgin.net> writes: >  >  > Rob Young wrote: > E >        NSK can't be safe if the era of proprietary systems is over.a > e > That's why it was singled out as an exception for its non-stop fault tolerance capabilities. And on ` > multiple occasions... Even got mentioned by Capellas himself who joked about having one in hisf > basement. IIRC it was shown to be around seven times as reliable as VMS in one slide - I think Eliasc > said based on information from Gartner!!!). You'll be glad to know Windows came bottom. Something % > called "Mainframe" came second top.h >   B 	I can handle that.  I can also handle the fact that IBM mainframeD 	sales are surprisingly up due to Linux and other factors.  In fact,+ 	IBM just announced a Linux only mainframe:     D http://www.eweek.com/article/0,3658,s%253D701%2526a%253D21696,00.asp  K In one such example, Boscov's Department Stores LLC, based in Reading, Pa.,eL last year decided to replace its 44 Windows NT servers with a single zSeries mainframe running Linux. -  L Harry Roberts, Boscov's chief information officer, said the increasing costsN associated with operating dozens of servers spurred the company to switch to a Linux-based mainframe. -  L "Over the last several years, we've added one server per month to our serverK farm," Roberts said. "This meant adding a new [NT] server administrator foriL every 10 servers. The increasing complexity and difficulty of backing up the= large and growing server farm was becoming a major concern." i    = 	So when Mr. Capellas rails against "proprietary" perhaps he  ? 	indirectly means Microsoft and just doesn't realize it -yet- !o  G 	Actually, all this talk about proprietary is getting seriously blurred B 	with Linux acting as de facto application "middleware".  At leastB 	in the zSeries case, no?  Is zSeries proprietary if it is running3 	Linux applications and native OS/390 applications?,  ? 	You see, everyone is waiting for the paradigm shift and IBM iso@ 	helping to create it.  The paradigm shift involves writing yourA 	applications with Linux in mind and then running them on any OS.c- 	Any OS that can virtualize Linux that is ;-)      				Robt   ------------------------------  % Date: Fri, 25 Jan 2002 16:34:35 -0500n% From: JF Mezei <jfmezei@videotron.ca> ( Subject: Re: Compaq financial conference, Message-ID: <3C51CF68.2CFBAE22@videotron.ca>   Rob Young wrote:M > In one such example, Boscov's Department Stores LLC, based in Reading, Pa.,eN > last year decided to replace its 44 Windows NT servers with a single zSeries > mainframe running Linux.  F IBM is not measured by the volume of wintel boxes it ships. Compaq is.  I So for Compaq to replace 44 of its wintel servers with a single Alpha, itiM won't look good for its image. Compaq has absolutely no motivation to replacerN its wintel boxes with real machines. Yes, Kerry Main will be able to provide aL few examples of his projects that converter wintel boxes into a consolidatedN VMS server. But Mr Main is a small island lost in the middle of a vast Compaq 
 wintel ocean.   N When Compaq was leaking information that it was going to transform itself intoK an enterprise company, one of the hoped-for accomplishements is that CompaqlL would send a strong message that it would no longer be measured in number ofJ wintel boxes sold. Unfortunatly, Compaq's TV campaign only re-enforced itsN wintel image by portraying wintel stuff in its "pretend to be enterprise" ads.   ------------------------------  % Date: Fri, 25 Jan 2002 16:45:48 -05000+ From: "Main, Kerry" <Kerry.Main@Compaq.com>l( Subject: RE: Compaq financial conferenceT Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF401AB1BE8@kaoexc01.americas.cpqcorp.net>   JF -=20   E >>> NSK is a sacred cow that runs the stock exchanges in New York.<<<u  H mmm.. If I recall correctly, of top 10 Exchanges in the world all 10 are
 Compaq.=20   5 are OpenVMS and 5 are NSK.  G NSK is certainly a great solution for this market, but then again so isr OpenVMS.  F I sense another purist discussion brewing, but OpenVMS also does faultH tolerant computing.  (Remember RTR is bundled as part of OpenVMS now ie. free)   , It depends on what level you are looking at.  D From a business perspective, NSK ensures a transaction is never lost once it is committed.=20  G It does this at the system/os level. Transactions are replicated to themD remote site for redundancy, so there is no read-write load balancingE across servers at multiple sites. However, at the server HW level, iti; definately has features that OpenVMS / Alpha does not have.e  C From a business perspective, OpenVMS (RTR) ensures a transaction is # never lost once it is committed.=20m  F It does this across servers, networks and site failures.  TransactionsC can be read-write load balanced across multiple servers at multipleoE sites. A network failure such as a router failure still does not stop % the transaction from being completed.o  G Both are great solutions and it boils down to price points, applicationaD availability, and how much redundancy the Customer is willing to pay for.  
 Reference:B http://www.compaq.com/products/software/ntenterprise/rtr/faqs.html. http://www.compaq.com/info/SP5104/SP5104PF.PDF   Customer testimonials:B http://www.openvms.compaq.com/openvms/brochures/ISE/ International Securities Exchange (ISE) G http://www.openvms.compaq.com/openvms/brochures/OMTESTIM/ OM Technology % (develops Exchange type applications)M  H Note - in both cases, custom application coding is required, but in this market space, that is expected.u   Regards,  
 Kerry Main Senior Consultant  Compaq Canada Corp.t Professional Servicesr Voice: 613-592-4660l Fax  :  819-772-7036 Email: Kerry.Main@Compaq.com     -----Original Message-----, From: JF Mezei [mailto:jfmezei@videotron.ca] Sent: January 25, 2002 1:25 PM To: Info-VAX@Mvb.Saic.Comk( Subject: Re: Compaq financial conference     Rob Young wrote:F >         NSK can't be safe if the era of proprietary systems is over.    B NSK is a sacred cow that runs the stock exchanges in New York. Too visible.E And Unix cannot replace it because Unix does not have fault tolerance B built-in, and more importantly, doesn't have applications that are written F specifically for those customers and applications that make use of the faultaG tolerance (an app needs to be fault tolerant aware to be fault tolerantn on NSK).   ------------------------------    Date: 25 Jan 2002 15:51:34 -0600+ From: young_r@encompasserve.org (Rob Young)s( Subject: Re: Compaq financial conference3 Message-ID: <wrR5sngthQv2@eisner.encompasserve.org>y  T In article <3C51CF68.2CFBAE22@videotron.ca>, JF Mezei <jfmezei@videotron.ca> writes: > Rob Young wrote:N >> In one such example, Boscov's Department Stores LLC, based in Reading, Pa.,O >> last year decided to replace its 44 Windows NT servers with a single zSeries  >> mainframe running Linux.- > H > IBM is not measured by the volume of wintel boxes it ships. Compaq is. > K > So for Compaq to replace 44 of its wintel servers with a single Alpha, itmO > won't look good for its image. Compaq has absolutely no motivation to replaceDP > its wintel boxes with real machines. Yes, Kerry Main will be able to provide aN > few examples of his projects that converter wintel boxes into a consolidatedP > VMS server. But Mr Main is a small island lost in the middle of a vast Compaq  > wintel ocean.s >   - 	Customers drive demand... if there is enough ; 	demand for Linux on IBM mainframes , it will sell more IBMrA 	mainframes.  Maybe Boscov's is a blip... maybe not.  Why pick ontC 	Kerry?  He isn't the only one attempting to consolidate.  CIOs andn, 	Technical Directors understand the concept.  P > When Compaq was leaking information that it was going to transform itself intoM > an enterprise company, one of the hoped-for accomplishements is that CompaqrN > would send a strong message that it would no longer be measured in number ofL > wintel boxes sold. Unfortunatly, Compaq's TV campaign only re-enforced itsP > wintel image by portraying wintel stuff in its "pretend to be enterprise" ads.  B 	And yet... and yet Compaq also acknowledges they are making money7 	on Enterprise outside of Industry Standard Enterprise:c  8 http://www.compaq.com/newsroom/pr/2002/pr2002011601.html  E Revenue from the Enterprise Computing segment-built on the pillars ofgN fault-tolerant and high performance computing, enterprise storage and industryJ standard servers-was up 15 percent from the third quarter to $2.7 billion.  L During the fourth quarter, the Business Critical Solutions Group-responsibleO for fault-tolerant and high performance systems-grew 31 percent sequentially onrC the strength of Compaq's NonStop Himalaya and AlphaServer systems.     				Robb   ------------------------------  # Date: Fri, 25 Jan 2002 22:02:17 GMTt* From: "Bill Todd" <billtodd@metrocast.net>( Subject: Re: Compaq financial conference@ Message-ID: <JBk48.27554$%b.1405989@bin4.nnrp.aus1.giganews.com>  6 "Main, Kerry" <Kerry.Main@Compaq.com> wrote in messageL news:BE56C50EA024184DAF48F0B9A47F5CF401AB1BE8@kaoexc01.americas.cpqcorp.net. .. JF -  E >>> NSK is a sacred cow that runs the stock exchanges in New York.<<<   H mmm.. If I recall correctly, of top 10 Exchanges in the world all 10 are Compaq.    5 are OpenVMS and 5 are NSK.  G NSK is certainly a great solution for this market, but then again so is  OpenVMS.  F I sense another purist discussion brewing, but OpenVMS also does fault tolerant computing.p   ***aK If it's 'purist' to use terms only according to their definition, so be it: H VMS hasn't provided 'fault-tolerant' computing since ftVAX disappeared -C unless you believe that 'fault-tolerant' and 'highly-available' arer> synonymous (which your subsequent comment about RTR suggests).   - bill   ------------------------------  % Date: Fri, 25 Jan 2002 17:14:22 -0500 + From: "Main, Kerry" <Kerry.Main@Compaq.com>8( Subject: RE: Compaq financial conferenceT Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF401AB1BE9@kaoexc01.americas.cpqcorp.net>   Rob,  < >>> 	I can handle that.  I can also handle the fact that IBM	 mainframes> 	sales are surprisingly up due to Linux and other factors.  In fact, . 	IBM just announced a Linux only mainframe:<<<  E Actually, I am willing to be corrected, but it appears some marketingnF words are at play here. These new servers run an "new" OS called z/VM.  
 Reference: http://www.vm.ibm.com/: " .... Built upon the solid VM/ESA base, z/VM exploits theC z/Architecture and helps enterprises meet their growing demands foreG multi-user server solutions with a broad range of support for operatingrE system environments such as z/OS, OS/390, TPF, VSE/ESA, CMS, or Linux.
 for zSeries."l  9 You don't need to understand mainframe lingo?  Check out: # http://www.vm.ibm.com/service/news/d   Great marketing though.e   :-)i   Regards,  
 Kerry Main Senior Consultant  Compaq Canada Corp.t Professional Servicesl Voice: 613-592-4660t Fax  :  819-772-7036 Email: Kerry.Main@Compaq.com     -----Original Message-----2 From: Rob Young [mailto:young_r@encompasserve.org] Sent: January 25, 2002 4:14 PM To: Info-VAX@Mvb.Saic.Com ( Subject: Re: Compaq financial conference    5 In article <3C51BB2C.B94E6A76@virgin.net>, Alan Greign <a.greig@virgin.net> writes: >=20 >=20 > Rob Young wrote: >=20E >        NSK can't be safe if the era of proprietary systems is over.- >=20F > That's why it was singled out as an exception for its non-stop fault tolerance capabilities. And onH > multiple occasions... Even got mentioned by Capellas himself who joked about having one in his.E > basement. IIRC it was shown to be around seven times as reliable as4  VMS in one slide - I think EliasD > said based on information from Gartner!!!). You'll be glad to know Windows came bottom. Something% > called "Mainframe" came second top.t >=20  8 	I can handle that.  I can also handle the fact that IBM	 mainframea> 	sales are surprisingly up due to Linux and other factors.  In fact, + 	IBM just announced a Linux only mainframe:s    D http://www.eweek.com/article/0,3658,s%253D701%2526a%253D21696,00.asp  F In one such example, Boscov's Department Stores LLC, based in Reading, Pa.,D last year decided to replace its 44 Windows NT servers with a single zSeries  mainframe running Linux.=20l  F Harry Roberts, Boscov's chief information officer, said the increasing costsdB associated with operating dozens of servers spurred the company to switch to as Linux-based mainframe.=20.  E "Over the last several years, we've added one server per month to ouro serverG farm," Roberts said. "This meant adding a new [NT] server administratorn for H every 10 servers. The increasing complexity and difficulty of backing up thei? large and growing server farm was becoming a major concern."=20o    ? 	So when Mr. Capellas rails against "proprietary" perhaps he=20 ? 	indirectly means Microsoft and just doesn't realize it -yet- !t  ? 	Actually, all this talk about proprietary is getting seriouslya blurredi< 	with Linux acting as de facto application "middleware".  At leastn: 	in the zSeries case, no?  Is zSeries proprietary if it is running 3 	Linux applications and native OS/390 applications?d  ? 	You see, everyone is waiting for the paradigm shift and IBM ise@ 	helping to create it.  The paradigm shift involves writing yourA 	applications with Linux in mind and then running them on any OS.D- 	Any OS that can virtualize Linux that is ;-)      				Rob    ------------------------------  % Date: Fri, 25 Jan 2002 17:21:59 -0500 + From: "Main, Kerry" <Kerry.Main@Compaq.com>r( Subject: RE: Compaq financial conferenceT Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF401AB1BEA@kaoexc01.americas.cpqcorp.net>  H >>> If it's 'purist' to use terms only according to their definition, so be it<<   F Well, it depends if one views it from a hardware / OS perspective or a. higher level business transaction perspective.  E Different approaches to the same objective i.e. how does one ensure ao. transaction is never lost once it is commited?  & Each has advantages and disadvantages.   Regards,  
 Kerry Main Senior Consultante Compaq Canada Corp.n Professional Serviceso Voice: 613-592-4660a Fax  :  819-772-7036 Email: Kerry.Main@Compaq.com     -----Original Message-----/ From: Bill Todd [mailto:billtodd@metrocast.net]  Sent: January 25, 2002 5:02 PM To: Info-VAX@Mvb.Saic.Comd( Subject: Re: Compaq financial conference      6 "Main, Kerry" <Kerry.Main@Compaq.com> wrote in messageH news:BE56C50EA024184DAF48F0B9A47F5CF401AB1BE8@kaoexc01.americas.cpqcorp. net. .e JF -  E >>> NSK is a sacred cow that runs the stock exchanges in New York.<<<c  H mmm.. If I recall correctly, of top 10 Exchanges in the world all 10 are Compaq.a   5 are OpenVMS and 5 are NSK.  G NSK is certainly a great solution for this market, but then again so isr OpenVMS.  F I sense another purist discussion brewing, but OpenVMS also does fault tolerant computing.r   ***uG If it's 'purist' to use terms only according to their definition, so be  it:uH VMS hasn't provided 'fault-tolerant' computing since ftVAX disappeared -C unless you believe that 'fault-tolerant' and 'highly-available' aren> synonymous (which your subsequent comment about RTR suggests).   - bill   ------------------------------  % Date: Fri, 25 Jan 2002 18:24:15 -0500:% From: JF Mezei <jfmezei@videotron.ca>g( Subject: Re: Compaq financial conference, Message-ID: <3C51E91C.AC58DFC0@videotron.ca>   "Main, Kerry" wrote:I > It does this at the system/os level. Transactions are replicated to theoF > remote site for redundancy, so there is no read-write load balancingG > across servers at multiple sites. However, at the server HW level, ite= > definately has features that OpenVMS / Alpha does not have.   G Note that the OS provides base services for automated failover, but the)H applications must take advantage of those services. Otherwise the tandemN behaves like a VMS cluster with just a failover of a process with no assurance( that current transactions are garanteed.  E > From a business perspective, OpenVMS (RTR) ensures a transaction isu" > never lost once it is committed.  K The problem with RTR is that if the machine(s) that run RTR fail, then thatC assurance is lost.    M Back in the 1990s when I had exposure to the Tandem machines, I witnessed the L tandem group next to me spend a lot of time planning which process was goingF to run on which processor and where its backup would be and that was aM critical step for load balancing since transactions only "ran" on the primarymN CPU with the designated backup CPU's performance wasted. That is where VMS has
 an advantage.a   ------------------------------  % Date: Fri, 25 Jan 2002 18:27:42 -0500 % From: JF Mezei <jfmezei@videotron.ca>w( Subject: Re: Compaq financial conference, Message-ID: <3C51E9EB.D850026F@videotron.ca>   Rob Young wrote:6 >         Customers drive demand... if there is enoughD >         demand for Linux on IBM mainframes , it will sell more IBM >         mainframes.a    K But you should also understand that customers are influenced a lot by their K vendor and advertising and trade rags as well as outfits such as gartner to . whom corporations look for guidance on trends.  L If IBM tells a potential customer it can consolidate all its PCs into a moreM cost effective mainframe and provide examples, and even hands-on testing at at: testing centre, then it is far more likely to make a sale.  L Compaq may have a few Kerry Mains worldwide, but the vast majority of CompaqL will not be presenting VMS solutions similar to what IBM is presenting, theyH will be presenting wintel solutions that are no different from Dell etc.   ------------------------------  # Date: Sat, 26 Jan 2002 00:53:36 GMTm* From: "Bill Todd" <billtodd@metrocast.net>( Subject: Re: Compaq financial conferenceB Message-ID: <k6n48.113239$QB1.8258126@bin3.nnrp.aus1.giganews.com>  6 "Main, Kerry" <Kerry.Main@Compaq.com> wrote in messageL news:BE56C50EA024184DAF48F0B9A47F5CF401AB1BEA@kaoexc01.americas.cpqcorp.net. ..  H >>> If it's 'purist' to use terms only according to their definition, so be it<<i  F Well, it depends if one views it from a hardware / OS perspective or a. higher level business transaction perspective.  E Different approaches to the same objective i.e. how does one ensure ax. transaction is never lost once it is commited?   ***t? No, they are not different approaches to the same objective.  A-K fault-tolerant system ensures that system operation will survive a hardwarefH fault without producing incorrect results, and by itself says absolutelyF nothing about the ability of a committed transaction to survive a siteI disaster (which is accomplished through high-availability mechanisms that G are orthogonal to fault-tolerance, such as remotely-mirrored storage orn	 systems).n  G A transaction *by definition* is never lost once it has been committed,pJ unless the storage on which it has been made persistent is lost.  That hasK nothing to do with fault-tolerance *or* availability:  it is the definitiontJ of how transactions work.  Increased availability means that the result ofB the committed transaction will have a greater probability of beingF accessible, and fault-tolerance means that the result of the committedF transaction will have a greater probability of being correct (and alsoI increases the probability that a local hardware failure will not make theoD result inaccessible - but this has nothing to do with surviving site disasters).e   - bill   ------------------------------    Date: 25 Jan 2002 20:43:03 -0600+ From: young_r@encompasserve.org (Rob Young)d( Subject: Re: Compaq financial conference3 Message-ID: <hFIIMOnEr+24@eisner.encompasserve.org>-  T In article <3C51E91C.AC58DFC0@videotron.ca>, JF Mezei <jfmezei@videotron.ca> writes: > "Main, Kerry" wrote:J >> It does this at the system/os level. Transactions are replicated to theG >> remote site for redundancy, so there is no read-write load balancing H >> across servers at multiple sites. However, at the server HW level, it> >> definately has features that OpenVMS / Alpha does not have. > I > Note that the OS provides base services for automated failover, but theoJ > applications must take advantage of those services. Otherwise the tandemP > behaves like a VMS cluster with just a failover of a process with no assurance* > that current transactions are garanteed. > F >> From a business perspective, OpenVMS (RTR) ensures a transaction is# >> never lost once it is committed.  > M > The problem with RTR is that if the machine(s) that run RTR fail, then thatc > assurance is lost. >   B 	If you are using RTR, you are *most likely* spending the time to > 	create the infrastructure (spending the money too) to ensure 7 	hardware failure doesn't cause loss of transactions.  dB 	The split VMS clusters (Cantor Fitzgerald for instance) paid off.  9 http://www.compaq.com/products/software/ntenterprise/rtr/t  I If your critical business application must be available 24x7, in spite ofrG network failures, system failures or complete site outages, then Compaqt0 Reliable Transaction Router (RTR) can help you.    				Robh   ------------------------------  # Date: Sat, 26 Jan 2002 03:57:01 GMTo* From: "Bill Todd" <billtodd@metrocast.net>( Subject: Re: Compaq financial conferenceA Message-ID: <hOp48.19633$vH6.1106472@bin6.nnrp.aus1.giganews.com>r  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:hFIIMOnEr+24@eisner.encompasserve.org...i7 > In article <3C51E91C.AC58DFC0@videotron.ca>, JF Mezei/ <jfmezei@videotron.ca> writes: > > "Main, Kerry" wrote:L > >> It does this at the system/os level. Transactions are replicated to theI > >> remote site for redundancy, so there is no read-write load balancing J > >> across servers at multiple sites. However, at the server HW level, it@ > >> definately has features that OpenVMS / Alpha does not have. > >eK > > Note that the OS provides base services for automated failover, but the L > > applications must take advantage of those services. Otherwise the tandemH > > behaves like a VMS cluster with just a failover of a process with no	 assurancec, > > that current transactions are garanteed. > > H > >> From a business perspective, OpenVMS (RTR) ensures a transaction is% > >> never lost once it is committed.n > >YJ > > The problem with RTR is that if the machine(s) that run RTR fail, then that > > assurance is lost. > >i >rB > If you are using RTR, you are *most likely* spending the time to> > create the infrastructure (spending the money too) to ensure6 > hardware failure doesn't cause loss of transactions.  " Let's try this once again, slowly:  F The *definition* of a transaction *includes* the requirement that onceE committed it cannot be 'lost', unless the storage on which its recordnJ resides is lost.  Thus the *only* issues involve loss of storage, and thisH can be guarded against by replicating said storage (typically mirroring,L remote if site-disaster-tolerance is desired) - or, if quick recovery is notH required, by maintaining a remote recovery log which can be used to roll0 forward an off-site backup copy of the database.  G There are two other orthogonal issues being talked about here that havepI *nothing to do* with 'losing' transactions (save insofar as they guard inlJ general against loss of persistent data).  One is high-availability, whichK is required for systems that cannot tolerate down-time and usually involveslJ the ability to continue processing on other hardware if failure of currentI hardware is detected (but does nothing to protect the integrity of systemt@ operation if hardware fails without being detected).  Another isI fault-tolerance, which is required to ensure (even if hardware fails in a L manner that allows it to continue to operate but produces incorrect results)E that the system produces correct results (at least if the software iseK correct) - and most fault-tolerant approaches result in higher availabilityaJ as well, since they typically use (local, not remote) hardware replication& to check results as they are produced.  E Fault-tolerance can be achieved without special hardware support, but K typically not with acceptable performance (since results must be checked at'K fairly fine grain - e.g., certainly prior to most disk writes).  VMS hasn't K offered fault-tolerance since the special ftVAX hardware went away, but VMSn; does offer high availability via its clustering mechanisms.u  J So, one more time:  there's nothing special about RTR in this respect - itJ just preserves the 'no loss after commit' guarantee that *any* transactionA monitor must provide.  It may also offer transaction-level (as antG alternative to system-level or storage-level) remote mirroring to guardrF against site-disaster data loss, but this is not fault-tolerance, just high-availability.   - bill   ------------------------------  % Date: Fri, 25 Jan 2002 14:34:08 -0500y> From: Charlie McCutcheon <charlie.mccutcheon@NOSPAMcompaq.com>L Subject: Re: compiler group (was: RE: Gartner Group lies about VMS ... loses0 Message-ID: <3C51B330.EF018E38@NOSPAMcompaq.com>  & --------------EC00B95ACFFB4581DCCD3361* Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit>   Steve Lionel wrote:y  E > On Wed, 23 Jan 2002 16:26:56 +0000, Alan Greig <a.greig@virgin.net>0 > wrote: >c2 > >>> Hm, "over time" there will not be any Alpha.5 > >>> So it more or less depends on the definition of  > >>> "over time", I suppose...o > >>= > >>   Just ask anyone waiting for Fortran 95 for their KL10.  >  > Or their VAX.  >cH > >Out of nothing more than curiosity did VAX Fortran and PDP-10 Fortran* > >ever share any common code anyone know? >bE > No, though they did share several common developers.  (And we stills8 > have two PDP-10/20 FORTRAN developers on the project!) > ...M  2 Fortran 10/20 only got to Fortran-77, as I recall.  L I helped add some of the features, first job out of college, "those were the
 days"...  8-)o   Charlieo    & --------------EC00B95ACFFB4581DCCD3361) Content-Type: text/html; charset=us-ascii  Content-Transfer-Encoding: 7bit   > <!doctype html public "-//w3c//dtd html 4.0 transitional//en"> <html> Steve Lionel wrote: \ <blockquote TYPE=CITE>On Wed, 23 Jan 2002 16:26:56 +0000, Alan Greig &lt;a.greig@virgin.net>
 <br>wrote:3 <p>>>> Hm, "over time" there will not be any Alpha.a7 <br>>>> So it more or less depends on the definition ofe! <br>>>> "over time", I suppose...n <br>>>I <br>>>&nbsp;&nbsp; Just ask anyone waiting for Fortran 95 for their KL10.t <p>Or their VAX.I <p>>Out of nothing more than curiosity did VAX Fortran and PDP-10 FortranV, <br>>ever share any common code anyone know?K <p>No, though they did share several common developers.&nbsp; (And we stilll: <br>have two PDP-10/20 FORTRAN developers on the project!)^ <br><a href="http://developer.intel.com/software/products/compilers/f50/">...</a></blockquote>2 Fortran 10/20 only got to Fortran-77, as I recall.F <p>I helped add some of the features, first job out of college, "those were the days"...&nbsp; 8-) 
 <p>Charlie <br>&nbsp;</html>-  ( --------------EC00B95ACFFB4581DCCD3361--   ------------------------------  % Date: Fri, 25 Jan 2002 12:14:39 -0800a# From: "Tom Linden" <tom@kednos.com>@L Subject: RE: compiler group (was: RE: Gartner Group lies about VMS ... loses9 Message-ID: <CIEJLCMNHNNDLLOOGNJIGEDPEAAA.tom@kednos.com>V  , This is a multi-part message in MIME format.  + ------=_NextPart_000_000A_01C1A599.DCFCAA10i Content-Type: text/plain;g 	charset="US-ASCII"e Content-Transfer-Encoding: 7bit:       -----Original Message-----G   From: Charlie McCutcheon [mailto:charlie.mccutcheon@NOSPAMcompaq.com]i)   Sent: Friday, January 25, 2002 11:34 AMf   To: Info-VAX@Mvb.Saic.Com H   Subject: Re: compiler group (was: RE: Gartner Group lies about VMS ... loses        Steve Lionel wrote: G     On Wed, 23 Jan 2002 16:26:56 +0000, Alan Greig <a.greig@virgin.net> 
     wrote:4     >>> Hm, "over time" there will not be any Alpha.7     >>> So it more or less depends on the definition ofm!     >>> "over time", I suppose...e     >>?     >>   Just ask anyone waiting for Fortran 95 for their KL10.I       Or their VAX.s  J     >Out of nothing more than curiosity did VAX Fortran and PDP-10 Fortran,     >ever share any common code anyone know?  G     No, though they did share several common developers.  (And we still :     have two PDP-10/20 FORTRAN developers on the project!)     ...t  4   Fortran 10/20 only got to Fortran-77, as I recall.  L   Well I remeber talking to some the 20 folks in Marlboro around 80-81 about
 moving our2   F77 and PL/I, but nothing ever came of it for usJ   I helped add some of the features, first job out of college, "those were the days"...  8-).  	   Charlie       + ------=_NextPart_000_000A_01C1A599.DCFCAA10h Content-Type: text/html; 	charset="US-ASCII"a+ Content-Transfer-Encoding: quoted-printableo  > <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD>1 <META content=3D"text/html; charset=3Dus-ascii" =j http-equiv=3DContent-Type>= <META content=3D"MSHTML 5.00.2920.0" name=3DGENERATOR></HEAD>d <BODY> <DIV>&nbsp;</DIV>  <BLOCKQUOTE=20J style=3D"BORDER-LEFT: #0000ff 2px solid; MARGIN-LEFT: 5px; MARGIN-RIGHT: = 0px; PADDING-LEFT: 5px">B   <DIV align=3Dleft class=3DOutlookMessageHeader dir=3Dltr><FONT = face=3DTahoma=20J   size=3D2>-----Original Message-----<BR><B>From:</B> Charlie McCutcheon =  H   [mailto:charlie.mccutcheon@NOSPAMcompaq.com]<BR><B>Sent:</B> Friday, =
 January=20#   25, 2002 11:34 AM<BR><B>To:</B> =a/ Info-VAX@Mvb.Saic.Com<BR><B>Subject:</B> Re:=20i>   compiler group (was: RE: Gartner Group lies about VMS ...=202   loses<BR><BR></DIV></FONT>Steve Lionel wrote:=20F   <BLOCKQUOTE TYPE=3D"CITE">On Wed, 23 Jan 2002 16:26:56 +0000, Alan = Greig=20,     &lt;a.greig@virgin.net&gt; <BR>wrote:=20C     <P>&gt;&gt;&gt; Hm, "over time" there will not be any Alpha.=20BG     <BR>&gt;&gt;&gt; So it more or less depends on the definition of=20t>     <BR>&gt;&gt;&gt; "over time", I suppose... <BR>&gt;&gt;=20I     <BR>&gt;&gt;&nbsp;&nbsp; Just ask anyone waiting for Fortran 95 for =h their=20     KL10.=20     <P>Or their VAX.=20fJ     <P>&gt;Out of nothing more than curiosity did VAX Fortran and PDP-10 =
 Fortran=206     <BR>&gt;ever share any common code anyone know?=20H     <P>No, though they did share several common developers.&nbsp; (And = we still=20sH     <BR>have two PDP-10/20 FORTRAN developers on the project!) <BR><A=20     =oJ href=3D"http://developer.intel.com/software/products/compilers/f50/">...<= /A></P></BLOCKQUOTE>G   <DIV>Fortran 10/20 only got to Fortran-77, as I recall.&nbsp;<SPAN=20 B   class=3D765241220-25012002><FONT color=3D#0000ff face=3DArial=20%   size=3D2>&nbsp;</FONT></SPAN></DIV>4;   <DIV><SPAN class=3D765241220-25012002></SPAN>&nbsp;</DIV>F>   <DIV><SPAN class=3D765241220-25012002><FONT size=3D2><FONT = color=3D#0000ff=20H   face=3DArial>Well I remeber talking to some the 20 folks in Marlboro =	 around=20r3   80-81 about moving our</FONT></FONT></SPAN></DIV>sE   <DIV><SPAN class=3D765241220-25012002><FONT color=3D#0000ff><FONT =t size=3D2><FONT=20oI   face=3DArial>F77 and PL/I</FONT><FONT face=3DArial>, but nothing ever ='
 came of it=20s0   for us</FONT></FONT></FONT>&nbsp;</SPAN></DIV>J   <P>I helped add some of the features, first job out of college, "those = were=20e   the days"...&nbsp; 8-)=20l7   <P>Charlie <BR>&nbsp; </P></BLOCKQUOTE></BODY></HTML>r  - ------=_NextPart_000_000A_01C1A599.DCFCAA10--l   ------------------------------  % Date: Fri, 25 Jan 2002 20:36:35 +0000:% From: Alan Greig <a.greig@virgin.net>gL Subject: Re: compiler group (was: RE: Gartner Group lies about VMS ... loses* Message-ID: <3C51C1D3.7FC9EE9C@virgin.net>   Charlie McCutcheon wrote:.   >> >i4 > Fortran 10/20 only got to Fortran-77, as I recall. > E > I helped add some of the features, first job out of college, "thoses > were the days"...  8-) >H   [posted and mailed]e  @ Well if you've got some spare time pop over to alt.sys.pdp10 andD volunteer your services to help upgrade  10/20 Fortran to Fortran-90: ;-) Somone's actually ported GCC to produce PDP-10 code...  A It's  fun having access to all the old layered products and DECUStC tapes running   TOPS-20 (or TOPS-10 or ITS) on an emulated KL-10 orbC KS-10. Paul Allen's even put a TOPS-20 system on the net running on  real hardware (XKL I think).   >r	 > CharlieR >i   --
 Alan Greig   ------------------------------  % Date: Fri, 25 Jan 2002 13:49:19 -08009 From: aek@spies.com (Al Kossow)nL Subject: Re: compiler group (was: RE: Gartner Group lies about VMS ... loses< Message-ID: <aek-2501021349190001@il0502a-dhcp193.apple.com>  J In article <3C51C1D3.7FC9EE9C@virgin.net>, Alan Greig <a.greig@virgin.net> wrote:  B > Well if you've got some spare time pop over to alt.sys.pdp10 andF > volunteer your services to help upgrade  10/20 Fortran to Fortran-90 > ;-)-  E heck, we'd be happy if someone still has the sources for the EXISTING<	 compiler.e   ------------------------------  # Date: Sat, 26 Jan 2002 04:41:28 GMTE0 From: William Barnett-Lewis <wlewis@mailbag.com>) Subject: DHCP as a client using Multinet? + Message-ID: <3C523375.B69AAFA1@mailbag.com>n  D I'm probably missing something obvious, if so please point me at the" right place in the documentation.   G I have a router that provides a DHCP server and I would like to have my D Vaxstation get it's IP from that. I have installed Multinet and it'sE PAK. I am not able to find in the documentation how to set it up as asG DHCP client. Again, if this is an FAQ, a pointer in the right direction  is appreciated.    Thanks,n   Williamh --  * You better watch out    What you wish for;+ It better be worth it   So much to die for.e-                                 Courtney Loveg   ------------------------------  % Date: Fri, 25 Jan 2002 20:32:53 +0000w% From: "a.carlini" <arcarlini@iee.org>rB Subject: Re: ethernet boot diskless VAX-11/730--the question first' Message-ID: <3C51C0F5.3DF3874E@iee.org>l   Joseph Ballantyne wrote: > D > I may have included too much contextual information in my originalJ >    query, which may have deflected attention from my questions.  So: forI >    a MOP boot (via DEUNA or DELUA), what should be the device type code.! >    in GR0<7:0> when VMB starts?r  8 I have not looked at the listings, but according to the 7 V5 internals manual, 97 (decimal) is the DEUNA. I nevere5 managed to hack a VAX-11/750 into netbooting and I'veo6 never heard of anyone managing to netboot a VAX-11/7xx machine of any sort.    C > I'm optimistic on this point.  VMS doesn't need anything from thehI >    console TU58 in normal operation, after it's booted.  I had a habit,s  6 Well you have a VAX-11/730 and I don't. I do, however,4 remember the DTJ article detailing the work done for4 the MicroVAX II port and that used the VAX-11/730 to0 test the requirements for instruction subsetting2 because it is a completely soft-loaded machine. It must load from *somewhere*.l  $ If you do get it to netboot (or even$ pump anything out on the network) be% sure to let us all know. (I remember  $ going through the same exercise with% a VAX 8350 - it was close but it camer$ a cropper somewhere in the secondary
 loader IIRC).t   Antonior   -- a   ---------------d- Antonio Carlini             arcarlini@iee.orgd   ------------------------------    Date: 26 Jan 2002 05:16:33 +0800, From: Paul Repacholi <prep@prep.synonet.com>: Subject: Re: Even Intel thinks Itanic isn't Proprietary...- Message-ID: <87y9imrv0u.fsf@prep.synonet.com>   / wspencer@ap.nospam.org (Warren Spencer) writes:t  B > It may be Intel itself that's promoting the notion that anythingA > they make is "industry standard".  Here's an excerpt from their ! > Itanic product brief, found at:-G > http://www.intel.com/ebusiness/products/itanium/overview/ds010401.htm   bA > "The powerful new processor family extends open-standards-basedIA > computing to the enterprise, delivering flexibility, choice andC$ > value over proprietary solutions."  ? Is this the new Yellow Terror, or the old itanic? The itanic iss about to turn 10 remember...   -- i< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.i@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Fri, 25 Jan 2002 21:40:43 GMTm4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk>' Subject: Re: fopen crashes while in ASTe0 Message-ID: <3C51CFE7.888C5E79@blueyonder.co.uk>   JF Mezei wrote:p > > > Program uses $QIO to connect to http server on the internet. > P > every 10 minutes, a timed AST is fired and it creates a new configuration file& > as a way to checkpoint its progress. >  > The problem: > O > The AST works fine the first time it is called (after about 500 transactions)nG > but upon its second invocation, it will generally fail at the fopen() P > statement. The filename has been verified to be valid. HOWEVER, there are someL > image activations where it won't fail (one completed succesfully, and some. > other failed after more than 2 invocations). > M > The AST does a complete "loop" (i.e one fopen, a few fprintf, then fclose).b- > This file is not used nor opened elsewhere.	 > L > The core of the program was used succesfully last week to build sequentialM > files and the AST ran without problems for a combined run time of about 334rW > hours. and close to 300,000 transactions (I ran multiple processes at the same time).y > P > However, this weekend, I need to run through all those records and update themN > with some additional information, also fetched from the internet. So I addedO > some RMS calls ($OPEN, $DISPLAY, $CONNECT, $GET, $UPDATE, $FREE) that fetch alO > record from the indexed file, use the existing code to get the infor from thesP > net (different URL) and then use $UPDATE to write it back to the file with the# > formerly empty fields now filled.t > O > This works fine. Except for the AST that causes image to crash after a while.i* > (about 1000 transactions or 20 minutes). > M > I have tested that the character string that holds the filename supplied tomP > the fopen statement is correct. And fopen does not create a new version of the# > file (so it doesn't get fer far).r > B > What additional steps could I do to  find out what is going on ? > N > If I run it with debugger, once it crashes, how can I find out what it wrongJ > since the routine will be inside the RTL and I won't have source code to > detect what is wrong ?     JF, a few comments  Q Some code you cannot mix between AST and user mode. I'm not sure about the C fileoS calls but I suspect not. You could try just setting and event flag in the AST then rP checking for it set in the user mode code loop and refreshing the logfile. That M way you AST does virtually nothing, all the C io calls are made at user mode.i  Q regarding the debugger, turn on the dsiplay of the actual instructions issued ands) get your instruction set manual out :-). d  S If you posted the actual screen dump info from debug maybe someone else could help,q if suggestion 1 doesn't.   regardsi -- i Tim.Llewellyn@cableinet.co.uk  n  C Standard disclaimer applies. My views in no way represent those of h! my employers or service provider.    ------------------------------  % Date: Fri, 25 Jan 2002 18:08:50 -0500o% From: JF Mezei <jfmezei@videotron.ca>e' Subject: Re: fopen crashes while in ASTe+ Message-ID: <3C51E580.81592BA@videotron.ca>    Tim Llewellyn wrote:S > Some code you cannot mix between AST and user mode. I'm not sure about the C filed > calls but I suspect not. a   How can I find out ? o  G What i do know is that the IO to that file is done entirely in that AST N routine and nowehere else, and the open->write->close cycle is done all within# a single invocation of the routine.n    I What I did find which baffles me is that 3 of the time quadwords had beentN allocated as single longwords. That checkpoint routine also calculates variousK seconds per transaction and completion time estimates and write it to a loge file that is always opened.o  I This program ran fine for over 260,000 transactions over about 300 hours.lF (actually, multiple processes running concurrently to amass the data).M Considering that a few time quadwords were not allocated long enough, I don'ty; quite understand why it ran and displayed correct results !S  M The worst part is that this week, I did not recompile/change the module which E contains that AST routine. So I am at a loss to explain what would beoH different in that module once linked that would have generated differentK results because it was linked to a main module that was slightly different.o  K The other puzzling thing is that the time routines would work fine, even onoK the second invocation of the faulty AST routine (they are done prior to thesL fopen() which caused the crash). BUT, when I compiled/DEBUG/NOOPTIMIZE, thenN the results produced were clearly out of whack. (at whcih point I added printfL statements to see the status codes for the time routines and that lead me toG reasile the bad allocation). But it would take half an hour before each-# failure so this is time consuming !o    J Anyhow, I have fixed the definition of the time variables and have begun 4K processes running in the hopes that the problem is fixed and that they willd continue to run to completion.  M Right now, it looks like 1.7 seconds per transaction, so if the net holds up,t. it may complete before the end of the weekend.   ------------------------------    Date: 25 Jan 2002 12:29:04 -0800+ From: davidc@montagar.com (David L. Cathey)hD Subject: Re: Gartner Group lies about VMS ... loses all credibility!= Message-ID: <e565ed03.0201251229.4ecfee84@posting.google.com>m  ` "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message news:<3C50CF59.E122C764@fsi.net>... > "David L. Cathey" wrote:
 > > [snip]d > > "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message news:<3C4F87E2.8F956AC3@fsi.net>... > > > [snip]M > > > I'm not asking for "free" - just don't take the food out of my family's-J > > > mouth, the clothes off their backs or the roof from over their head. > > " > > Yes, you are asking for free.  >  > No, I'm not. Read it again.   $ Then how much can you afford to pay?  l- > > > Why is that so difficult to understand?2 > > + > > Because you appear to want a free ride.e > E > No, I don't Read it again, and again, and again... until you get it> > right.  $ Then how much can you afford to pay?  N > > Nope...  Still no magic bullet.  Maybe a large number of differently sized% > > and aimed slightly-magic bullets.  > : > Well, yeah, there is. You're just not willing to see it.  H I see things very well.  You think things can be solved very simply, but, we pressed for details, you go on a tangent.   > Q > > > > > Getting 10, 50 or 100 machines respectively, fully outfitted with o.s.,)P > > > > > warranty and three years of platinum-level support. Third-party app.'s4 > > > > > are, of course, another question entirely. > > > >l8 > > > >         Here we go - free hardware and software. > > >e( > > > Try $10,000 each. Check your math. > > ( > > Nope, you're wanting them for free,  >  > No, I'm not.  G So, are you telling me, that if I can deliver to your boss for $500,0000F 50 Alphas with OpenVMS, he'd buy them?  Give me his phone number, I'll  get him touch with some folks...  S) > > after all, I did ask you want kind ofJL > > marketing program you wanted - you wanted a 10+ boxes.  You neglected to > > mention the price, >  > See your example, above.  K Give me your bosses phone number.  Since you're unwilling to make the sale, I I'll do it.  Of course, if you're not telling the truth that your boss ist5 willing to put up the cash, then don't waste my time.0  17 > > but given your continued comments, I very seriouslym/ > > doubt that you could afford a $10,000 box. S > 4 > You asked about my boss - you didn't ask about me.  M Let me ask him, then.  Here, I'll even give you that parts list, you can omit G almost $700 per system if you drop the monitor/keyboard/mouse and use an= serial console (I pulled this from ActiveAnswers last night):   I Quantity Part Number     Description        Estimated Retail Price* (USD) 8 1        DY-74BAA-DA     AlphaServer DS10, 600, OpenVMS,H                          256 MB, no cage, CD                   $7,587.00H 1        3X-BA10A-AA     DS10 Internal Storage Cage             $ 200.00H 1        BN39C-02        1.8 m video extension cable            $  25.001 2        3X-BC34A-06     6 ft. keyboard or mouse IH                          extension cable                        $  40.00: 1        3R-A2936-AA     P710 17 inch (16.0 inch viewable)H                          1280X1024 Monitor (Northern Hemisphere)$ 544.00H 1        3X-DEPVZ-AA     PCI LVD SCSI, 2D Graphics, Ethernet    $ 999.00H 1        DS-RZ3DD-WA     9.1 GB 10K RPM Ultra3 SCSI Disk        $ 306.001 1        LK461-A2        VMS/VT style Keyboard - oH                          North American/English                 $  75.00H 1        BN26J-1K        Power Cord - North America             $  11.00H 1        3X-BN51A-SE     SCSI Cable kit                         $  75.00  *                 Estimated Total $ 9,862.00  I Knock off the monitor and stuff, at by the time you add up the difference J savings under $10,000 for quantity 10/50, you can add in extra support andA CONDIST.  There's your system.  I'd love to see a copy of the PO.e  - > > After all, you are totallyI > > unwilling to believe that a $1000 CSA is "affordable" to a developer.i > E > ...who has a $1000 budget. I will, too - but not until I get my tax.	 > refund.u  N I'm a developer, and have $1000 in my budget.  I know several other developersK with $1000 in their budget, too.  Just because you don't have $1000 in yourI8 budget does not mean that the program itself is lacking.  M > > > Based on what criterion? Size? Business? Use of IT within the business?  > > R > > Based on your "Ready, Fire, ... Aim" strategy which you believe that marketing > > should employ.   > I > In context, the reference was specifically to that element of marketings* > we call "advertising", as you'll recall.  K I'm just using your rules.  You don't like them, let's try another scenario  and run the numbers.  tB > > It doesn't really matter, as long as you get as many people to+ > > hear about OpenVMS as possible, right? o > G > That's the general idea. We're so far behind the curve at this point, G > many of them were "ascendant" while VMS was declining. It's news to a $ > bunch of them. "VMS? What's that?"  J Again, you don't believe that matters.  Just target anyone, doesn't matter% if they are a likely customer or not.a  9 > > > A bit random, but I'll play along for the moment...3 > > B > > So is "Ready, Fire, Aim" as your suggested marketing approach. > 1 > There's a difference. *THINK*, man, *THINK* !!!   5 Hey, your scenario - I'm just seeing where it goes...T   > > Do you not get it? a >  > No.W > 8 > > Giving away these machines IS the marketing program, > H > It is? Who suggested that? You suggested an example which I equated toI > 10 machines for $100,000, 50 machines for $500,000 and 100 machines for 
 > a megabuck.F >  > What did I miss?  F I guess what you missed is that every time you talk about pricing, youI start talking about $1M glasses of water.  I asked what marketing programoH would be successful in attracting your boss as a customer, and likely toH spend money on Compaq products.  I've never heard you be happy about anyK price every discussed, so when you want 10+ machines dropped in your bosses J lap - it seems obvious that you expected them gratis as a "lost leader" in; order to get something into a customers hands to play with.l  J The $5B was the (incorrect) cost of this marketing effort.  You're correct$ that it's $500B - way too expensive.  I So, let's ask the question again - what's going to light a fire under youbJ boss to get him to buy Compaq products - what message, what magazine, what	 anything.i  N I don't think it's a $10K system, since they've had that (and cheaper) before.  K > > F'-ing-A is right.  To accomplish your marketing program, it would costC > > Compaq over $500 Billion.  > J > Oops! You flipped the sign bit! That's dollars *TO* Compaq, not from it! >O4 > > A little steep.  If you think Compaq was selling! > > them - well, they already do.0 > # > Exactly what I was talking about.w  J Okay, let's assume that if Compaq had a $10,000 box, they would sell $500B of them.  K To who?  Not your boss.  He's not going to cough up $500,000 bucks.  Why donI I know this?  Because it's out there, and he hasn't coughed it up.  If hesM would, please give me his phone number, because I'm going to make a few bucks<
 really quick.9  Q > > Yup, I underestimated it quite a bit - it would cost Compaq DECADES of annualp: > > profit to recover from this little marketing exersize. > H > How so? Seems to me they'd reap a decade or two of profits from such a > "fire sale".  J (Hard to resist) Evidence, please.  You have not shown any level of demandK that if Compaq had a $10,000 Alpha/OpenVMS box.  I can show evidence of thewI contrary, since you can buy a $10,000 Alpha/OpenVMS box TODAY!  NOW!  And C that's at Standard List!  Yet, they are not flying off the shelves.v  O Show me the evidence by having your boss cut a PO for the above system outlinedr above.  N > > Do I need to ask the question again?  What magazines do you read?  What TVG > > shows do you watch?  Where would an OpenVMS advertizement find you?  > I > Let me put it this way, though it does seem futile (your willingness to,I > answer questions is, well, don't go there): what magazoids does OpenVMShF > advertise in? What audio/video streams does OpenVMS buy ad. time on?  G I don't know all of them.  However, I have seen the prints of the Ads,  E pictures of the Ads in place (even at bus stops!), and more.  You sawwI there is no evidence.  Yet I've seen evidence.  I can't help that you aresK either not looking, unwilling to even see it in the obvious places (OpenVMS I publications which mention stuff like this).  I've told you where you cano. find it - any futility is solely on your part.  dJ > OpenVMS ads would find me in the mainstream: CBS Radio network news (viaE > their local affiliate), major market newspapers (Chgo. Tribune, SunMD > Times), CNBC and MSNBC, The Weather Channel - y'know, all the same( > places Wall Street Journal advertises.  J No trade journals or industry journals, though?  Interesting.  How do you 4 keep up on what is happening in the industry itself?  n= > > I get mailings, e-mailings, brochures, and more.  Do you?t > D > Suffice it to say, no. What mailing list yields this? What are the# > requirements to be so privileged?u  D I don't know why.  I don't feel privileged.  Did you at least get a H bouncy, blinky rubber ball?  If not, then you are appearently completelyK off the radar screen - Compaq Marketing appears to not know you even exist.o  O I guess I need to ask the question again - have YOU bought any Compaq products.tM I mean you - where your name shows up on the invoice (even if your name isn'toI on the check)?  Are you a designated contact point for Compaq support for L your business as part of a service contract?  Anything where your name wouldJ show up anywhere as part of an actual Compaq Customer database, especially an OpenVMS customer?  M > > Oh boy, have you never been involved in a procurement cycle...  I had onepQ > > last year where we were the best product, best price - and still they awardedsL > > the contract to some other company.  "Best Offer" doesn't always get the > > job. > J > If they bouight from the other guy, then you were not the "best offer" -G > in *THEIR* eyes! The prospect's opinion is what matters, not yours orc- > mine. You won't win them all - get over it.c  I Hey, maybe you're getting the clue, if even a little bit.  The concept of G "best offer" is very political.  In my case, the client had decided whorJ they wanted even before the Dog-And-Pony shows.  Really didn't matter much what our "offer" was.e  N Compaq has to deal with this quite a bit, too.  Now, how does Compaq deal withN putting out a best offer that they will accept, if their consultant (you) tell> them Compaq is killing OpenVMS and will not keep a commitment?  P > > > >         Customers are the leverage, not the O/S.  Driving away customersO > > > > undermines the leverage to attracting more customers.  That is what you P > > > > want, isn't it?  More OpenVMS customers?  That is what you've stated youD > > > > want to do to achieve that, isn't it?  Drive away customers? > > >e5 > > > Do you consistently interpret things backwards?u > > ' > > It's not backwards.  Read it again.  >  > I did - it's still backwards.r  L Again, back to the procurement cycle - often you need reference sites.  MoreL customers gets you leverage as part of the offer - you not only can talk theN talk, but these customers demonstrate and can be refered so that your prospect knows you can walk the walk.  O > > You really need to decide what affordable means.  In one place you say thatf > > $10,000 is affordable, r > J > I did? i only recall saying I'd go along the example - I never commentedC > any further. Be careful about reading things into what's written.S  L Okay - then $10,000 isn't affordable.  How much can you afford?  Are you nowL telling me that your boss will NOT buy a $10,000 box?  Much less 10, or even! 50 of them?  I thought as much...u  7 > > other places you say that a $1000 cost puts you outiO > > of the market.  Frankly, I can't take any "affordability" comments from youtA > > until you figure it out for yourself.  OpenVMS is affordable.n > E > To whom? Certainly not "startups" - we can't even get into the CSA!l  E J. H. Christ - what do you think I was went I first got into the ASAPsH program?  A multinational corporation?  You're not a startup, either, so don't start with that nonsense.x   > > Alpha isn't dead,  > > > Has "the Q" recanted their proclamations of June 25th, 2001?  I Alpha isn't dead until they make the last one.  This hasn't happened yet.wF Actually, when they first produced the Alpha, they refered to it as anF architecture for the next 25 years (read the really early Alpha info).K The Alpha came out in late 80's, first OpenVMS commercial ship in Nov 1991.nK The current end-of-Roadmap is about at 20 years lifespan for the Alpha - ofoH which there will be Alpha's likely for that extra 5 year period of time.K Not too bad considering the number of technilogical revolutions in the chip D making industry.  The original chip designers called it about right.  M > > > Is Cerner Millenium certified to work with that as well as with Oracle?  > > M > > non-sequitor - you stated porting software was illegal, I demonstrated it L > > isn't.  I'll take this diversion as a statement that you were incorrect. > H > Wrong again. The statement was intended to illustrate the viability ofJ > attempting a port - i.e., is there a market where porting costs could be > recouped.i  L Then you're still wrong.  You think Oracle 9i isn't ported to OpenVMS?  Your own statement proved you wrong.m  
 > Get it now?   M I get it.  You can't afford to port your applications, since $1000 investmente to the game is too big.s  I > > Commercial software will have to ported by the company that owns it. e > $ > ...or their (sub-)contractor(s)...  I True.  And if their subcontractor can't afford the $1000 investment, then-K either they've chosen the wrong subcontractor, or the subcontractor did not. price the job right.  	 > > AfteriD > > all, a Hobbyist is not going to be able to get the source code.  > G > Rather a brash assumption. One might, but under the terms of the OVMSeG > hobby license, commercial work is verboten. Thus, our quandry, unlessoC > one can cajole the vendor into a cash advance to acquire the CSA.a  K You include that as part of your bill (God, I feel like I'm trying to teach H you how to run your business...).  That or you set your rate high enoughI to insure that you recoup your costs.  Or you do it as a fixed-bid, wheregB you know that your bid-price is sufficient to cover those costs.    D > Oh, but there's that whole "years in business" artificial obstacle > blocking the way.o  F Didn't stop me.  The "years in business" obstacle I saw was in the SunE Developer.  You've been in business for "years", at least that's theyaD way it appears by looking at your domain name registration.  If not,F I can try to talk to the CSA Program on your behalf.  Or you can get a0 local OpenVMS Ambassador do plea on your behalf.  ; > > Okay, not only do you want free hardware and software, . > , > No, I don't - never said that, never will.   How much can you afford?  d > > you want Compaq toN > > buy your family clothes and food so you can play on your free hardware and
 > > software.h > I > Jeez, Dave, maybe you should go get a coffee or dinner or something and E > send me the bill. You're just all bent outta shape. You're not evensG > making twisted sense anymore. Take John and Pat with you. I'll borrow = > against my next paycheck to cover my bills and your dinner.   J Then what kind of "financial incentives" are you talking about?  Or do you@ have a different definition of "financial incentives" than I do?  hM > > But that assumes you're an ISV (i.e. CSA), which you're not, since that'sa > > too expensive. > A > Well, everyone gets SOMEthing right - even you - now and again.p  K The key point is that it's too expesive for you.  But, in your example, you.G can't afford a 5 cent glass of water, can you?  Does that mean that theo< glass of water is truly too expensive, or you're flat broke?  < > > Buy hey, it's not like your being unreasonable, right??? >  > No, I'm not. Prove me wrong.  L I've appearently shown that a $1000 inventment into a the computing industryN is too high of a price for *you* to afford.  I've yet to determine, and you'veJ still never answered, what price can you afford?  Actually, does it matterB what platform we're talking about?  You can't afford to get in theI Windows market, either (MSCE is more than $1000, too, not to mention box,n software, MSDN on top).   M > > I've said it before, and it's what Compaq is doing.  They are going after R > > those customers where they have a strong history, and demonstrated superiorityL > > in the market.  The medical markets, telecom, banking, etc.  And working > > from there.t > I > What about the rest? What about he ISVs? ...VARs? ...OEMs? ...any othero > partners?a  H Who do you think provides the applications that run on OpenVMS for theseK markets?  Hell, you mentioned Cerner yourself - they are an OEM for Compaq.e  < > > You don't work in any of these markets, I guess, do you? > $ > Healthcare? Been there since 1997.  H Good!  Then you know the market there is good, especially with security, HIPA and other things.  iM > > > So far, all you're doing is arguing. Stop being part of the problem andaF > > > become part of the solution: propose answers to these questions. > >  > > I have . >  > Hhmmm... can't find that...y  I No surprised...  Everything I've mentioned, you claim is unaffordable, or ' simply doesn't exist (even if it does).    G > > - for every answer I proposed, you shot back a "won't work", "can't0 > > do that".  > I > Huh... can't find that either. Are we both reading the same thread? The F > IE minions insist that Netscape is buggy - maybe that's the problem. > ' > > The porting of apps is one of them.e > E > Which brings us back to the question: how do we get ISVs to port so C > there'll be app.'s so there'll be demand so there'll be a market?y  F Well, that's why Compaq has been working with ISV's that they consider< strategic to the OpenVMS business.  Or didn't you know that?  Q > > > >         Let me ask an obvious question.  David, do you have an idea for anO > > > > commercial software product that you're wanting to create, and the onlyiS > > > > thing standing in your way is that you cannot afford to purcahse an Alpha +uN > > > > CSA Membership + OpenVMS SDK in order to create this software product? > > > > V > > > >         You seem to be in a real pickle, then.  You can't develop a commercialQ > > > > application using the Hobbyist Program.  But you can't afford to join the( > > > > CSA Program, either. > > > >iU > > > >         Assuming you do have an idea, maybe I can lease you time on one of my 9 > > > > systems for you to develop your software product.  > > > N > > > Hold that thought - and be prepared to offer it to anyone else who lurks- > > > or actively participates in this group.  > > N > > So let's be clear on this - assuming adequate hardware/software - how muchL > > could you afford to lease a development environment?  And let's not hearP > > anymore of that "dying of thirst without a penny to your name" crap and want > > it for free. > $ > No, *YOU* said free. Get it right!  N I said I wouldn't accept that as an answer.  You're dodging the question.  How much can you afford?  I > I'm already paying a 1st mortgage, a 2nd mortgage and an outrageous car0H > payment (the wife *INSISTED* on that Ford Explorer!). So, you tell me:I > what's the "pound of flesh" worth to you? (...and remember, you may get J > other takers by token of this newsgroup. Say "marketing" - it takes many > forms, does it not?)  G You're still not answering the question.  If I could provide an OpenVMSnH system with a couple Gig's of disk and some basic software licenses, howI much could you afford?  You never really want to say free, but can't seem K to tell me what level above that would satisfy you, either.  No matter whatf9 cost I place on this, you'll still claim "Unaffordable!".1  K I can't give it to you for less than my cost.  And I suppect that even justlJ paying the electic bill would be too much for you to afford.  I can't takeH food out of my kids mouth to provide you a system just because your wife wanted an Explorer.b  M I can't take your claims of unaffordable seriously.  You base those claims onfK your personal finances, not on a legitimate business plan.  Then you try toiP paint everyone (all potential ISV's) with your financial brush, which is absurd.N Sorry David, but there doesn't appear to be a Compaq, Sun, HP, Dell system outL there that you can afford.  Anywhere.  That's not the fault of the industry,( and you can't blame the industry for it.  O Isn't irony fun - I just got off the phone with someone who would be interestedtH in buying a development system just like the $10K system I quoted above.   ------------------------------  # Date: Sat, 26 Jan 2002 03:46:21 GMT 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>mD Subject: Re: Gartner Group lies about VMS ... loses all credibility!' Message-ID: <3C5227A6.C2B83D98@fsi.net>d  E I snipped the rest of it because we're just going to have to agree torA disagree on those points. You're so deeply entrenched in your ownlB position that you'll never be open enough to see the world from myF perspective. So, I'll have to chalk that up as another case where whatH I'm doing isn't getting the right result, so change my approach - exceptE I'm convinced that there just is no "right" approach at this point ins time.    "David L. Cathey" wrote: > I > You're still not answering the question.  If I could provide an OpenVMSiJ > system with a couple Gig's of disk and some basic software licenses, how > much could you afford? ?  C I'm going to have to make some brash assumptions here - getting theyG details out of you would take too long and maintain too much noise hereg
 in the ng.  H I'll assume you're talking about a subscription to a timesharing service@ or it's equivalent, including access to compilers, SDKs, etc. as available on your setup there.  3 > You never really want to say free, but can't seemsM > to tell me what level above that would satisfy you, either.  No matter whats; > cost I place on this, you'll still claim "Unaffordable!".   G Careful about jumping to conclusions, reading in what isn't there, etc.5G I snipped a prime example of that latter point, so I'll not retrieve itn now.  0 > I can't give it to you for less than my cost.   E If you can itemize your costs, I'd offer you cost + 20% as a startingnE point in negotiations, with the stipulation that I start getting some H respect. You can't know where I'm at unless you've been where I've been.B I've been more than patient and I've swallowed a lot of undeservedE insults. Had that not happened, I'd probably have offered a full 100%I	 mark up.    E You'll notice I refrain from such affronts as much as I possibly can,t2 but don't always succeed as well as I might like.    > And I suppect that even justL > paying the electic bill would be too much for you to afford.  I can't takeJ > food out of my kids mouth to provide you a system just because your wife > wanted an Explorer.h  E There you go again. Best to stay out of people's family life (word toeE the wise). (Yes, my patience is wearing extremely thin.) ...and don'ti$ you *DARE* say that I brought it up!  O > I can't take your claims of unaffordable seriously.  You base those claims ono= > your personal finances, not on a legitimate business plan.    C Doesn't matter. A business can't commit funds it has no source for,vD either. If you don't have the cash or the credit, you're toast. JustH because the entity is only a legal entity instead of a live person makes no fundamental difference.   > Then you try to R > paint everyone (all potential ISV's) with your financial brush, which is absurd.  G ...or so you claim. I speak only for myself and on behalf of anyone whotG is in or can identify with my position. Others are excluded implicitly.u  P > Sorry David, but there doesn't appear to be a Compaq, Sun, HP, Dell system outN > there that you can afford.  Anywhere.  That's not the fault of the industry,* > and you can't blame the industry for it.  , Then whose "fault" is it? Who set the price?  A You might do well to consider the percentage of gross income that H "enterprise" companies allocate for IT spending, then try to apply thoseH same criteria to smaller "enterprises" ($10 million gross or less). ThenH you might begin to understand. Dunno. I'm sure you've got a comeback for that, as well.  Q > Isn't irony fun - I just got off the phone with someone who would be interestedhJ > in buying a development system just like the $10K system I quoted above.  D Then I guess all these posts aren't so bad after all! You just got aE prospect! Had we not been discussing such systems, you would probablyeF have thought of it eventually anyway, but aren't you glad it was fresh
 in your mind?u   -- v David J. Dachterae dba DJE Systemss http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/?   ------------------------------  # Date: Sat, 26 Jan 2002 04:39:42 GMT 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>hD Subject: Re: Gartner Group lies about VMS ... loses all credibility!' Message-ID: <3C523428.D080A22D@fsi.net>a   "David L. Cathey" wrote: > [snip]& > Then how much can you afford to pay?  5 This is exactly the point I can't seem to drive home.r  ; What *I* (dba DJE Systems) can afford to pay is immaterial.s  F The issue at hand is what the *END USER* (read: my (or your) customer)E can afford or cost-justify, and VMS ain't it 95% or more of the time.?  H Could I scrape up a grand for the CSA? Sure. It'd be a lean month or so,3 but I could without harming the household too much.a  F But what would it matter? I could devote the rest of my useful life toH porting the kitchen sink to VMS - but if no one (or very few) can afford VMS, what's the point?   What can your customer afford?   Ask *THEM* !!!   -- u David J. Dachterad dba DJE Systemse http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/t   ------------------------------  % Date: Fri, 25 Jan 2002 14:03:53 -0500*- From: "John Eisenschmidt" <jeisensc@aaas.org>t* Subject: Re: I/O Performance on Alpha 4100+ Message-ID: <sc5165e5.018@AAASMTA.aaas.org>e   Brian,  J I ran into the same thing last year. While I'm not familiar specifically =J with the 4100, some people mentioned looking at HSG controllers (like an =C HSG-80). Our problem was that those controllers are really pretty =o2 expensive (more than we spent on our whole setup).  I I think a lot of it depends on how much capacity you need, what kind of =tI performance you want, and how much you want to spend. We ended up going =pI with KZPCA-AA controllers (stupid LVD SCSI controllers) and Compaq 4254 = J disk shelves (split bus U160). We threw a bunch of 10K RPM Compaq drives =H in them and went with volume shadowing. We chose this setup because we =F needed the absolute fastest disk I/O we could get for as cheap as we =I possibly could. For what we paid, the performance has been great - much =pF better than we would have gotten with an RA-3000 or probably even an = RA-7000.  C On the other hand, that kind of setup may not suit you. Some more =t information would be helpful.o  9 >>> BrianNFO <briannfo@aol.com> 01/25/2002 1:35:54 PM >>>tI I haven't posted here in a while, but I'm looking to speed up I/O on an =d AlphatD 4100 5/400 running VMS 7.2-1.  Most of the devices are things like = RZ40-VA, andF I believe the VA indicates narrow scsi?  I don't know exactly what PCI controllers are in use.e  I Applications on this system are I/O bound, and I would think I can make =a someJ big improvements through either a caching controller, or a combination new) controller, wide shelves and wide drives.n  % Any suggestions would be appreciated!y   Brian0   ------------------------------  + Date: Fri, 25 Jan 2002 18:40:22 +0000 (UTC)c From: david20@alpha2.mdx.ac.uk2 Subject: Re: Inquirer : "Intel *could* can Itanic"+ Message-ID: <a2s8qm$gfu$1@aquila.mdx.ac.uk>   m In article <Zph48.10559$vH6.584445@bin6.nnrp.aus1.giganews.com>, "Bill Todd" <billtodd@metrocast.net> writes:t >e3 >"JF Mezei" <jfmezei@videotron.ca> wrote in messagei' >news:3C519B04.534DE065@videotron.ca...  >> Alan Greig wrote: > M >That would certainly be Intel's easiest (and quite possibly *only*) route toxJ >obtaining a platform that could compete head-to-head with POWER:  even ifM >Itanic sinks, it's not clear how Yamhill (and Hammer) would fare in the high'4 >end, given their need to cover the low end as well. >AM >But from a political standpoint ditching Itanic for Alpha (rather than usingaI >the Alpha team to patch it up as much as possible, or even to completelyeJ >replace the underlying technology but retain the Itanic ISA, as some have$ >suggested) would be very difficult. >i .b .l .'L >It makes a lot of sense, but I wouldn't hold your breath.  Especially sinceL >Intel does not own *exclusive* rights to Alpha as it does to Itanic, and myG >impression is that they feel exclusivity is important here.  And sincefL >Compaq has already licensed Alpha technology to others, obtaining exclusive; >rights may now be impossible (though that's just a guess).s >. >- billk >h  L Of course if the Compaq/HP Merger fails and Intel wanted to obtain exclusiveK rights they might just purchase Compaq. As an additional benefit they would.# then own the OS running their fabs.tH (Whether Intel would then do anything with VMS other than tweaking it to> improve its performance when running a fab is another matter).  
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  % Date: Fri, 25 Jan 2002 12:42:47 -0800l% From: Dean Woodward <deanw@rdrop.com>m2 Subject: Re: Inquirer : "Intel *could* can Itanic") Message-ID: <3C51C347.BE109AED@rdrop.com>c  G I'm a Portland, OR local- You have no idea the giggle fits I go throughnF when I see some of the Intel code-names.  "Tualatin" and "Yamhill" areG both "rivers"- narrow, muddy, twisty bits of slothful watercourse. Draw   your own conclusions, there. ;-)  A Tualatin is also a suburb of Portland, while there's a town namedcG Yamhill- population < 1000, I'd guess- the kind of place you don't wantt6 to blink as you drive through, for fear you'd miss it.  G As for "Rolner Acres", decades ago it was a failed housing development, G full of rolling piles of dirt, and used by the locals as an after-hours G trash dump.  In particular, it was a favorite place to abandon vehiclesgD while they could still get there under their own power.  Again, draw your own conclusions...w   ------------------------------  # Date: Sat, 26 Jan 2002 04:22:54 GMTn1 From: "David J. Dachtera" <djesys.nospam@fsi.net> 2 Subject: Re: Inquirer : "Intel *could* can Itanic"& Message-ID: <3C523000.2C6B1E5@fsi.net>   "Dr. Dweeb" wrote: > I > While busily writing my processor independent DCL, I was tempted to addm > 9 > $ If f$getsyi("arch_type") .eq. 3 Then arch_type := IPF  > I > But perhaps I also should add this as well - to be on the safe side :-)r > 9 > $ If f$getsyi("arch_type") .eq. 4 Then arch_type := AMDu  D Why not just use F$GETSYI( "ARCH_NAME" )? It works across nodes now, doesn't it?s   -- a David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------    Date: 25 Jan 2002 15:47:41 -0800. From: hemingway_n@hotmail.com (Neil Hemingway) Subject: Kerberos in VMS= Message-ID: <4c5b7963.0201251547.1b2c848c@posting.google.com>w  C In message <8m9bn4$39g$3@mailint03.im.hou.compaq.com>, Hoff Hoffmana wrotef  H >   The next protocol being added for external authentication support isM >   Kerberos (both the client and server), and this is expected to be part ofv >   the OpenVMS V7.3 release.O  C 7.3 is now out, and I haven't found any mention of Kerberos in it. sF Did it slip from this release, or am I looking in completely the wrong place?   ------------------------------    Date: 25 Jan 2002 18:50:18 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)  Subject: Re: Kerberos in VMS3 Message-ID: <OMNffThUfq6a@eisner.encompasserve.org>r  n In article <4c5b7963.0201251547.1b2c848c@posting.google.com>, hemingway_n@hotmail.com (Neil Hemingway) writes:E > In message <8m9bn4$39g$3@mailint03.im.hou.compaq.com>, Hoff Hoffmand > wrotei > I >>   The next protocol being added for external authentication support is?N >>   Kerberos (both the client and server), and this is expected to be part of >>   the OpenVMS V7.3 release. > E > 7.3 is now out, and I haven't found any mention of Kerberos in it. dH > Did it slip from this release, or am I looking in completely the wrong > place?  E I thought it was a layered product for now, but distributed on the OSl CDROMs.k   ------------------------------  % Date: Fri, 25 Jan 2002 17:09:52 -0500b% From: Sarah Page <Sarah.Page@sas.com>tO Subject: Re: MORE!!  Problems with IO$_SETMODE | IO$M_WRTATTN I/O Function CodeV2 Message-ID: <=9VRPJAaZWRhlDYbc87oYyRPcKno@4ax.com>  E OK, now that I've got the arguments in the right place and the AST ise= delivered at the correct time, can anyone tell me why the ASTe: argument, which is a long pointer (8 bytes), ends up beingD sign-extended by the time the AST gets hit?  The QIO() doc says that@ p1 and p2 may be 64-bit values on Alpha.  FYI, my application is compiled w/ pointer_size=long.   I.e.,s   char *testarg; i .. .  . 2 /* At this point, testarg is 0x0000000080079b70 */C status = SYS$QIOW (0, nmbx_chan, IO$_SETMODE | IO$M_WRTATTN, &iosb,,# 0, 0, &myAST, testarg, 0, 0, 0, 0);t  " static void myAST( char *testarg ) {o3 /* At this point, test arg is 0xFFFFFFFF80079b70 */t }"    
 Thanks again!n  C On Thu, 24 Jan 2002 09:51:13 -0500, Sarah Page <Sarah.Page@sas.com>n wrote:  D >Thanks a ton.  Yes, I'm embarassed about having the arguments mixedD >up, it should've been obvious based on the results I was getting ;) > F >On Wed, 23 Jan 2002 17:32:12 -0500, John Santos <JOHN@egh.com> wrote: >s( >>On Wed, 23 Jan 2002, Sarah Page wrote: >>J >>> I create a temporary mailbox (nmbx_chan) and then immediately make the >>> following call:' >>> G >>> status = SYS$QIOW (0, nmbx_chan, IO$_SETMODE | IO$M_WRTATTN, &iosb,y* >>> &myAST,  nmbx_chan, 0, 0, 0, 0, 0, 0); >>> F >>> The way I read the doc, this call says, "Call the myAST() functionG >>> when something is written to the mailbox and no reader is present."  >>> J >>> myAST() is hit immediately (even though I *know* that NOTHING has beenJ >>> written to the mailbox).  myAST() does a QIOW() on nmbx_chan using theH >>> IO$_READVBLK function code, and the call hangs indefinately, furtherJ >>> proving that there is nothing in the mailbox!  So why is my AST hit asH >>> soon as I establish it for unsolicited writes to the mailbox, beforeF >>> anything is written to the mailbox?  I got the same results when I9 >>> tried with the READATTN function code just for kicks.y >>>  >>> Thanks.  >>B >>The AST arg to sys$qio[w] is the completion AST for the i/o, not7 >>the attention AST that might ultimately be delivered.u >>? >>The completion AST is delivered as soon as the i/o completes.'= >>This happens synchronously, since you are doing a sys$qiow.y= >>Delivery of this AST just means the driver has received the 9 >>io$_setmode function and processed it (by setting up ans" >>attention AST, but see below...) >>= >>You seem to be wanting myAST() to be the attention AST, notg< >>the completion AST, and you also seem to want nmbx_chan to= >>be the attention AST parameter.  These are passed as P1 ands: >>P2 to the QIOW (you are passing 0's for both these.)  So >>you should be using: >>E >>status = SYS$QIOW (0, nmbx_chan, IO$_SETMODE | IO$M_WRTATTN, &iosb,t( >>0, 0, &myAST,  nmbx_chan, 0, 0, 0, 0); >>8 >>If you want both a completion AST and an attention AST> >>(different routines, triggered under different circumstances; >>which presumably do different things) then you would use:e >>E >>status = SYS$QIOW (0, nmbx_chan, IO$_SETMODE | IO$M_WRTATTN, &iosb,w? >>&myAST_comp,  nmbx_chan, &myAST_attn, nmbx_chan, 0, 0, 0, 0);  >>@ >>where myAST_comp() is the completion (of the QIOW) routine and( >>myAST_attn() is the attention routine. >>< >>As someone else pointed out, be sure that any data that is? >>needed by an AST routine is statically allocated, or it couldl; >>be gone by the time the AST is triggered, and the routine A >>could end up reading or writing random memory!  This especiallym< >>applies to IOSB's used with SYS$QIO (non-"W"), but also toA >>anything else, such as the channel number you are using for thei >>AST parameter. >>? >>See Section 4.3.4 of the I/O User's Reference Manual (sectionnE >>cited is in the V7.3 manual; could be different in other versions.)' >> >>Hope this helps.   ------------------------------  % Date: Fri, 25 Jan 2002 16:08:51 -0500e2 From: "Sue Skonetski" <susan.skonetski@compaq.com>- Subject: new partner also sent to openvms.org-/ Message-ID: <aOj48.57$am1.892@news.cpqcorp.net>:  D IMSure Network, Inc. is a technology company that provides eCommerceG Internet technology products that make businesses more profitable.   We I provide eCommerce infrastructure services, enterprise level software, and I end-2-end eCommerce service provider (ESP) support. Our products, B2BflexeC and B2Cflex, enable eCommerce applications to be deployed, managed,tJ supported and upgraded from centrally located data centers, rather than on individual company servers.   G IMSure Network's  B2B/B2C e-commerce package is designed for hosting intJ multiple cross-platform environments.  The package was written entirely inK Java for that very reason.  In mid-summer 2001, it was decided to port thisiK package to the OpenVMS environment.  Ensuring that the software package waslH pure Java resulted in an application that is easy to run on any properlyE implemented Java virtual machine.  The key result is the applicationsoI ability to run on a wide variety of platforms with ease, including almosttG all operating systems and platforms currently available.  When platform I independence is needed, Java is clearly an outstanding technology to work. with.o       Regards, Mark Muehleisenb markm@imsurenetwork.com  iMSure Network, Inc. 480.456.3811 x 205   ------------------------------  # Date: Fri, 25 Jan 2002 18:53:00 GMTn2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: RE: New VMS newsgroup/ Message-ID: <gQh48.47$am1.671@news.cpqcorp.net>p   In article <35666012DF4CD411BE940090279FA240010BF1BA@ppnt41.physics.ox.ac.uk>, John Macallister <J.Macallister1@physics.ox.ac.uk> writes: L :As for the content, I don't believe subdividing the VMS list/group would be	 :useful. m  C   I concur.  I seriously doubt any subdivision also would have any bA   affect on sorting the torrent -- short of implementing and thene:   enforcing some level of newsgroup moderation, of course.  E   The OpenVMS newsgroups are already subdivided, BTW.  comp.os.vms*, sK   comp.sys.dec*, vmsnet.alpha*, vmsnet.announce, vmsnet.announce.newusers, e>   vmsnet.decus.journal, vmsnet.decus.lugs, vmsnet.employment, ?   vmsnet.epsilon-cd, vmsnet.groups, vmsnet.infosystems.gopher,  ?   vmsnet.infosystems.misc, vmsnet.internals, vmsnet.mail.misc, a/   vmsnet.mail, vmsnet.mail.pmdf*, vmsnet.misc,  C   vmsnet.networks.desktop.misc, vmsnet.networks.desktop.pathworks*,nE   vmsnet.networks.management.decmcc, vmsnet.networks.management.misc,u8   vmsnet.networks.misc, vmsnet.networks.tcp-ip.cmu-tek, @   vmsnet.networks.tcp-ip.misc, vmsnet.networks.tcp-ip.multinet*,?   vmsnet.networks.tcp-ip.tcpware*, vmsnet.networks.tcp-ip.ucx*,c1   vmsnet.networks.tcp-ip.wintcp, vmsnet.pdp-11*, fA   vmsnet.sdk.openvms.fieldtest, vmsnet.sources, vmsnet.sources.d, ?   vmsnet.sources.games, vmsnet.sysmgt, vmsnet.test, vmsnet.tpu, A   vmsnet.uucp, vmsnet.vms-posix, de.comp.os.vms*, fr.comp.os.vms,(:   umn.comp.os.vms, uw.vms, cz.comp.vms, pitt.comp.sys.vms,E   utexas.cc.sysmod.vms, is.tolvur.vms, and probably more I've missed.y  H     *Groups with more than 500 postings, as seen by the local newserver.  E   comp.os.vms has over 70,000 more postings (as seen by the the news ,F   server) than the next nearest comp.sys.dec, and orders of magnitude E   more posting traffic than any of the other newsgroups listed above.     N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Fri, 25 Jan 2002 19:09:01 GMTp2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: RE: New VMS newsgroup/ Message-ID: <h3i48.50$am1.713@news.cpqcorp.net>s   In article <2795B75EF003D311801A00A0C906B511011C6C7E@cucexec.gbc.getronics.nl>, "Dijk, Jeroen van" <Jeroen.vanDijk@Getronics.com> writes:t : H :Killfile is something from the unix/midrange world, but basicly it is a :filter.  J   Most Windows-based newsreaders apparently have this capability, as well.L   (I know that every one I've ever looked at on any recent platform version !   has had some form of killfile.)0  I :My filter makes overtime, because of the number of postings in INFO-VAX.:  E   Playing devil's advocate, we should change because your environment E   is underpowered and underconfigured.  (I can't say I've noticed anyh;   problems with MXRN killfile processing on OpenVMS Alpha.)n   M :Switching over to digest-format is not a option, the digest would get to bigmN :if it's done on a daily bases. Digest works only at low-volume newsgroups and :maillists.   G   Various newsreader clients also offer batch-like processing, meaning cI   you could implement a digest format yourself -- off-line, or as needed.    ..J :There has be done something and moderating is not a option. The technicalH :questions needs be seen and spread fast and no moderator can work 24/7 % :with a response time of 10 minutes. w  J   Sounds like you want it all -- fast posting turn-around, properly routedJ   postings, low dreck counts, and no local filtering overhead.  The norms C   within a newsgroup may prove somewhat disappointing, however. :-)-  F   There are programs which implement moderation, BTW.  Users known to G   post dreck or unknown or new users get moderated, and see some delay oF   in posting.  The users that are more "trusted" are processed throughG   more automated moderating means.  These can also include cancellation=#   features and mechanisms, as well.d  J   If you want a discussion forum for OpenVMS, well, those certainly exist.J   If you want volume -- pun intended -- this is it.  If you want and need K   answers, contact CSC.  If you want calmer and more structured discussion dH   fora, consider Encompasserve.  There are other options -- I have heardI   rumor that there was at least one website that was considering startingqH   up and OpenVMS discussion forum.  That said, I believe that trying to )   structure a usenet newsgroup is futile.a     Yours,   Sancho Panza.e  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Fri, 25 Jan 2002 12:09:40 -0800:0 From: Mark Berryman <mark.berryman@mvb.saic.com> Subject: Re: New VMS newsgroup+ Message-ID: <3C51BB84.5040904@mvb.saic.com>M   See comments below.t   Dijk, Jeroen van wrote:   	  >> ../..   >>d>  >>> I'd suggest switching over to digest-format mailings fromD  >>> INFO-VAX (this assumes that INFO-VAX provides digest-format) or#  >>> using a newsreader killfile tov  >>>  >> reduce the  >>h8  >>> newsgroup traffic -- my own killfile already gets a  >>>  >> regular workout.  >>g7  >> What is a killfile? A filter for newsgroup readers?-  >>-  >>-  >E  > Killfile is something from the unix/midrange world, but basicly itm  > is a filter.:  >A  > My filter makes overtime, because of the number of postings inq  > INFO-VAX.    G There are a large number of junk and spam postings that are filtered atTH the Info-VAX gateway (the filter is, unfortunately, not perfect and someD junk still makes it in).  However, subscribers to Info-VAX see a lot& less junk than those reading via news.      >D  > Switching over to digest-format is not a option, the digest would,  > get to big if it's done on a daily bases.    = The digest is created and mailed twice daily.  There are many)G subscribers to it and I have not heard anyone complain that its size isw	 unwieldy.B  <  > Digest works only at low-volume newsgroups and maillists.    G High-volume digests exist so there are certainly some who disagree withcH this opinion.  You could try subscribing to the digest and trying it out> rather than simply assuming it would not be practical for you.    E  > Secondly because my filters can only work with single messages ands  > not digests.      Umm, improve your filters?    C  > There has be done something and moderating is not a option.  Thet  > technicalD  > questions needs be seen and spread fast and no moderator can work+  > 24/7 with a response time of 10 minutes.e    D Actually, making comp.os.vms a moderated group is the only practicalB option (other than leaving things as they are, of course) and yourA assumptions about how moderation works are not necessarily valid.e    Consider the following scenario:  4 Comp.os.vms is reorganized as a moderated newsgroup.* The moderator sets up a "subscriber list".A Any post coming from someone on the "subscriber list" is approved / automatically, no manual intervention involved. ; Any other posting is treated the same as posting to a fully:F manually-moderated group (i.e., it requires the moderator's approval).I The moderator adds the authors of such postings to the "subscriber list" c as appropriate.dC Anyone the denizens of the newsgroup feels is abusing their postingeG privilege is simply removed from the "subscriber list".  (Which doesn'tsC mean they *can't* post, just that their postings now require manualn
 approval).  F Just some food for thought.  I am sure there are other viable answers 3 that are better than trying to split the newsgroup.k  
 Mark Berrymand Mark.Berryman@mvb.saic.com Info-VAX administrator   ------------------------------  % Date: Fri, 25 Jan 2002 16:37:44 -0500g% From: JF Mezei <jfmezei@videotron.ca>f Subject: Re: New VMS newsgroup, Message-ID: <3C51D024.102C4B2D@videotron.ca>  E Considering that a proxy vote may happen in March, any discussions ontN restructuring of VMS newsgroups should wait until the issue of the HP takeoverN of Compaq is settled. A couple of months won't make much of a different in the grand scale of things.   ------------------------------    Date: 25 Jan 2002 15:38:46 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)r Subject: Re: New VMS newsgroup3 Message-ID: <yyr1SyI40$OV@eisner.encompasserve.org>   ^ In article <3C51BB84.5040904@mvb.saic.com>, Mark Berryman <mark.berryman@mvb.saic.com> writes:  I > There are a large number of junk and spam postings that are filtered at)J > the Info-VAX gateway (the filter is, unfortunately, not perfect and someF > junk still makes it in).  However, subscribers to Info-VAX see a lot( > less junk than those reading via news.  B Please reposition the filter, then, because 95% of all spam that I+ see in the newsgroup _comes_from_ Info-VAX.e  F > Actually, making comp.os.vms a moderated group is the only practicalD > option (other than leaving things as they are, of course) and yourC > assumptions about how moderation works are not necessarily valid.o > " > Consider the following scenario: > 6 > Comp.os.vms is reorganized as a moderated newsgroup., > The moderator sets up a "subscriber list".C > Any post coming from someone on the "subscriber list" is approveda1 > automatically, no manual intervention involved.t= > Any other posting is treated the same as posting to a fullyIH > manually-moderated group (i.e., it requires the moderator's approval).  & How does this defend against forgery ?  B Or is the idea that spammers would not care enough about one group( to borrow the name of a regular poster ?   ------------------------------  # Date: Fri, 25 Jan 2002 21:22:26 GMT.4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk> Subject: Re: New VMS newsgroup/ Message-ID: <3C51CB9E.1A0769C@blueyonder.co.uk>i   Terry C Shannon wrote: > , > On Thu, 24 Jan 2002, Didier Morandi wrote: >  > > "Dijk, Jeroen van" wrote:t > > >r@ > > > Please split this newsgroup in to comp.os.vms.advocacy and > > > comp.os.vms.technical 0 > > > the volume of messages is getting too big. > >  > > Add my vote. > >e >  > Mine as well.g  3 I'm with Hoff etc and against you guys on this one.a  4 It seems most of the problem stuff at present is due" to inappropriate troll crossposts.   regardse   -- 2 Tim.Llewellyn@cableinet.co.uk     C Standard disclaimer applies. My views in no way represent those of $! my employers or service provider.n   ------------------------------  # Date: Fri, 25 Jan 2002 21:24:36 GMTt4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk> Subject: Re: New VMS newsgroup0 Message-ID: <3C51CC2A.A07310FC@blueyonder.co.uk>   Didier Morandi wrote:b >  > Hoff Hoffman wrote:e > >e > ../..t# > >   I'd suggest switching over tosF > >   digest-format mailings from INFO-VAX (this assumes that INFO-VAXJ > >   provides digest-format) or using a newsreader killfile to reduce theJ > >   newsgroup traffic -- my own killfile already gets a regular workout. > 5 > What is a killfile? A filter for newsgroup readers?T >   ; yup, you can usually filter posts based on sender or topic.   B Doesn't stop you seeing filtered persons posts quotes in followups though.    regardsJ > D.   -- : Tim.Llewellyn@cableinet.co.uk  s  C Standard disclaimer applies. My views in no way represent those of e! my employers or service provider.    ------------------------------  % Date: Fri, 25 Jan 2002 13:55:00 -080020 From: Mark Berryman <mark.berryman@mvb.saic.com> Subject: Re: New VMS newsgroup+ Message-ID: <3C51D434.3060702@mvb.saic.com>m   Larry Kilgallen wrote:  ` > In article <3C51BB84.5040904@mvb.saic.com>, Mark Berryman <mark.berryman@mvb.saic.com> writes: >  > I >>There are a large number of junk and spam postings that are filtered at J >>the Info-VAX gateway (the filter is, unfortunately, not perfect and someF >>junk still makes it in).  However, subscribers to Info-VAX see a lot( >>less junk than those reading via news. >> > D > Please reposition the filter, then, because 95% of all spam that I- > see in the newsgroup _comes_from_ Info-VAX.n    O What you see is what manages to leak through (which I think is in the range of nL an average of 1 or 2 posts a day).  But, yes, I do try to refine the filter , every so often to reduce the number further.     >  > F >>Actually, making comp.os.vms a moderated group is the only practicalD >>option (other than leaving things as they are, of course) and yourC >>assumptions about how moderation works are not necessarily valid.  >>" >>Consider the following scenario: >>6 >>Comp.os.vms is reorganized as a moderated newsgroup., >>The moderator sets up a "subscriber list".C >>Any post coming from someone on the "subscriber list" is approvede1 >>automatically, no manual intervention involved.c= >>Any other posting is treated the same as posting to a fullysH >>manually-moderated group (i.e., it requires the moderator's approval). >> > ( > How does this defend against forgery ? > D > Or is the idea that spammers would not care enough about one group* > to borrow the name of a regular poster ? >   N Spam mail frequently uses bogus addresses but rarely forged addresses.  There H are a number of reasons for this, one of which is the fact that spam is O blasted out to as many destinations as possible and forging an address for one tM destination doesn't help for the other destinations.  Of all the spam I have rM filtered out of Info-VAX, not one has ever had a forged address of a regular T poster.$    
 Mark Berryman    ------------------------------    Date: 25 Jan 2002 17:29:56 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)Q Subject: Re: New VMS newsgroup3 Message-ID: <rT$8EeXGv9YP@eisner.encompasserve.org>n  ^ In article <3C51D434.3060702@mvb.saic.com>, Mark Berryman <mark.berryman@mvb.saic.com> writes: >  >  > Larry Kilgallen wrote: > a >> In article <3C51BB84.5040904@mvb.saic.com>, Mark Berryman <mark.berryman@mvb.saic.com> writes:  >> l >> oJ >>>There are a large number of junk and spam postings that are filtered atK >>>the Info-VAX gateway (the filter is, unfortunately, not perfect and some G >>>junk still makes it in).  However, subscribers to Info-VAX see a lot3) >>>less junk than those reading via news.n >>>  >> >E >> Please reposition the filter, then, because 95% of all spam that I . >> see in the newsgroup _comes_from_ Info-VAX. >  > Q > What you see is what manages to leak through (which I think is in the range of 8N > an average of 1 or 2 posts a day).  But, yes, I do try to refine the filter . > every so often to reduce the number further.  
 But you said:s  G >>>junk still makes it in).  However, subscribers to Info-VAX see a lote) >>>less junk than those reading via news.C  < Are you saying there are non-Info-VAX spams to comp.os.vms ? I don't see them.   G >>>Actually, making comp.os.vms a moderated group is the only practicaloE >>>option (other than leaving things as they are, of course) and yourrD >>>assumptions about how moderation works are not necessarily valid. >>>i# >>>Consider the following scenario:o >>>s7 >>>Comp.os.vms is reorganized as a moderated newsgroup.y- >>>The moderator sets up a "subscriber list". D >>>Any post coming from someone on the "subscriber list" is approved2 >>>automatically, no manual intervention involved.> >>>Any other posting is treated the same as posting to a fullyI >>>manually-moderated group (i.e., it requires the moderator's approval).v >>>y >> l) >> How does this defend against forgery ?p >> nE >> Or is the idea that spammers would not care enough about one groupn+ >> to borrow the name of a regular poster ?t >> l > P > Spam mail frequently uses bogus addresses but rarely forged addresses.  There J > are a number of reasons for this, one of which is the fact that spam is Q > blasted out to as many destinations as possible and forging an address for one aO > destination doesn't help for the other destinations.  Of all the spam I have tO > filtered out of Info-VAX, not one has ever had a forged address of a regular k	 > poster.a  ; Well, we will just hope they don't read this discussion :-)    ------------------------------  % Date: Fri, 25 Jan 2002 16:51:19 -0800m0 From: Mark Berryman <mark.berryman@mvb.saic.com> Subject: Re: New VMS newsgroup+ Message-ID: <3C51FD87.5000708@mvb.saic.com>i   Larry Kilgallen wrote:  ` > In article <3C51D434.3060702@mvb.saic.com>, Mark Berryman <mark.berryman@mvb.saic.com> writes: >  >> >>Larry Kilgallen wrote: >> >>a >>>In article <3C51BB84.5040904@mvb.saic.com>, Mark Berryman <mark.berryman@mvb.saic.com> writes:, >>>e >>>l >>>tK >>>>There are a large number of junk and spam postings that are filtered attL >>>>the Info-VAX gateway (the filter is, unfortunately, not perfect and someH >>>>junk still makes it in).  However, subscribers to Info-VAX see a lot* >>>>less junk than those reading via news. >>>> >>>>E >>>Please reposition the filter, then, because 95% of all spam that It. >>>see in the newsgroup _comes_from_ Info-VAX. >>>m >>Q >>What you see is what manages to leak through (which I think is in the range of rN >>an average of 1 or 2 posts a day).  But, yes, I do try to refine the filter . >>every so often to reduce the number further. >> >  > But you said:s >  > H >>>>junk still makes it in).  However, subscribers to Info-VAX see a lot* >>>>less junk than those reading via news. >>>> > > > Are you saying there are non-Info-VAX spams to comp.os.vms ? > I don't see them.p    M Yes, SPAM does come in from the USENET side of the house and I catch most of hM it (hence my comment above).  However, I also see cancel messages frequently hL come in for said SPAM (but only because my gateway logs the presence of the M cancel message then tosses it since it would be meaningless to forward it to  N the Info-VAX mailing list).  I'm assuming these are generated by a Cancel-Bot M running somewhere but have never checked.  Since I grab all messages as soon  J as they come in for gateway processing, I see them even if they later get F cancelled.  Perhaps, by the time you read comp.os.vms, they have been ! cancelled and you don't see them?-  N I have to admit, however, that the most recent SPAM messages have all come in @ via email and I am currently refining my filter to address that.  
 Mark Berryman@   ------------------------------  # Date: Sat, 26 Jan 2002 01:22:00 GMTEL From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr") Subject: RE: New VMS newsgroup8 Message-ID: <00A0894D.432AE382@SSRL04.SLAC.STANFORD.EDU>  c In article <3YrTtc21ZyD$@eisner.encompasserve.org>, Kilgallen@SpamCop.net (Larry Kilgallen) writes:e >In article <2795B75EF003D311801A00A0C906B511011C6C90@cucexec.gbc.getronics.nl>, "Dijk, Jeroen van" <Jeroen.vanDijk@Getronics.com> writes: >> r; >>> >> 1. Taking all information out of comp.os.vms (as is s >>> required in suchJ >>> >>    splits) would make it hard for those with borderline interest to >>> >>    find the discussion. >>> >  >>> > That is the down part, >>> ? >>> And getting more people enthusiastic about VMS is our most w >>> crucial job. >>> E >>> > but there are enough people that crosspost or place a followup d >>> > in the right group.  >>> , >>> Crossposts are the worst part of Usenet. >> iD >> You can filter that you only read posts that are not crossposted. >e% >Which VMS newsreader provides that ?i  L NEWSRDR does.  (By default, it keeps track of message ids and won't show youL the same message twice in two different newsgroups.  But you can also put inK an ignore specification that won't show you anything that's posted to more oN than one newsgroup. [Mine is set to not show anything posted to three or more,, and this keeps the SPAM down to a minimum.])   -- Alanh    O =============================================================================== 0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056pM  Physical mail to: SSRL -- SLAC BIN 69, PO BOX 4349, STANFORD, CA  94309-0210nO ===============================================================================r   ------------------------------    Date: 25 Jan 2002 23:12:49 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)e Subject: Re: New VMS newsgroup3 Message-ID: <bxL2xijfuoVa@eisner.encompasserve.org>o  ^ In article <3C51FD87.5000708@mvb.saic.com>, Mark Berryman <mark.berryman@mvb.saic.com> writes:  O > Yes, SPAM does come in from the USENET side of the house and I catch most of  O > it (hence my comment above).  However, I also see cancel messages frequently  N > come in for said SPAM (but only because my gateway logs the presence of the O > cancel message then tosses it since it would be meaningless to forward it to -P > the Info-VAX mailing list).  I'm assuming these are generated by a Cancel-Bot O > running somewhere but have never checked.  Since I grab all messages as soon fL > as they come in for gateway processing, I see them even if they later get H > cancelled.  Perhaps, by the time you read comp.os.vms, they have been # > cancelled and you don't see them?-  8 Quite likely.  I know my news server does honor cancels.   ------------------------------  # Date: Fri, 25 Jan 2002 21:22:17 GMTi4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk>$ Subject: Re: Newbie Help on Security0 Message-ID: <3C51CB82.919E5E1E@blueyonder.co.uk>   Paul Repacholi wrote:e > 8 > Tim Llewellyn <tim.llewellyn@blueyonder.co.uk> writes: > C > > yeah, why do you need a firewall anyway. It it wasn't for thoserG > > pesky windows and unix boes on your LAN you could dispense with it.n >  > > with tongue in cheek.i > H > Remove the tongue! With SSH, having the VMS machine in parellel to theE > firewall so you can sort out the wreckage via VMS. :) Plus, if theysG > find an exposed mmachine, with luck they will stub their toes on thatu: > before they try your FW. And the alarms can tip you off. >  > But SSH is a MUST for this!!  % I have no doubt you are correct Paul,a  C Selling the idea to corporate network security people etc could be   interesting though.t   regards'    e --   Tim.Llewellyn@cableinet.co.uk  r  C Standard disclaimer applies. My views in no way represent those of  ! my employers or service provider.    ------------------------------  # Date: Fri, 25 Jan 2002 21:42:54 GMT 4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk>P Subject: Re: Origins and dated nature of passord dictionary (was Re: bkjfn is an0 Message-ID: <3C51D074.7A334086@blueyonder.co.uk>   Larry Kilgallen wrote: > R > In article <3C5194D8.CF2D11EB@ACM.org>, "C.W.Holeman II" <cwhii@ACM.org> writes: > > Nic Clews wrote:* > >> We recently had a case of someone whoL > >> change their password so many times, it blew the password history! What8 > >> I can't get over was they used generated passwords! > F > The field test of password history did not have that feature -- thatE > is why it is always good to seek wider input (as in field test) for  > your software. > A > > Maybe they rolled through them until they got one they liked.r > B > No, I think the point is that a normal user who tries to use theB > overflow the history buffer so they can choose the same passwordE > again will find that once the history buffer is full their usernamei; > is automatically changed to require a generated password.  > @ > A user who was already using generated passwords should see no/ > difference when they fill the history buffer.h  Q yup, I actually had someone try that once because I refused to just mcr authorizesF and give them back their old password. Eventually I gave in and turnedL of generated passwords for that user after he enabled them by filling up the# history buffer in an afternoon :-).  -- t Tim.Llewellyn@cableinet.co.uk  n  C Standard disclaimer applies. My views in no way represent those of e! my employers or service provider.h   ------------------------------    Date: 26 Jan 2002 04:55:07 +0800, From: Paul Repacholi <prep@prep.synonet.com>4 Subject: Re: Problem with TSM and ETS-Terminalserver- Message-ID: <873d0utal0.fsf@prep.synonet.com>e  = "Thomas Egenberger" <thomas.egenberger@technidata.de> writes:3   > Hello, > B > I have problems to add a Lantronix ETS16PR Terminalserver to the? > TSM-Database on an Alpha OpenVMS V7.1-1H2 with TSM 2.1 ECO 7.   iH > According to the Lantronix manual, The ETS16PR supports the management > software TSM and NCP.rH > according to the help of TSM, the server type ETS is supported by TSM.   G > Before trying this, I copied the Terminal Firmwarefile ETS16pr.sys tor > mom$load:AIT03TSV.SYS.I > The logical name mom$load includes the Directory SYS$COMMON:[DECSERVER]<  1 > Following problem exists:a  B > A35M1_TEX_3> tsm! > Terminal Server Manager V2.1-07 L > Copyright  Digital Equipment Corporation. 1994, 1995 All Rights Reserved. > Usage is DIRECTORY > TSM> add servery@ > Name                                                    :ait03L > Address                                                 :00-08-A3-23-99-14@                                                            ^^^^^  C That is a DEC address, not a Lantronix one. It should be 00-80, nots 00-08...   -- g< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.o@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Sat, 26 Jan 2002 03:55:47 GMTe1 From: "David J. Dachtera" <djesys.nospam@fsi.net>d. Subject: Re: What is unix for SEARCH /WINDOW=0' Message-ID: <3C5229DD.B20462A6@fsi.net>g   "David J. Dachtera" wrote: >  > "C.W.Holeman II" wrote:s > >  > > What is unix for:? > >r > >         $ SEARCH /WINDOW=0 > I > On Solaris 8, I believe it's "find -name ..." Not sure about other UN*X- > systems...  C Sorry - that's wrong. That will only find all the files and put thelH path+name to stdout. Somehow you gotta use find in a pipeline to recurse8 the directory tree (if that's your intent), "grep" for aE string/expression, then echo the filespec if the string/expression is  found in that file.o  G I know there's a way to do it - just can't recall the syntax right now.a- Has to do with the "-exec" option, I think...u   -- e David J. Dachterao dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/t   ------------------------------    Date: 26 Jan 2002 05:30:57 +0800, From: Paul Repacholi <prep@prep.synonet.com>: Subject: Re: Wierd File Reading Behaviour - Dec C on Alpha- Message-ID: <87u1tarucu.fsf@prep.synonet.com>e  ' robert.james@uk.ngrid.com (rob) writes:-  F > The only way I can get it to be as slow as the default (recommended)C > behaviour is to do a setbuf(fp,NULL), which turns buffered IO offe
 > completely!C  . > What is going on ?!u  0 You have turned on buffering. Don't ask, VaxC...    D > FYI I want to sequentially read thousands of these files every day, > with as little resource usage as possible.  ? See the changes I did to gzip on the last freeware, in Taylor.hl
 and adapt.   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.g@                                              West Australia 6076. 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, 25 Jan 2002 23:26:49 -0000w, From: "pocketmoon" <pocketmoon@ntlworld.com>: Subject: Re: Wierd File Reading Behaviour - Dec C on AlphaC Message-ID: <eIl48.24693$Ph2.4199207@news2-win.server.ntlworld.com>r  7 Paul Repacholi <prep@prep.synonet.com> wrote in message-' news:87u1tarucu.fsf@prep.synonet.com...a) > robert.james@uk.ngrid.com (rob) writes:B >5H > > The only way I can get it to be as slow as the default (recommended)E > > behaviour is to do a setbuf(fp,NULL), which turns buffered IO offi > > completely!  >o > > What is going on ?!i >u2 > You have turned on buffering. Don't ask, VaxC... >i >n <SNIP>   Eek!  ) So it's doing unbuffered IO by default ?!h  A Hmm of to try IO_PERFORM if I can get the Ops Manager to grant beh VMS$BUFFER_THINGY_USER privs...I   ------------------------------  # Date: Sat, 26 Jan 2002 03:09:11 GMTt- From: "Richard L. Dyson" <rickdyson@home.com>h- Subject: Re: WTB:Freeware 4 or 5 CD for sale?.( Message-ID: <3C521DD6.D9E7BAB6@home.com>  J > I don't know if I have a spare but I've be willing to burn copies of theK > CDs. I don't believe there to be any problem with making a copy as all ofaJ > the software is freeware and free to distribute.  Unless, of course, theG > CD "compilation" is considered Copyrighted by Compaq (d|i|g|i|t|a|l).a  F 	As Hoff pointed out, Compaq doesn't care if you make copies of the FW discs.  C 	However, I would point out that not everything on all of the discsbM was indeed "free".  Digital did want that to be true and then Compaq too, but K some of the software that has been included on several of the distributionsdH is actually shareware where the author did ask for $$ if you used it for non-personal use.a   Rick   ------------------------------  % Date: Sat, 26 Jan 2002 01:13:36 +0000g0 From: Roberta Davies <roberta.davies@virgin.net>D Subject: Re: Younger recruits versus experienced veterans  ( was The* Message-ID: <3C5202C0.C26EA822@virgin.net>   Bruce Hoult wrote: > 6 > In article <3C4E5AB1.C75106D@videotron.ca>, JF Mezei > <jfmezei@videotron.ca> wrote:s >  > > Pat Durkin wrote:sD > > > Oh, the Vegemite website says J.L. Kraft and his brothers wereI > > > originally from Canada.  Don't know about that.  Do Canadians enjoyo > > > Marmite and Vegemite?  > >-L > > Nop. My theory is that baby australians are spoon fed some of that stuff > > 5.K > > seconds after their umbilical cord is cut. Their system feels instantlyvI > > deprived so it learns to accept any substance as food, hence the baby> > > becomesaM > > accustomed to that yest infection stuff. But if the baby doesn't get that>C > > treatment at birth, he/she is not likely to be able to eat thaty > > substance. > $ > I don't recall such an experience. > D > Also, my partner is from Europe (came to NZ in her early 30's) and( > probably eats more Vegemite than I do. > * > > As for myself, I am partial to Nutella >  > Ick!!s > 7 > "Less sugar than some jams, less fat than pure lard".n > I > Gimme peanut butter any day.  And not that US stuff with loads of sugar I > either.  NZ peanut butter has peanuts, a little salt, and nothing else.    Not on scones, surely!   Robbie   ------------------------------  % Date: Sat, 26 Jan 2002 01:22:50 +0000a0 From: Roberta Davies <roberta.davies@virgin.net>D Subject: Re: Younger recruits versus experienced veterans  ( was The* Message-ID: <3C5204EA.DDB8C718@virgin.net>   JF Mezei wrote:  >  > Brian {Hamilton Kelly} wrote:nM > > be sweetish, whereas a crumpet is more likely to be thought of as savory.  > > L > > Wonderful when toasted, spread with butter, and just a smear of Marmite. > K > Was all fine until you mentioned Marmite. Never quite figured out why the N > brits (Marmite), the Aussies and Kiwis (Vegemite) would have such an apetiteO > for yeast infection. I think that offering Vegemite to anyone under 18 shoudlt" > be considered child molestation. > O > And what I don't quite get is why the USA is guilty of inflicting Aussies andiJ > Kiwis with Vegemite (it is made by Kraft, a USA corporation). And poepleO > complain about a few talibans being held by americans in a tropical "resort".iK > Interesting that nobody would have complained if they had been sent to anf > alaskan prison.   H The Marmite TV ads in Britain are based on the theme "you either love itG or hate it".  There is no middle of the road when it comes to Marmite!  1 (Personally, I can't stand to be near the stuff.)d  C The first ad showed a Marmite promotion outside a supermarket, withhE pretty girls distributing samples of Marmite on toast to passers-by.  H Some grabbed eagerly, some practically mobbed the girls; others ran awayB screaming, and one man (after my own heart) threw a piece down and ground it under foot.a  E Another showed a young couple cuddling on the sofa.  They kiss.  NextyE scene he's in the bathroom, gagging and swilling down mouthwash.  SheaG has a Marmite-spread crumpet on the table next to her side of the sofa,m with one bite taken out of it.  D Our favourite showed a 20-something man in an obviously shared flat,G staggering down to find something for breakfast.  He sniffs at the milkaH and recoils in disgust.  He makes a piece of toast and butters it, usingH a knife he picks up from the draining board.  He takes a bite -- arrgh! F Spits the toast out, looks around desperately.  Grabs the curdled milkC and gulps it down.  The knife has a smear of Marmite on the side heo didn't notice.   Robbie   ------------------------------    Date: 25 Jan 2002 17:33:49 -1000/ From: Jim Thomas <thomas@atlas.cfht.hawaii.edu>eD Subject: Re: Younger recruits versus experienced veterans  ( was The2 Message-ID: <wwn0z1n5uq.fsf@atlas.cfht.hawaii.edu>  > >>>>> "Bob" == Bob Koehler <koehler@encompasserve.org> writes:  i  Bob> In article <slrna50mvs.j9.stanb@citadel.metropolis.local>, stanb@dial.pipex.com (Stan Barr) writes:y  >> J  >> There's an old country saying: "You can eat every part of a pig exceptH  >> the grunt!".  If you live in farming country, you learn _nothing_ is
  >> wasted...c  L  Bob>    They way I heard it was "You can use everything except the squeal".2  Bob>    When I woked at McD's we used the squeal.  ) I didn't know McD's did Chinese meals :-),   Notheadc   ------------------------------  + Date: Fri, 25 Jan 2002 19:07:13 +0000 (UTC)e" From: greenaum@BOLLOCKSyahoo.co.ukY Subject: Re: Younger recruits versus experienced veterans ( was The demise    of      comn5 Message-ID: <3c7aab61.343198970@news.btopenworld.com>d  7 On 24 Jan 2002 13:22:17 +0100, Martin Hyer Kristiansens <martin@netimage.dk> sprachen:  E >He spoke danish with english grammar and scottish accent; completelyI >incomprehensible.  B Yeah my German grandmother speaks with a Yorkshire accent acquiredB from being in the UK longer than I have (and I'm native). But I doD notice the odd consonant being a little bit sharp, and the odd vowelB being a bit to the left or right (vowel shift, yknow). And since IE learned German at school I've also noticed her grammar is sometimes akB bit Deutsche, but to everyone else who hasn't studied German you'd never perceive it.  F Meanwhile she often doesn't know what her German relatives are talkingD about, because she only visits the Vaterland every few years, so her German is 45 years out of date.e  H ------------------------------------------------------------------------  H if love is a drug, then, ideally, it's a healing, healthful drug... it's@ kind of like prozac is supposed to work (without the sexual side5 effects and long-term damage to the brain and psyche)t   ------------------------------  + Date: Fri, 25 Jan 2002 19:08:36 +0000 (UTC)a" From: greenaum@BOLLOCKSyahoo.co.ukY Subject: Re: Younger recruits versus experienced veterans ( was The demise    of      come5 Message-ID: <3c7bac5a.343448300@news.btopenworld.com>   C On Thu, 24 Jan 2002 18:23:47 GMT, CBFalconer <cbfalconer@yahoo.com>t	 sprachen:   B >You are making a common mistake.  Only Canadians originating from0 >the western suburbs of Montreal speak English,   E I'm from God's Own County, matey-boy, nothing I've ever done has been 	 "common".a  5 WE speak English, everyone else is a soft Southerner.s  H ------------------------------------------------------------------------  H if love is a drug, then, ideally, it's a healing, healthful drug... it's@ kind of like prozac is supposed to work (without the sexual side5 effects and long-term damage to the brain and psyche)k   ------------------------------   Date: 25 Jan 2002 19:06:55 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)=B Subject: Re: [Q] Why is there the limit of 8 levels of directories, Message-ID: <a2sacf$14ei$1@info.cs.uofs.edu>  D All this reminds me of my days SysAdmining Pr1mes.  It seems someoneC who had never worked worked with tree style directories was curious=> about how far down you could actually go.  He never found out.D He wrote a little program that created a directory, moved to it, andD repeated.  It was impossible to tell how deep he was when it crashedB the system.  It was also impossible to do anything with this superA deep tree.  You couldn't even delete it.  We ended out backing upoC everything else ont he disk and re initializing the whole thing anduD then restoring all the user files.  In those days, on that hardware,> it took more than a a day and the other users were not amused.  9 Maybe having a limit below critical mass is a good idea!!l   bill   -- oJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   o   ------------------------------  # Date: Fri, 25 Jan 2002 19:39:36 GMTe* From: "Bill Todd" <billtodd@metrocast.net>B Subject: Re: [Q] Why is there the limit of 8 levels of directoriesB Message-ID: <Yvi48.111141$TC1.7723928@bin5.nnrp.aus1.giganews.com>  > "Bill Gunshannon" <bill@triangle.cs.uofs.edu> wrote in message& news:a2sacf$14ei$1@info.cs.uofs.edu...F > All this reminds me of my days SysAdmining Pr1mes.  It seems someoneE > who had never worked worked with tree style directories was curiousl@ > about how far down you could actually go.  He never found out.F > He wrote a little program that created a directory, moved to it, andF > repeated.  It was impossible to tell how deep he was when it crashedD > the system.  It was also impossible to do anything with this superC > deep tree.  You couldn't even delete it.  We ended out backing up E > everything else ont he disk and re initializing the whole thing andeF > then restoring all the user files.  In those days, on that hardware,@ > it took more than a a day and the other users were not amused. > ; > Maybe having a limit below critical mass is a good idea!!l  I Reminds me of the first test program I wrote for RMS-11.  Figured I'd see@H how well it handled max-size (255-byte) keys.  Unfortunately, no one hadI thought to ensure that the bucket size would always be made sufficient to-G include more than a single key:  when the default (single-block) buckettK turned out to be too small to contain two keys (including bucket overhead),oL the first insert immediately drove the tree depth to 255 (IIRC there *was* a1 limit check there, albeit of the BPT variety...).<   - bill   ------------------------------   End of INFO-VAX 2002.049 ************************